A Proteomic Study on Liver Impairment in Rat Pups Induced by Maternal Microcystin-LR Exposure*

Environmental Pollution 212 (2016) 197e207

Contents lists available at ScienceDirect

Environmental Pollution

journal homepage: www.elsevier.com/locate/envpol

A proteomic study on liver impairment in rat pups induced by maternal microcystin-LR exposure*

* Sujuan Zhao, Ping Xie , Jun Chen, Luyi Liu, Huihui Fan

Donghu Experimental Station of Lake Ecosystems, State Key Laboratory of Freshwater Ecology and Biotechnology of China, Institute of Hydrobiology, Chinese Academy of Sciences, Donghu South Road 7, Wuhan 430072, PR China article info abstract

Article history: There is mounting evidence indicating that microcystins (MCs) are heptapeptide toxins. Recent studies Received 1 September 2015 have also shown that MCLR can transfer from mother to offspring, but it is unclear whether maternal Received in revised form MCLR can influence the liver of offspring or not. In this study, pregnant SD rats were injected intra- 14 December 2015 peritoneally with a saline solution (control) or 10 mg/kg MCLR per day from gestational day 8 (GD8) to Accepted 23 December 2015 postnatal day 15 (PD15) for a total of 4 weeks. 2-DE and MALDI-TOF-TOF mass spectrometry were used to Available online xxx screen for MCLR target proteins in the livers of rat pups. Our results demonstrated that MCLR could accumulate in the livers of neonatal rats. Proteomics studies also showed that MCLR significantly Keywords: fl MCLR in uenced many proteins, including those involved in the cytoskeleton, metabolism and particularly Rat pups oxidative stress. In addition, MCLR induced cellular structural damage and resulted in the production of Proteomics intracellular reactive oxygen species (ROS) and lipid peroxidation. Moreover, protein phosphatase (PP) Liver activity was inhibited and some serum biochemistry parameters were altered. These results suggest an Oxidative damage early molecular mechanism behind the hepatotoxicity induced by maternal MC exposure and highlight the importance of monitoring MC concentrations in new-born mammals. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction fishermen living around Lake Chaohu, China (Chen et al., 2009), which demonstrates that chronic exposure is a health risk. Microcystins (MCs) are a family of cyclic heptapeptide toxins The mechanisms of MC toxicity have not been fully elucidated naturally produced by freshwater cyanobacteria, and their pro- but most likely involve protein phosphatase inhibition. Micro- duction has become a worldwide ecological issue that results in cystins act by inhibiting Ser/Thr protein phosphatase-1 and -2A water eutrophication and health threats to livestock and humans (PP1, PP2A) (Mackintosh et al., 1990), which leads to cytoskeletal (Paerl and Huisman, 2009). More than 80 structural analogues of disruption (Eriksson et al., 1989) and subsequent cell death (Boe MCs have been identified; of these, MCLR is the most toxic and et al., 1991). It has also been widely reported that MCs can induce most commonly encountered cyanobacterial toxin (Figueiredo de the production of reactive oxygen species (ROS), which may cause et al., 2004). There is evidence that prolonged exposure to MC-LR DNA damage (Nong et al., 2007). may induce neoplastic nodular formations that can be precursors Currently, mounting evidence has indicated that MCs have of primary liver cancer (PLC) (Yu, 1995; Ueno et al., 1996; Li et al., embryonic toxicity in both fish and mammals. The main effects of 2009). MCLR is classified as possibly carcinogenic to humans exposure to MCs during the early life stages of fish are interference (Group 2B) by the International Agency for Research on Cancer with developmental processes and organ functions (Palikova et al., (IARC) (2010). Recently, MCs were detected in the serum of 2003). In mammals, it has been demonstrated that MCLR can damage the placental barrier directly, allowing MCLR to enter the embryo (Wei et al., 2002). There is some evidence that MCs can inhibit growth, affect actin and microtubule organization and alter * This paper has been recommended for acceptance by David Carpenter. the morphology of mammalian embryos (Rao et al., 1998; * Corresponding author. Donghu Experimental Station of Lake Ecosystems, State Key Laboratory for Freshwater Ecology and Biotechnology of China, Institute of Sepulveda et al., 1992; Frangez et al., 2003). Moreover, Bu et al. Hydrobiology, the Chinese Academy of Sciences, Donghu South Road 7, Wuhan (2006) found that foetal livers exhibit petechial haemorrhaging 430072, PR China. Fax: þ86 186 27 68780622. and hydropic degeneration when exposed to 6 and 12 mg/kg E-mail address: [email protected] (P. Xie). http://dx.doi.org/10.1016/j.envpol.2015.12.055 0269-7491/© 2016 Elsevier Ltd. All rights reserved. 198 S. Zhao et al. / Environmental Pollution 212 (2016) 197e207 microcystin from GD 6 to GD15. Recently, Li et al. (2011b) suggested delivered dose of MCLR decreased as the pregnancy progressed that chronic exposure to microcystin may be associated with because the weight of the mother at GD8 was used to calculate the increased serum enzyme levels and liver damage among the MCLR dose, and the body weight increased from this point studied children from the Three Gorges Reservoir Region. Although throughout pregnancy. liver is the target organ and is therefore the most extensively In rats, blastocyst implantation in the uterus occurs until the 5th studied organ in research on MCs, the molecular mechanisms of gestational day, and organogenesis occurs starting from the 6th day MC-induced susceptibility to liver damage in offspring are largely (Manson and Kang, 1989). Thus, GD8 (during the stable organo- unknown. genesis stage) was chosen for the start of exposure. Given that both Pregnancy is an important and critical phase, during which gestation and lactation are more sensitive periods of life than the organogenesis can occur (Manson and Kang, 1989). Because subtle normal physiological phase, the MCLR exposure period included 14 anomalies that can appear during animal development may be days of pregnancy and 14 days of lactation. 10 mg MCLR/kg BW/day undetectable using conventional cytopathology and biochemical was chosen for this study based on the protocol of Li et al. (2012). analysis, a proteomics-based approach has been employed to study During the 4 weeks (28 days) of maternal MCLR exposure, the dams the effects of MC exposure (Li et al., 2012; Li et al., 2011a; Zhao et al., delivered naturally, and litters were killed on postnatal day 15. 2012a; Fu et al., 2005). However, there are few reports on the effects of MCLR during gestation and lactation (Chernoff et al., 2000). Thus, 2.3. MCLR concentration and effect of MCLR on pup livers in this study, we investigate hepatotoxicity of MCLR in rat pups, aiming to understand the mechanisms of MCLR-induced liver After blood collection (approximately 0.35 mL per rat), the pups damage in offspring. were sacrificed, and the livers were removed and weighed. The liver/body weight ratio was calculated. The liver index was calcu- 2. Materials and methods lated as follows: liver weight (g)/body weight (g) 100%. The livers of 10 pups from different dams in both groups were randomly 2.1. Chemicals chosen and analysed. The extraction and quantitative analysis of the MCLR content in the rat liver (0.2 g lyophilized sample) was The cyanobacterial toxin MCLR was isolated and purified from performed as published by Wang et al. (2008). freeze-dried surface blooms collected from Lake Dianchi using the methods previously published by Dai et al. (2008). MCLR was 2.4. Light microscopy, electron microscopy and serum biochemical separated using semi-preparative high-performance liquid chro- analyses matography (Waters 600E, USA), and pure MCLR was obtained. The MC content was determined using liquid chromatography- In preparation for light microscopy, liver samples were fixed in eelectrospray ionizationemass spectrometry (LCeESIeMS, 10% buffered formalin for 24 h at 4e8 C and then immediately Thermo Electron Corporation, Waltham, MA). The MCLR (purity dehydrated in a graded series of ethanol, immersed in xylol and >95%; the remaining components were primarily pigments) con- embedded in paraffin wax using an automatic processor. 4-mm centration was determined using UV spectra and retention times. A sections were mounted. Following deparaffinization, the sections commercial microcystin-LR standard (Wako Pure Chemical In- were rehydrated, stained with hematoxylin and eosin, and subse- dustries, Japan, purity > 95%) was used to compare the peak areas quently subjected to pathological assessment. of test samples (Fig. S1AeD in the Supporting Information). All In preparation for transmission electron microscopy, samples other chemicals used in this study were of analytical grade, and the were fixed in 2.5% glutaraldehyde fixative (in pH 7.4 phosphate chemicals used for electrophoresis were obtained from Bio-Rad buffer for 10 h at 4 C) and post-fixed in 1% osmium tetroxide Laboratories (Hercules, California, USA). fixative (in pH 7.4 phosphate buffer for 0.5 h at 4 C). After rinsing with phosphate buffer, the specimens were dehydrated in a graded 2.2. Animals ethanol series of 50e100% and then embedded in Epon 812. Ultra- thin sections were prepared with glass knives on an LKB-V ultra- Ten female SpragueeDawley [Crl:CD(SD)] rats and 10 male SD microtome (Nova, Sweden), stained with uranyl acetate and lead rats (both 12 weeks of age) were supplied by the Wuhan University citrate and examined by transmission electron microscopy using a Laboratory Animal Research Center (Hubei, China). The rats were Tecnai G2 20 TWIN (FEI, USA). Three samples from each dam were housed under controlled conditions of a 12 h light/dark cycle, randomly chosen for histological and transmission electron mi- 50 ± 5% humidity and 20 ± 1 C. The animals were allowed free croscopy analysis. access to food and water and were treated humanely to minimize Three pups from each dam were randomly chosen for plasma suffering. All animal procedures were approved by the Institutional biochemical analyses, and blood samples were collected and Animal Care and Use Committee (IACUC) and were in accordance centrifuged at 850 g for 10 min. An automated analyser (Beckman with the National Institutes of Health Guide for the Care and Use of coulter LX-20, USA) was used for all serum analyses. The harvested Laboratory Animals (permit