ETS2 Regulating Neurodegenerative Signaling Pathway of Human

May 7, 2009 - Down-regulation of genes resulting from repeated exposure to sarin may .... down-regulation of the Notch homology gene and PSEN1 seen in...
0 downloads 0 Views 3MB Size
Download PDF
990

Chem. Res. Toxicol. 2009, 22, 990–996

Chemical Profile ETS2 Regulating Neurodegenerative Signaling Pathway of Human Neuronal (SH-SY5Y) Cells Exposed to Single and Repeated Low-Dose Sarin (GB) Arjunan Pachiappan,†,‡ Maung Maung Thwin,† Loke Weng Keong,§ Fook Kay Lee,§ Jayapal Manikandan,| Viswanathan Sivakumar,† and Ponnampalam Gopalakrishnakone*,† Departments of Anatomy and Physiology, Yong Loo Lin School of Medicine, National UniVersity of Singapore, Singapore 117597, DSO National Laboratories, Singapore 118230, and Porter Neuroscience Research Center, National Institutes of Health, Bethesda, Maryland 20892 ReceiVed September 15, 2008

The mechanistic understanding of low-level sarin-induced neurotoxicity after single or repeated doses has yet to be explored at a cellular level. Using the microarray (Affymetrix-GeneChips) transcription profiling approach, the present study examined gene expression in human SH-SY5Y cells exposed to single (3 and 24 h) or repeated (2 × 24 h) doses of sarin (5 µg/mL) to delineate the possible mechanism. Two hundred twenty-four genes whose expression was significantly (P < 0.01) altered by at least 3-fold were selected by GeneSpringGX analysis. The comparative gene expression data confirmed the transcriptional changes to be related to dose and exposure time of sarin. The effect of a single noncytotoxic sarin dose on gene transcription was variable, whereas repeated doses over 48 h persistently downregulated genes linked to neurodegenerative mechanisms. Thirty persistently altered genes were validated using real-time quantitative reverse transcription-polymerase chain reaction (qRT-PCR). Similar qRTPCR profiles obtained in sarin-treated SH-SY5Y and HCN-1A cells confirmed the cell-independent alterations in expression levels. Genes (ETS2, APOE, PSEN1, DDC, and CD9) implicated mainly in the regulation of sarin-induced neuropathogenesis were further confirmed by Western blot and doubleimmunofluorescence assays. The regulome pathway suggests a new feasible mechanism by which sarin increases ETS2 expression and takes control over other genes involved in the neurodegenerative pathway. The overall data delineate an in vitro experimental model suitable for studying the neuropathology of cells and may provide novel insights into therapeutic interventions. Introduction Sarin (GB) irreversibly inhibits acetylcholinesterase (AChE)1 (1), leading to cholinergic hyperactivity and death at high-level exposures due to acetylcholine (ACh) accumulation and subsequent overstimulation of the nicotinic and muscarinic receptors (2). In patients, high exposure to sarin can cause muscle fasciculation, respiratory deficiency, reduced consciousness, and flaccid paralysis (3). Neurological abnormalities (4) or chronic memory decline persists in sarin victims for several years after exposure (3). Sarin-induced neuronal cell death in the brain has also been confirmed by several in vitro and in vivo animal studies (5).

* To whom correspondence should be addressed. Tel: (65)6516 3207. Fax: (65)6778 7643. E-mail: [email protected]. † Department of Anatomy, National University of Singapore. ‡ National Institutes of Health. § DSO National Laboratories. | Department of Physiology, National University of Singapore. 1 Abbreviations: ACh, acetylcholine; AChE, acetylcholinesterase; ASChI, acetylthiocholine iodide; ANOVA, analysis of variance; ATCC, American Type Culture Collection; DAPI, 4′,6-diamidino-2-phenylindole; FBS, fetal bovine serum; PCA, principle component analysis; PVDF, polyvinylidene difluoride; qRT-PCR, quantitative reverse transcription-polymerase chain reaction; TBST, tris-buffered saline tween-20.

Differential expression of AChE in rat brain tissues after sarin administration has previously been demonstrated by microarray profiling studies (6-9). Despite the limited gene expression studies done on animal brain tissues, the mechanistic understanding of the signaling pathways involved after a single or repeated low dose sarin exposure in human neuronal cells has yet to be explored at a molecular level. Utilizing Affymetrix HG-U133A arrays for transcriptional profiling of human neuronal SH-SY5Y and HCN-1A cell lines exposed to single and repeated doses of sarin, we confirmed the results of the gene array by quantitative reverse transcriptionpolymerase chain reaction (qRT-PCR) and identified the key pathway involved in sarin-induced neurodegeneration with the aid of PathwayStudio analyses. The present data, for the first time, provided the molecular evidence on an ETS2-regulated mitochondrial death pathway as the key neurodegenerativesignaling pathway in response to sarin exposure. Understanding these genes and pathways will be useful for identification of early molecular markers, which may provide novel insights into therapeutic interventions.

