Chromatin Alterations Associated with Down-Regulated Metabolic Gene Expression in the Prefrontal Cortex of Subjects with Schizophrenia

Home , Translocase

ORIGINAL ARTICLE Chromatin Alterations Associated With Down-regulated Metabolic Gene Expression in the Prefrontal Cortex of Subjects With Schizophrenia

Schahram Akbarian, MD, PhD; Martin G. Ruehl, Dipl Ing; Erin Bliven, BS; Lori A. Luiz, BS; Amy C. Peranelli, BS; Stephen P. Baker, MScPH; Rosalinda C. Roberts, PhD; William E. Bunney, Jr, MD; Robert C. Conley, MD; Edward G. Jones, MD, PhD; Carol A. Tamminga, MD; Yin Guo, MD

Background: Schizophrenia is frequently accompanied cal analysis, custom-made complementary DNA arrays, by hypometabolism and altered gene expression in the pre- and quantitative real-time reverse transcriptase– frontal cortex. Cellular metabolism regulates chromatin polymerase chain reaction. structure, including covalent histone modifications, which are epigenetic regulators of gene expression. Results: Subjects with schizophrenia, as a group, showed no significant alterations in histone profiles or gene ex- Objective: To test the hypothesis that down-regulated pression. In a subgroup of 8 patients with schizophrenia, metabolic gene expression is associated with histone modi- levels of H3-(methyl)arginine 17, H3meR17, exceeded con- fication changes in the prefrontal cortex of subjects with trol values by 30%, and this was associated with the de- schizophrenia. creased expression of 4 metabolic transcripts.

Design and Subjects: Histones and gene transcripts Conclusions: High levels of H3-(methyl)arginine 17 are were profiled in the postmortem prefrontal cortex of 41 associated with down-regulated metabolic gene expres- subjects with schizophrenia and 41 matched controls. The sion in the prefrontal cortex of a subset of subjects with phosphorylation, acetylation, and methylation of 6 schizophrenia. Histone modifications may contribute to the lysine, serine, and arginine residues of histones H3 and pathogenesis of prefrontal dysfunction in schizophrenia. H4 were examined together with 16 metabolic gene transcripts using serial immunoblotting, immunohistochemi- Arch Gen Psychiatry. 2005;62:829-840

YSFUNCTION, HYPOACTIV- core histones, H2A, H2B, H3, and H4, ity, and hypometabolism together with 147 base pairs of genomic Author Affiliations: Brudnick Neuropsychiatric Research of the prefrontal DNA wrapped around them, compose Institute, Department of cortex (PFC) may con- the nucleosomes, the basic units of chro- Psychiatry (Drs Akbarian and tribute to the negative matin. Chromatin fibers are composed of Guo, Mr Ruehl, and Mss Bliven, symptoms and cognitive deficits of schizo- arrays of nucleosomes connected by D1-6 26,27 Luiz, and Peranelli) and phrenia. The molecular pathogenesis of linker histones and DNA. Dynamic Bioinformatics Unit, prefrontal dysfunction in schizophrenia is changes in chromatin conformation and Information Services still not clear, but the down-regulated ex- accessibility of transcription factors are (Mr Baker), University of pression of a subset of metabolic genes is highly regulated by the N-terminal his- Massachusetts Medical School, thought to be involved.7-9 Alterations in cel- tone tails.28,29 Covalent modifications at Worcester; Maryland Psychiatric Research Center, lular metabolism may then further histone N-terminal tails are differentially University of Maryland, compromise orderly gene expression, af- regulated in chromatin at sites of active 10-20 Baltimore (Drs Roberts and fecting neurotransmission, myelina- gene expression compared with inactive Conley); Department of tion,10,21,22 and other functions.23,24 There is and silenced chromatin.28,29 For example, Psychiatry and Human evidence that in individual schizophrenic the methylation of histone H3 at the Behavior, University of patients, PFC hypometabolism persists for arginine 17 position (H3meR17), or the California at Irvine decades,25 suggesting that the underlying phosphoacetylation of H3 at the serine (Dr Bunney); Center for molecular mechanisms, including dysregu- 10/lysine 14 position (H3pS10-acK14), Neuroscience, Department of lated metabolic gene expression, remain defines the open chromatin state and Psychiatry, University of 28-34 California at Davis (Dr Jones); stable for extended periods. actual or potential transcription. and Department of Psychiatry, In eukaryotes, the rate-limiting bio- Studies in rodents showed that treatment University of Texas chemical response that leads to the acti- with (1) antipsychotic drugs that block 35 Southwestern Medical Center, vation of gene expression involves alter- dopamine D2 -like receptors, (2) 26 36 Dallas (Dr Tamminga). ations in chromatin structure. The 4 D1-agonists, or (3) the anticonvulsant

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 829

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/02/2021 and mood-stabilizing drug valproate sodium37-39 edge, this is the first evidence that histone modifica- induces histone modification changes not only at tions as epigenetic regulators of gene expression may defined genomic sequences but also on a global and contribute to prefrontal dysfunction in schizophrenia. genome-wide level in selected areas of the forebrain. The drug-induced, coordinated regulation of histone METHODS chemistry in whole or “bulk” chromatin is thought to have profound effects on nuclear signaling and chroma- RATIONALE AND STUDY DESIGN tin function, including transcription and cellular differ- 28,29,40 entiation. The D2-like antagonist drugs and valpro- According to a recent study by Middleton and colleagues,7 a ate sodium are frequently prescribed for the treatment subset of 10 metabolic genes are expressed at decreased levels of schizophrenia and other major psychiatric in the PFC of subjects with schizophrenia. This set of 10 tran- diseases,41-43 which raises the question of whether chro- scripts includes malate dehydrogenase I (MDH), for which 4 matin modifications are involved in the molecular studies reported either a significant decrease7-9 or increase10 in mechanism of action of these drugs. Furthermore, in the PFC of subjects with schizophrenia. Therefore, we sought eukaryotes, cellular metabolism is tightly coupled to (1) to examine the expression of these metabolic genes in a larger chromatin structure because many chromatin- cohort of schizophrenic subjects and controls and (2) to de- remodeling complexes require adenosine triphos- termine whether altered metabolic gene expression is linked phatase activity.44,45 In addition, caloric restriction to an abnormal histone modification profile. This study was conducted in 2 parts (Figure 1). First, 6 different histone modi- up-regulates histone deacetylase activity, which in turn 46,47 fications and 16 metabolic gene transcripts were profiled in 21 leads to transcriptional silencing of chromatin. matched pairs of schizophrenic subjects and controls. For the Given this background, we hypothesized that hypome- overall cohort of 21 subjects with schizophrenia, we did not find tabolism and decreased metabolic gene expression in significant changes in histone or transcript levels. Therefore, as the PFC of subjects with schizophrenia is accompanied the next step, we defined subgroups of subjects with schizophre- by abnormal levels of 1 or several covalent histone nia based on histone modification levels. Then, the association modifications. These changes may reflect an adaptive between modifications and down-regulated metabolic gene ex- response to prefrontal hypoactivity, may play a caus- pression was further tested by conducting additional studies on ative role in prefrontal dysfunction, or both. In any another 20 pairs of subjects with schizophrenia and controls col- case, chromatin-related abnormalities are likely to have lected independently of the 21 matched pairs in part 1. profound and lasting implications for cellular function and PFC circuitry. Furthermore, molecular studies on POSTMORTEM TISSUE chromatin from diseased brain is pivotal to gain further insight into epigenetic factors that are thought to be The present study used the brains of 82 subjects, or 41 matched involved in the etiology of the schizophrenia.48-52 pairs of subjects with schizophrenia and controls (Table 1). All subjects were matched for age and autolysis time (±15%), Herein, we identify a subgroup of subjects with schizo- and 38 of 41 pairs were also matched for sex, as described pre- phrenia affected by down-regulated expression of 4 of viously.53 The entire matching process was finished before any 16 metabolic genes in conjunction with high levels of experimental procedures were performed. Of the 41 pairs, 21 histone H3 methylation (H3meR17). To our knowl- were obtained from a brain collection at the University of Cali- fornia and were used in previous studies12,53 of GAD67 gene expression and white matter neuron distribution. Key findings reported for this postmortem collection were independently rep- 21 Patients With Schizophrenia and I. Histone Profiling 15-17,54-58 21 Matched Controls (California II. Gene Expression Profiling licated. Therefore, this postmortem cohort seems to be Brain Collection) III. Subgrouping of Subjects With representative of the disease. Another 20 pairs were obtained Schizophrenia by Histone Profile from a brain bank at the University of Maryland that includes + a tissue collection from subjects diagnosed as having schizophrenia.55,59,60 For both brain banks, procedures for tissue col- 20 Subjects With Schizophrenia and IV. Confirm Histone–Gene Expression lection, neuropathologic examination (to rule out degenera- 20 Matched Controls (Maryland Association in Larger Cohort Brain Collection) tive and neurologic disease), diagnosing schizophrenia using DSM-IV–based diagnostic criteria, and selection of subjects with schizophrenia and controls were described in detail in previ- Figure 1. Study design. ous publications.12,53,59-61 Controls were defined as people with-

Table 1. Diagnostic Characteristics, Medication Status, and Postmortem Confounds of Postmortem Collection

