Altered Expression of Regulators of the Cortical Chloride Transporters NKCC1 and KCC2 in Schizophrenia

ORIGINAL ARTICLE

ONLINE FIRST Altered Expression of Regulators of the Cortical Chloride Transporters NKCC1 and KCC2 in Schizophrenia

Dominique Arion, PhD; David A. Lewis, MD

Context: Disturbances in markers of cortical ␥- parison subjects. Brain specimens from 18 macaque aminobutyric acid neurotransmission are a common find- monkeys exposed to haloperidol, olanzapine, or sham ing in schizophrenia. The nature of ␥-aminobutyric acid long-term. neurotransmission (hyperpolarizing or depolarizing) depends on the local intracellular chloride concentration. Main Outcome Measures: Relative expression lev- In the central nervous system, the intracellular chloride els for NKCC1, KCC2, OXSR1, STK39, WNK1, WNK3, level is determined by the activity of 2 cation-chloride and WNK4 transcripts compared with the mean expres- transporters, NKCC1 and KCC2. The activities of these sion level of 3 housekeeping transcripts. transporters are in turn regulated by a network of serine- threonine kinases that includes OXSR1, STK39, and the Results: OXSR1 and WNK3 transcripts were substan- WNK kinases WNK1, WNK3, and WNK4. tially overexpressed in subjects with schizophrenia relative to comparison subjects. In contrast, NKCC1, KCC2, Objective: To compare the levels of NKCC1, KCC2, STK39, WNK1, and WNK4 transcript levels did not dif- OXSR1, STK39, WNK1, WNK3, and WNK4 transcripts fer between subject groups. OXSR1 and WNK3 tran- in prefrontal cortex area 9 between subjects with schizo- script expression levels were not changed in antipsychotic- phrenia and healthy comparison subjects. exposed monkeys and were not affected by potential confounding factors in the subjects with schizophrenia. Design: Real-time quantitative polymerase chain reaction technique was used to measure transcript levels in Conclusion: In schizophrenia, increased expression lev- the prefrontal cortex. els, and possibly increased kinase activities, of OXSR1 and WNK3 may shift the balance of chloride transport by Setting: Human brain specimens were obtained from NKCC1 and KCC2 and alter the nature of ␥-aminobu- autopsies conducted at the Allegheny County Medical Ex- tyric acid neurotransmission in the prefrontal cortex. aminer’s Office, Pittsburgh, Pennsylvania. Arch Gen Psychiatry. 2011;68(1):21-31. Participants: Postmortem brain specimens from 42 sub- Published online September 6, 2010. jects with schizophrenia and 42 matched healthy com- doi:10.1001/archgenpsychiatry.2010.114

CORE FEATURE OF SCHIZO- depends on the amount of GABA available phrenia is the impair- at the synapse; (2) the kinetics of GABA neu- ment of certain cognitive rotransmission, which is determined, in part, functions, such as work- by the subunit composition of GABAA re- ing memory,1 that are de- ceptors; and (3) the nature of GABA neu- pendent on the circuitry of the dorsolat- rotransmission, which can be hyperpolar- A 2 eral prefrontal cortex (DLPFC). These izing, depolarizing, or shunting depending cognitive deficits in schizophrenia are as- on the flow of chloride ions when GABAA sociated with altered neural network syn- receptors are activated.6 Presynaptic and chrony3,4 that may be attributable, at least postsynaptic markers of the strength and ki- Author Affiliations: in part, to functional abnormalities in ␥- netics of GABA neurotransmission are Departments of Psychiatry aminobutyric acid (GABA) neurotrans- known to be altered in the DLPFC in sub- (Drs Arion and Lewis) and 5 7,8 Neuroscience (Dr Lewis), mission in the DLPFC. jects with schizophrenia, but measures of Translational Neuroscience Disturbed GABA neurotransmission in the nature of GABA neurotransmission have Program, University of schizophrenia might arise from alterations not been evaluated. Pittsburgh, Pittsburgh, in 1 or more of the following factors: (1) the The flow of chloride ions through Pennsylvania. strength of GABA neurotransmission, which GABAA receptors depends on intracellu-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 21