Experimental Procedures Noncytotoxic Sarin Concentration. Sarin (90% purity; {P}NMR, DSO National Laboratories, Singapore) was applied to

10.1021/tx8003467 CCC: $40.75  2009 American Chemical Society Published on Web 05/07/2009

Neuropathogenesis of sarin cultured SH-SY5Y [CRL-2266, American Type Culture Collection (ATCC), Manassas, VA] cell lines at varying concentrations (0.1-2500 µg/mL) and time intervals (3, 6, 24, 48, and 72 h). Cell proliferation was measured as previously described (10) using trypan blue (Sigma, MO) and XTT assays (Cell Proliferation Kit-II, Roche, Singapore). Exposure of SH-SY5Y to Sarin. Human SH-SY5Y cell lines were selected based on the pathologic role of neuronal cells in neurological diseases (11). Briefly, 1 × 107 SH-SY5Y cells grown under standard conditions in 12 mL of EMEM [F12-K medium containing 10% fetal bovine serum (FBS), 100 units mL-1 penicillin, and 100 µg mL-1 streptomycin] were exposed to a single (3 or 24 h) or repeated (2 × 24 h) noncytotoxic sarin dose (5 µg/ mL) in a specially defined BSL-II incubator. Six replicates were used for each time point and exposure conditions. The untreated cells of the same passage served as controls. The harvested (by sterile cells scrapper) cell pellets were snap-frozen with RNAlater (Ambion, TX) and stored at -80 °C for RNA isolation. RNA Extraction and Validation. Total RNA was extracted from the cells using RNeasy mini kit, treated with RNase-free Dnase-I (QiageN, CA) at room temperature (RT) for 20 min, and stored at -80 °C until use. The isolated RNA was used for microarray hybridization and qRT-PCR experiments as described (10). RNA purity (A260/A280) and quantity (A260) were checked by Biophotometer (Eppendorf, CA), and the RNA integrity (18S:28S ratio) assessed using denaturing agarose gels was visualized with ethidium bromide under UV light. RNAs were further subjected to an Agilent 2100 bioanalyzer (Agilent Technologies, CA). AChE Assay Using Cell Lysates. Cell pellets treated with sarin (0.1-50 µg/mL) and controls (n ) 5) were resuspended in 0.1 M phosphate buffer, pH 7.4, containing 0.2% (v/v) Triton X-100 (final 3.0 × 105 cells/mL). A 1:10 dilution of cell lysate was incubated for 20 min with 1.0 mM DTNB (5,5′-dithiobis-2-nitrobenzoic acid), and the OD was read at 412 nm for 20 min at 30 s intervals initially and for 5 min at 10 s intervals after adding 10 µL of 100 mM acetylthiocholine iodide (ASChI). One unit of AChE activity is defined as the amount of enzyme required to convert 1 µmol of substrate per minute. Microarray GeneChip Analysis and Criteria for Gene Selection. Each time point had its own control sets to compare with corresponding sarin-treated groups (3, 24, and 48 h; n ) 3:3 chips/time point). Amplification of cDNA, fragmentation of cRNA for hybridization on the HG-U133A (Affymetrix, Santa Clara, CA) GeneChip, and concurrent assessment of gene expression were carried out as described (11). Scanned images were analyzed using Affymetrix MicroArray Suite (MAS) v5.0. GeneSpring GX v7.3.1 (Silicon Genetics, CA) software was used for parametric analysis based on a cross-gene error model (PCGEM). Besides gene categorization and filter criteria applications (11), the fifth filter of significant statistical associations and the Benjamini and Hochberg False Discovery Rate multiple testing corrections were also applied. A conservative analytical approach was used to limit the number of false positives. The 224 differentially expressed genes were annotated with the AmiGO-Biological Process functional classification, the unification tool (http://smd.stanford.edu/cgibin/source/ sourceSearch), and Online Mendelian Inheritance in Man (http:// www.ncbi.nlm.nih.gov/sites/entrez?db)OMIM). The data were deposited in NCBIs Gene Expression Omnibus (GEO; http://www. ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE2732. Principal Component Analysis (PCA). For quantifying gene expression data from the signal-processing viewpoint, the raw data were preprocessed prior to cluster pattern analysis. To stabilize the variation across the expression ranges associated with the probe concentration (12), the data were log transformed. The noise in the signal was filtered, based on reliable hybridizations of probes and variation within each chip among the experiment groups. The multidimensional data matrix was then performed for PCA using GeneSpringGX software. Gene Validation by Quantitative Real-Time PCR Assay. RNA reverse transcription, DNA amplification and detection, data

