Expression Profiling of Fibroblasts Identifies Cell Cycle Abnormalities in Schizophrenia L. Wang,†,‡ H. E. Lockstone,†,‡ P. C. Guest,‡ Y. Levin,‡ A. Palota´s,§ S. Pietsch,‡ E. Schwarz,‡ H. Rahmoune,‡ L. W. Harris,‡ D. Ma,‡ and S. Bahn*,‡ Institute of Biotechnology, University of Cambridge, Cambridge, U.K., and Asklepios-Med Bt. (private practice and research centre) H-6722 Szeged, Kossuth Lajos sgt. 23, Hungary Received September 26, 2009
Many previous studies have attempted to gain insight into the underlying pathophysiology of schizophrenia by studying postmortem brain tissues of schizophrenia patients. However, such analyses can be confounded by artifactual features of this approach such as lengthy agonal state and postmortem interval times. As several aspects of schizophrenia are also manifested at the peripheral level in proliferating cell types, we have studied the disorder through systematic transcriptomic and proteomic analyses of skin fibroblasts biopsied from living patients. We performed comparative transcriptomic and proteomic profiling to characterize skin fibroblasts from schizophrenia patients compared to healthy controls. Transcriptomic profiling using cDNA array technology showed that pathways associated with cell cycle regulation and RNA processing were altered in the schizophrenia subjects (n ) 12) relative to controls (n ) 12). LC-MSE proteomic profiling led to identification of 16 proteins that showed significant differences in expression between schizophrenia (n ) 11) and control (n ) 11) subjects. Analysis in silico revealed that these proteins were also associated with proliferation and cell growth pathways. To validate these findings at the protein level, fibroblast protein extracts were analyzed by Western blotting which confirmed the differential expression of three key proteins associated with these pathways. At the functional level, we confirmed the decreased proliferation phenotype by showing that cultured fibroblasts from schizophrenia subjects (n ) 5) incorporated less 3H-thymidine into their nuclei compared to those from controls (n ) 6) by day 4 over an 8 day time course study. Similar abnormalities in cell cycle and growth pathways have been reported to occur in the central nervous system in schizophrenia. These studies demonstrate that fibroblasts obtained from living schizophrenia subjects show alterations in cellular proliferation and growth pathways. Future studies aimed at characterizing such pathways in fibroblasts and other proliferating cell types from schizophrenia patients could elucidate the molecular mechanisms associated with the pathophysiology of schizophrenia and provide a useful model to support drug discovery efforts. Keywords: Schizophrenia • gene profiling • proteomics • fibroblasts cell cycle
Introduction Schizophrenia is a severe and common psychiatric disorder, affecting approximately 1% of the world’s population. Twin and family studies have shown that complex interactions between genetic and environmental factors appear to precipitate the disorder, although the etiology and pathogenesis remain mostly unknown. Schizophrenia is characterized by psychosis, hallucinations and disordered thought, and accordingly, the majority of research has focused on identifying abnormalities within the brain. However, our laboratory and others have also * To whom correspondence should be addressed. Dr Sabine Bahn, Director Cambridge Centre for Neuropsychiatric Research, Institute of Biotechnology, University of Cambridge, Tennis Court Road, Cambridge CB21QT, U.K., e-mail
[email protected], Tel 01223 76 7799, Fax 01223 334 162. † These authors contributed equally to this work. ‡ University of Cambridge. § Asklepios-Med Bt. 10.1021/pr900867x
2010 American Chemical Society
detected differences in peripheral systems in schizophrenia patients. These include effects on T-cells,1 other immunological abnormalities,2 decreased levels of apolipoprotein A1 in peripheral tissues and brain,3 and fiber atrophy in the neuromuscular system.4 These findings support the hypothesis that a systemic biochemical defect may exist that affects the function of the brain and other systems throughout the body.5 Studies of psychiatric disorders at the molecular level have often been carried out using postmortem tissue but this has several limitations. These limitations include potentially poor RNA integrity after lengthy agonal state and/or postmortem intervals, and confounding factors associated with the chronic nature of the disorder and treatment with antipsychotic medications. As many molecular pathways in neuronal cells are shared with peripheral cell types, the study of such cells can circumvent the problems associated with postmortem tissue and provide a useful model for investigating the disorder. Journal of Proteome Research 2010, 9, 521–527 521 Published on Web 11/16/2009
research articles
Wang et al.