Age, Sex, PMI, Hemisphere, Brain pH, Schizophrenia Antipsychotic Group Mean ± SEM, y F/M, No. Mean ± SEM, h R/L, No. Mean ± SEM Subtype, % Medication, % Suicide, % Subjects with 50.7 ± 2.8 13/28 15.1 ± 1.4 20/21 6.55 ± 0.06 CPS: 36 Typical: 73 12.2 schizophrenia CUS: 54 Atypical: 5 (n = 41) NOS: 9 Typical ϩ Atypical: 7.5 None: 14.5* Controls (n = 41) 51.1 ± 2.8 12/29 15.1 ± 1.2 18/23 6.52 ± 0.05 NA NA 0

Abbreviations: CPS, chronic paranoid schizophrenia; CUS, chronic undifferentiated schizophrenia; L, left; NA, not applicable; NOS, not otherwise specified; PMI, postmortem interval; R, right. *Two months before death.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 830

cohol abuse or substance abuse or dependence; 4 controls and Me 3 subjects with schizophrenia had positive toxicologic findings for alcohol, narcotics, or both. None of the cases in this H3 me R17 ARTKQTARKSTGGKAPRKQL H3 study were known to be individuals with human immunode- 12010 ficiency virus or hepatitis B or who were in comas or receiving P Ac artificial respiration before death. All tissue samples were fresh- H3 p S10- ac K14 ARTKQTARKSTGGKAPRKQL H3 frozen and kept at –75°C until further processing. All proce- 12010 dures were approved by the institutional review boards of the Ac Ac University of Massachusetts, the University of Maryland, and the University of California at Davis. From each brain, small H3 ac K9/14 ARTKQTARKSTGGKAPRKQL H3 (0.5- to 1.0-cm3) blocks of tissue were obtained from the ros- 12010 tral pole of the left and right frontal lobes (Brodmann area A10), Me which is part of the PFC and is affected by functional hypoac- ARTKQTARKSTGGKAPRKQL 1-6 H3 me K4 H3 tivity and hypometabolism. 12010

Ac HISTONE PROFILING H4 ac K8 SGRGKGGKGLGKGGAKRHRK H4 12010

Tissue samples of PFC were homogenized in 0.2M sulfuric acid Ac and incubated on wet ice in the same solution for 45 minutes at a concentration of approximately 2 mg DNA/mL. The samples H4 ac K12 SGRGKGGKGLGKGGAKRHRK H4 12010 were then pelleted by centrifugation, and a supernatant containing acid-soluble proteins was added to one-third volume of 100% trichloroacetic acid, mixed well to precipitate his- Figure 2. The 6 site-specific covalent modifications at the N-terminal tails of tones, pelleted, washed in 100% acetone/0.05M hydrochlo- histones H3 and H4 studied in the postmortem cohorts. Ac indicates acetylation; Me, methylation; P, phosphorylation; single-letter amino acid ride, resuspended in double-distilled H20, and stored at –75°C codes: K, lysine; S, serine. H3meR17 is histone H3 methylated at arginine until use. Sample concentrations were determined in tripli- 17; H3pS10-acK14 is a dimodified histone, defined by phosphorylation of cate using a protein assay (Pierce Biotechnology Inc, Rock- serine 10 and acetylation of lysine 14. H3acK9/14 is defined by the ford, Ill) in conjunction with a microtiter plate reader (Bench- acetylation of lysines 9 and 14 on histone H3. H3meK4 is defined by the mark; Bio-Rad Laboratories, Hercules, Calif), adjusted methylation of H3-lysine 4. H4acK8 is defined by the acetylation of H4-lysine accordingly and eluted in 1ϫ Laemmli buffer to a concentra- 8. H4acK12 is defined by the acetylation of H4-lysine 12. tion of 10 µg/µL. In addition, equal loading was controlled by gel Coomassie blue stain. Samples were analyzed using sodium dodecyl sulfate–polyacrylamide gel electrophoresis and cation (Upstate Biotechnology) to the primary antibody immunoblotting, using polyvinyl difluoride membranes (Bio- incubation solution. Immunoreactivity was selectively abol- Rad Laboratories) with a 0.2-µm pore size to ensure efficient ished on immunoblots and in tissue sections by peptides blotting of the histones, which have molecular weights in the (final concentration, 1.5 µg/mL) carrying the epitope recog- range of 10 to 15 kDa. For immunoblotting, membranes were nized by the site- and modification-specific antibody. preblocked in Tris-buffered saline, with 0.1% Tween-20 (TBS-T) (pH, 7.5), with 5% dry milk (Bio-Rad Laboratories), incubated overnight in primary antibody in TBS-T, washed repeat- GENE EXPRESSION STUDIES edly, and incubated for 1 hour in peroxidase-conjugated sec- ondary goat anti–rabbit antibody (1:200) (Amersham Total RNA from approximately 100 to 200 mg of PFC tissue Biosciences, Piscataway, NJ), and immunoreactive bands were samples was prepared using the Trizol Reagent (Invitrogen, then visualized by chemiluminescence using a charged- Carlsbad, Calif), and 10 µg was used as the template for biotin- coupled device camera–based system (ChemiDoc; Bio-Rad Labo- labeled complementary DNA (cDNA) probe synthesis by an- ratories) and film autoradiograms. All data were normalized to nealing to GEAprimer (SuperArray Biosciences Corporation, Frederick, Md) at 70°C and then incubating for 10 minutes at input (optical density per micrograms of protein). The follow- Ј Ј ing primary antibodies were used (working dilution): (1) anti- 42°C with biotin-16-2 -deoxy-uridine-5 -triphosphate (GIBCO, H3-[(phospho)Ser10,(acetyl)Lys14]=H3pS10-acK14 (1: Invitrogen Corporation, Grand Island, NY) and Moloney mu- 500); (2) anti-H3-(acetyl)Lys9-Lys14=H3acK9/14 (1:1500); (3) rine leukemia reverse transcriptase (RT) (Promega, Madison, anti-H3-(methyl)Lys4=H3meK4 (1:500); (4) anti-H3-(methyl) Wis). This procedure omits polymerase chain reaction (PCR)– Arg17=H3meR17 (1:150); (5) anti-H4-(acetyl)Lys8=H4acK8 based amplification of the cDNA template, which could result (1:6000); and (6) anti-H4-(acetyl)Lys12=H4acK12 (1:1000). in skewing and distortion of differences between samples. Cus- All primary antibodies were polyclonal (rabbit) and were pur- tom-made GEArray membranes (SuperArray Biosciences Cor- chased from Upstate Biotechnology Inc (Waltham, Mass). The poration) were placed in prehybridization buffer (GEAhyb) with site-specific histone modifications are shown in Figure 2. salmon sperm DNA, 100 µg/mL, then labeled probe was added for overnight hybridization at 68°C. The membranes were washed in sodium chloride/sodium citrate (SSC)/1% sodium IMMUNOHISTOCHEMICAL ANALYSIS dodecyl sulfate with increasing stringency (2ϫ–0.1ϫSSC) at 68°C and then were further processed for chemilumines- Procedures for tissue fixation and immunohistochemical analy- cence with alkaline phosphatase–conjugated streptavidin sis on free-floating sections have been described previously.53 and then incubated with chemiluminescence substrate The final working dilution of primary antibodies ranged from (CDP-Star; PerkinElmer, Boston, Mass). Chemiluminescence 1:250 (anti-H3pS10-acK14 and anti-H3meR17 antibodies) to was detected by using electrochemiluminescence autoradi- 1:1000 (anti-H4acK12 antibody). We tested antibody specific- ography film (Amersham), and hybridization signals were ity by adding synthetic peptides containing the first 21 resi- quantified by densitometry using Quantity One software dues of histone H3 or H4 together with a site-specific modifi- (Bio-Rad Laboratories).

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 831

Gene GenBank cDNA Fragment Gene Abbreviation Accession No. (5؅–3؅ bp) Metabolic Pathway ATP synthase† ATP5A1 NM004046 61-500 Mitochondrial electron transport Ca2ϩ/calmodulin-dependent protein kinase II CAMIIK NM015981 227-624 NA Cytochrome c oxidase subunit VIIa polypeptide 2-l COX7RP AB007618 32-399 Mitochondrial electron transport Crystallin, µ† CRYM NM001888 281-581 Ornithine-polyamine Cytochrome c CYTOC/CYC1 X06994 351-774 Mitochondrial electron transport Glutamic-oxaloacetic transaminase 2† GOT2 NM002080 1348-1757 Malate shuttle; aspartate-alanine (aspartate aminotransferase) Hypoxanthine-guanine phosphoribosyltransferase HPRT1 M31642 203-669 Purines Malate dehydrogenase I (cytosolic)‡ MDH NM005917 575-1027 Malate shuttle, TCA cycle Ornithine aminotransferase† OAT NM000274 217-645 Ornithine-polyamine Ornithine decarboxylase antizyme OAZ1 NM004152 324-739 Ornithine-polyamine Ornithine decarboxylase antizyme inhibitor† OAZIN NM015878 1233-1663 Ornithine-polyamine Ornithine decarboxylase ODC M16650 1112-1486 Ornithine-polyamine 3-Oxoacid coenzyme A transferase† OXCT NM000436 170-608 Keton body Pyruvate dehydrogenase kinase, isoenzyme 2 PDK2 NM002611 781-1174 Carbohydrate Ribosomal protein L13a RPL13A NM012423 173-428 NA Translocase of inner mitochondrial membrane 17, homologue A† TIM17A NM006335 628-1065 Mitochondrial translocase Ubiquitin carboxy-terminal esterase L1 (ubiquitin thiolesterase)† UCHL1 NM004181 37-365 Ubiquitin Ubiquitin-specific protease 14† USP14 NM005151 929-1385 Ubiquitin ␤-Actin BACTIN X00351 402-887 NA Glyceraldehyde-3-phosphate dehydrogenase GAPDH M33197 71-748 Carbohydrate pUC 18 (plasmid DNA as negative control) pUC18 NA NA NA