number SCXK 2008-0004). serum was used to determine the serum activity of alanine Female and male rats were randomly assigned to the treatment aminotransferase (ALT), aspartate aminotransferase (AST), cholin- and control groups. Each group has five female and five male SD esterase (CHE), albumin (ALB), pre-albumin (PA) and bile acid rats. They were allowed to acclimate for 5 days before being paired (TBA). for mating. The day on which a vaginal plug was detected in the morning and sperm were found on a vaginal smear was considered 2.5. ROS production, lipid peroxidation (LPO), GSH content and gestational day (GD) 1. On the evening of GD8, the dams were protein phosphatase (PP) activity analysis implanted intraperitoneally with Alzet osmotic pumps (model 2004; Alza Corp., Palo Alto, CA, USA) following the manufacturer's Reactive oxygen species (ROS) were measured as previously recommendations. The pumps were designed to deliver vehicle described with some modifications (Tedesco et al., 2001). The ROS alone (0.9% saline solution) or an MCLR solution (10 mg MCLR/kg kit was purchased from the Beyotime Institute of Biotechnology BW/day). These pumps continued to release solution at a constant (Jiangsu, China). Briefly, the liver tissue was broken down into small rate (0.25 mL/h) for 4 weeks. It should be noted that the actual pieces and then digested to give a single-cell suspension using 0.1% S. Zhao et al. / Environmental Pollution 212 (2016) 197e207 199 collagenase II. The cells were then incubated with 6-carboxy- analysed with a PDQuest software package (BioRad) according to 20,70dichloro-dihydrofluorescein diacetate (DCFH-DA) at a final the following procedures: spot detection, spot edition, background concentration of 10 mM for 20 min and washed three times with subtraction, spot matching, and normalization. The protein spots saline (0.9%) solution. ROS production in the liver was measured were detected automatically and edited manually to remove fluorometrically (FACS Aria TMⅢ, Becton and Dickinson Company, streaks, speckles and artifacts. The spots that changed in the control USA) with excitation and emission at 488 and 525 nm, respectively. groups or exposure groups were then selected for further MS The results were expressed as arbitrary units. analysis. The concentration of malondialdehyde (MDA) was measured Spots that changed by about two-fold were selected for analysis using the TBA method to assess the extent of LPO (Du and Bramlage, with a MALDI TOF/TOF mass spectrometer (4800 Proteomics 1992). MDA assay kits were supplied by the Nanjing Jiancheng Analyzer, Applied Biosystems, USA). Mass spectra (m/z Bioengineering Institute in China. The level of the MDA-TBA adduct, 800e4,000Da) were acquired in a positive ion mode. GPS Explorer which was formed by the reaction of MDA with TBA at a high software (v3.6, default parameters, Applied Biosystems) was used temperature (90e100 C) and under acidic conditions, was to generate the peak lists for the database search against the NCBInr measured colourimetrically at 532 nm. 20120805 Rattus norvegicus protein database (19668816 se- The GSH content was measured according to the method pub- quences) using Mascot search engine (v2.2). The identification lished by Griffith (1980), and the kits were supplied by the Nanjing parameters were set as follows: species, R. norvegicus; enzyme, Jiancheng Bioengineering Institute, China. trypsin; allow for one missed cleavage site; fixed modification, The PP activity was analysed according to the method of Fontal carbamidomethyl (C); variable modification, oxidation (M); peptide et al. (1999) with some modifications. 0.2 mg liver tissue was charge, 1þ; monoisotopic; mass tolerance, ±50 ppm for the pre- sonicated in 5 volumes of homogenizing buffer containing 5 mM cursor ions and ±0.5 Da for fragment ions. The mascot protein score TriseHCl (pH 7.5), 0.5 mM EGTA, 1 mM EDTA, 1 mM 2- was calculated as 10*Log(P), where P was the probability that the mercaptoethanol, and 1 mM phenylmethylsulfonyl fluoride, and observed match was a random event. Protein scores greater than the tissue was homogenized with a glass homogenizer (Isamu et al., 61, indicating a low probability (<5%) that the observed match was 2010). The supernatant was recovered by centrifugation (20,000 g) a random event, were identified as proteins. The specific processes at 4 C for 30 min. 35 mL of liver homogenate was mixed with 5 mLof or functions of the identified proteins were then identified by 1 NiCl2 (40 mM), 5 mL of 5 mg mL bovine serum albumin (Sigma) searching Gene Ontology (http://www.Geneontology.org). Meta- and 35 mL of phosphatase assay buffer (50 mM TriseHCl, 0.1 mM bolic pathways were identified using KEGG Pathway and GenMAPP CaCl2, pH 7.4). These samples were incubated at 37 C for 10 min. Pathway. (http://www.genome.jp/kegg/pathway.html; http:// Then, 120 mLof100mM 6,8-difluoro-4-methylumbelliferyl phos- www.genmapp.org/). phate (Gibco, Molecular Probes) was added, and the samples were incubated at 37 C for another 30 min. The PP activity was analysed 2.7. Western blot analysis using a fluorescence microplate reader at 355 nm (excitation) and 460 nm (emission). Three pups from every dam were randomly Western blot (WB) analysis was performed as follows: approx- chosen for this analysis. imately 50 mg of protein from each sample was denatured, elec- trophoresed, and transferred onto a PVDF membrane (Millipore 2.6. Proteome analysis Corporation, Billerica, MA, USA). The membrane was blocked, and the blots were incubated with specific antibodies against Tgm2, Proteins were extracted from the livers of rat pups and two- PPm1a, Mat1a, Pc (San Ying Biotechnology, Wuhan, China) and dimensional gel electrophoresis (2-DE) analysis was performed as GAPDH (Bioworld Technology, Minnesota, USA) and then with previously published with slight modifications (Zhao et al., 2012a). secondary antibodies following the manufacturer's instructions. Briefly, the frozen liver samples were homogenized in lysis buffer ECL reagent (Millipore Corporation, Billerica, MA, USA) was applied containing 2 M thiourea, 7 M urea, 50 mM DTT, 4% CHAPS, 50 mM to the membrane for 1 min. A FluorChem Q (Alpha Innotech, San Tris base, 0.2% Bio-Lytes 4/6, 1 mM protease inhibitor cocktail, 1% Leandro, CA, USA) system was used to evaluate the protein signal. RNase and 1% DNase, and the supernatant was obtained at The bands on the western blot were quantified with Quantity One 12,000 mg for 1 h at 4 C. Then, the supernatant was dialysed for software (BioRad, Hercules, California, USA). Three samples from desalination and freeze-dried. The protein powder was recon- each dam in the control and treated groups were randomly chosen stituted in a rehydration buffer for a final volume of 300 mL. Three for WB analysis. pools were prepared. Protein concentration was determined by the Bradford assay using BSA standards. 2.8. Statistical analysis The first dimension was performed on IPG strips with a pH gradient of 4e7 over 17 cm (BioRad, Hercules, California, USA) in a All of the data were expressed as the mean ± standard deviation PROTEAN IEF cell (BioRad, Hercules, California, USA) using the (SD). The homogeneity of the variance was assessed using Levene's following program: 30 min at 250 V, 30 min at 1000 V, 5 h at test. If the data failed to pass the test, then a logarithmic trans- 10,000 V and 10,000 V constant, for a total of 60,000 Vh. After formation was used. An independent-samples t-test was used to completion of IEF, strips were incubated in equilibration buffer (6 M identify significant differences between the control and exposure urea, 0.375 M TriseHCl, pH 8.8, 20% glycerol, 2% SDS) containing 2% groups. The statistical analysis was performed using SPSS 13.0 DTT for 15 min at room temperature, followed by the same buffer software (SPSS, Chicago, IL), and differences were considered to be with 2.5% iodoacetamide for 15 min. Second-dimension vertical significant at p < 0.05. 10% polyacrylamide gels slabs were freshly prepared between glass plates. Electrophoresis was carried out at 15 mA/gel for 30 min, and 3. Results separation was performed at 25 mA/gel until the bromophenol blue reached the gel bottom. The gels were then visualized using the 3.1. The accumulation of MCLR in the liver of rat pups Silver Strain Plus kit (BioRad, Hercules, California, USA). Following electrophoresis, a GS-800 (BioRad, Hercules, Califor- There were no statistically significant differences in the body nia, USA) was used to scan the gels. The digitized images were weight, liver weight and liver indexes between the MCLR exposure 200 S. Zhao et al. / Environmental Pollution 212 (2016) 197e207 group and the control group (Table S1 in the supporting informa- and the appearance of fat droplets (Fig. 1E and F). tion). However, some petechiae were found in the MCLR-treated 2 livers (Fig. S1). The ESI LC/MS spectra for MCLR detection in the 3.3. Serum biochemical analysis livers of rat pups after 28 days of 10 mg/kg MCLR exposure are shown in Fig. S2E, F. The levels of MCLR reached 6.94 ± 0.833 ng/g The biochemical examinations contain analyses of 6 blood pa- dry weight (DW) in the livers of rat pups in the MCLR-treated rameters (ALT, AST, CHE, TBA, ALB, and PA) that roughly reflect liver group. function and protein metabolism. The significant differences in these parameters between the two groups are presented in Fig. 2 3.2. Histopathological and ultrastructural examination (p < 0.05). Compared with the control group, there was a significant increase in AST activity up to 183% of that in the control group. Liver tissue preparations from the MCLR-treated group showed However, the level of TBA decreased to 11% of that in the control apparent broad hepatocellular gaps and cytoplasmic vacuolization group, whereas ALB and PA decreased by 17% and 99%, respectively. (Fig. 1B and C). The transmission electron microscopy images showed that the sections derived from the control group had intact 3.4. ROS production, lipid peroxidation, GSH content and PP activity plasmalemma, distinct cell junctions and intact nuclei with com- plete nuclear membranes (Fig. 1D). However, there were prominent As shown in Fig. 3A, ROS production significantly increased by morphologic alterations in the MCLR-treated group. The most 43.8% after MCLR exposure. The formation of MDA is an indicator of conspicuous change was the widening of the intercellular spaces the lipid peroxidation level and the development of oxidative