Table 1. Demographic Details of Fibroblast Samples Used for Transcriptomic, Proteomic, and Proliferation Studies transcriptomic profiling
Number Age Gender (m/f)
proteomic profiling
schizophrenia
control
schizophrenia
control
schizophrenia
control
n ) 12 47.9 ( 12.7 5/7
n ) 12 47.5 ( 14.0 6/6
n ) 11 59.8 ( 13.4 3/8
n ) 11 50.5 ( 12.3 4/7
n)5 59.8 ( 17.1 2/3
n)6 51.7 ( 11.9 3/3
A small number of studies have used fibroblast cultures to study schizophrenia and these have identified alterations in growth and morphology,6 decreased cellular adhesion,7 and altered apoptotic pathways.8 However, there has been only one study so far which has used high-throughput transcriptomic profiling to analyze fibroblast cultures from schizophrenia patients.9 Here, we have employed a multiplatform approach to identify disease-related abnormalities in schizophrenia fibroblasts by analyses at the transcriptomic, proteomic and functional levels.
Materials and Methods Clinical Samples. Forearm skin biopsies were obtained at Asklepios-Med Bt, Hungary, from schizophrenia patients and control subjects. Controls were recruited from the same geographical area and confirmed as controls by screening for any medical conditions, psychological disorders or drug usage. All subjects provided written agreement to participate in this Institutional Review Board (IRB)-approved study (Hungary). Subjects were assessed using the Diagnostic Interview for Psychosis10 and diagnosed according to DSM-III-R. Samples were obtained from antipsychotic-treated schizophrenia patients in accordance with clinical visitations and those from control subjects were collected at approximately the same time periods on different days. There was no significant difference in storage times between the schizophrenia and control groups. In some cases, samples were missing due to losses during preparation procedures, although the sample numbers (N) given (Table 1) reflect those actually used in the experimental analyses. All studies were designed to ensure that there was sufficient overlap of samples in the different experiments for cross comparison purposes (see Supplementary Table 1, Supporting Information). Fibroblast Cell Culture. Cells were isolated from tissue in supplemented RPMI 1640 medium (Sigma; Poole, Dorset, UK) containing 10% fetal calf serum (FCS), 1% L-Glutamine, Penicillin, Streptomycin solution (Sigma) in 25 cm2 culture flasks. After 2 weeks, fibroblasts were detached with 0.25% Trypsin/0.02% EDTA-solution (Sigma) and cultivated further with a starting density of 5 × 105 in 75 cm2 flasks in 5% CO2 at 37 °C in supplemented RPMI 1640. Fibroblasts were detached with 0.25% Trypsin/0.02% EDTA-solution when cells reached 80-90% confluence. Media were changed every 3 days. Microarray Hybridization. Fibroblasts from 12 schizophrenia and 12 controls (passage 3) were used for transcriptomic analysis (Table 1). Fibroblasts were harvested for microarray analysis after reaching 80-90% confluence in 75 cm2 flasks. Total RNA was extracted using RNeasy Plus Mini kit (Qiagen, West Sussex, U.K.) and purified using the Qiaquick PCR purification kit (Qiagen). The Agilent 2100 Bioanalyser “lab on a chip” (Agilent technologies, Palo Alto, CA) was used to confirm RNA integrity according to the manufacturer’s instructions and all samples were included in the study (data not shown). Total RNA (1 µg) from each sample was used to prepare biotinylated fragmented cRNA according to the GeneChip Expression Analysis Protocol (GE Healthcare; Little Chalfont, 522
cell proliferation
Journal of Proteome Research • Vol. 9, No. 1, 2010
Bucks, UK) using the Codelink Expression Assay kit. Each cRNA sample was hybridized to separate CodeLink Human Whole Genome Bioarrays for 18 h in a 37 °C incubator with constant rotation of 300 rpm. Chips were washed and stained with Cy5Streptavidin and scanned using the GenePix 4000B, according to the Scanning CodeLink Bioarrays protocol (GE Healthcare). Microarray Data Analysis. Image analysis and feature extraction were performed using the Codelink Bioarray software (GE Healthcare). This software outputs signal intensity data and assigns a virtual reporter flag for each probe, indicating the quality of the expression measurement. Features free from background contamination, shape irregularity or pixel saturation were flagged “good”. Background corrected intensity data were read into the R statistical software package11 and analyzed using BioConductor (http://www.R-project.org) packages.12 Data were filtered to retain only those probes which were flagged ‘good’ in at least 75% of the samples (16 294 probes) before quantile normalization was performed. Quality control checks revealed one outlier sample (schizophrenia group), which was removed and data were renormalized. Principal component analysis (PCA) revealed that the main source of variation in this data set was a batch effect [samples were hybridized in two batches of 12 samples (6 schizophrenia and 6 controls) due to capacity of the incubator]. The batch effect was accounted for using the Prediction Analysis for Microarrays (PAMR) package from BioConductor. PCA performed after this adjustment confirmed that the batch effect was no longer a source of variation (data not shown). The Linear Models for Microarray Analysis (LIMMA) package13 from BioConductor was used to assess differential gene expression between fibroblasts derived from schizophrenia patients and healthy controls. Raw P-values were adjusted for multiple testing using the false discovery rate (FDR) method of Benjamini and Hochberg.14 Gene Set Enrichment Analysis (GSEA)15,16 was used to perform a functional category-based analysis, which assesses the entire ranked list of genes to identify gene ontology (GO) biological process categories for enrichment at