Chem. Res. Toxicol., Vol. 22, No. 6, 2009 991 acquisition, and analysis were performed to verify the selected genes (10). Genes that were significantly (P < 0.001) and consistently (g3-fold) expressed between three time points were selected for qRT-PCR. On the basis of GeneOntology classification, 30 sets of transcripts persistently altered in the functional clusters were identified by multiple statistical approaches. The specificity of qRTPCR primers was checked by BLAST: http://www.ncbi.nlm.nih. gov/genome/seq/HsBlast.html (Supporting Information, Table ST1). Genes in each sample were normalized to that of β-actin and GAPDH, and the data were quantified by relative quantification (2-∆∆CT) method (13). The same procedure was followed for sarintreated HCN-1A (CRL-10442) cells to validate potential target genes. Automatic Regulome Networks Building by ResNet PathwayStudio. For data analysis and prediction of biological pathways, the ResNet database (PathwayStudio v5.0, Ariadne, MD) was used. The complete set of rules and ResNet data produced by the Medscan were followed (14) to autogenerate good quality signaling networks. Genes and encoding proteins in the ResNet database in 46 functional groups were classified by using Gene Ontology (http://www.geneontology.org) and Entrez (NCBI) gene annotation. Protein Isolation and Western Blot Assay. Control and sarintreated SH-SY5Y cells were lysed with Mammalian Protein Extraction Reagent containing protease inhibitor cocktail (Pierce, IL). After separation by 10% SDS-PAGE under reducing conditions, proteins were transferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad). The membranes were blocked with 5% skimmed milk in tris-buffered saline tween-20 (TBST) (20 mM Tris-HCl, pH 7.6, containing 137 mM NaCl, 3 mM KCl, and 0.05% Tween 20) and incubated overnight at 4 °C with one of the affinitypurified polyclonal antibodies in blocking solution: ETS2 (1:1000), PSEN1 (1:2000), CD9 (1:1000) (Santa Cruz, CA), APOE (1:1000), or DDC (1:5000) (Chemicon, CA). Membranes incubated with monoclonal mouse antiactin (1:3000) (Sigma) served as the control. After incubation for 1 h at 25 °C with an appropriate dilution (1: 5000) of secondary antibody HRP-conjugated antirabbit IgG (GE Healthcare, United Kingdom), the immunoblots were detected by enhanced chemiluminescence using an ECL kit (SuperSignal, Pierce) and scanned with a computer-assisted G-710 densitometer, and labeling intensities of the bands were quantified using Quantity One Software (Bio-Rad). Double-Immunofluorescence Staining. Four percent paraformaldehyde fixed cells were blocked with 3% normal goat serum in PBS for 1 h at RT, incubated with indicated dilutions of the primary antibodies (APOE, 1:1000; ETS2, 1:100; DDC, 1:500; CD9, 1:50; or PSEN1, 1:50) at 4 °C for overnight, and detected with a secondary antibody FITC-conjugated goat antirabbit IgG. Washed cells were nuclear counter stained with 4′,6-diamidino-2phenylindole (DAPI) and mounted in DAKO fluorescent mounting medium (DAKOCytomation, Denmark), and cellular colocalization was studied with a laser scanning confocal microscopy (Fluoview 1000, Olympus, Tokyo, Japan). Statistical Analysis. Statistical analyses were performed with the Statistical Package for Social Sciences using paired Student’s t test, (v11.5). One-way analysis of variance (ANOVA) was used to compare differences between the means of the various treatment groups, and results were represented as means ( SEMs. A statistically significant (*P < 0.05, **P < 0.01, or ***P < 0.001) level was considered wherever applicable. Graphs, EC50 determinations, and statistical analysis were performed using GraphPad Prism Software v4.01.

Results Noncytotoxic Challenge of Sarin. At 50 µg/mL, sarin caused considerable cell death (Supporting Information, Figure S1) but had no significant (P < 0.05) affect on cell viability at e5 µg/ mL, with most cells remaining intact without any swelling and membrane disruption (Figure 1A). Although no changes were noted in the growth kinetics of SH-SY5Y cells at 5 µg/mL,

992

Chem. Res. Toxicol., Vol. 22, No. 6, 2009

Figure 1. Effect of sarin on toxicity and AChE activity of SH-SY5Y cells. Cultured human SH-SY5Y cells were incubated for 3-48 h with (A) 0.1-2500 and (B) 0.1-50 µg/mL of sarin, respectively, for toxicity and AChE activity experiments. The cell viability was quantified by XTT and trypan blue assays. The minimal direct acute cytotoxicity of sarin was e5 µg/mL. As compared to controls, sarin inhibited ∼70% (P < 0.01) of cholinesterase activity of the cells. Values are means ( SEMs, with each measurement performed in triplicate, and were repeated thrice.