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/23/2021 tween subject groups. Thus, increased expression levels, Table. Summary of Demographic Characteristics of Subjects and possibly increased kinase activities, of OXSR1 and WNK3 in schizophrenia would be predicted to both in- qPCR, Mean (SD) crease NKCC1 and decrease KCC2 activities, producing Comparison Subjects With greater intracellular chloride concentration and thus alter- Subjects Schizophrenia ing the nature of GABA neurotransmission in the DLPFC. Sample size 42 42 Sex, No. METHODS M3131 F1111 Race, No. SUBJECTS White 34 29 Black 8 13 Brain specimens (n=84) were obtained during autopsies con- Age, y 48.5 (13.1) 47.0 (12.8) ducted at the Allegheny County Medical Examiner’s Office (Pitts- PMI, h 17.7 (5.7) 18.7 (8.8) burgh, Pennsylvania) after consent was obtained from the next Brain pH 6.7 (0.20) 6.6 (0.4) of kin. An independent committee of experienced research cli- RIN 8.2 (0.7) 8.1 (0.8) nicians made consensus DSM-IV diagnoses for each subject on Storage time (months at −80°C) 74.2 (42.9) 78.5 (42.5) the basis of medical records and structured diagnostic inter- views conducted with the decedent’s family members, as de- Abbreviations: qPCR, quantitative polymerase chain reaction; scribed previously.30 To control for experimental variance and PMI, postmortem interval; RIN, RNA integrity number; SD, standard to reduce biological variance between subject groups, each sub- deviation. ject with schizophrenia (n=42) was matched with 1 healthy comparison subject for sex and as closely as possible for age and postmortem interval (PMI); samples from a given subject lar levels of chloride, which are regulated by the relative pair were always processed together. The subject groups (Table; activities of the sodium-potassium-chloride cotrans- see eTable 1 [http://www.archgenpsychiatry.com] for details porter 1 (NKCC1; SLC12A2), which mediates chloride on individual subjects) did not significantly differ in mean age, uptake, and the potassium-chloride cotransporter 2 PMI, brain pH, RNA integrity number (RIN), or tissue storage Ͻ Ͼ ␹2 (KCC2; SLC12A5), which mediates chloride extrusion time at −80°C (all t82 1.67; all P .10) or in race ( =1.587; from the cell. In the central nervous system, NKCC1 is P=.21). All procedures were approved by the University of Pitts- found in both neurons and glial cells, but KCC2 is strictly burgh’s Committee for the Oversight of Research Involving the Dead and Institutional Review Board for Biomedical Research. neuronal.9-11 The activities of both transporters are sensitive to intracellular chloride levels, and their activation depends on their phosphorylation (NKCC1) or de- TISSUE PREPARATION phosphorylation (KCC2) status.12-15 Oxidative stress The right hemisphere of each brain was blocked coronally, im- response kinase (OXSR1; also known as OSR1) and Ste mediately frozen, and stored at −80°C, as described previ- 20–related, proline-alanine–rich kinase (STK39; also ously.31 Area 9 of the DLPFC was identified cytoarchitectoni- 16-19 known as SPAK) are highly expressed in the brain and cally from Nissl-stained coronal sections31 spanning the bind to NKCC1 where they phosphorylate residues pres- rostrocaudal axis of the superior frontal sulcus. The cortical ent in the N-terminal domain, resulting in an increase gray matter was dissected from cryostat sections (40 µm) in a of NKCC1 activity.17,20-23 In addition, NKCC1 can be phos- manner that ensured limited white matter contamination and phorylated by the coexpression of STK39 and WNK4 excellent RNA preservation, as described previously.32 Total RNA (with no K [lysine] protein kinase 4) kinases in vitro.24 was isolated from tissue homogenates using the TRIzol proto- Conversely, STK39-mediated phosphorylation has a domi- col from Invitrogen (Invitrogen Corporation, Carlsbad, Cali- nant negative effect on KCC2 function.24 Another WNK fornia) and further purified using the RNeasy kit (QIAGEN, Valencia, California). RNA integrity was assessed by measur- kinase, WNK1, activates both STK39 and OXSR1 via phos- 23,25-27 ing the RIN using the Agilent 2100 bioanalyzer (Agilent Tech- phorylation. Finally, a fifth kinase, WNK3, both ac- nologies, Santa Clara, California). tivates NKCC1 and inhibits KCC2, regardless of cellu- 28,29 lar tonicity. Consequently, alterations in the relative QUANTITATIVE REAL-TIME POLYMERASE expression levels of NKCC1 and KCC2, or of their regu- CHAIN REACTION latory kinases, in the DLPFC of subjects with schizophrenia could, by shifting intracellular chloride levels, Total RNA was converted to complementary DNA using the alter the nature of GABA transmission and thereby con- High-Capacity cDNA Archive Kit from Applied Biosystems (Fos- tribute to impaired neural network synchrony and cog- ter City, California). For each reaction, we used 50 ng of total nitive dysfunction in affected individuals. RNA from each subject. Priming was performed with random To examine this possibility, we quantified messenger hexamers, according to the manufacturer’s recommendations. RNA (mRNA) expression levels of NKCC1 and KCC2 and The efficiency for each primer set (eTable 2) was between 92% their associated regulatory kinases STK39, OXSR1, WNK1, and 100%, and the amplified product resulted in a specific single WNK3, and WNK4 in the DLPFC from subjects with product in dissociation curve analysis. Given the high level of homology between certain domains of OXSR1 and STK39, the schizophrenia and matched healthy comparison subjects. quantitative real-time polymerase chain reaction (qPCR) primer In subjects with schizophrenia, OXSR1 and WNK3 tran- sets for these 2 transcripts were designed within a unique se- scripts were substantially overexpressed, relative to com- quence in the 3Ј untranslated region of each mRNA. parison subjects. In contrast, NKCC1, KCC2, STK39, Samples from each matched pair of subjects with schizo- WNK1, and WNK4 transcript levels did not differ be- phrenia and comparison subjects were always assayed on the