the top (increased expression in schizophrenia) or bottom (decreased in schizophrenia) of the list. Probes were ranked by t-statistic and analyzed at the gene level using default settings for analyzing preranked lists in GSEA. All data is MIAME compliant and the raw data has been deposited in a MIAME compliant database as detailed on the MGED Society Web site (http://www.mged. org/Workgroups/MIAME/miame.html). Proteomic Profiling Using Liquid Chromatography Mass Spectrometry (LC-MSE). Fibroblasts from 11 schizophrenia and 11 controls (passage 3) were used for proteomic analysis (Table 1). These included 6 schizophrenia and 9 control samples analyzed in the microarray experiment. The ProteoExtract subcellular Proteome Extraction kit (S-PEK) (Merck, U.K.) was used to separate the total lysate into four cellular fractions: cytosolic, membrane, nuclear and cytoskeletal. The cytosolic fraction was analyzed in this study because it contains the highest number of abundant soluble proteins to facilitate robust
research articles
Cell Cycle Abnormalities in Schizophrenia E
LC-MS analysis. Protein concentrations were determined for the cytosolic fraction using the detergent-compatible DC protein assay (Bio-Rad, Hercules, CA). Contaminating salts and mild nonionic detergents were removed by buffer exchange into 50 mM ammonium bicarbonate and trypsin digestion was carried out as described previously.17 Prior to LC-MSE analysis, 25fmol/µL Saccharomyces cerevisiae enolase tryptic digest was added to each sample for normalization purposes. Samples were analyzed using split-less nano Ultra Performance Liquid Chromatography (10kpsi nanoAcquity, Waters, Milford MA) coupled to a Quadrupole Time-of-Flight mass spectrometer (Qtof Premier, Waters Milford MA) in MSE (Expression) mode as described previously.17 A reference compound (Glu-Fibrinopeptide B) was infused using the LockSpray and scanned every 30 s to maintain mass accuracy. Analyte spectra were corrected based on the Glu-Fibrinopeptide B (Sigma) reference spectra. Raw data were processed using the ProteinLynx Global Server software version 2.3 (Waters). Quantitative and qualitative information was produced automatically by the software according to user defined thresholds as described.18 Intensity measurements were obtained and protein identification achieved as described previously.19 Ions detected in at least 2 out of 3 injections of each sample and at least 60% of the samples were included in the analysis. Protein identifications were based on at least two peptides and the maximum random false identification rate was set to 4%. Proteomic Data Analysis. Statistical data analysis of proteomic data was performed using the R software package (www.r-project.org). Student’s t-test was applied to identify differences in the level of protein expression between schizophrenia and control groups. Proteins reaching the p < 0.05 threshold were considered significant and this corresponded to a false discovery rate (FDR) of less than 20%. In Silico Pathway Analysis. The Ingenuity Pathways Analysis (IPA) software (http://www.ingenuity.com) was used to identify the functions of differentially expressed proteins and their interaction networks. IPA classifies data in a systematic way using a comprehensive ontology of functional annotations and protein-protein interaction data derived from the published literature. The most significant interaction networks, biological functions and pathways associated with the differentially expressed proteins were identified. Western Blot Validation. Total protein extracts of each sample (20 µg) were separated by SDS/PAGE electrophoresis (180 V for 45 min.) using 4-12% NuPAGE Novex Bis-Tris gels (Invitrogen, Paisley, U.K.). Proteins were transferred electrophoretically onto nitrocellulose membranes (Invitrogen) at room temperature using a semidry BioRad transfer system. Membranes were blocked by incubation for 1 h in PBS containing 5% skimmed milk powder and 0.1% Tween 20. After washing, membranes were incubated with primary antibodies at 4 °C overnight. The membranes were washed three times by gentle agitation in PBS containing 0.1% Tween 20. Secondary antibodies (HRP-conjugated horse antimouse or goat antirabbit IgG) were diluted at 1:1000 in PBS. Membranes were incubated for 90 min at room temperature. The immune complexes were visualized using the ECL Plus kit and exposure to Amersham Hyperfilm ECL (GE Healthcare). The intensity of the protein bands on films was quantified using the Scion imaging software (NIH scion image, U.S.A.). Student’s t test was applied for the statistical analysis with a significance threshold of P < 0.05. Fold changes were calculated
by comparing the density of bands between schizophrenia and control subjects.13 Cell Proliferation Assay. Fibroblasts (2.5 × 104) were plated in 12-well plates (Fisher Scientific) in 5% CO2 at 37 °C in supplemented RPMI 1640. After 24 h, the cells were washed twice with PBS (0.5 mL) and then incubated in step down medium (RPMI 1640 containing 1% FCS) for 10 h. Cells were pulsed with 0.037MBq 3H-thymidine (Amersham Biosciences) allowing incorporation into DNA for a 24 h incubation period at the indicated time points. Cells were washed twice in PBS and then 1 mL of 5% tricloroacetic acid (TCA) was added and the mixture incubated at 4 °C for 2 h. The cells were washed twice with 5% TCA/95% ethanol and then dried for 15 min at 37 °C. After this, 0.5 mL 0.2 M NaOH was added for a 3 h incubation at 37 °C and the mixtures were added to 5 mL vials containing 3.5 mL scintillation liquid and left to sit overnight. Incorporation of 3H-thymidine into progeny DNA was determined using a liquid scintillation counter (Packard). Student’s t-test was used to compare the schizophrenia and control groups at each time point.