pairwise comparison analysis indicated alterations in gene signatures. Thus, 5 µg/mL sarin was selected as a noncytotoxic dose to see the differential gene expression in these cells. Sarin Inactivation of AChE Activity. Varying doses (0.1-50 µg/mL) of sarin decreased AChE activity, indicating that a very little amount (0.1 µg/mL) of sarin is sufficient to inactivate AChE activity (∼53%) in neuronal (SH-SY5Y) cell lysates (Figure 1B). Ninety-one percent inhibition of cholinesterase activity was observed in neuronal cell lysates spiked with 5 µg/mL of sarin, indicating that the selected optimum dose is nontoxic to SH-SY5Y cells yet effective in inactivating AChE enzyme activity. Microarray GeneChip Analysis and Data Mining. A total of 224 genes, 53 up- and 23 down-regulated (at single dose) and 49 up- and 175 down-regulated (at repeated doses), passed the 0.01 confidence level in the statistical tests. The number of gene changes increased with repeated doses at 48 h (148 genes) as compared to a single dose (76 genes) treatment. Overall, there was a decrease (22.3%) in up-regulation and an increase (77.7%) in down-regulation gene levels at repeated doses (Supporting Information, Table ST1). Down-regulation of genes resulting from repeated exposure to sarin may imply that lesions in the neuronal cells are either irreversible or slow-repairing types. PCA. The three-dimensional graph (Figure 2A) shows differences between the controls (C-3, 24, and 48 h) and treated groups at corresponding time points (T-3, 24, and 48 h). Each replicate in the graph clearly demonstrates variations among the groups and the reproducibility of microarray data. Gene expression pattern differences obviously exist between the single (T-3 and 24 h) and the repeated dose (T-48 h) groups. The first two components (PC1 and PC2) indicate 5.96% variation within the same groups, while the third component (PC3) indicates

Pachiappan et al.

94.04% variation in the total gene expression, thus confirming the toxicity-related gene expression as a function of dose and time. Conditional Tree View and Hierarchical Clustering Analysis. Conditional clustering was performed for 224 selected genes with GeneSpringGX software using Pearson correlation as similarity matrix. The homogeneity of the coloring pattern of each chip validates reproducibility of the replicate data obtained from control and treated groups (Supporting Information, Figure S2). The global and differential expression patterns (Figure 2B) demonstrate clustering of most of the up-regulated genes at 3 and 24 h after a single dose and the down-regulated genes at 48 h after repeated doses of sarin in a dose- and time-dependent manner. Expression patterns at 3 and 24 h post-treatment (single dose) groups are almost similar, whereas a variation in expression patterns is evident at 48 h with the repeated doses. The conditional and hierarchical clustergrams clearly indicate that the clustered samples (per experiment with each time point) in the column and the highly up- (red color) and down- (blue color) regulated genes in the rows may probably be associated with the neurodegenerative process. Gene Expression Validation by qRT-PCR. Table 1 shows the genes that were significantly altered by sarin treatment. APOE, CHRNE, GRIN2B, GABRG3, GADD45A, GRIK5, ADORA1, BAX, and CFLAR genes were significantly (P < 0.001-0.05) up-regulated, whereas BCHE, APP, BAChE, ADORA2A, β2M, PRSS12, SNCA, ABCB1, CLU, CALM3, ITM2B, BCL2L1, CALM1, MAPT, AK3L1, PSEN2, and CHRNB4 genes were significantly (P < 0.01 and P < 0.05) down-regulated at various time points with both doses. PSEN1, CD9, and DDC were significantly (P < 0.001) down-regulated (g10 fold) with repeated doses (at 48 h) but not (g5-fold) with a single dose (at 3 or 24 h). Among the up-regulated genes examined by qRT-PCR, ETS2 was the gene most significantly (P < 0.001) and consistently (>10-fold) increased at all time (3, 24, and 48 h) points with both doses of sarin. qRT-PCR performed on HCN-1 cells at 48 h sarin postexposure (Supporting Information, Figure S7) indicated that ETS2 was also highly expressed in these cell lines. Similar gene expression profiles found in both SHSY5Y and HCN-1 cell lines support the idea of a functional synergy between the two neuronal cell lines. The gene expression changes quantified by qRT-PCR correlated well with the microarray results (Figure 2C), thus validating the results of microarray analysis. Western Blot and Quantification Analysis. Western blots revealed bands corresponding to ETS2, APOE, DDC, CD9, and PSEN1 proteins with MW ∼ 55, 42, 53, 24, and 47 kDa, respectively (Figure 3A-E). Except for APOE at 3 h, sarin administration increased ETS2 and APOE levels significantly at all three time points, while DDC, CD9, and PSEN1 levels were significantly reduced as compared to controls (Figure 3A-E). Interestingly, two proteins (CD9 and PSEN1) were significantly (P < 0.01) decreased at all time points, irrespective of the doses applied (Figure 3C-E). Overall, the protein levels were significantly (P < 0.01) altered at 48 h postexposure to repeated doses of sarin as compared to their respective controls (Figure 3A-E, lanes 5 and 6). Obviously, there was a considerable overlap in the results obtained for both mRNA and protein changes for all above-mentioned five proteins in sarin-treated human neuronal cells. Immunofluorescence and Western blot showed consistent results for colocalized proteins. The initial increase (ETS2 and APOE) or decrease (DDC, CD9, and PSEN1) in protein levels at earlier time points (3 and 24 h) were followed by a gradual increase or decline over the later