Fig. 1. Pathological alterations in the livers of rat pups after 4 weeks MCLR exposure at 10 mg/kg/day. (A) Conrol liver, 600; (B) Treated liver with apparent broad hepatocellular gaps (arrow)and fatty degeneration (green triangle), 600; (C) at high magniﬁcation from (A), indicated by frame. Treated liver with cytoplasmic vacuolization (green triangle), 1500; Ultrastructure changes in livers of after rat pups after 4 weeks MCLR exposure at 10 mg/kg (D) Control, 17,000 (E) Treated 17,000 showing the widening of the cell junctions (arrow heads) and aggregated ribosomes (yellow arrow heads). (F) Treated 17,000 showing the widening of the cell junctions (black arrow), large lipid droplets (yellow triangle) and compaction of the heterochromatin (white arrow) found in the livers of exposure groups. (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.) S. Zhao et al. / Environmental Pollution 212 (2016) 197e207 201

3.5. Proteome analysis

The image analysis detected over 1000 protein spots in the control group, whereas over 1200 protein spots were detected in the MCLR-treated group (Fig. 4). The comparison of the 2-DE gels of the control and treated groups revealed 106 differentially expressed protein spots. Then, 45 protein spots with altered expression levels (1.9-fold or 0.5-fold; p < 0.05) were manually excised and submitted for identification using MALDI-TOF/TOF MS analysis. From these differentially expressed protein spots, forty proteins were successfully identified with high confidence based on high scores and sequence coverage (Table 1), and six of these spots were identified as the same protein. The remaining five protein spots were not identified. Spots with CI values greater than 95% were accepted. Using the Gene Ontology (GO) hierarchy and the KEGG database, the proteins identified were assigned into Fig. 2. The average values of serum parameters for the litters in two groups and the several functional groups. Of the identified proteins, two proteins results of independent-samples T test. The values are expressed as mean ± S.D. The were characterized as cell cytoskeleton proteins. Six proteins were significance levels observed are *p < 0.05, compared with the values in control group.