Results Transcriptomic Profiling. Transcriptomic analysis of fibroblasts from 12 schizophrenia and 12 control subjects was carried out using Codelink Whole Genome Bioarrays. Over 400 transcripts were differentially expressed between schizophrenia samples and controls at a significance threshold of p < 0.05 although none of these remained significant after adjusting for multiple testing (false discovery rate >0.99). This suggested the need for a larger sample size to increase confidence in detection of significant differences. Nevertheless, to investigate whether a biological signal could be detected in the data, the list of differentially expressed transcripts was analyzed to identify those that showed functional relationships. Thirteen categories with FDR corrected p-values below 0.05 were identified among the transcripts that showed increased expression in schizophrenia, although no significant categories were found among the transcripts with decreased expression (Table 2). Categories with the highest significance were those containing several closely related terms within the GO hierarchy, which arises due to overlapping genes. The two most significantly enriched biological processes among the genes showing higher expression in schizophrenia were cell cycle/mitosis and RNA processing (Table 2). Proteomic Profiling. High-throughput protein profiling of fibroblast samples from 11 schizophrenia subjects and 11 controls was performed using quantitative label-free LC-MSE analysis. A total of 580 proteins were identified based on peptides that were detected in at least 2 out of the 3 injections for each sample. These were filtered further to result in 389 proteins detected in at least 60% of the samples. Statistical analysis identified 16 proteins that showed significantly different expression between schizophrenia and control samples (Table 3). After correcting for multiple testing, the false discovery rate was below 20%, which was considered an acceptable level. The differentially expressed proteins included those involved in in regulation of the cell cycle and cellular development (annexin A5, caspase recruitment domain-containing protein 6, actin, cytoplasmic 1 β-actin, hsp90 cochaperone cdc37 and calpain small subunit 1), RNA regulation (the snRNA activating protein complex subunit1, UV excision repair protein RAD23 homologue B and isoleucyl-tRNA synthetase, cytoplasmic) and Journal of Proteome Research • Vol. 9, No. 1, 2010 523
research articles
Wang et al. a
Table 2. Functional Categories Significantly Enriched among the Top-Ranking Transcripts in the Microarray Analysis increase in schizophrenia
a
Gene Ontology ID
Gene Ontology category
size
ES
NES
FDR q-val
GO:0022402 GO:0065004 GO:0000278 GO:0000087 GO:0000279 GO:0006325 GO:0007067 GO:0051276 GO:0006396 GO:0008380 GO:0022403 GO:0007049 GO:0000226
Cell cycle process Protein-DNA complex assembly Mitotic cell cycle M phase of mitotic cell cycle M phase Estab/ maint chromatin architecture Mitosis Chromosome organization RNA processing RNA splicing Cell cycle phase Cell cycle Microtubule cytoskeleton organization
107 25 86 43 58 45 43 69 119 71 92 184 21
0.44 0.59 0.46 0.54 0.50 0.52 0.54 0.50 0.42 0.45 0.43 0.39 0.65
1.91 1.91 1.93 1.95 1.97 1.91 1.99 1.99 1.82 1.80 1.81 1.79 1.99
0.010 0.011 0.011 0.012 0.012 0.012 0.013 0.019 0.029 0.031 0.031 0.033 0.038
Abbreviations: ES, enrichment score; NES, normalised enrichment score; FDRqval, false discovery rate q-value.