Abbreviations: ATP, adenosine triphosphate; bp, base pair; NA, not applicable; TCA, tricarboxylic acid. *The 20 cDNAs and the control DNA spotted in duplicate on the nylon membrane–based array. For each cDNA, the length and position of the cDNA fragment used for spotting is provided. †Transcripts were expressed at decreased levels in the prefrontal cortex of another cohort of subjects with schizophrenia.7 ‡Transcripts were expressed at decreased levels in the prefrontal cortex of other cohorts of subjects with schizophrenia.7-9

cDNA ARRAYS Expression of each gene was calculated to correct for potential differences in RNA input and PCR primer efficiency us- The custom-made GEArray membranes contained 20 cDNAs ing the following equation62,63: spotted in duplicate and a plasmid (pUC18) as negative con- V=(Ereference)CT-reference/(Etarget)CT-target trol (Table 2). The array included cDNAs of 17 metabolic genes, including the 10 metabolic genes that in a recent study7 where V indicates the relative value of the target transcript nor- were decreased in the PFC of patients with schizophrenia. malized to reference transcript (␤-actin); E, primer slopes; and Our array included 3 control cDNAs: (1) ␤-actin for internal CT, the threshold crossing cycle. normalization; (2) the ␣-subunit of calcium/calmodulin- dependent protein kinase II (CAMIIK)12,61; and (3) the “housekeeping” ribosomal protein (RPL13A). Because the ANIMAL STUDIES hybridization signal for 1 metabolic gene, ATP5A1, wasinall cases and controls less than 2ϫbackground, it was not con- Adult 6- to 8-week-old outbred mice from a mixed genetic back- sidered for further analysis. Thus, the present study reports ground, predominantly Sv129 /J and C57BL/6J, were treated for data for 16 metabolic genes. 14 days with (1) the conventional antipsychotic drug haloperidol, (2) the atypical drug risperidone, or (3) saline as a control. For each treatment, 6 animals were selected, and sex- QUANTITATIVE PCR matched littermates were matched to different treatment arms. The male-female ratio was 1:1 for each treatment group. Halo- Total RNA (0.5 µg) was used for cDNA synthesis and real- peridol and risperidone were administered intraperitoneally with time RT-PCR using the Platinum Quantitative RT-PCR Ther- saline as vehicle (injection volume, 2 mL/kg body weight) at a moScript One-Step System (Invitrogen), an Opticon 2 cycler, dose of 1 mg/kg twice daily at 8 AM and 4 PM. Two hours after and Opticon software (MJ Research, Waltham), together with the last treatment, the animals were killed and the brain was FAM dye-labeled TaqMan MGB probes specific for removed, and the medial rostral cortex was dissected and fro- coactivator-associated arginine methyltransferase 1 (CARM1), zen until homogenized and further processed for histone im- crystallin (CRYM), cytochrome c1 (CYTOC/CYC1), malate munoblotting as described previously herein. dehydrogenase 1 (MDH), ornithine aminotransferase (OAT), and peptidyl arginine deiminase (PADI4/PAD4) and for ␤-actin for normalization (Taqman Assays-on-Demand Gene STATISTICAL ANALYSIS Expression Products; Applied Biosystems, Foster City, Calif) and according to the manufacturer’s instructions using 6mM Associations between histone modification or transcript levels and magnesium sulfate and the following cycling protocol: 50°C age, brain pH, and postmortem interval were evaluated for 50 minutesϫ1, 95°C for 3 minutesϫ1, and 95°C for 15 using the Pearson correlation coefficient. All histone and gene ex- seconds followed by 60°C for 60 secondsϫ45 cycles. Data pression data from each subject with schizophrenia are ex- were expressed relative to custom-made standards. pressed relative to the matched control to test for significant

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 832

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/02/2021 changes in gene expression and histone modifications in the total RESULTS cohort or in subgroups of subjects with schizophrenia. The matched-pair design was necessary because the study was conducted for 3 years in a large number of postmortem samples HISTONE PROFILING (N=82), with data being collected from many different immunoblots, cDNA array membranes, and real-time RT-PCR experi- We profiled histone modifications in the PFC using site- ments. For each experiment, the sample from a subject with schizo- and modification-specific antibodies against the tail se- phrenia was processed in parallel with the sample from the matched quence of 2 core histones, H3 and H4. We focused on control. Subgroups of schizophrenic subjects were defined by dif- the 6 antibodies that fulfilled the criterion of consistent ferences in histone modification levels compared with controls. and robust immunoreactivity in our postmortem mate- Because nothing is known about the regulation of histone modifications in human brain, our subgrouping rationale was guided rial: (1) an antibody against dimodified H3 molecules, by studies64-71 that used histone acetylation levels to subgroup cer- defined by the phosphorylation of serine 10 and acety- tain types of carcinomas and adenomas. The significance of dif- lation of lysine 14, H3pS10-acK14; (2) an antibody that ferences in metabolic gene transcripts in defined subgroups of sub- binds to H3 acetylated at lysine 9 or lysine 14, H3acK9/ jects with schizophrenia were examined using a nonparametric 14; (3) an antibody against H3 methylated at lysine 4, Mann-Whitney test and an exact permutation distribution.72 The H3meK4; (4) an antibody against H3 methylated at ar- significance of case-control differences was evaluated using ginine 17, H3meR17; (5) an antibody that recognizes his- 1-sample t tests of the hypothesis that the mean difference was tone H4 acetylated at lysine 8, H4acK8; and (6) an an- zero. Analyses were conducted using a statistical software pro- 73,74 tibody against H4 acetylated at lysine 12, H4acK12. The gram (StatXact version 6). 6 histone modifications are illustrated in Figure 2, and representative examples of immunoblots are shown in Figure 3. To confirm that these antihistone antibodies specifically label nuclei, we examined the cellular and C8 S8 C4 S4 C6 S6 laminar distribution pattern of immunoreactivities us- H3meR17 ing histochemical analysis. Each of the 6 antihistone an- H4acK12 tibodies resulted in robust immunoreactivity of nuclei H3meK4 in the PFC (Figure 4). Immunoreactivity for H3meR17

H4acK8 was primarily confined to cortical gray matter. At higher magnification power, intense H3meR17 labeling was ap- H3pS10-acK14 parent in large, presumably neuronal nuclei (Figure 4E

H3ack9/14 and F). In contrast, robust immunoreactivity for other H4acK12 types of histone modifications, including H4acK12, was expressed in large and small, presumably nonneuronal nuclei (Figure 4I and J). Figure 3. Histone profile of the human prefrontal cortex. Representative Levels for each of the 6 histone modifications, includ- examples of histone immunoblots for matched schizophrenia (S) and control ing H3meR17, differed less than 15% between subjects (C) pairs 4, 6, and 8. Note the robust immunoreactivity for each of the 6 with schizophrenia and controls, and these changes were histone modifications. Note the increased levels of H3meR17 in S8 compared with C8. The approximate sizes of the immunoreactive bands are not statistically significant (Figure 5A). We conclude as follows: H3, 14.5 kDa; and H4, 10.5 kDa. that subjects with schizophrenia, as a group, do not show

I Nissl I H3meR17 Nissl H3meR17 H3pS10-acK14 H4acK12 No Primary Antibody II II

III III

C E G I K IV IV

V V

VI VI

WM WM A B D F H J L

Figure 4. Laminar and cellular distribution of histone immunoreactivity in the postmortem prefrontal cortex. A and B, Sections through the full thickness of the prefrontal cortex stained for Nissl (A) and H3meR17 (B) immunoreactivity. Notice in (B) the punctate staining pattern in cortical layers I through VI but the very weak labeling in white matter (WM). C-L, Digitized images from prefrontal layer III of control 4 (C, E, G, I, and K) and matched subject with schizophrenia 4 (D, F, H, J, and L) stained for Nissl (C and D), H3meR17 (E and F), H3pS10-acK14 (G and H), and H4acK12 (I and J) immunoreactivity. Two sections (K and L) were processed without primary antibody. Notice the robust histone immunoreactivity in the nuclei. The images were taken at the following magnifications: 2.5ϫ10 (A and B) and 40ϫ10 (C-L).

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 833

20 Relative Optical Density 10

0 H3pS10-acK14 H3acK9/14 H4acK8 H4acK12 H3medK4 H3meR17 Histone Modifications B 0

–0.5

–1.0

Relative Optical Density –1.5

–2.0 CRYM COX7RP CYTOC GAPDH GOT2 HPRT MDH PDK2 OAT OAZ1 OAZIN ODC OXCI TIM17A UCHL1 USP14 CAMIIK RPL13A Metabolic Genes

Figure 5. Mean levels of histone immunoreactivity (A) and metabolic gene transcripts (B) in 21 subjects with schizophrenia and 21 matched controls. Metabolic gene transcripts (B) are shown after internal normalization to ␤-actin and log-transformation. There are no statistically significant differences between the 2 cohorts. Error bars indicate standard error of the mean.