Fig. 3. (A) Column diagram of ROS levels expressed as DCF fluorescence intensity in hepatocytes exposed to MCLR. (B) MDA in livers exposed to MCLR. (C) GSH content in livers of rat pups exposed to MCLR. PPs activity (D) in the hepatic tissue of rat pups after 4 weeks of MCLR exposure (control, 10 mg/kg/day). Data are expressed as mean values ± SD. Asterisk indicate a statistically significant difference at p < 0.05. stress. As shown in Fig. 3B, MDA significantly increased by found to be part of the oxidative stress response. Ten proteins were approximately 137%. The hepatic GSH levels fell by 60e65% characterized as metabolism-related proteins, and four of them are (Fig. 3C). MCLR is a selective serine/threonine PP inhibitor. In this associated with liver damage. Seven proteins were related to the study, MCLR inhibited PPs by 24.4 ± 10.8% (Fig. 3D). translation, maturation and degradation of proteins. The other 10 proteins were categorized into other functional groups, including a protein-binding group. The GO annotations for biological processes 202 S. Zhao et al. / Environmental Pollution 212 (2016) 197e207

adverse effects on the embryonic development of mammals and can cause embryonic death and malformations or retarded growth (Bu et al., 2006). We hypothesized that MCLR could transfer from mother to offspring and that the liver of offspring was also the primary target of MCs. In this study, we analysed changes in the livers of rat pups after 28 days of low-dose perinatal MCLR exposure. We found that MCLR caused liver structural damage and functional impairment in the offspring. In addition, the 2-DE results showed that several adverse changes in the livers of rat pups were induced by oxidative stress and mitochondrial dysfunction. In particular, MCLR signiﬁcantly altered the expression levels of several proteins that are associated with cellular metabolism, and these changes may lead to liver injury.

4.1. Toxin transmission, cellular structural damage and serum biochemical indicator changes

MC content in the aquatic environment is generally low. How- ever, MCs can bioaccumulate in fish, mammals, and even humans through the food chain (Chen et al., 2009). Fishermen who fish on Lake Chaohu were found to have serum MC levels of approximately 0.39 ng/mL (Chen et al., 2009). In this study, the level of MCLR in the livers of rat pups was 6.94 ng/g dry weight. This is the first analysis indicating that substantial levels of MCLR were transferred from mother to offspring. Consistent with the level of MCLR in the pup livers, histological and ultrastructural observations suggested liver injury after MC exposure. In this study, loss of adherence between cells was prominent in the MCLR-treated group. The widening of intercellular spaces is presumed to be a critically important effect of MCs and is related to the disruption of microtubules, cytokeratin intermediate filaments and microfilaments (Ding et al., 2000). Thus, the disassociation of hepatic cells in the present study is indicative of cytoskeletal damage in the livers of rat pups exposed to MCLR. In the proteomic study, we also found that the levels of two cytoskeletal proteins, actin-related protein 3 and the intermediate filament protein keratin 8, dramatically decreased in the treated group. These changes might be involved in the disruption of liver metabolism and function. Fig. 4. Image of 2-DE gel stained with silver staining. (A) Control; (B) 2-DE gel image Liver dysfunction was also evidenced by several parameters in with proteins expressed in the 10 mg/kg/day MCLR exposure condition. Protein spots the serum analysis. The levels of serum AST increased significantly, that were altered by MCLR exposure are labelled with characters. Each gel is repre- and the levels of TBA, PA and ALB decreased (Table 1). In addition, sentative of three independent replicates. the level of the blood protein albumin (ALB) was lower according to the 2-DE results. Other studies have reported that the mean AST and molecular functions revealed that several proteins are impli- value is higher in children exposed to MCs through drinking water cated in binding, in regulation of cellular processes and in catalytic and aquatic food in the Three Gorges Reservoir Region Li et al. activity and metabolism (Fig. 5A). The proteins involved in the (2011b). In another study, the serum MC concentrations of chron- oxidationereduction pathway are shown in Fig. 5B in the red circle. ically exposed fishermen had stronger positive relationships with ALT, AST, GGT, and ALP levels than with other biochemical indices, suggesting that MC accumulation might influence the activity of 3.6. Verification of differentially expressed proteins by western blot these serum enzymes, which are indicators of liver function (Chen analysis et al., 2009).

From the candidates, Tgm2, MAT1a, PPm1a and pyruvate 4.2. Redox balance disruption and mitochondrial dysfunction carboxylase (Pc) were selected for western blot analysis, as shown in Fig. 6. The western blot results for the selected proteins were The liver has been described as the most important organ in the consistent with the 2-DE and silver staining results. These results regulation of redox metabolism. In addition to cytoskeletal damage, indicate that the proteomics analysis used in the study was oxidative damage was observed, as evidenced by augmented levels accurate. of ROS and MDA and the reduction of the antioxidant GSH. Furthermore, mitochondrial impairment is considered to be a sig- 4. Discussion niﬁcant contributor to the formation of ROS (Indo et al., 2007). As the primary site of cellular energy generation and oxygen con- The liver is the most important target organ of microcystin sumption, the mitochondrion is the main target for various toxins (Dahlem et al., 1989). MC also affects the heart, kidney, nervous and threats, which may explain the resulting cellular toxicity system and gastrointestinal tract and exhibits genotoxicity (Liu (Binukumar et al., 2010). In this study, MCLR exposure altered the et al., 2006). Several studies have indicated that MCs have levels of proteins associated with oxidative stress in the S. Zhao et al. / Environmental Pollution 212 (2016) 197e207 203

Table 1 A detailed list of protein spots identiﬁed by MALDI-TOF/TOF MS from the livers of rat pups after exposure to MC-LR.