Table 3. Proteins Significantly Changed in Fibroblasts from Schizophrenia Patients Identified by LC-MSEa
a
acc no
gene name
protein name
fold change
P-value
P47895 P80723 Q99536 P08758 Q16543 Q16533 Q9HC38 Q96HP4 P60709 Q9Y5Z4 P54727 Q9BX69 P04632 P78417 P41252 Q9UI15
ALDH1 BASP1 VAT1 ANXA5 CDC37 SNAPC1 GLOD4 OXNAD1 ACTB HEBP2 RAD23B CARD6 CAPNS1 GSTO1 IARS TAGLN3
Aldehyde dehydrogenase 1A3 Brain acid soluble protein 1 Synaptic vesicle membrane protein VAT-1 Annexin A5 Hsp90 cochaperone Cdc37 snRNA activating protein complex subunit 1 Glyoxalase domain-containing protein 4 Oxidoreductase NAD-binding domain-containing 1 Actin, cytoplasmic 1 Heme-binding protein 2 UV excision repair protein RAD23 homologue B Caspase recruitment domain-containing protein 6 Calpain small subunit 1 Glutathione transferase omega-1 Isoleucyl-tRNA synthase, cytoplasmic Transgelin-3
2.16 1.82 1.54 1.50 1.49 1.45 1.44 1.43 1.36 1.35 1.34 1.32 1.32 1.30 1.24 0.56
0.018 0.048 0.009 0.006 0.047 0.018 0.022 0.021 0.042 0.042 0.032 0.018 0.048 0.045 0.041 0.015
The Q-value was 0.185 for all proteins.
Table 4. In Silico Pathway Analysis of Differentially Expressed Proteinsa molecules in network
score
focus molecules
top function
ACTB, Akt, ALDH1A3, ANXA5, ascorbic acid, BASP1, CAPNS1, CARD6, CCT5, COTL1, DAPK1, CDC37, CLIC4, DARS, EIF2AK1, EPRS, ERCC5, F2, GSTO1, HRAS,IARS (includes EG:3376), JUN, NFkB, PROS1, RAD23A, RAD23B, RAGE, RB1, SNAPC1, SNAPC3, TAGLN3, TDG, TGFB1, TP53
34
12
Cell Death, Cellular Development, Cellular Growth and Proliferation
a Twelve out of the 16 differentially expressed proteins were assembled into a single interaction network using the Ingenuity Pathway Analysis tool. The network was associated significantly with cell death, cellular development, cellular growth and proliferation pathways. The 12 focus proteins are indicated in bold. Interacting proteins assigned by the Ingenuity Pathway Analysis tool are shown in lightface. Abbreviations: Akt, protein kinases B; CCT5, T-complex protein1 subunit epsilon; COTL1, Coactosin-like protein; DAPK1, Death-associated protein kinase 1; CLIC4, Chloride intracellular channel protein 4; DARS, Aspartyl-tRNA synthetase, cytoplasmic; EIF2AK1, Eukaryotic translation initiation factor 2-alpha kinase 1; EPRS, Bifunctional aminoacyl-tRNA synthetase; ERCC5, DNA repair protein complementing XP-G cells; F2, Prothrombin; HRAS, GTPase HRas; JUN, Transcription factor AP-1; NFkB, Nuclear factor NF-kappa-B p105 subunit, PROS1, Vitamin K-dependent protein S; RAD23A, UV excision repair protein RAD23 homolog A; RAGE, Advanced glycosylation end product-specific receptor; RB1, Retinoblastoma-associated protein; SNAPC3, snRNA-activating protein complex subunit 3; TDG, G/T mismatch-specific thymine DNA glycosylase; TGFB1, Transforming growth factor beta-1; TP53, Cellular tumor antigen p53.
the oxidative stress response (glutathione transferase omega-1 and calpain small subunit 1). Finally, synaptic vesicle membrane protein VAT-1 homologue and transgelin-3 are involved in the regulation of nervous system development. The transcripts corresponding to these proteins did not show any evidence for being altered in schizophrenia in the microarray data (not shown). Ingenuity Pathway Analysis was used to find biological functions and networks significantly associated with the 16 altered proteins and their interacting proteins. The most 524
Journal of Proteome Research • Vol. 9, No. 1, 2010
significant network contained 12 of the differentially expressed proteins. The most significant biological functions associated with this network were cell death, cellular development, cellular growth and proliferation (Table 4). These results are consistent with the biological functions identified in the transcriptomic profiling data. Western Blot Validation. Western blot analysis was used to validate 3 differentially expressed proteins, selected based on their roles in cell death, cellular development, cellular growth and proliferation functions. This analysis showed that calpain
research articles
Cell Cycle Abnormalities in Schizophrenia
Figure 1. Confirmation of changes in selected proteomic markers by Western blot analysis. (a) Western blot images showing significantly increased expression levels of calpain small subunit 1, Cdc37 and annexin A5 in fibroblasts from schizophrenia patients (n ) 5) compared to healthy controls (n ) 5). (b) Scatter plots showing results of densitometric scanning of the band intensities. P-values were calculated using a one tailed t test as increases in all proteins were being tested. Note: one sample was missing for annexin A5 in the control set.