First, we wanted to find out whether the reported alterations in metabolic gene expression7-10 are representa- PDK2 tive of a larger postmortem cohort. Second, we wanted to examine whether changes in metabolic gene expres- UCHL1 sion are associated with histone modification changes. Examples of hybridization signals on the arrays are GAPDH shown in Figure 6. When the total group of 21 subjects with schizophrenia was compared with the 21

CRYM controls, none of the metabolic and control genes showed a statistically significant difference between groups (Figure 5B). We conclude that in larger cohorts, HPRT the PFC of subjects with schizophrenia is not consistently affected by down-regulated metabolic gene OAZ1 expression.

Figure 6. Metabolic gene transcripts in the prefrontal cortex. Digitized HIGH LEVELS OF H3meR17 ARE ASSOCIATED images from film autoradiograms of complementary DNA (cDNA) arrays. WITH DECREASED EXPRESSION Representative examples of hybridization signals for 6 genes on the cDNA OF 4 METABOLIC GENES arrays. Note the intense signal for the gene expressed at high levels (GAPDH) and the lower-intensity signals for the remaining metabolic cDNAs (CRYM, HPRT, OAZ1, PDK2, and UCHL1) (magnification ϫ1). Although little is known about the regulation of histone modifications in human brain, alterations in histone modi- consistent alterations in global levels of methylated, acety- fication levels, including H3 and H4 acetylation, were used lated, and phosphoacetylated histones in the PFC. to define subgroups for certain diseases, including adenomas and carcinomas.64-71 In analogy to these ex- GENE EXPRESSION STUDIES amples from clinical pathology, we conducted subgroup analyses for each histone modification separately. We measured, in the postmortem cohort, transcript lev- We dichotomized the cohort of subjects with schizo- els of metabolic genes that were thought to be involved phrenia into subgroups based on modified histone lev- in PFC dysfunction (Table 2). Our rationale was 2-fold. els of 1.3 or greater relative to matched controls. After

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 834

0.5 0.25 log log ∆ ∆

0 0 Schizophrenia-Control Schizophrenia-Control Metabolic Gene Transcript, Metabolic Gene Transcript, –0.5 Metabolic Gene Transcript, –0.25

–1.0 –0.5 0 0.51.0 1.5 2.0 2.5 CRYM CYTOC MDH OAT H3meR17, Schizophrenia/Control

Figure 7. Decreased metabolic gene expression in the prefrontal cortex of subjects with schizophrenia is associated with high levels of H3meR17. A, For each of the 41 matched pairs, relative levels of H3meR17 and levels of metabolic gene transcripts. Dotted vertical line indicates an H3meR17 level of 1.3, the cutoff point for subjects with schizophrenia with H3meR17 levels of 1.3 or greater relative to matched controls (Schiz-H3meR17Ն1.3C). The dashed horizontal line indicates zero difference in gene expression between subjects with schizophrenia and controls. Note that the Schiz-H3meR17Ն1.3C subgroup (n=8) lacks subjects with high levels of gene expression. B, Mean levels of metabolic gene transcripts in subjects with schizophrenia relative to matched controls. Subjects with schizophrenia (n=41) were divided into the Schiz-H3meR17Ն1.3C subgroup (n=8) and a subgroup with H3meR17 levels of less than 1.3 relative to matched controls (Schiz-H3meR17Ͻ1.3C) (n=33). The decrease in CRYM, CYTOC/CYC1, MDH, and OAT levels in the Schiz-H3meR17Ն1.3C subgroup compared with matched controls is significant (P=.04). Error bars represent standard error of the mean.

the first part of the study, which was conducted on 21 gene expression (Figure 7A). Each case was scored from matched pairs, we identified 6 subjects with schizophre- –4 to 0, with –4 defining cases that showed a decrease in nia who showed H3meR17 levels of 1.3 or greater rela- 4 of 4 transcripts relative to matched controls. The Schiz- tive to their matched controls (Schizo-H3meR17Ն1.3C). H3meR17Ն1.3C subgroup (n=8) scored significantly lower This subgroup of subjects with schizophrenia (n=6) compared with the Schiz-H3meR17Ͻ1.3C subgroup (n=33) showed a significant decrease in the expression of 3 meta- (mean±SEM, –3.63±0.26 vs –1.7±0.25; Mann- bolic transcripts, CRYM, CYTOC/CYC1, and MDH (t test, Whitney z=–3.24; PϽ.005, 2-tailed). We conclude that P=.03), and a tendency for a decrease in OAT tran- expression of a set of 4 metabolic genes (CRYM, scripts (PϽ.1) compared with the remaining 15 pa- CYTOC/CYC1, MDH, and OAT) is significantly de- tients with schizophrenia. Next, we sought to confirm creased in the Schiz-H3meR17Ն1.3C subgroup. The 4 tran- the association between H3 methylation and down- scripts (CRYM, MDH, CYTOC/CYC1, and OAT) were regulated gene expression for CRYM, CYTOC/CYC1, MDH, strongly correlated (r=0.55-0.83; PϽ.005-.0001), which and OAT transcripts in the PFC of subjects with schizo- indicates that they are not independently regulated. To phrenia. To this end, we studied an additional set of 40 examine whether the 4 metabolic transcripts are signifi- samples consisting of 20 matched pairs of subjects with cantly decreased in the Schiz-H3meR17Ն1.3C subgroup, schizophrenia and controls. Immunoreactivity for we calculated, for each subject and matched control, the H3meR17 was again measured using immunoblotting. mean difference in expression levels for CRYM, MDH, The messenger RNA levels for each of the 4 genes were CYTOC/CYC1, and OAT. The Schiz-H3meR17Ն1.3C sub- measured using real-time RT-PCR and FAM dye- group (n=8) showed a significant deficit in expression labeled TaqMan MGB probes. Of these additional 20 sub- of the 4 transcripts compared with matched controls (⌬ jects with schizophrenia, 2 showed H3meR17 levels of log mean[Schiz-H3meR17Ն1.3C–Control]±SEM, 1.3 or greater relative to their matched controls. Thus, 8 –0.25±0.09; t=2.61; P=.04, 2-tailed t test) (Figure 7B). (20%) of 41 subjects with schizophrenia showed levels In contrast, the remaining 33 schizophrenic subjects of H3meR17 of 1.3 or greater relative to their matched showed no significant deficits in expression of the 4 tran- controls. We first compared gene expression differ- scripts compared with matched controls (⌬ log mean- ences between the Schiz-H3meR17Ն1.3C subgroup (n=8) [Schiz-H3meR17Ͻ1.3C–Control]±SEM, 0.02±0.06; t=0.36; and the remaining 33 subjects with schizophrenia P=.7) (Figure 7B). We conclude that the set of 4 meta- (Figure 7A). For each of the 41 subjects with schizo- bolic transcripts is significantly decreased in the Schiz- phrenia, levels of metabolic transcript were normalized H3meR17Ն1.3C subgroup of subjects with schizophrenia to those of matched controls. The 8 subjects in the Schiz- compared with other subjects with schizophrenia and H3meR17Ն1.3C subgroup, but not the remaining sub- matched controls. jects with schizophrenia (the Schiz-H3meR17Ͻ1.3C sub- To further test the association between H3meR17 and group), were consistently affected by decreased metabolic decreased metabolic gene expression, we dichotomized

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 835

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/02/2021 Table 3. Diagnostic and Postmortem Data for the Schiz-H3meR17Ն1.3C Subgroup and Comparison Groups

Subjects, Age, Sex, PMI, Hemisphere, Brain pH, Schizophrenia Antipsychotic Group No. Mean ± SEM y F/M, No. Mean ± SEM, h R/L, No. Mean ± SEM Subtype, % Medication, % Suicide, % Schiz-H3meR17Ն1.3C 8 51.3 ± 6.1 2/6 12.3 ± 3.3 2/6 6.49 ± 0.15 CPS: 28 Typical: 75 12.5 CUS: 71 Atypical: 12.5 None: 12.5 Schiz-H3meR17Ͻ1.3C 33 51.6 ± 3.3 11/22 16.0 ± 1.5 18/15 6.55 ± 0.06 CPS; 37.5 Typical: 65 12.1 CUS: 50 Atypical: 3 NOS: 12.5 Typical ϩAtypical: 9 None: 15 Controls 1 8 49.6 ± 6.6 2/6 13.0 ± 2.8 1/7 6.44 ± 0.11 NA NA 0 Controls 2 33 51.5 ± 3.1 10/23 15.7 ± 1.4 17/16 6.53 ± 0.06 NA NA 0

Abbreviations: CPS, chronic paranoid schizophrenia; CUS, chronic undifferentiated schizophrenia; NA, not applicable; NOS, not otherwise specified; PMI, postmortem interval; Schiz-H3meR17Ն1.3C, subjects with schizophrenia with levels of H3meR17 of 1.3 or greater relative to their matched controls; Schiz-H3meR17Ͻ1.3C, subjects with schizophrenia with levels of H3meR17 of less than 1.3 relative to their matched controls.