Spot Identiﬁcation Abbreviation Fold Accession no. MW (KDa)/pI MASCOT Expectation SCb Functional category ID changea experimental score

Cytoskeleton A1 Cytokeratin 8 polypeptide Krt8 0.27 gij203734 52,678/5.49 232 4.3E-19 36% Cytoskeleton A2 Actin-related protein 3 homolog Actr3 0.43 gij23956222 47,652/5.61 65 0.021 17% Cytoskeleton Response to oxidative stress A3 Sulfite oxidase Suox d0.000001 gij74024923 54,606/5.79 186 1.7E-14 19% Oxidative phosphorylation A4 2-Oxoisovalerate dehydrogenase Bckdha 0.46 gij77736548 41,845/6.15 211 5.4E-17 36% Energy metabolism subunit alpha A5 3-Hydroxyisobutyrate Hibadh 0.39 gij83977457 35,679/8.73 172 4.3E-13 16% Oxidoreductases, energy dehydrogenase metabolism B1 Aldehyde dehydrogenase 2 Aldh2 3.59 gij14192933 56,078/7.63 97 0.000015 16% Oxidoreductase activity B2 Aldehyde dehydrogenase 1b1 Aldh1b1 6.04 gij58865518 58,102/6.62 131 5.4E-09 25% Oxidoreductase activity B3 Biliverdin reductase A precursor Blvra 2.44 gij16758714 33,491/5.81 145 2.2E-10 24% Oxidoreductase activity Metabolism A6 Protein-glutamine gamma- Tgm2 0.49 gij42476287 77,983/4.98 96 1.60E-05 14% Gamma-glutamyltransferase glutamyltransferase 2 activity A7 Chain A, Methionine Mat1a 0.48 gij77157805 41,620/6.35 155 2.2E-11 9% Amino acid metabolism Adenosyltransferase 1a A8 Protein phosphatase 1A Ppm1a 0.36 gij8394012 42,960/5.19 93 0.000031 26% Phosphatase activity, MAPK signalling pathway A9 Adenosine kinase Adk 0.51 gij52345435 30,481/6.6 123 0.000000074 8% Energy metabolism A10 Sedoheptulokinase Shpk 0.23 gij76096312 51,120/5.56 196 1.7E-15 20% Phosphotransferases A11 Ketohexokinase Khk 0.34 gij13994119 33,299/6.24 67 0.012 23% Fructose and mannose metabolism A12 Fructose-1,6-bisphosphatase 1 Fbp1 0.52 gij51036635 41,313/6.31 128 0.000000039 16% Gluconeogenesis, lipid biosynthesis B4 NADH dehydrogenase 1 alpha Ndufa10 c10,000 gij170295834 38,366/6.8 92 0.000039 15% Oxidative phosphorylation, subcomplex 10 Alzheimer's disease B5 Glycerol-3-phosphate Gpd1 2.74 gij57527919 38,112/6.16 82 0.00041 19% Glycolysis dehydrogenase 1 B6 Pyruvate carboxylase Pc 2.03 gij31543464 130,436/6.34 67 0.014 8% Oxaloacetate metabolism Translation, maturation and degradation of proteins A13 40S ribosomal protein SA Rpsa d0.000001 gij8393693 32,917/4.80 71 0.0058 21% Epithelial cell differentiation A14 Nitrilase homolog 1 isoform b Nit1 d0.000001 gij128485844 32,640/5.93 207 1.4E-16 42% Hydrolases B9 Elongation factor 2 Eef2 1.99 gij8393296 96,192/6.41 62 0.042 14% Protein synthesis B10 Ubiquitin thioesterase OTUB1 Otub1 2.02 gij19527388 31,478/4.85 66 0.018 18% Hydrolases B11 Transthyretin precursor Ttr 1.91 gij7305599 15,880/5.77 118 0.00000023 25% Transport thyroxine and retinol B7 Protein disulfide-isomerase A3 Pdia3 4.58 gij8393322 57,044/5.88 283 3.4E-24 40% Protein disulfide isomerase activity B8 Protein disulfide-isomerase A3 Pdia3 2.19 gij8393322 57,044/5.88 142 4.3E-10 24% Protein disulfide isomerase activity Binding B12 Fetuin precursor Ftu 4.8 gij56140 45,056/6.05 104 0.0000027 21% Plasma binding proteins B13 Heterogeneous nuclear Hnrpk 2.01 gij38197650 51,281/5.19 117 0.00000014 23% RNA binding proteins ribonucleoprotein K isoform a þ B14 Calreticulin precursor Calr 1.96 gij6680836 53,248/4.57 218 2.3E-17 13% Binds Ca2 ions, signal transduction Blood A15 Albumin Alb 0.49 gij158138568 70,670/6.09 151 5.40E-11 29% Lipid binding B15 Fibrinogen gamma chain precursor Fgg 2.41 gij61098186 50,247/5.85 66 0.018 19% Inflammatory response Other functions A16 rCG62047, isoform CRA_b CRA_b 0.47 gij149050792 28,169/6.24 64 0.03 21% Unknown B16 Alpha-2-HS-glycoprotein, isoform Ahsg 4.15 gij6978477 38,757/6.05 112 0.00000043 19% Endocytosis CRA_e B17 59-kDa bone sialic acid-containing Bsac 5 gij220676 59,736/6.25 107 0.0000014 21% Unknown protein precursor B18 Alpha-2-HS-glycoprotein Ahsg 3.44 gij60552688 45,056/6.43 122 0.000000043 21% Endocytosis B19 Serine/arginine-rich splicing factor 2 Sasf2 2.23 gij6755478 25,461/11.86 79 0.00086 25% mRNA splicing

a The fold changes (mean values ± SD, n ¼ 3) are indicated as compared to the controls. Only the fold changes (1.9-fold or 0.5-fold) are shown with their corresponding spot on the other gel. Values > 1 indicate up-regulations, and< 1 down-regulations. b SC indicates the sequence coverage of the protein in percentage obtained by MS/MS identification. c 10,000, the spot disappeared in the control group. d 1E-06, the spot disappeared in the MC-LR exposed group. mitochondria of the liver. Gene correlations and KEGG and Gen- carboxylic acids (Marchitti et al., 2007). Aldehyde dehydrogenase 2 MAPP pathway analyses showed that seven proteins [spots A3, A4, (Aldh2) and aldehyde dehydrogenase 1b1 (Aldh1b1), described as A5, B1, B2, and B3 (Suox, Bckdha, Hibadh, Aldh2, Aldh1b1 and MC-sensitive (Chen et al., 2006), are highly reactive and cytotoxic. Blvra)] identified in the study are associated with oxidative stress They are involved in various physiological processes such as responses. We found that all proteins except Blvra are active in the enzyme inactivation, protein modification, and DNA damage mitochondria. In particular, aldehyde dehydrogenases are known to response (O'Brien et al., 2005). These proteins increased dramati- participate in aldehyde metabolism by oxidizing aldehydes to cally in the MCLR-treated group in this study. In addition, the levels 204 S. Zhao et al. / Environmental Pollution 212 (2016) 197e207