Neuropathogenesis of sarin

Chem. Res. Toxicol., Vol. 22, No. 6, 2009 993

Figure 2. Principal component, hierarchical, and conditional cluster analyses and data comparison. (A) Loadings for the first, second, and third principal components are plotted on the X-, Y-, and Z-axes, respectively. Controls for different time points (C-3, C-24, and C-48 h) are separated distinctly from treated (T-3, T-24, and T-48 h) groups in a dose- and time-dependent manner. PC ) principal component. (B) The dendogram shows the relationship among gene expression levels (rows) over a series of conditions (columns). Single and repeated doses are placed under separate clusters. Pseudo color scales for fold changes are shown on the right. Saturated blue (-5.0 and above) and saturated red (5.0 and above) indicate decreased and increased gene expression levels, respectively, as compared to controls. The highly up- and down-regulated genes are closely clustered and are interrelated to sarin-induced neurodegeneration and neurodegenerative disorders. (C) The bar diagram shows the qRTPCR expression pattern of 30 genes selected from microarray data. Gene expression of individual samples was normalized to GAPDH and β-actin mRNA levels at respective time points. The overall validation data show the same trend obtained from microarray analysis. All experiments were performed in triplicate and repeated at least twice. Values are expressed as (SE: *P < 0.05, **P < 0.01, and ***P < 0.001.

time point (48 h) (Supporting Information, Figures S3-S7). It is interesting to note that ETS2 immunoreactivity increased in sarin-treated cells at 3 h, followed by a more prominent increase, which became evident at 48 h as compared to untreated control (Supporting Information, Figure S3). Correlation of Microarray Data with qRT-PCR, Western Blot, and Immunofluorescence. Out of several selected genes that correlate well with the ones predicted as potential targets for sarin-induced neurotoxicity, five genes (ETS2, APOE, PSEN1, CD9, and DDC) were selected for confirmation by Western blot and double-immunofluorescence. With both methods, consistent expression profiles were obtained for cells at 48 h postexposure to sarin. The correlation between protein and gene expressions was even more robust for ETS2 and PSEN1. The results of qRT-PCR, Western blot, and immunofluorescence

further corroborate the microarray data, suggesting that ETS2 has the ability to control the expression of several defined target genes delineated in the pathways. Summary of Molecular Pathways Involved. The regulome network diagram (Supporting Information, Figure S9) outlines the possible influence of sarin on gene expression and the subsequent effects on signal transduction. Low-level sarin treatment altered several major pathways in the human neuronal cells related to cholinergic, purinergic, NMDA-glutamatergic, GABAergic, catecholaminergic, and serotogenic signalings, at both single and repeated doses for all time points studied. Other molecules and processes that were differentially altered by repeated doses of sarin exposure at 48 h include AChE/BuChE, neurodegeneration, learning and memory, dementia/ataxia, immune response, inflammatory cascade, mitochondrial-dysfunction, and apoptosis (Supporting Information, Tables ST1-ST4).

994

Chem. Res. Toxicol., Vol. 22, No. 6, 2009

Pachiappan et al.

Table 1. Differentially Expressed Genes Involved in Neurodegeneration and Neuropathogenesis of Human Neuronal Cells Induced by Sarin gene description

single dose 3 or 24 h

repeated doses 48 h

gene symbol

sequence derived from

up-regulated genes apolipoprotein E BCL2-associated X protein cholinergic receptor, nicotinic, epsilon CASP8 and FADD-like apoptosis regulator gutamate receptor, ionotropic, kainate 5 V-ets erythroblastosis E26 homologue 2 GABAA receptor γ 3 subunit N-methyl-D-aspartate receptor subunit NR3 growth arrest and DNA-damage-inducible, R45 adenosine A1 receptor

APOE BAX CHRNE CFLAR GRIK5 ETS2 GABRG3 GRIN2B GADD45A ADORA1

NM_000041 NM_004324 NM_000080 NM_003879 NM_002088 NM_005239 NM_033223 NM_000834 NM_001924 NM_000674

5.02 ( 0.47 2.79 ( 0.06 2.66 ( 0.64 3.89 ( 0.47 2.62 ( 0.50 2.17 ( 0.47 2.59 ( 0.50 2.41 ( 0.52 2.83 ( 0.51 2.76 ( 0.56

11.52 ( 0.98 7.57 ( 0.89 5.56 ( 0.67 5.50 ( 0.97 4.05 ( 0.99 3.99 ( 0.95 3.87 ( 1.01 3.39 ( 0.68 3.37 ( 1.01 3.05 ( 0.74

down-regulated genes adenosine A2a receptor BCL2-like 1 synuclein, R (non-A4 component of amyloid precursor) adenylate kinase 3 brain AChE presenilin 2 butyrylcholinesterase microtubule-associated protein τ amyloid β (A4) precursor protein (protease nexin-II, AD) cholinergic receptor, nicotinic, β polypeptide 4 protease, serine, 12 (neurotrypsin, motopsin) presenilin 1 integral membrane protein 2B calmodulin 1 (phosphorylase kinase, δ) β-2-microglobulin ATP-binding cassette, subfamily B DDC clusterin (apolipoprotein J) calmodulin 3 (phosphorylase kinase, δ) CD9 antigen (p24)