the entire group of subjects with schizophrenia (N=41) phrenia, 8 (20%) showed H3pS10-acK14 levels greater into subgroups defined by modified histone levels of 0.7 than 1.3 relative to controls. Of these 8 patients in the or less and greater than 0.7 compared with controls. We Schiz-H3pS10-acK14Ն1.3C subgroup, 2 were also part of found no significant differences in metabolic tran- the Schiz-H3meR17Ն1.3C subgroup. scripts between subgroups or relative to matched con- Compared with the 8 matched controls, the trols. Furthermore, levels of H3meR17 and metabolic gene Schiz-H3pS10-acK14Ն1.3C subgroup (n=8) showed a transcripts showed no correlation (R2 =0.007-0.11) significant deficit in OAT expression (⌬ log mean[Schiz- (Figure 7A). From these experiments, we conclude that H3pS10-acK14Ն1.3C–Control]±SEM: –0.39±0.12; significant deficits in metabolic gene transcripts in the t =–3.21; P =.02, 2-tailed t test). However, schizo- PFC are limited to a subgroup of subjects with schizophrenic subjects with H3pS10-acK14 levels of less than phrenia defined by high levels of H3meR17. 1.3 (n=33) showed no significant differences compared Regulation of H3R17 methylation includes coactivator- with matched controls (⌬ log=0.00±0.05). Further- associated arginine methyltransferase 1 (CARM1)75 and more, we dichotomized the entire group of schizo- peptidylarginine deiminase 4 (PADI4 or PAD4), which phrenic subjects (n=41) into subgroups defined by modi- converts methyl-Arg to citrulline, releasing methyl- fied histone levels of 0.7 or less and greater than 0.7 amine.76,77 We measured CARM1 and PAD4 transcript relative to controls, and no statistically significant dif- levels using quantitative, real-time RT-PCR in the Schiz- ferences were observed. Furthermore, we did not find a H3meR17Ն1.3C subgroup and matched controls and did statistically significant association between the H3acK9/ not observe significant differences (⌬ log mean [Schiz- 14, H3meK4, H4acK8, and H4acK12 histone modifica- H3meR17Ͻ1.3C–Control]±SEM; CARM1: 0.03±0.11; tions and metabolic gene expression. We conclude that PAD4: 0.14±0.12). In all case and control brains, rela- significant deficits in multiple metabolic gene tran- tive levels of CARM1 transcript were much higher than scripts in the PFC of subjects with schizophrenia are lim- those of PAD4 (CARM1/␤-actin: 0.21±0.06; PAD4/␤-actin: ited to a subgroup of cases defined by high levels (Ն1.3 0.01±0.013). We conclude that CARM1, which methyl- relative to controls) of H3 methylation at arginine 17. ates H3 at arginine 17,75 is expressed at robust levels in the human PFC. CONFOUNDING VARIABLES

HIGH LEVELS OF H3pS10-ACK14 ARE The alterations in 4 of 16 metabolic transcripts in the ASSOCIATED WITH DECREASED EXPRESSION Schiz-H3meR17Ն1.3C subgroup were not accompanied by OF 1 METABOLIC GENE an overall change in messenger RNA levels because levels for 12 of 16 metabolic transcripts were not statisti- After the first part of the study, we identified 6 (29%) of cally significantly altered in these subgroups; further- 21 of the subjects with schizophrenias who showed more, levels of 2 nonmetabolic transcripts, CAMIIK and H3pS10-acK14 levels of 1.3 or greater relative to matched RPL13A, differed less than 15% between schizophrenia controls. The H3pS10-acK14 Ն1.3C subgroup of subjects subgroups or compared with matched controls. with schizophrenia (n=6) showed a significant de- Owing to the matching process, the differences in age, crease in expression of the OAT transcript compared with postmortem interval, and female-male ratio between the remaining 15 subjects with schizophrenia (P=.02). schizophrenia cases and controls were less than 10%, and To confirm the association between H3 phosphoacety- the differences in pH between the 2 cohorts was less than lation, H3pS10-acK14, and down-regulated OAT gene ex- 1% (Table 1). The 8 subjects who composed the Schiz- pression, we measured in the additional set of 20 matched H3meR17Ն1.3C subgroup did not show significant differ- pairs H3pS10-acK14 immunoreactivity and OAT cDNA ences in age, sex, postmortem interval, and tissue pH com- levels. Of these additional 20 patients with schizophre- pared with their matched controls and other groups nia, 2 showed H3pS10-acK14 levels of 1.3 or greater rela- (Table 3). The schizophrenia subgroups did not show tive to matched controls. Thus, of 41 patients with schizo- significant differences in ratios of medicated and un-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 836

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/02/2021 Haloperidol Risperidone Saline A B Haloperidol-Treated Animals H3pS10-acK14 Risperidone-Treated 2 All Subjects With H4acK12 Animals Schizophrenia

1 P Ac Me H3meR17 (NH2)ARTKQTARKSTGGKAPRKQL H3 H4acK12 Subgroup

Optical Density, Mean Optical Density, 0 H3pS10-acK14 H3meR17

Figure 8. Long-term antipsychotic drug treatment does not affect H3 methylation and phosphoacetylation in whole chromatin of the rodent Subjects With Decreased Metabolic prefrontal cortex. A, Western blot from acid-extracted proteins from the Gene Expression rostral medial cortex of mice treated for 14 days with haloperidol or Subjects With Normal or Elevated risperidone. Immunoreactivity for H3pS10-acK14, H3meR17, and H4acK12 Metabolic Gene Expression in drug-treated animals is comparable to that in their saline-treated littermates. B, Quantitative results for H3pS10-acK14 and H3meR17 immunoreactivity in the rostral medial cortex. Levels in haloperidol- and Figure 9. Subgroup model for prefrontal cortex (PFC) hypometabolism in risperidone-treated animals were normalized to the levels in their schizophrenia. Of all subjects diagnosed as having schizophrenia, only a saline-treated littermates. Note that differences between drug- and portion show down-regulated metabolic gene expression in the PFC. This saline-treated animals are less than 15%. Error bars represent standard error hypothesis is indirectly supported by findings from in vivo imaging of the mean. studies1-6,80 because the severity of prefrontal hypoactivity and hypometabolism shows considerable variability among subjects diagnosed as having schizophrenia. Among schizophrenic subjects with PFC medicated patients, schizophrenia subtypes, or fre- hypometabolism/down-regulated metabolic gene expression, there is a subgroup that shows high levels of open chromatin-associated H3 quency of suicide. None of the 8 cases in the Schiz- methylation, H3meR17. Alterations in other histone modifications, including Ն H3meR17 1.3C subgroup and none of their matched H3 phosphoacetylation, H3pS10-acK14, may also be associated with controls were diagnosed as having alcohol or substance prefrontal hypometabolism and dysregulated metabolic gene transcription. Ac indicates acetylation; H3, histone H3; K, lysine; Me, methylation; abuse or dependence, and none of these 16 subjects had P, phosphorylation; R, arginine; S, serine. positive results on toxicologic screening.

Furthermore, levels of H3meR17, the histone marker 78,79 on which the subgrouping of the schizophrenic cohort closely related to the primate PFC, levels of methyl- is based, were not significantly associated with tissue pH, ated H3, H3meR17, and phosphoacetylated H3, autolysis time, or age. Tissue pH was significantly cor- H3pS10-acK14, remained unchanged in haloperidol- and related with 2 histone modifications (H3acK9/14 and risperidone-treated animals compared with saline- H4acK12; r=0.54 and 0.52, respectively; PϽ.01) but not treated controls (Figure 8). We conclude that antipsy- with gene transcripts. There was a correlation between chotic drug treatment does not alter modified histone lev- age at the time of death and glyceraldehyde-3- els in whole chromatin from human and rodent cerebral phosphate dehydrogenase messenger RNA levels cortices. (r=–0.39; P=.01) but no statistically significant corre- lations between age and other gene transcripts or his- COMMENT tones. Autolysis time showed a weak correlation with levels of pyruvate dehydrogenase kinase, isoenzyme 2 (PDK2) We show that down-regulated metabolic gene expres- transcript (r=–0.26; PϽ.1). Taken together, these re- sion in the PFC is not a consistent feature of subjects di- sults indicate that the down-regulation of CRYM, agnosed as having schizophrenia. A subgroup of subjects CYTOC/CYC1, MDH, and OAT transcripts in the Schiz- with schizophrenia shows decreased expression of mul- H3meR17Ն1.3C subgroup is not attributable to confound- tiple metabolic transcripts in conjunction with high lev- ing postmortem factors, clinical variables, or medica- els of open chromatin-associated histone H3 methyl- tion status. ation, H3meR17 (Figure 9). These alterations were not explained by clinical characteristics, medication, or post- ANTIPSYCHOTIC DRUG TREATMENT mortem factors. In schizophrenic subjects with high levels of H3meR17 in prefrontal chromatin, multiple meta- Li et al35 recently showed that short-term, but not long- bolic pathways, including ornithine-polyamine metabolism term, treatment with D2-like antagonists selectively in- (CRYM and OAT), mitochondrial electron transport duces H3 phosphoacetylation in whole chromatin from (CYTOC/CYC1), and the tricarboxylic acid cycle (MDH) striatal extracts. However, levels of H3 methylation, in- were affected. Molecular alterations in these metabolic path- cluding H3meR17, remained unchanged in the striatum ways were recently reported for other cohorts with schizo- of drug-treated animals.35 The results of the present post- phrenia7-9 and subjects with bipolar disorder.81 mortem study suggest that histone modifications in whole The methylation of histone H3 at arginine 17 is asso- PFC chromatin are not affected by drug treatment. To ciated with the open chromatin state and transcrip- further confirm this hypothesis, we treated adult mice tional activation.28,29,31 Therefore, dysregulation of for 14 days with (1) the conventional antipsychotic drug H3meR17 in prefrontal chromatin, as observed in a sub- haloperidol, (2) the atypical drug risperidone, or (3) sa- group of subjects with schizophrenia, could affect wide- line. In the medial rostral cortex, which in rodents is spread chromosomal regions. Presently, it remains un-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 837