Fig. 5. (A) Distribution of GO annotations of cellular component, biological processes and function by GO for identified proteins. (B) The network of proteins showing an enrichment of genes involved in oxidative stress pathways (indicated by a red circle) by KEGG and GenMAPP pathways. Abbreviations: Suox, sulfite oxidase; Blvra, biliverdin reductase A precursor; Bckdha, 2-oxoisovalerate dehydrogenase subunit alpha; Hibadh, 3-hydroxyisobutyrate dehydrogenase; Aldh1b1, aldehyde dehydrogenase 1b1; Aldh2, aldehyde dehydrogenase 2; Pc, pyruvate carboxylase; others seen in Table 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) of Bckdha and Hibadh, which participate in energy metabolism, MCLR treatment. Spots A6, A7, A8, A9, A10, A11, and A12 (Tgm2, were reduced by MCLR through oxidoreductase activity. Suox and Mat1a, Ppm1a, Adk, Shpk, Khk, and Fbp1) were down-regulated, Blvra are also involved in oxidationereduction. All of these proteins and spots B4, B5, and B6 (Ndufa10, Gpd1 and Pc) were up- were found to be significantly changed, indicating that there was a regulated in the MCLR-treated group. Of these proteins, several disturbance in the redox balance after exposure to MCLR. proteins (Tgm2, Mat1a, PPm1a and Adk) that may be associated with liver injury were found to be affected by MCLR in the livers of rat pups (Fig. 7). 4.3. Proteins related to liver metabolism Tgm2 is a cross-linked protein that can create proteinaceous structures that are resistant to proteolytic and mechanical The metabolic proteins in this study changed in response to S. Zhao et al. / Environmental Pollution 212 (2016) 197e207 205

Fig. 6. Representative 2-DE and Western blot analysis. (A) The magniﬁed images of protein spots from the 2-DE gels are shown in the upper part and the gel pictures presented here were from two independent experiments (n ¼ 3). Western blot showed that equal protein amounts of livers from the control groups and the MC-LR exposure groups were separated by gel electrophoresis and immunoblotted with antibodies against Pc, Tgm2, MAT1a and PPm1a. One membrane out of the three independent replicates is presented. GAPDH was used as internal loading control. (B) Quantiﬁcation of protein content for Pc, Tgm2, MAT1a and PPm1a using scanning densitometry. Each bar represents the mean (SD) of three independent experiments. (p < 0.05).