ADORA2A BCL2L1 SNCA AK3 BAChE PSEN2 BCHE MAPT APP CHRNB4 PRSS12 PSEN1 ITM2B CALM1 β2M ABCB1 DDC CLU CALM3 CD9

NM_000675 NM_001191 NM_000345 NM_013410 NM_015921 NM_000447 NM_000055 NM_016841 NM_000484 NM_000750 NM_003619 NM_007318 NM_021999 NM_006888 NM_004048 NM_000927 NM_000790 NM_001831 NM_005184 NM_001769

-2.69 ( 0.49 -2.19 ( 1.04 -3.05 ( 0.32 -4.52 ( 0.51 -2.72 ( 0.13 -3.65 ( 0.51 -3.05 ( 0.61 -3.17 ( 0.47 -3.67 ( 0.47 -2.96 ( 0.63 -4.08 ( 0.47 -3.78 ( 0.45 -3.61 ( 0.49 -2.02 ( 0.43 -3.46 ( 0.92 -2.62 ( 0.44 -2.42 ( 0.44 -2.32 ( 0.48 -3.07 ( 0.62 -2.05 ( 0.41

-2.40 ( 0.75 -2.65 ( 0.96 -3.27 ( 0.94 -4.99 ( 0.96 -4.65 ( 0.68 -5.05 ( 0.96 -5.51 ( 0.72 -5.78 ( 0.95 -5.92 ( 0.92 -6.36 ( 0.78 -6.82 ( 0.96 -7.78 ( 0.95 -11.22 ( 0.94 -20.63 ( 0.95 -21.37 ( 0.94 -23.37 ( 1.04 -23.85 ( 1.04 -39.57 ( 0.92 -39.97 ( 0.98 -99.80 ( 1.04

Fold changes are means ( SD of 18 chips with cross-comparisons between nine sarin-treated groups (three chips per time point) and nine respective control groups (3 × 3 chips). The genes were selected based on their involvement in neurodegenerative pathways and on the persistent (P < 0.001) expression documented after both (single and repeated) doses of sarin in at least two time points (3, 24, or 48 h).

Discussion A mechanistic explanation of possible longer term neurobehavioral consequences at low level exposures is largely lacking at present. Although covalent binding with organophosphates at low level exposures has been suggested (15), brain proteins sensitive to some anticholinesterases cannot be excluded as target(s) for such low-level effects. Among the five proteins (APOE, DDC, CD9, PSEN1, and ETS2) verified by Western blot and immunofluorescence, APOE is associated with increased risk of developing Alzheimer’s disease (AD) and other neurodegenerative disorders (16). DOPA decarboxylase (DDC) synthesizes dopamine and serotonin, and its defect can result in neurodegenerative diseases due to the combined serotonin and catecholamine dearth (17). Signal transduction events that play a role in the regulation of cell development, activation, growth, and motility are mediated by CD9, which is drastically down-regulated by sarin exposure at 48 h in our study. This may suggest association of this gene to neuronal growth abnormalities and motilities. PSEN1 is important in the development of neurodegenerative disease as indicated by its role played in the signaling pathway in AD-related neuronal cell death. Because PSEN1 and Notch-1 are linked with AD risk (18), down-regulation of the Notch homology gene and PSEN1 seen in our study might probably lead to an increase in neuronal cell death. Alterations in PSEN1 are the cause of frontotemporal dementia (19), and PSEN1 down-regulation can contribute to neuronal pathogenesis (20). In our current study, PSEN1 transcript and protein levels were

drastically and significantly down-regulated by sarin treatment at all time points, and that could probably lead to neurodegeneration. Up-regulation of ETS-2, a transcription factor implicated in mental retardation and neurodegenerative disorders, recapitulates several features of Down’s syndrome (DS) pathology (21). In DS/AD and sporadic AD brains, higher ets-2 immunoreactivity was evident in cells associated with BAX, intracellular Aβ, and hyperphosphorylated τ, suggesting that up-regulation of ETS2 promotes activation of mitochondrial death pathway leading to increased neuronal vulnerability and degeneration (22). Strong Ets-2 immunoreactivity observed by us in neuronal cells linked with degenerative markers such as BAX, APP, and microtubuleassociated τ (MAPT) proteins may possibly lead to neuronal damage. ETS2, in particular, was induced 13-fold above the control level. In this scenario, we propose that overexpression of ETS2 is an important contributor to the increased susceptibility of neuropathogenesis through activation of a mitochondrial death pathway. Among the 224 genes found to be differentially regulated, 30 of them were functionally involved in the neurodegenerative process and pathways. Neurodegeneration initiates a cascade of new gene expression, as evidenced herein by the large number of down-regulated transcripts (180 out of 224) seen at 48 h in neuronal cells induced with repeated doses of sarin. They are involved and interlinked with one another, leading to neurodegenerative process. Interestingly, genes believed to be involved in neuronal damage were differentially expressed at 48 h with repeated doses but not with the single dose. Moreover, real-