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/02/2021 clear whether this increase in H3meR17 reflects a O’Leary DS. Hypofrontality in neuroleptic-naive patients and in patients with chronic compensatory mechanism to overcome the deficit in ex- schizophrenia: assessment with xenon 133 single-photon emission computed tomography and the Tower of London. Arch Gen Psychiatry. 1992;49: pression of metabolic and other gene transcripts. 943-958. Because decreased metabolism leads to a deficit in high- 4. Wolkin A, Sanfilipo M, Wolf AP, Angrist B, Brodie JD, Rotrosen J. Negative symp- energy phosphate molecules in the PFC of subjects with toms and hypofrontality in chronic schizophrenia. Arch Gen Psychiatry. 1992; schizophrenia,80 a breakdown in orderly metabolic ac- 49:959-965. tivity may further affect the chromatin-remodeling ma- 5. Heckers S, Goff D, Schacter DL, Savage CR, Fischman AJ, Alpert NM, Rauch SL. Functional imaging of memory retrieval in deficit vs nondeficit schizophrenia. chinery of the nucleus that depends on adenosine tri- Arch Gen Psychiatry. 1999;56:1117-1123. 44,45,82,83 phosphate. Thus, in a vicious cycle, decreased 6. Callicott JH, Bertolino A, Egan MF, Mattay VS, Langheim FJ, Weinberger DR. chromatin-remodeling could render genomic DNA less Selective relationship between prefrontal N-acetylaspartate measures and nega- accessible for the large multienzyme transcription com- tive symptoms in schizophrenia. Am J Psychiatry. 2000;157:1646-1651. 7. Middleton FA, Mirnics K, Pierri JN, Lewis DA, Levitt P. Gene expression profiling plexes and could alter the pattern of acetylation, meth- reveals alterations of specific metabolic pathways in schizophrenia. J Neurosci. ylation, and phosphorylation at the NH2-terminal tails 2002;22:2718-2729. of the core histones that define the functional state of chro- 8. Vawter MP, Shannon Weickert C, Ferran E, Matsumoto M, Overman K, Hyde TM, matin.26-35 These maladaptive changes may result in ad- Weinberger DR, Bunney WE, Kleinman JE. Gene expression of metabolic en- ditional chromatin defects,84 including abnormal DNA zymes and a protease inhibitor in the prefrontal cortex are decreased in 85 86 schizophrenia. Neurochem Res. 2004;29:1245-1255. methylation and repair. Therefore, hypoactivity and 9. Prabakaran S, Swatton JE, Ryan MM, Huffaker SJ, Huang JJ, Griffin JL, Way- 1-6,25,80,87 hypometabolism and dysregulated metabolic gene land M, Freeman T, Dudbridge F, Lilley KS, Karp NA, Hester S, Tkachev D, Mim- expression in the PFC of subjects with schizophrenia may mack ML, Yolken RH, Webster MJ, Torrey EF, Bahn S. Mitochondrial dysfunc- eventually lead to long-lasting molecular imprints in pre- tion in schizophrenia: evidence for compromised brain metabolism and oxidative frontal chromatin, resulting in transcriptional dysregu- stress. Mol Psychiatry. 2004;9:684-697. 7-24,52 88-91 10. Hakak Y, Walker JR, Li C, Wong WH, Davis KL, Buxbaum JD, Haroutunian V, lation, alterations in cytoarchitecture and neu- Fienberg AA. Genome-wide expression analysis reveals dysregulation of my- 92-95 96 ronal morphology, and oligodendrocyte loss. It elination-related genes in chronic schizophrenia. Proc Natl Acad SciUSA. 2001; remains the subject of future studies to elucidate the mo- 98:4746-4751. lecular mechanisms that link histone modifications and 11. Tsai G, Coyle JT. Glutamatergic mechanisms in schizophrenia. Annu Rev Phar- other epigenetic determinants of gene expression to cor- macol Toxicol. 2002;42:165-179. 12. Akbarian S, Kim JJ, Potkin SG, Hagman JO, Tafazzoli A, Bunney WE Jr, Jones tical dysfunction in schizophrenia. EG. Gene expression for glutamic acid decarboxylase is reduced without loss of neurons in prefrontal cortex of schizophrenics. Arch Gen Psychiatry. 1995; Submitted for Publication: December 5, 2003; final re- 52:258-266. vision received November 10, 2004; accepted January 14, 13. Benes FM, Vincent SL, Marie A, Khan Y. Up-regulation of GABAA receptor bind- ing on neurons of the prefrontal cortex in schizophrenic subjects. Neuroscience. 2005. 1996;75:1021-1031. Correspondence: Schahram Akbarian, MD, PhD, Brud- 14. Meador-Woodruff JH, Haroutunian V, Powchik P, Davidson M, Davis KL, Wat- nick Neuropsychiatric Research Institute, Department of son SJ. Dopamine receptor transcript expression in striatum and prefrontal cor- Psychiatry, University of Massachusetts Medical School, tex and occipital cortex: focal abnormalities in orbitofrontal cortex in schizophrenia. Worcester, MA 01604 (Schahram.akbarian@umassmed Arch Gen Psychiatry. 1997;54:1089-1095. 15. Mirnics K, Middleton FA, Marquez A, Lewis DA, Levitt P. Molecular character- .edu). ization of schizophrenia viewed by microarray analysis of gene expression in pre- Funding/Support: This work was supported by the Na- frontal cortex. Neuron. 2000;28:53-67. tional Alliance for Research on Schizophrenia and De- 16. Guidotti A, Auta J, Davis JM, Di-Giorgi-Gerevini V, Dwivedi Y, Grayson DR, Im- pression, Great Neck, NY; the National Alliance for Au- pagnatiello F, Pandey G, Pesold C, Sharma R, Uzunov D, Costa E, DiGiorgi Gerevini tism Research, Princeton, NJ; and the Worcester V. Decrease in reelin and glutamic acid decarboxylase67 (GAD67) expression in schizophrenia and bipolar disorder: a postmortem brain study. Arch Gen Psychiatry. Foundation for Biomedical Research (Dr Akbarian). The 2000;57:1061-1069. brain collection at Maryland Psychiatric Research Cen- 17. Volk DW, Austin MC, Pierri JN, Sampson AR, Lewis DA. Decreased glutamic acid ter was supported by grant MH60744 from the National decarboxylase67 messenger RNA expression in a subset of prefrontal cortical Institutes of Health, Bethesda, Md (Dr Roberts). ␥-aminobutyric acid neurons in subjects with schizophrenia. Arch Gen Psychiatry. 2000;57:237-245. Previous Presentation: This study was presented in part 18. Weickert CS, Webster MJ, Hyde TM, Herman MM, Bachus SE, Bali G, Wein- at the 2003 International Congress on Schizophrenia Re- berger DR, Kleinman JE. Reduced GAP-43 mRNA in dorsolateral prefrontal cor- search; March 30, 2003; Colorado Springs, Colo. tex of patients with schizophrenia. Cereb Cortex. 2001;11:136-147. Acknowledgment: We thank the members of the Mary- 19. Dracheva S, Marras SA, Elhakem SL, Kramer FR, Davis KL, Haroutunian V. land Brain Collection: Terri U’Prichard, MA, for permis- N-methyl-D-aspartic acid receptor expression in the dorsolateral prefrontal cortex of elderly patients with schizophrenia. Am J Psychiatry. 2001;158:1400- sions and interviews, Sami Daoud, MD, for dissections, and 1410. David Fowler, MD, and Jack Titus, MD (Office of the Chief 20. Perrone-Bizzozero NI, Sower AC, Bird ED, Benowitz LI, Ivins KJ, Neve RL. Lev- Medical Examiner), for their support and cooperation. els of the growth-associated protein GAP-43 are selectively increased in association cortices in schizophrenia. Proc Natl Acad SciUSA. 1996;93:14182- 14187. REFERENCES 21. Flynn SW, Lang DJ, Mackay AL, Goghari V, Vavasour IM, Whittall KP, Smith GN, Arango V, Mann JJ, Dwork AJ, Falkai P, Honer WG. Abnormalities of myelina- 1. Weinberger DR, Berman KF, Zec RF. Physiologic dysfunction of dorsolateral pre- tion in schizophrenia detected in vivo with MRI, and post-mortem with analysis frontal cortex in schizophrenia, I: regional cerebral blood flow evidence. Arch Gen of oligodendrocyte proteins. Mol Psychiatry. 2003;8:811-820. Psychiatry. 1986;43:114-124. 22. Tkachev D, Mimmack ML, Ryan MM, Wayland M, Freeman T, Jones PB, Starkey 2. Tamminga CA, Thaker GK, Buchanan R, Kirkpatrick B, Alphs LD, Chase TN, Car- M, Webster MJ, Yolken RH, Bahn S. Oligodendrocyte dysfunction in schizophre- penter WT. Limbic system abnormalities identified in schizophrenia using posi- nia and bipolar disorder. Lancet. 2003;362:798-805. tron emission tomography with fluorodeoxyglucose and neocortical alterations 23. Polesskaya OO, Haroutunian V, Davis KL, Hernandez I, Sokolov BP. Novel pu- with deficit syndrome. Arch Gen Psychiatry. 1992;49:522-530. tative nonprotein-coding RNA gene from 11q14 displays decreased expression 3. Andreasen NC, Rezai K, Alliger R, Swayze VW II, Flaum M, Kirchner P, Cohen G, in brains of patients with schizophrenia. J Neurosci Res. 2003;74:111-122.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 838