Figure 2. Functional validation of transcriptomic/proteomic results showing decreased proliferation of fibroblasts from schizophrenia subjects. Scatter plots showing the incorporation of 3H-thymidine into fibroblast nuclei from schizophrenia (N ) 5) and control (N ) 6) subjects over an 8 day time course. P-values were calculated using a two-tailed t test.
small subunit 1, hsp90 cochaperone cdc37 (cdc37) and annexin A5, were increased significantly in schizophrenia patients compared with healthy controls consistent with the profiling results (Figure 1). Cell Proliferation Assay. Based on the cell cycle findings at the transcriptomic and proteomic levels, and on previous evidence from our laboratory showing a decreased proliferative response of T-cells to stimulation in schizophrenia patients,1 we measured the rate of proliferation in fibroblasts by incorporation of 3H-thymidine for functional validation of the transcriptomic and proteomic findings. No difference was observed between schizophrenia and control subjects at day 1. However, by day 4 and day 8, significant decreases were observed in schizophrenia samples compared to controls (P ) 0.038 and P ) 0.046, respectively) (Figure 2). It should be noted that the schizophrenia patients used in this study were slightly older than controls (59.8 ( 17.1 years and 51.7 ( 11.9 years, respectively), although this difference was not significant and only modest correlations were found between cell viability and age within each group (Pearson correlation test: R ) 0.14 for
controls and R ) 0.30 for schizophrenia patients). Therefore, it is not likely that the difference in subject age can fully explain the reduced proliferation observed in fibroblasts from schizophrenia patients.
Discussion In the present study, we have used transcriptomic and proteomic profiling, together with cell proliferation assays, to characterize fibroblast cultures from schizophrenia patients compared to healthy controls. Although the effects were modest, all three levels of analysis indicated that cell cycle and survival abnormalities are present in fibroblasts from schizophrenia patients. It should be noted that these samples were obtained from patients who were receiving antipsychotic medication at the time of biopsy. Therefore, it can not be excluded that some of the findings could be due to residual drug effects. However, it is likely that any such effects were eradicated during the cell culture procedure. In the microarray data, gene ontology categories related to the cell cycle (parJournal of Proteome Research • Vol. 9, No. 1, 2010 525
research articles ticularly mitosis) and RNA processing were significantly enriched among the top-ranking transcripts increased in schizophrenia. Increased expression of some of these genes would have the functional effect of suppressing cell proliferation (CHEK1, CHFR) or inducing apoptosis (ZAK, CDC6) and cell cycle arrest (PCBP4). A slower growth rate in fibroblasts from schizophrenia patients compared to healthy controls was reported in an early study characterizing this cell type6 and similar results were found using olfactory cells obtained from schizophrenia subjects.21 These results support the recent finding from our laboratory that T-cells obtained from schizophrenia patients show reduced proliferation in response to stimulation compared to those from healthy controls.1 Taken together, these findings suggest a dysregulation of the cell cycle in schizophrenia subjects. The protein profiling results provided additional support for cell cycle/proliferation and survival defects in schizophrenia as several of the altered proteins are associated with regulation of these pathways (annexin A5, caspase recruitment domaincontaining protein 6, β-actin, hsp90 cochaperone cdc37and calpain small subunit 1). The differential expression of three of these proteins (annexin A5, hsp90 cochaperone cdc37 and calpain small subunit 1) was confirmed by Western blot analysis, providing validation of these findings. One notable feature of these analyses was a strong bias toward increased expression of transcripts and proteins in schizophrenia. In the transcriptomic study, all 13 of the categories were increased. In the proteomic study, 15 out of 16 proteins showed increased expression. The reason for this and the implications are not clear. There was not a strong concordance between the altered proteins and the gene expression data. A similar lack of concordance between protein and mRNA expression levels has been demonstrated previously in a landmark study.22 This lack of overlap could arise through regulatory mechanisms such as post-translational modification. Several studies report evidence of cell cycle and proliferation abnormalities in the brain itself. For example, a recent microarray study identified expression differences in genes associated with cell cycle pathways in the anterior cingulate gyrus, which was hypothesized to contribute to the myelin and oligodendrocyte deficits observed in schizophrenia.23 Reduced proliferation of hippocampal neural stem cells, which may relate to abnormalities in adult neurogenesis, has also been reported in schizophrenia patients.24 There is also evidence for a reduced incidence of cancer in schizophrenia patients and an increase in apoptosis has been proposed as a mechanism that may play a central role in both this phenomenon and neurodevelopment.25,26 This is consistent with our results (unpublished data) and those from other researchers27 showing decreased levels of epidermal growth factor (EGF)28 in serum from drug-naive schizophrenia patients compared to healthy controls. It is well established that schizophrenia patients have a shorter life expectancy compared to healthy controls. This is most likely due to the high comorbidity with late onset diseases such as type II diabetes mellitus and cardiovascular disease.29 Numerous studies have reported an increased incidence of insulin-resistance and diabetes in schizophrenia patients30,31 including first onset, antipsychotic-naı¨ve patients.32 Thus, it would be of interest to investigate the role of insulin signaling pathways in cell cycle abnormalities in schizophrenia. Thus, it is critical to identify and implement novel diagnostic strategies 526
Journal of Proteome Research • Vol. 9, No. 1, 2010
Wang et al. to facilitate early identification of schizophrenia for improved disease management before the onset of such complicating factors. In summary, our results show alterations in expression of mRNA transcripts and proteins involved in the cell cycle and concomitant abnormalities in the growth response of fibroblasts from schizophrenia patients. These changes may have functional ramifications in neuronal cells and consequently in the pathophysiology of schizophrenia. Further studies aimed at characterizing these pathways in fibroblasts and other proliferating cell types could result in a better understanding of the onset and etiology of schizophrenia. In addition, these studies demonstrate that fibroblasts obtained from living schizophrenia subjects could serve as a useful model of pathophysiological mechanisms to support drug discovery efforts.