Neuropathogenesis of sarin

Chem. Res. Toxicol., Vol. 22, No. 6, 2009 995

and neuronal cell destruction (23). The early transcriptional response at 3 h to acute sarin treatment might be attributed to (Figure 1B) increasing ACh levels arising from AChE inhibition and to the corresponding induction of signal transduction processes. The cholinergic pathway then proceeds to play a primary role in regulating the ability to perceive and assign resources for the processing of competing stimuli in the nervous system. This is followed by another cascade involving the reduction of muscarinic receptors and activation of glutamatergic and GABAergic receptors, leading to the release of the excitatory neurotransmitter. In many neurodegenerative and neurologic disorders, neuronal injury and death are caused by the overstimulation of the excitatory amino acids receptors. This in turn activates the NMDA subtype of glutamate receptor and the opening of NMDA calcium channels, resulting in massive Ca2+ influx of postsynaptic cell causing neuronal degeneration. Third, alterations in gene transcription in turn are thought to be an indirect effect of sarin affecting activation of a second signaling system, including AChRs and the ETS-related transcription factors. The sarin-induced signals transmitted through nAChRE then arbitrate the indirect activation of transcription factor ETS2, which regulates and drives PSEN1, PSEN2, APP, SNCA, APOE, DDC, and ABCB1, and promotes the activation of mitochondrial death pathway toward neurodegeneration. In addition, PSEN1 appears to play a crucial role in the normal metabolism of APPand ETS2-controlled pathological developments. In summary, the neurodegenerative effect of sarin is mediated via mitochondrial death pathway in which increased expression of ETS2 regulates the expression of numerous neurodegenerative and cell death-related genes. To our knowledge, no similar neurodegenerative signaling pathway has been described before on a molecular level for sarin-induced neurotoxicity. AChE inhibition plays a direct and consistent role in the development of sarin-induced neurotoxicity. The provided information may aid in the discovery of therapeutic targets for mitigation of neurodegeneration in human. Figure 3. Western blot analysis and quantification. Changes in the protein levels of (A) V-ets erythroblastosis (ETS2); (B) apolipoproteinE (APOE); (C) dopa-decarboxylase (DDC); (D) CD9 molecule (CD9); and (E) presenilin1 (PSEN1) in SH-SY5Y cell homogenates at 3, 24, and 48 h after sarin (5 µg/mL) treatment. Equal amounts of proteins were loaded. Control (lanes 1, 3, and 5) and sarin treated (lanes 2, 4, and 6) for all time points were examined. Hybridizing with anti-ETS2, APOE, DDC, CD9, and PSEN1 was detected as 42, 55, 53, 24, and 47 kDa bands, respectively. Single dose, lanes 1-4; repeated doses, lanes 5 and 6. Protein expression changes were quantified by densitometry analysis as represented by bars in the diagram. Values are expressed as means ( SDs. Level of significance: *P e 0.05, and **P e 0.01 as compared to controls. C ) control, and T ) treated.

time PCR data obtained with HCN-1A cells further confirmed that sarin treatment at 48 h did significantly affect the expression level of these neurodegenerative genes discussed above (Supporting Information, Figure S8). Hence, in the human model of neurodegeneration, the specific genes expressed within 48 h may be ascribed to sarin induction with repeated doses. The late responses and the extent of neurodegeneration indicated in the pathway lead to neuropathogenesis of the disease (Supporting Information, Figure S9). These pathways show that sarininduced neurotoxicity possibly involves multiple molecular mechanisms including neuromuscular and interneuronal transmission, which may operate through a variety of receptor subtypes and activate different intracellular responses in SHSY5Y cells. Here, we postulated three distinct and consequential mechanisms that may contribute to sarin-induced AChE inhibition

Acknowledgment. We are thankful to the National University of Singapore and DSO National Laboratories. This research was supported by research grant #R 181-000-066-123/232 from DSTA, Singapore. Supporting Information Available: Photomicrographs of SH-SY5Y cells (Figure S1), overall conditional and hierarchical cluster analysis (Figure S2), confocal images (Figures S3-S7), qRT-PCR expression pattern (Figure S8), and neurodegenerative-signaling pathway (Figure S9). Gene-specific primer sequences used in qRT-PCR validation (Table ST1), list of differentially expressed genes (Table ST2), and entity, relation, and the major groups of relation of genes involved in the pathways, respectively (Tables ST3-ST5). This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Niijima, H., Nagao, M., Nakajima, M., Takatori, T., Matsuda, Y., Iwase, H., and Kobayashi, M. (1999) Sarin-like and soman-like organophosphorous agents activate PLCgamma in rat brains. Toxicol. Appl. Pharmacol. 156, 64–69. (2) Taylor, P., and Brown, J. H. (1999) Acetylcholine. In Basic Neurochemistry: Molecular Cellular and Medical Aspects (Siegel, G. J., Agranoff, B. W., Alberts, R. W., Fisher, S. K., and Uhler, S. K., Eds.) pp 213-242, Lippincott-Raven Publishers, Philadelphia. (3) Yokoyama, K., Araki, S., Murata, K., Nishikitani, M., Okumura, T., Ishimatsu, S., Takasu, N., and White, R. F. (1998) Chronic neurobehavioral effects of Tokyo subway sarin poisoning in relation to posttraumatic stress disorder. Arch. EnViron. Health 53, 249–256.