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/02/2021 24. Eastwood SL, Harrison PJ. Interstitial white matter neurons express less reelin DNA methylation and its contribution to complex diseases: from cancer to and are abnormally distributed in schizophrenia: towards an integration of mo- schizophrenia. Clin Genet. 2003;64:451-460. lecular and morphologic aspects of the neurodevelopmental hypothesis. Mol 50. Lewis SJ, Zammit S, Gunnell D, Smith GD. A meta-analysis of the MTHFR C677T Psychiatry. 2003;8:769, 821-831. polymorphism and schizophrenia risk. Am J Med Genet B Neuropsychiatr Genet. 25. Cantor-Graae E, Warkentin S, Franzen G, Risberg J, Ingvar DH. Aspects of sta- 2005;135:2-4. bility of regional cerebral blood flow in chronic schizophrenia: an 18-year fol- 51. Petronis A, Gottesman II, Kan P, Kennedy JL, Basile VS, Paterson AD, Popen- low-up study. Psychiatry Res. 1991;40:253-266. dikyte V. Monozygotic twins exhibit numerous epigenetic differences: clues to 26. Felsenfeld G, Groudine M. Controlling the double helix. Nature. 2003;421:448-453. twin discordance? Schizophr Bull. 2003;29:169-178. 27. Hayes JJ, Hansen JC. Nucleosomes and the chromatin fiber. Curr Opin Genet 52. Aston C, Jiang L, Sokolov BP. Microarray analysis of postmortem temporal cor- Dev. 2001;11:124-129. tex from patients with schizophrenia. J Neurosci Res. 2004;77:858-866. 28. Turner BM. Cellular memory and the histone code. Cell. 2002;111:285-291. 53. Akbarian S, Kim JJ, Potkin SG, Hetrick WP, Bunney WE Jr, Jones EG. Maldis- 29. Berger SL. Histone modifications in transcriptional regulation. Curr Opin Genet tribution of interstitial neurons in prefrontal white matter of the brains of schizo- Dev. 2002;12:142-148. phrenic patients. Arch Gen Psychiatry. 1996;53:425-436. 30. Li J, Lin Q, Yoon HG, Huang ZQ, Strahl BD, Allis CD, Wong J. Involvement of 54. Knable MB, Barci BM, Bartko JJ, Webster MJ, Torrey EF. Molecular abnormali- histone methylation and phosphorylation in regulation of transcription by thy- ties in the major psychiatric illnesses: Classification and Regression Tree (CRT) roid hormone receptor. Mol Cell Biol. 2002;22:5688-5697. analysis of post-mortem prefrontal markers. Mol Psychiatry. 2002;7:392-404. 31. Bauer UM, Daujat S, Nielsen SJ, Nightingale K, Kouzarides T. Methylation at argi- 55. Kirkpatrick B, Conley RC, Kakoyannis A, Reep RL, Roberts RC. Interstitial cells nine 17 of histone H3 is linked to gene activation. EMBO Rep. 2002;3:39-44. of the white matter in the inferior parietal cortex in schizophrenia: an unbiased 32. Cheung P, Tanner KG, Cheung WL, Sassone-Corsi P, Denu JM, Allis CD. Syn- cell-counting study. Synapse. 1999;34:95-102. ergistic coupling of histone H3 phosphorylation and acetylation in response to 56. Rioux L, Nissanov J, Lauber K, Bilker WB, Arnold SE. Distribution of microtubule- epidermal growth factor stimulation. Mol Cell. 2000;5:905-915. associated protein MAP2-immunoreactive interstitial neurons in the parahippo- 33. Salvador LM, Park Y, Cottom J, Maizels ET, Jones JC, Schillace RV, Carr DW, campal white matter in subjects with schizophrenia. Am J Psychiatry. 2003; Cheung P, Allis CD, Jameson JL, Hunzicker-Dunn M. Follicle-stimulating hor- 160:149-155. mone stimulates protein kinase A–mediated histone H3 phosphorylation and acety- 57. Ikeda K, Ikeda K, Iritani S, Ueno H, Niizato K. Distribution of neuropeptide Y in- lation leading to select gene activation in ovarian granulosa cells. J Biol Chem. terneurons in the dorsal prefrontal cortex of schizophrenia. Prog Neuropsycho- 2001;276:40146-40155. pharmacol Biol Psychiatry. 2004;28:379-383. 34. Li J, Chen P, Sinogeeva N, Gorospe M, Wersto RP, Chrest FJ, Barnes J, Liu Y. 58. Hashimoto T, Volk DW, Eggan SM, Mirnics K, Pierri JN, Sun Z, Sampson AR, Arsenic trioxide promotes histone H3 phosphoacetylation at the chromatin of Lewis DA. Gene expression deficits in a subclass of GABA neurons in the pre- CASPASE-10 in acute promyelocytic leukemia cells. JBiolChem. 2002;277: frontal cortex of subjects with schizophrenia. J Neurosci. 2003;23:6315-6326. 49504-49510. 59. Gao XM, Sakai K, Roberts RC, Conley RR, Dean B, Tamminga CA. Ionotropic 35. Li JH, Guo Y, Schroeder FA, Youngs RM, Schmidt TW, Ferris C, Konradi C, Ak- glutamate receptors and expression of N-methyl-D-aspartate receptor subunits barian S. Dopamine D2-like antagonists induce chromatin remodeling in striatal in subregions of human hippocampus: effects of schizophrenia. Am J Psychiatry. neurons through cyclic AMP-protein kinase A and NMDA receptor signaling. 2000;157:1141-1149. J Neurochem. 2004;90:1117-1131. 60. Roberts RC, Knickman J. The ultrastructural organization of the patch matrix com- 36. Crosio C, Heitz E, Allis CD, Borrelli E, Sassone-Corsi P. Chromatin remodeling partments in the human striatum. J Comp Neurol. 2002;452:128-138. and neuronal response: multiple signaling pathways induce specific histone H3 61. Albert KA, Hemmings HC Jr, Adamo AI, Potkin SG, Akbarian S, Sandman CA, modifications and early gene expression in hippocampal neurons. J Cell Sci. 2003; Cotman CW, Bunney WE Jr, Greengard P. Evidence for decreased DARPP-32 in 116:4905-4914. the prefrontal cortex of patients with schizophrenia. Arch Gen Psychiatry. 2002; 37. Tremolizzo L, Carboni G, Ruzicka WB, Mitchell CP, Sugaya I, Tueting P, Sharma 59:705-712. R, Grayson DR, Costa E, Guidotti A. An epigenetic mouse model for molecular 62. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real- and behavioral neuropathologies related to schizophrenia vulnerability. Proc Natl time quantitative PCR and the 2–⌬⌬CT. Methods. 2001;25:402-408. Acad Sci U S A. 2002;99:17095-17100. 63. Mimmack ML, Brooking J, Bahn S. Quantitative polymerase chain reaction: vali- 38. Yildirim E, Zhang Z, Uz T, Chen CQ, Manev R, Manev H. Valproate administra- dation of microarray results from postmortem brain studies. Biol Psychiatry. 2004; tion to mice increases histone acetylation and 5-lipoxygenase content in the 55:337-345. hippocampus. Neurosci Lett. 2003;345:141-143. 64. Timmermann S, Lehrmann H, Polesskaya A, Harel-Bellan A. Histone acetylation 39. Gould TD, Quiroz JA, Singh J, Zarate CA, Manji HK. Emerging experimental thera- and disease. Cell Mol Life Sci. 2001;58:728-736. peutics for bipolar disorder: insights from the molecular and cellular actions of 65. Ono S, Oue N, Kuniyasu H, Suzuki T, Ito R, Matsusaki K, Ishikawa T, Tahara E, current mood stabilizers. Mol Psychiatry. 2004;9:734-755. Yasui W. Acetylated histone H4 is reduced in human gastric adenomas and 40. Gurvich N, Tsygankova OM, Meinkoth JL, Klein PS. Histone deacetylase is a tar- carcinomas. J Exp Clin Cancer Res. 2002;21:377-382. get of valproic acid–mediated cellular differentiation. Cancer Res. 2004;64: 66. Toh Y, Yamamoto M, Endo K, Ikeda Y, Baba H, Kohnoe S, Yonemasu H, Hachitanda 1079-1086. Y, Okamura T, Sugimachi K. Histone H4 acetylation and histone deacetylase 1 41. Casey DE, Daniel DG, Wassef AA, Tracy KA, Wozniak P, Sommerville KW. Effect expression in esophageal squamous cell carcinoma. Oncol Rep. 2003;10: of divalproex combined with olanzapine or risperidone in patients with an acute 333-338. exacerbation of schizophrenia. Neuropsychopharmacology. 2003;28:182-192. 67. Toh Y, Ohga T, Endo K, Adachi E, Kusumoto H, Haraguchi M, Okamura T, Nicol- 42. Bowden CL. Clinical correlates of therapeutic response in bipolar disorder. J Affect son GL. Expression of the metastasis-associated MTA1 protein and its relation- Disord. 2001;67:257-265. ship to deacetylation of the histone H4 in esophageal squamous cell carcinomas. 43. Wassef AA, Hafiz NG, Hampton D, Molloy M. Divalproex sodium augmentation Int J Cancer. 2004;110:362-367. of haloperidol in hospitalized patients with schizophrenia: clinical and economic 68. Faure AK, Pivot-Pajot C, Kerjean A, Hazzouri M, Pelletier R. Peoc’h M, Sele B, implications. J Clin Psychopharmacol. 2001;21:21-26. Khochbin S, Rousseaux S. Misregulation of histone acetylation in Sertoli cell- 44. Ezhkova E, Tansey WP. Proteasomal ATPases link ubiquitylation of histone H2B only syndrome and testicular cancer. Mol Hum Reprod. 2003;9:757-763. to methylation of histone H3. Mol Cell. 2004;13:435-442. 69. Yasui W, Oue N, Ono S, Mitani Y, Ito R, Nakayama H. Histone acetylation and 45. Santos-Rosa H, Schneider R, Bernstein BE, Karabetsou N, Morillon A, Weise C, gastrointestinal carcinogenesis. Ann N Y Acad Sci. 2003;983:220-231. Schreiber SL, Mellor J, Kouzarides T. Methylation of histone H3 K4 mediates as- 70. Kawahara K, Kawabata H, Aratani S, Nakajima T. Hyper nuclear acetylation (HNA) sociation of the Isw1p ATPase with chromatin. Mol Cell. 2003;12:1325-1332. in proliferation, differentiation and apoptosis. Ageing Res Rev. 2003;2: 46. Matuoka K, Chen KY, Takenawa T. Rapid reversion of aging phenotypes by nico- 287-297. tinamide through possible modulation of histone acetylation. Cell Mol Life Sci. 71. Mahlknecht U, Hoelzer D. Histone acetylation modifiers in the pathogenesis of 2001;58:2108-2116. malignant disease. Mol Med. 2000;6:623-644. 47. Imai S, Johnson FB, Marciniak RA, McVey M, Park PU, Guarente L. Sir2: an NAD- 72. Good P. Permutation Tests: A Practical Guide to Resampling Methods for Test- dependent histone deacetylase that connects chromatin silencing, metabolism, ing Hypotheses. New York, NY: Springer-Verlag; 1994. and aging. Cold Spring Harb Symp Quant Biol. 2000;65:297-302. 73. SAS Institute Inc. The Mixed Procedure: SAS/STAT Software: Changes and En- 48. Abdolmaleky HM, Smith CL, Faraone SV, Shafa R, Stone W, Glatt SJ, Tsuang hancement Through Release 6.12. Cary, NC: SAS Institute Inc; 1997:571-702. MT. Methylomics in psychiatry: modulation of gene-environment interactions may 74. Mehta CR, Patel N. StatXact Version 6. Cambridge, Mass: Cytel Software Corp; be through DNA methylation. Am J Med Genet B Neuropsychiatr Genet. 2004; 2004. 127:51-59. 75. Schurter BT, Koh SS, Chen D, Bunick GJ, Harp JM, Hanson BL, Henschen- 49. Singh SM, Murphy B, O’Reilly RL. Involvement of gene-diet/drug interaction in Edman A, Mackay DR, Stallcup MR, Aswad DW. Methylation of histone H3 by