Fig. 7. The putative model showing three metabolic proteins (Tgm2, MAT1a and PPm1a) involved in MCLR induced apoptosis or tumorigenesis. degradation (Ai et al., 2008). It has been proposed that Tgm2 ac- principal biological methyl donor and plays a critical role in tivity establishes a barrier against tumour progression and metas- maintaining normal hepatic function (Maria et al., 2002). It can also tasis (Fesüs and Piacentini, 2002). Tgm2 often displays aberrant regulate GSH levels through the transsulfuration pathway gene methylation and reduced protein expression (Ai et al., 2008). (Finkelstein et al., 1975). He et al. (2012) found that in SD rats, MCLR Interestingly, Zhao et al., 2012b detected that the reduction in Tgm2 exposure induced a disruption in transsulfuration from SAM to was less pronounced in the liver of mice treated with 20 mg/kg glutathione due to a SAM deficiency. In addition, it has been widely MCLR for 28 days, which suggested that aberrant Tgm2 expression reported that MCs induce ROS overproduction, and GSH constitutes may be a potential marker of MC-induced liver injury. In this study, the first line of defence against ROS (Rao and Bhattacharya, 1996; the expression of Tgm2 significantly decreased in MC-LR-treated Zegura, 2008). In this study, the MAT1a and glutathione (GSH) livers and further implicated Tgm2 in liver injury. levels were both significantly decreased in MCLR-treated livers, The methionine adenosyltransferase (MAT) reaction adds ATP to indicating that detoxification systems might be inhibited by MCLR, methionine to generate S-adenosylmethionine (SAM). SAM is the making the cells more susceptible to oxidative stress. Thus, the 206 S. Zhao et al. / Environmental Pollution 212 (2016) 197e207 reduction of MAT1a may result in SAM deficiency and increasing of monitoring microcystin concentrations in mammalian offspring. susceptibility to liver injury. We also need to consider the potential for foetal exposure to The phosphatase PPm1a belongs to the family of protein serine/ microcystin during the perinatal period. threonine phosphatases (PS/TPs). PPm1a is able to not only de- phosphorylate some proteins, but it also plays a critical role in Acknowledgements terminating transforming growth factor b (TGF-b) signalling (Lin et al., 2006; Bu et al., 2008). The loss of TGF-b may result in This work was funded by the National Natural Science Foun- increased probability of accumulating further mutations and cy- dation of China (31070457). togenetic changes (Akhurst, 2004). In this study, we found that MCLR significantly inhibited PP activity and decreased PPm1a Appendix A. Supplementary data levels, which further indicated that MCLR could perturb the dy- namic equilibrium of protein phosphorylation/dephosphorylation Supplementary data related to this article can be found at http:// and activate a number of signalling pathways that can lead to cell dx.doi.org/10.1016/j.envpol.2015.12.055. damage. The enzyme adenosine kinase (Adk) catalyses the phosphory- References lation of adenosine to produce AMP. Adenosine is produced in the methionine cycle with homocysteine via hydrolysis of S-adeno- Ai, L., Kim, W.J., Demircan, B., 2008. The transglutaminase 2 gene (TGM2), a po- sylhomocysteine (SAH) by SAH hydrolase (SAHH). The failure to tential molecular marker for chemotherapeutic drug sensitivity, is epigeneti- e fi cally silenced in breast cancer. Carcinogenesis 29, 510 518. ef ciently remove adenosine (by phosphorylation by Adk) can Akhurst, R.J., 2004. TGF-b signaling in health and disease. Nat. Genet. 36, 790e792. result in an accumulation of SAH, which is a potent inhibitor of all Binukumar, B.K., Amanjit, B., Kandimalla, R., Sunkaria, A., Gill, K.D., 2010. Mito- transmethylation reactions. (Lawrence de Koning et al., 2003)An chondrial energy metabolism impairment and liver dysfunction following chronic exposure to dichlorvos. Toxicology 270, 77e84. Adk knockout mouse has been generated; this mouse has no pre- Bøe, R., Gjertsen, B.T., Vintermyr, O.K., Houge, G., Lanotte, M., Døskeland, S.O., 1991. natal phenotype but develops progressive microvesicular liver The protein phosphatase inhibitor okadaic acid induces morphological changes steatosis from postnatal day two (Boison et al., 2002). In this study, typical of apoptosis in mammalian cells. Exp. Cell Res. 195 (1), 237e246. the low expression of Adk protein may indicate that it can interfere Boison, D., Scheurer, L., Zumsteg, V., Rülicke, T., Litynski, P., Fowler, B., Brandner, S., Mohler, H., 2002. Neonatal hepatic steatosis by disruption of the adenosine with a wide range of methyltransferase reactions after MCLR kinase gene. Proc. Natl. Acad. Sci. U. S. A. 99, 6985e6990. exposure. Bu, S., Kapanadze, B., Hsu, T., Trojanowska, M., 2008. Opposite effects of dihy- In addition, the decrease in the proteins Shpk, Khk, and Fbp1 drosphingosine 1-phosphate and sphingosine 1-phosphate on transforming growth factor-beta/Smad signaling are mediated through the PTEN/PPM1A- may be related to a MCLR-induced stress response. These proteins dependent pathway. J. Biol. Chem. 283, 19593e19602. involved in carbohydrate metabolism, which indicates that MCLR Bu, Y.Z., Li, X.Y., Zhang, B.J., Chung, Ik-Kyo, 2006. Microcystins cause embryonic exposure can influence glycogen metabolism. toxicity in mice. Toxicon 48, 966e972. Chen, J., Xie, P., Li, L., Xu, J., 2009. First identification of the hepatotoxic microcystins Three proteins increased dramatically in this study. Ndufa10 (an in the serum of a chronically exposed human population together with indi- NADH dehydrogenase) was expressed only in the MCLR-exposed cation of hepatocellular damage. Toxicol. Sci. 108, 81e89. group. Ndufa10 participates in oxidative phosphorylation and Chen, T., Cui, J., Liang, Y., Xin, X.B., Young, D.O., Chen, C., Shen, P.P., 2006. Identifi- cation of human liver mitochondrial aldehyde dehydrogenase as a potential may be involved in several neurodegenerative and cardiovascular target for microcystin-LR. Toxicology 220, 71e80. diseases (Indo et al., 2007; Stohs, 1990). Glycerol-3-phosphate de- Chernoff, N., Hunter, E.S., Hall, L.L., Rosen, M.B., Brownie, C.F., Malarkey, D., 2000. hydrogenase (Gpd) plays a major role in lipid biosynthesis. Through Lack of teratogenicity of microcystin-LR in the mouse and toad. J. Appl. Toxicol. 22 (1), 13e17. the reduction of dihydroxyacetone phosphate into glycerol 3- Dahlem, A.M., Hassan, A.S., Swanson, S.P., Carmichael, W.W., Beasley, V.R., 1989. phosphate, Gpd allows for the dephosphorylation of glycerol 3- A model system for studying the bioavailability of intestinally administered phosphate into glycerol (Harding et al., 1975). The increase in microcystin-LR, a hepatotoxic peptide from the cyanobacterium Microcystis e Gpd1 expression is consistent with the increased number of lipid aeruginosa. Pharmacol. Toxicol. 64, 177 181. Dai, M., Xie, P., Liang, G., Chen, J., Lei, H., 2008. Simultaneous determination of droplets observed in the ultrastructural analysis. During gluco- microcystin-LR and its glutathione conjugate in fish tissues by liquid neogenesis, pyruvate carboxylase (Pc) is involved in the synthesis chromatography-tandem mass spectrometry. J. Chromatogr. B 862 (12), e of phosphoenolpyruvate from pyruvate (Joseph et al., 1997). He- 43 50. Ding, W., Shen, H., Ong, C., 2000. Microcystic cyanobacteria extract induces cyto- patic glucose production is sustained by increased pyruvate and has skeletal disruption and intracellular glutathione alteration in hepatocytes. En- been shown to be accompanied by increased Pc concentrations viron. Health Perspect. 108, 605e609. (Joseph et al., 1997). In our study, the increase in Pc protein indi- Du, D., Bramlage, W.J., 1992. Modified thiobarbituric acid assay for measuring lipid oxidation in sugar-rich plant tissue extracts. J. Agric. Food Chem. 40, cated that the organism needs more energy to resist stress after 1566e1570. MCLR treatment. Eriksson, J.E., Paatero, G.I.L., Meriluoto, J.A.O., Codd, G.A., Kass, G.E.N., Nicotera, P., Other biological functions, such as translation, maturation and Orrenius, S., 1989. Rapid microfilament reorganization induced in isolated rat hepatocytes by microcystin-LR, a cyclic peptide toxin. Exp. Cell Res. 185, degradation of proteins, seem to be affected as well. This may be 86e100. explained by the induction of oxidative stress, which results in Fesüs, L., Piacentini, M., 2002. Transglutaminase 2: an enigmatic enzyme with oxidation of proteins, thus preventing their activity or leading to diverse functions. Trends Biochem. Sci. 27, 534e539. Figueiredo de, D.R., Azeiteiro, U.M., Esteves, S.M., Goncalves, F.J., Pereira, M.J., 2004. protein misfolding. Microcystin-producing bloomsea serious global public health issue. Ecotoxicol. In sum, the data support the hypothesis that maternal Environ. Saf. 59 (2), 151e163. microcystin-LR exposure can affect the liver in offspring. These Finkelstein, J.D., Kyle, W.E., Martin, J.L., Pick, A.M., 1975. Activation of cystathionine results suggest that MCLR can increase ROS, disturb the redox synthase by adenosylmethionine and adenosylethionine. Biochem. Biophys. Res. Commun. 66, 81e87. balance and result in oxidative damage in the liver. Importantly, Fontal, O.I., Vieytes, M.R., Baptista de Sousa, J.M., Louzao, M.C., Botana, L.M., 1999. some proteins related to methionine metabolism, carbohydrate A fluorescent microplate assay for microcystin-LR. Anal. Biochem. 269, e metabolism and lipid metabolism had an early response to MC 289 296. Frangez, R., Zuzek, R.M.C., Mrkun, J., Sedmak, B., Kosec, M., 2003. Microcystin-LR exposure; these proteins may contribute to MC-induced liver affects cytoskeleton and morphology of rabbit primary whole embryo cultured impairment and dysfunction. In addition, the inhibition of protein cells in vitro. Toxicon 41, 999e1005. phosphatases, leading to the hyperphosphorylation of cellular Fu, W.Y., Xu, L.H., Yu, Y.N., 2005. Proteomic analysis of cellular response to microcystin in human amnion FL cells. J. Proteome Res. 4, 2207e2215. proteins (e.g. disruption of cytoskeleton proteins), is associated Griffith, O., 1980. Determination of glutathione and glutathione disulfide using with liver injury. The results of this study highlight the importance glutathione reductase and 2-vinylpyridine. Anal. Biochem. 106, 207e212. S. Zhao et al. / Environmental Pollution 212 (2016) 197e207 207