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 22

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/23/2021 same plate. For each sample, amplified product differences for body specificity was demonstrated by the absence of labeling each transcript were measured with 4 replicates using SYBR in the thymus40 and by the absence of specific labeling when Green chemistry-based detection.33 the primary antibody was omitted. ␤-Actin, cyclophilin A, and glyceraldehyde-3-phosphate dehydrogenase were used as endogenous reference genes. These WESTERN BLOTTING AND SDS GEL 3 transcripts were selected based on their previously demon- ELECTROPHORESIS strated stable expression across both subjects with schizophre- 32 nia and healthy comparison subjects. The qPCR reactions were Western blots were conducted using brain tissue lysate pre- carried out in an ABI Prism 7000 thermal cycler (Applied Bio- pared as previously described41 (eAppendix). The specificity systems) using the ABI Prism 7000 SDS software with the au- of the anti-OXSR1 antibody was confirmed by detection of a tomatic baseline and threshold detection options selected. These band of identical molecular weight (58 kDa) compared with data were exported to Microsoft Excel (Microsoft, Redmond, cloned human OXSR1 expressed in vitro and by lack of cross- Washington) and delta cycle thresholds (dCTs) were calcu- reactivity with a cell lysate of in vitro–expressed STK39 (Ori- lated for each sample by using the geometric mean of the 3 en- Gene, Rockville, Maryland) (data not shown). dogenous reference genes as the normalization factor (ie, cycle To assess the effect of PMI on OXSR1 immunoreactivity, 4 threshold [CT] for each transcript in a sample minus the geo- adjacent coronal tissue blocks (2-3 mm thick) were prepared metric mean of ␤-actin, cyclophilin A, and glyceraldehyde-3- 34 from the DLPFC of the same adult macaque monkey, stored phosphate dehydrogenase CTs for the same sample). The ex- in room-temperature artificial cerebrospinal fluid for 0, 6, 12, pression level for each transcript was then calculated as the 42 −dCTs or 24 hours, and then flash frozen. The OXSR1 protein con- expression ratio value (where expression ratio=2 ), and all tent was reported as percentage relative to the 0-hour PMI. results are reported as the expression ratio. Two isoforms of the WNK3 gene are expressed in the brain: isoform 1, which is brain specific, and isoform 2, which is ubiq- LASER MICRODISSECTION ANALYSES uitous. Isoform 1 differs from isoform 2 only by the presence of an additional stretch of 47 amino acids in the protein.35 Given To assess the laminar and cellular patterns of the expression the differential expression pattern of both isoforms, we de- differences in OXSR1 and WNK3-1&2, we used laser micro- signed 1 primer set detecting WNK3 isoform 1 (primer set dissection techniques to obtain mRNA samples from layer 3 WNK3-1) and 1 primer set detecting both isoforms 1 and 2 tissue and from individually dissected layer 3 pyramidal cells (primer set WNK3-1&2). WNK3-1 was used on the 42 pairs from a subset of 10 subject pairs (eTable 1A and B, pairs with •). of subjects, and WNK3-1&2 was used only on an arbitrary sub- Using the results from qPCR experiments conducted on the en- set of 12 pairs of subjects with schizophrenia and healthy com- tire area 9 gray matter for all 42 subject pairs, we selected the parison subjects (eTable 1A, pairs with *) that had been used 10 pairs with mean expression differences between subjects with in a previous