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 839

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/02/2021 coactivator-associated arginine methyltransferase 1. Biochemistry. 2001;40: 86. Benes FM, Walsh J, Bhattacharyya S, Sheth A, Berretta S. DNA fragmentation 5747-5756. decreased in schizophrenia but not bipolar disorder. Arch Gen Psychiatry. 2003; 76. Wang Y, Wysocka J, Sayegh J, Lee YH, Perlin JR, Leonelli L, Sonbuchner LS, 60:359-364. McDonald CH, Cook RG, Dou Y, Roeder RG, Clarke S, Stallcup MR, Allis CD, Coon- 87. Buchsbaum MS, Haier RJ, Potkin SG, Nuechterlein K, Bracha HS, Katz M, Lohr J, rod SA. Human PAD4 regulates histone arginine methylation levels via Wu J, Lottenberg S, Jerabek PA. Frontostriatal disorder of cerebral metabolism in demethylimination. Science. 2004;306:279-283. never-medicated schizophrenics. Arch Gen Psychiatry. 1992;49:935-942. 77. Cuthbert GL, Daujat S, Snowden AW, Erdjument-Bromage H, Hagiwara T, Yamada 88. Benes FM, Davidson J, Bird ED. Quantitative cytoarchitectural studies of the ce- M, Schneider R, Gregory PD, Tempst P, Bannister AJ, Kouzarides T. Histone deimi- rebral cortex of schizophrenics. Arch Gen Psychiatry. 1986;43:31-35. nation antagonizes arginine methylation. Cell. 2004;118:545-553. 89. Selemon LD, Rajkowska G, Goldman-Rakic PS. Abnormally high neuronal den- 78. Birrell JM, Brown VJ. Medial frontal cortex mediates perceptual attentional set sity in the schizophrenic cortex: a morphometric analysis of prefrontal area 9 shifting in the rat. J Neurosci. 2000;20:4320-4324. and occipital area 17. Arch Gen Psychiatry. 1995;52:805-818. 79. Uylings HB, Groenewegen HJ, Kolb B. Do rats have a prefrontal cortex? Behav 90. Rajkowska G, Selemon LD, Goldman-Rakic PS. Neuronal and glial somal size in Brain Res. 2003;146:3-17. the prefrontal cortex: a postmortem morphometric study of schizophrenia and 80. Pettegrew JW, Keshavan MS, Panchalingam K, Strychor S, Kaplan DB, Tretta Huntington disease. Arch Gen Psychiatry. 1998;55:215-224. MG, Allen M. Alterations in brain high-energy phosphate and membrane phos- 91. Buxhoeveden D, Roy E, Switala A, Casanova MF. Reduced interneuronal space pholipid metabolism in first-episode, drug-naive schizophrenics: a pilot study in schizophrenia. Biol Psychiatry. 2000;47:681-683. of the dorsal prefrontal cortex by in vivo phosphorus 31 nuclear magnetic reso- 92. Kalus P, Muller TJ, Zuschratter W, Senitz D. The dendritic architecture of pre- nance spectroscopy. Arch Gen Psychiatry. 1991;48:563-568. frontal pyramidal neurons in schizophrenic patients. Neuroreport. 2000;11: 81. Konradi C, Eaton M, MacDonald ML, Walsh J, Benes FM, Heckers S. Molecular 3621-3625. evidence for mitochondrial dysfunction in bipolar disorder. Arch Gen Psychiatry. 93. Rosoklija G, Toomayan G, Ellis SP, Keilp J, Mann JJ, Latov N, Hays AP, Dwork 2004;61:300-308. AJ. Structural abnormalities of subicular dendrites in subjects with schizophre- 82. Imbalzano AN. Energy-dependent chromatin remodelers: complex complexes and nia and mood disorders: preliminary findings. Arch Gen Psychiatry. 2000;57: their components. Crit Rev Eukaryot Gene Expr. 1998;8:225-255. 349-356. 83. Becker PB, Horz W. ATP-dependent nucleosome remodeling. Annu Rev Biochem. 94. Glantz LA, Lewis DA. Decreased dendritic spine density on prefrontal cortical py- 2002;71:247-273. ramidal neurons in schizophrenia. Arch Gen Psychiatry. 2000;57:65-73. 84. Issidorides MR, Stefanis CN, Varsou E, Katsorchis T. Altered chromatin ultra- 95. Costa E, Davis J, Grayson DR, Guidotti A, Pappas GD, Pesold C. Dendritic spine structure in neutrophils of schizophrenics. Nature. 1975;258:612-614. hypoplasticity and downregulation of reelin and GABAergic tone in schizophre- 85. Veldic M, Caruncho HJ, Liu WS, Davis J, Satta R, Grayson DR, Guidotti A, Costa nia vulnerability. Neurobiol Dis. 2001;8:723-742. E. DNA-methyltransferase 1 mRNA is selectively overexpressed in telencephalic 96. Hof PR, Haroutunian V, Friedrich VL Jr, Byne W, Buitron C, Perl DP, Davis KL. GABAergic interneurons of schizophrenia brains. Proc Natl Acad Sci U S A. 2004; Loss and altered spatial distribution of oligodendrocytes in the superior frontal 101:348-353. gyrus in schizophrenia. Biol Psychiatry. 2003;53:1075-1085.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, AUG 2005 WWW.ARCHGENPSYCHIATRY.COM 840