Harding Jr., Joseph, W., Pyeritz, Eric A., Copeland, Eric S., White III, Harold, B., 1975. 59, 125e150. Role of glycerol 3-phosphate dehydrogenase in glyceride metabolism - effect of Maria, L.M., Fernando, J.C.L., Alfonso, M.C., Elena, R.G.T., 2002. Spontaneous oxida- diet on enzyme activities in chicken liver. Biochem. J. 146, 223e229. tive stress and liver tumors in mice lacking methionine adenosyltransferase 1A. He, J., Chen, J., Wu, L.Y., Li, G.Y., Xie, P., 2012. Metabolic response to oral microcystin- FASEB J. 16 (8), 1292e1314. LR exposure in the rat by NMR-based metabonomic study. J. Proteome Res. 12, Nong, Q., Komatsu, M., Izumo, K., Indo, H.P., Xu, B., Aoyama, K., Majima, H.J., 5934e5946. Horiuchi, M., Morimoto, K., Takeuchi, T., 2007. Involvement of reactive oxygen IARC, 2010. Monographs on the Evaluation of Carcinogenic Risks to Humans, species in Microcystin-LR induced cytogenotoxicity. Free Radic. Res. 24, 1e12. Ingested Nitrate and Nitrite and Cyanobacterial Peptide Toxins, p. 94. O'Brien, P.J., Siraki, A.G., Shangari, N., 2005. Aldehyde sources, metabolism, mo- Indo, H.P., Davidson, M., Yen, H.C., Suenaga, S., Tomita, K., Nishii, T., Higuchi, M., lecular toxicity mechanisms, and possible effects on human health. Crit. Rev. Koga, Y., Ozawa, T., Majima, H.J., 2007. Evidence of ROS generation by mito- Toxicol. 35, 609e662. chondria in cells with impaired electron transport chain and mitochondrial Paerl, H.W., Huisman, J., 2009. Climate change: a catalyst for global expansion of DNA damage. Mitochondrion 7, 106e118. harmful cyanobacterial blooms. Environ. Microbiol. Rep. 1, 27e37. Isamu, K., Hiromi, B., Yoshinori, U., Noriyuki, S., 2010. In-gel protein phosphatase Palikova, M., Navratil, S., Marsalek, B., Blaha, L., 2003. Toxicity of crude extract of assay using fluorogenic substrates. Anal. Biochem. 400, 118e122. cyanobacteria for embryos and larvae of carp (Cyprinus carpio L.). Acta Vet. Joseph, J.B., Masafumi, M., Ralph, A.D., Rubin, B., 1997. In vitro and in vivo sup- Brno 2, 437e443. pression of gluconeogenesis by inhibition of pyruvate carboxylase. Biochem. Rao, P.R., Bhattacharya, 1996. The cyanobacterial toxin microcystin-LR induced DNA Pharmacol. 53, 67e74. damage in mouse liver in vivo. Toxicology 114 (1), 29e36. Lawrence de Koning, A.B., Werstuck, G.H., Zhou, J., Austin, R.C., 2003. Hyper- Rao, P.V.L., Bhattacharya, R., Parida, M.M., Jana, A.M., Bhaskar, A.S.B., 1998. Fresh- homocysteinemia and its role in the development of atherosclerosis. Clin. water cyanobacterium Microcystis aeruginosa (UTEX 2385) induced DNA Biochem. 36, 431e441. damage in vivo and in vitro. Environ. Toxicol. Pharmacol. 5, 1e6. Li, G.Y., Chen, J., Xie, P., Jiang, Y., Wu, L., Zhang, X., 2011a. Protein expression profiling Sepulveda, M.S., Rojas, M., Zambrano, F., 1992. Inhibitory effect of a Microcystis sp. in the zebrafish (Danio rerio) embryos exposed to the microcystin-LR. Prote- (cyanobacteria) toxin on development of preimplantation mouse embryos. omics 11, 2003e2018. Comp. Biochem. Physiol. C 102 (3), 549e553. Li, G.Y., Cai, F., Yan, W., Li, C.R., Wang, J.H., 2012. A proteomic analysis of MCLR- Stohs, S.J., 1990. Oxidative stress induced by 2,3,7,8-tetrachlo-rodibenzo-p-dioxin induced Neurotoxicity: implications for Alzheimer's disease. Toxicol. Sci. 127, (TCDD). Free Radic. Biol. Med. 9, 79e90. 485e495. Tedesco, I., Russo, G.L., Nazzaro, F., 2001. Antioxidant effect of red wine anthocya- Li, H.Y., Xie, P., Li, G.Y., Hao, L., Xiong, Q., 2009. In vivo study on the effects of nins in normal and catalase-inactive human erythrocytes. J. Nutr. Biochem. 12, microcystin extracts on the expression profiles of proto-oncogenes (c-fos, c-jun 505e511. and c-myc) in liver, kidney and testis of male Wistar rats injected i.v. with Ueno, Y., Nagata, S., Tsutsumi, T., Hasegawa, A., Watanabe, M.F., Park, H.D., toxins. Toxicon 53, 169e175. Chen, G.C., Chen, G., Yu, S.Z., 1996. Detection of microcystins, a blue green algal Li, Y., Chen, J., Zhao, Q., Pu, C.W., Qiu, Z.Q., 2011b. A cross-sectional investigation of hepatotoxins, in drinking water sampled in Haimen and Fusui, endemic areas of chronic exposure to microcystin in relationship to childhood liver damage in primary liver cancer in China, by highly sensitive immunoassay. Carcinogenesis the three gorges reservoir region, China. Environ. Health Perspect. 119, 17, 1317e1321. 1483e1488. Wang, Q., Xie, P., Chen, J., Liang, G.D., 2008. Distribution of microcystins in various Lin, X., Duan, X., Liang, Y.Y., Su, Y., Wrighton, K.H., Long, J., Hu, M., Davis, C.M., organs (heart, liver, intestine, gonad, brain, kidney and lung) of Wistar rat via Wang, J., Brunicardi, F.C., Shi, Y., Chen, Y.G., Meng, A., Feng, X.H., 2006. PPM1A intravenous injection. Toxicon 52, 721e727. functions as a Smad phosphatase to terminate TGFbeta signaling. Cell 125, Wei, G.R., Zhang, Z.Y., Lian, M., 2002. Damage effect of microcystin on placenta in SD 915e928. pregnant rats. China Public Health 18 (8), 921e922. Liu, L.L., Zhong, R.G., Zeng, Y., 2006. Advances in study on microcystins and their Yu, S.Z., 1995. Primary prevention of hepatocellular carcinoma. J. Gastroenterol. toxicology. J. Hyg. Res. 35, 247e249. Hepatol. 10, 674e682. Mackintosh, C., Beattie, K.A., Klumpp, S., Cohen, P., Codd, G.A., 1990. Cyanobacterial Zegura, B., 2008. Patterns of microcystin-LR induced alteration of the expression of microcystin-LR is a potent and specific inhibitor of protein phosphatase-1 and genes involved in response to DNA damage and apoptosis. Toxicon 51 (4), phosphatase-2A from both mammals and higher-plants. FEBS Lett. 264, 615e623. 187e192. Zhao, S.J., Xie, P., Li, G.Y., Chen, J., 2012a. The proteomic study on cellular responses Manson, J., Kang, Y.S., 1989. Test methods for assessing female reproductive and of the testes of zebrafish (Danio rerio) exposed to microcystin-RR. Proteomics developmental toxicology. In: Hayes, A.W. (Ed.), Principles and Methods of 12, 300e312. Toxicology. Raven Press, New York, pp. 311e359. Zhao, Y.Y., Xie, P., Fan, H.H., 2012b. Genomic profiling of microRNAs and proteomics Marchitti, S.A., Deitrich, R.A., Vasiliou, V., 2007. Neurotoxicity and metabolism of the reveals an early molecular alteration associated with tumorigenesis induced by catecholamine-derived 3,4-dihydroxyphenylacetaldehyde and 3,4-dihydroxy MC-LR in mice. Environ. Sci. Technol. 46, 34e41. -phenylglycolaldehyde: the role of aldehyde dehydrogenase. Pharmacol. Rev.