Acknowledgment. The research was funded by the Stanley Medical Research Institute and by Psynova Neurotech Ltd. We thank Rachel Craddock and John Goodall for helpful discussion and comments on the manuscript. Supporting Information Available: Supplementary Table 1. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Craddock, R. M.; Lockstone, H. E.; Rider, D. A.; Wayland, M. T.; Harris, L. J.; McKenna, P. J.; et al. Altered T-cell function in schizophrenia: a cellular model to investigate molecular disease mechanisms. PLoS ONE 2007, 2 (1), e692. (2) Schwarz, M. J.; Muller, N.; Riedel, M.; Ackenheil, M. The Th2hypothesis of schizophrenia: a strategy to identify a subgroup of schizophrenia caused by immune mechanisms. Med. Hypotheses 2001, 56 (4), 483–486. (3) Huang, J. T.; Wang, L.; Prabakaran, S.; Wengenroth, M; Lockstone, H. E.; Koethe, D.; et al. Independent protein-profiling studies show a decrease in apolipoprotein A1 levels in schizophrenia CSF, brain and peripheral tissues. Mol. Psychiatry 2008, 13 (12), 1118–1128. (4) Flyckt, L.; Borg, J; Borg, K.; Ansved, T.; Edman, G.; Bjerkenstedt, L.; et al. Muscle biopsy, macro EMG, and clinical characteristics in patients with schizophrenia. Biol. Psychiatry 2000, 47 (11), 991– 999. (5) Mahadik SPaM, S. Cultured Skin Fibroblasts as a Cell Model for Investigating Schizophrenia. J. Psychiatry Res. 1996, 30 (6), 421– 439. (6) Mahadik, S. P.; Mukherjee, S.; Laev, H.; Reddy, R.; Schnur, D. B. Abnormal growth of skin fibroblasts from schizophrenic patients. Psychiatry Res. 1991, 37 (3), 309–320. (7) Mahadik, S. P.; Mukherjee, S.; Wakade, C. G.; Laev, H.; Reddy, R. R.; Schnur, D. B. Decreased adhesiveness and altered cellular distribution of fibronectin in fibroblasts from schizophrenic patients. Psychiatry Res. 1994, 53 (1), 87–97. (8) Catts, V. S.; Catts, S. V.; McGrath, J. J.; Feron, F.; McLean, D.; Coulson, E. J.; et al. Apoptosis and schizophrenia: a pilot study based on dermal fibroblast cell lines. Schizophr. Res. 2006, 84 (1), 20–28. (9) Matigian, N. A.; McCurdy, R. D.; Feron, F.; Perry, C.; Smith, H; Filippich, C.; et al. Fibroblast and lymphoblast gene expression profiles in schizophrenia: are non-neural cells informative? PLoS ONE 2008, 3 (6), e2412. (10) Jablensky, A. The concept of schizophrenia: pro et contra. Epidemiol. Psichiatry Soc. 1999, 8 (4), 242–247. (11) R Development Core Team: A language and environment for statistical computing; R Foundation for Statistical Computing: Vienna, Austria, 2007 (http://www.R-project.org). (12) Gentleman, R. C.; Carey, V. J.; Bates, D. M.; Bolstad, B.; Dettling, M.; Dudoit, S. et al. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004, 5 (R80). (13) Smyth, G. K.; Michaud, J.; Scott, H. S. Use of within-array replicate spots for assessing differential expression in microarray experiments. Bioinformatics 2005, 21 (9), 2067–2075.