real-time qPCR study.32 schizophrenia and control subjects that were closest to the mean of the entire group. Cryostat sections (12 µm) were cut and thaw-mounted onto HIERARCHICAL CLUSTERING glass polyethylene naphthalate membrane slides (Leica Micro- systems, Bannockburn, Illinois) that had been previously UV- Two-way cluster analyses were performed using delta-delta CTs treated at 254 nm for 30 minutes, dried, and stored at −80°C. (subject with schizophrenia dCT minus matched healthy com- On the day of the microdissection, slides were immersed in an parison subject dCT) across all 42 subject pairs for all tran- ethanol–acetic acid fixation solution, stained with thionin, de- scripts using Pearson mean analysis in Genes@Work (IBM, Ar- 36 hydrated through 100% ethanol, and air dried. Using a Leica monk, New York). microdissection system (LMD 6500), 2 independent samples of layer 3 pyramidal neurons (eFigure 3A) per subject from 2 ANTIPSYCHOTIC-TREATED MONKEYS different slides (ϫ40 objective; power, 15; aperture, 9; speed, 12; balance, 14; and offset, 120) or 3 strips of tissue from layer The effect of long-term exposure to antipsychotic medication 3, as previously described43 (ϫ5 objective; power, 43; aper- on the levels of transcripts showing altered expression in schizo- ture, 18; speed, 13; balance, 25; and offset, 60), were ob- phrenia was examined using macaque monkeys exposed to halo- tained. An average of 50 single pyramidal cells or 4.1 mm2 of peridol, olanzapine, and placebo, as described before37,38 (eAp- layer 3 tissue were collected per sample in 0.5-mL microtube pendix). All housing and experimental procedures were caps (Ambion/Applied Biosystems) and lysed by vortexing for conducted in accordance with National Institutes of Health 30 seconds in 200 µL of RLT Buffer Plus (QIAGEN). The RNA guidelines and with approval of the University of Pittsburgh was extracted and PCR analyses were conducted, and cell type Institutional Animal Care and Use Committee. specificity and absence of glial contamination of the single pyramidal cell samples were confirmed, as described in the eAp- IMMUNOCYTOCHEMICAL EXPERIMENTS pendix. We elected to measure both isoforms of WNK3 using the WNK3-1&2 primer set because of the lower expression of To assess the distribution and relative levels of OXSR1 pro- WNK3-1, although the expression difference between schizo- tein, we conducted immunocytochemical experiments using phrenia and comparison samples using WNK3-1&2 was only immersion-fixed cryostat sections (40 µm) of DLPFC area 9 half of the expression difference obtained using WNK3-1. from 6 white male human subjects (17-62 years of age; 3 subjects with a short PMI [5-8.2 hours] and 3 subjects with a long STATISTICAL ANALYSES PMI [14.5-37 hours]). Tissue sections were processed for OXSR1 immunoreactivity using a 1:200 dilution of a goat anti-OXSR1 The effect of diagnosis and the influence of potential con- (SC49473; Santa Cruz Biotechnology, Santa Cruz, California) founding variables on the transcript expression ratios were and the Vectastain ABC kit (Vector Laboratories, Burlingame, assessed using 2 analyses of covariance (ANCOVA) models. California), according to a previously described protocol.39 The In the first model, transcript expression ratio was used as the OXSR1 antibody was raised against the C-terminal portion dependent variable; diagnostic group, as the main effect; and (amino acids 370-420) of the human OXSR1 protein. Anti- subject pair, as a blocking factor. The RIN, pH, and freezer