996

Chem. Res. Toxicol., Vol. 22, No. 6, 2009

(4) Duffy, F. H., and Burchfield, J. L. (1980) Long term effects of the organophosphate sarin on EEGs in monkeys, and humans. Neurotoxicology 1, 667–689. (5) Kadar, T., Shapira, S., Cohen, G., Sahar, R., Alkalay, D., and Raveh, L. (1995) Sarin-induced neuropathology in rats. Hum. Exp. Toxicol. 14, 252–259. (6) Afshari, C. A. (2002) Perspective: Microarray technology, seeing more than spots. Endocrinology 143, 1983–1989. (7) Damodaran, T. V., Jones, K. H., Patel, A. G., and Abou-Donia, M. B. (2003) Sarin (nerve agent GB)-induced differential expression of mRNA coding for the acetylcholinesterase gene in the rat central nervous system. Biochem. Pharmacol. 65, 2041–2047. (8) Damodaran, T. V., Greenfield, S. T., Patel, A. G., Dressman, H. K., Lin, S. K., and Abou-Donia, M. B. (2006) Toxicogenomic studies of the rat brain at an early time point following acute sarin exposure. Neurochem. Res. 31, 367–381. (9) Damodaran, T. V., Patel, A. G., Greenfield, S. T., Dressman, H. K., Lin, S. M., and Abou-Donia, M. B. (2006) Gene expression profiles of the rat brain both immediately and 3 months following acute sarin exposure. Biochem. Pharmacol. 71, 497–520. (10) Pachiappan, A., Thwin, M. M., Manikandan, J., and Gopalakrishnakone, P. (2005) Glial inflammation and neurodegeneration induced by candoxin, a novel neurotoxin from Bungarus candidus venom: Global gene expression analysis using microarray. Toxicon 46, 883–899. (11) Jamsa, A., Hasslund, K., Cowburn, R. F., Backstrom, A., and Vasange, M. (2004) The retinoic acid and brain-derived neurotrophic factor differentiated SH-SY5Y cell line as a model for Alzheimer’s diseaselike tau phosphorylation. Biochem. Biophys. Res. Commun. 319, 993– 1000. (12) Bobashev, G., Das, S., and Das, A. (2002) In Methods of Microarray Data Analysis II (Lin, S. M., and Johnson, K. F., Eds.) Kluwer Academic Publishers, Boston, MA. (13) Livak, K. J., and Schmittgen, T. D. (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-delta delta C(T)) method. Methods 25, 402–408.

Pachiappan et al. (14) Yuryev, A., Mulyukov, Z., Kotelnikova, E., Maslov, S., Egorov, S., Nikitin, A., Daraselia, N., and Mazo, I. (2006) Automatic pathway building in biological association networks. BMC Bioinf. 7, 171–183. (15) Richards, P., Johnson, M., Ray, D., and Walker, C. (1999) Novel protein targets for organophosphorus compounds. Chem.-Biol. Interact. 119-120, 503–511. (16) Huang, Y. (2006) Molecular and cellular mechanisms of apolipoprotein E4 neurotoxicity and potential therapeutic strategies. Curr. Opin. Drug DiscoVery DeV. 9, 627–641. (17) Christenson, J. G., Dairman, W., and Udenfriend, S. (1972) On the identity of DOPA decarboxylase and 5-hydroxytryptophan decarboxylase (immunological titration-aromatic L-amino acid decarboxylaseserotonin-dopamine-norepinephrine). Proc. Natl. Acad. Sci. U.S.A. 69, 343–347. (18) Redmond, L., Oh, S. R., Hicks, C., Weinmaster, G., and Ghosh, A. (2000) Nuclear Notch1 signaling and the regulation of dendritic development. Nat. Neurosci. 3, 30–40. (19) Raux, G., Gantier, R., Thomas-Anterion, C., Boulliat, J., Verpillat, P., Hannequin, D., Brice, A., Frebourg, T., and Campion, D. (2000) Dementia with prominent frontotemporal features associated with L113P presenilin 1 mutation. Neurology 55, 1577–1578. (20) Pastorcic, M., and Das, H. K. (2000) Regulation of transcription of the human presenilin-1 gene by ets transcription factors and the p53 protooncogene. J. Biol. Chem. 275, 34938–34945. (21) Sumarsono, S. H., Wilson, T. J., Tymms, M. J., Venter, D. J., Corrick, C. M., Kola, R., Lahoud, M. H., Papas, T. S., Seth, A., and Kola, I. (1996) Down’s syndrome-like skeletal abnormalities in Ets2 transgenic mice. Nature 379, 534–537. (22) Helguera, P., Pelsman, A., Pigino, G., Wolvetang, E., Head, E., and Busciglio, J. (2005) Ets-2 promotes the activation of a mitochondrial death pathway in Down’s syndrome neurons. J. Neurosci. 25, 2295– 2303. (23) McDonough, J. H., Jr., and Shih, T. M. (1997) Neuropharmacological mechanisms of nerve agent-induced seizure and neuropathology. Neurosci. BiobehaV. ReV. 21, 559–579.

TX8003467