research articles
Cell Cycle Abnormalities in Schizophrenia (14) Benjamini YH., Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. 1995, B 57, 289–300. (15) Li GZG, D. Gorenstein, M. V. Silva, J. C. Vissers, and J. P. C. a. G., S. J. A novel ion accounting algorithm for protein database searches Human Proteome Organisation (HUPO) 5th Annual World Congress: Long Beach, CA 2006. (16) Aravind, S. P. T.; Mootha, V. K.; Mukherjee, S.; Ebert, B. L.; Gillette, M. A.; Paulovich, A.; Pomeroy, S. L.; Golub, T. R.; Lander, E. S.; Mesirov, J. P. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (43), 15545–15550. (17) Levin, Y.; Wang, L.; Ingudomnukul, E.; Schwarz, E; Baron-Cohen, S.; Palotas, A.; et al. Real-time evaluation of experimental variation in large-scale LC-MS/MS-based quantitative proteomics of complex samples. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2008, Nov 13. (18) Vissers JPCL, J. I.; Aerts, J. Analysis and Quantification of Diagnostic Serum Markers and Protein Signatures for Gaucher Disease. Mol. Cell. Proteomics 2007, 6. (19) Schwarz, E.; Levin, Y.; Wang, L.; Leweke, F. M.; Bahn, S. Peptide correlation: a means to identify high quality quantitative information in large-scale proteomic studies. J. Sep. Sci. 2007, 30 (14), 2190–2197. (20) Mootha CML, V. K.; Eriksson, K.-F.; Subramanian, A.; Sihag, S.; Lehar, J.; Puigserver, P.; Carlsson, Emma; Ridderstråle, M.; Laurila, E.; Houstis, N.; Daly, M. J.; Patterson, N.; Mesirov, J. P.; Golub, T. R.; Tamayo, P.; Spiegelman, B.; Lander, E. S.; Hirschhorn, J. N.; Altshuler, D.; Groop, L. C. PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat. Genet. 2003, 34, 267–273. (21) McCurdy, R. D.; Feron, F.; Perry, C.; Chant, D. C.; McLean, D.; Matigian, N.; et al. Cell cycle alterations in biopsied olfactory neuroepithelium in schizophrenia and bipolar I disorder using cell culture and gene expression analyses. Schizophr Res 2006, 82 (23), 163–173.
(22) Gygi, S. P.; Rochon, Y.; Franza, B. R.; Aebersold, R. Correlation between protein and mRNA abundance in yeast. Mol. Cell. Biol. 1999, 19 (3), 1720–1730. (23) Katsel, P.; Davis, K. L.; Li, C.; Tan, W; Greenstein, E.; Kleiner Hoffman, L. B.; et al. Abnormal indices of cell cycle activity in schizophrenia and their potential association with oligodendrocytes. Neuropsychopharmacology 2008, 33 (12), 2993–3009. (24) Reif, A; Fritzen, S.; Finger, M.; Strobel, A.; Lauer, M; Schmitt, A.; et al. Neural stem cell proliferation is decreased in schizophrenia, but not in depression. Mol. Psychiatry 2006, 11 (5), 514–522. (25) Catts, V. S. C. S. V. Apoptosis and schizophrenia: is the tumour suppressor gene, p53, a candidate susceptibility gene? Schizophr. Res. 2000, 41 (3), 405-415. (26) Jarskog, L. F.; Glantz, L. A.; Gilmore, J. H.; Lieberman, J. A. Apoptotic mechanisms in the pathophysiology of schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry 2005, 29 (5), 846– 858. (27) Futamura, T.; Toyooka, K.; Iritani, S.; Niizato, K; Nakamura, R.; Tsuchiya, K.; et al. Abnormal expression of epidermal growth factor and its receptor in the forebrain and serum of schizophrenic patients. Mol. Psychiatry 2002, 7 (7), 673–682. (28) Xian, C. J.; Zhou, X. F. Roles of transforming growth factor-alpha and related molecules in the nervous system. Mol. Neurobiol 1999, 20 (2-3), 157–183, Oct-Dec. (29) Meyer, S. M. The metabolic syndrome and schizophrenia. Acta Psychiatry Scand. 2009, 119 (1), 4–14. (30) Ish-Shalom, D CC; Vorwerk, P.; Sacerdoti-Sierra, N.; Shymko, R. M.; Naor, D.; De Meyts, P. Mitogenic properties of insulin and insulin analogues mediated by the insulin receptor. Diabetologia 1997, 40 (Suppl 2), S25–S31. (31) Bushe, C.; Holt, R. Prevalence of diabetes and impaired glucose tolerance in patients with schizophrenia. Br. J. Psychiatry Suppl. 2004, 47, 67–71. (32) Ryan, M. C.; Collins, P.; Thakore, J. H. Impaired fasting glucose tolerance in first-episode, drug-naive patients with schizophrenia. Am. J. Psychiatry 2003, 160 (2), 284–289.
PR900867X
Journal of Proteome Research • Vol. 9, No. 1, 2010 527