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 23

0.120

0.160

0.090

0.130

0.060 0.100 Expression Ratio in Comparison Subjects

0.030 0.070 0.030 0.060 0.090 0.120 0.070 0.100 0.130 0.160 0.190

C D 0.040 OXSR1 WNK3-1

0.006 0.035

0.030 0.004

0.025

0.002

0.020

Expression Ratio in Comparison Subjects 24-Pair cohort 18-Pair cohort

0.015 0.000 0.015 0.0200.025 0.030 0.035 0.040 0.000 0.002 0.004 0.006 Expression Ratio in Subjects With Schizophrenia Expression Ratio in Subjects With Schizophrenia

Figure 1. Expression ratio values (where expression ratio=2−delta cycle thresholds) for subjects with schizophrenia and comparison subjects in the 24-pair and 18-pair cohorts. A, NKCC1. B, KCC2. C, OXSR1. D, WNK3-1. Data points below the diagonal unity line represent subject pairs in which the expression level for the gene of interest is greater in the subjects with schizophrenia relative to the matched healthy comparison subject.

storage time were entered as covariates because they may A 1-way analysis of variance model with expression ratios affect RNA integrity. Subject pairing may be considered an as the dependent variable and treatment group as the main effect attempt to control for experimental variance by the parallel was used to compare transcript expression levels in the halo- processing of tissue samples from each subject pair and not a peridol-, olanzapine-, and placebo-exposed monkeys. true statistical pair design. Therefore, to validate the first model, a second unpaired ANCOVA model was performed, using diagnostic group as the main effect and with age, sex, RESULTS PMI, RIN, pH, and freezer storage time as covariates. Because the 42 pairs of subjects were used in 2 successive qPCR runs TRANSCRIPT LEVELS FOR CHLORIDE of 24 and 18 pair cohorts (eTable 1A) and 18 (eTable 1B) that involved different lots of reagents, qPCR run was also entered TRANSPORTERS AND RELATED TRANSCRIPTS as a covariate in both models. Tissue storage time did not have a significant effect in either model for any transcript and was ex- After corrections for multiple comparisons, neither the cluded in the reported analyses. The reported P values for each mean level of NKCC1 mRNA (Figure 1A) nor KCC2 model were corrected for multiple comparisons (n=21) using mRNA (Figure 1B) was significantly altered in subjects the Bonferroni procedure. Ͻ with schizophrenia (paired ANCOVA, both F1,38 3.8 The influences of potential confounding variables on the ex- Ͼ Ͻ and P .43; unpaired ANCOVA, both F1,75 4.1 and pression ratio values in subjects with schizophrenia were as- PϾ.33). For the chloride transporter–related kinases, sessed with ANCOVA models using each confounding variable as the main effect and sex, age, PMI, pH, RIN, and run as group differences in gene expression levels were not covariates. Pearson correlation was used to assess the relation- significant for STK39, WNK1, or WNK4 (paired, all Ͻ Ͼ Ͻ ships of the expression ratios for all pairs of transcripts; re- F1,38 5.1 and all P .20; unpaired, all F1,75 4.11 and ported P values were also corrected for multiple comparisons all PϾ.32). In contrast, the mean expression level of (n=21) using the Bonferroni procedure. OXSR1 mRNA (Figure 1C) was significantly (paired,

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 24

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/23/2021 Ͻ F1,38=54.3 and P .001; unpaired, F1,75=38.1 and Ͻ P .001) 19.4% greater in the subjects with schizophre- 0.06 Placebo nia. In addition, mean WNK3-1 mRNA levels Haloperidol Olanzapine 0.05 (Figure 1D) were significantly (paired, F1,38=50.9 and Ͻ Ͻ P .001; unpaired, F1,75=45.6 and P .001) 48.7% greater in the subjects with schizophrenia. Relative to 0.04 their matched healthy comparison subjects, OXSR1 and 0.03 WNK3-1 mRNA levels were higher in 41 and 37, ER respectively, of the 42 subjects with schizophrenia (Figure 1C and D). None of the covariates (age, sex, 0.02 PMI, RIN, brain pH, or freezer storage time) were significant in any of the analyses for OXSR1 and WNK3-1 0.01 Ͻ Ͼ mRNA (paired, all F1,38 2.69 and all P .11; unpaired, Ͻ Ͼ 0.00 all F1,75 3.76 and all P .06). OXSR1 WNK3-1&2 WNK3-1

WNK3 ISOFORMS 1 AND 2 Figure 2. Effect of antipsychotic medication on the transcript expression for OXSR1, WNK3 isoform 1 (WNK3-1), and WNK3 isoforms 1 and 2 Because of the low expression level of the brain-specific (WNK3-1&2). Expression ratio (ER) values were calculated for each transcript for 6 triads of monkeys with long-term exposure to placebo, WNK3-1 mRNA (primer set WNK3-1) in the DLPFC, haloperidol, or olanzapine. we designed a new set of primers binding to both isoforms 1 and 2 of WNK3 in the human brain (primer set WNK3-1&2). This primer set produced lower CT val- way with WNK1, WNK4, NKCC1, and STK39, whereas ues than the WNK3-1 primer set in the same subjects WNK3-1 and KCC2 clustered in a distinct pathway. (n=12 pairs), reflecting the higher level of expression Pearson correlations were performed using the dCT for both WNK3 isoforms combined than for isoform 1 values obtained for each transcript for each of the 84 sub- alone. Consistent with the findings for isoform 1 alone, jects (Figure 4 and eTable 3); the reported P values were the combined expression of both WNK3 isoforms was calculated using Bonferroni correction for multiple com- ϩ also higher ( 27.0%; paired, F1,8=3.97 and P =.08; parisons. Expression levels of OXSR1 were significantly unpaired, F1,19=7.1 and P=.01) in the subjects with positively correlated with those for WNK3-1 (r=0.744; schizophrenia, and the expression levels from both PϽ10−7) and WNK1 (r=0.478; PϽ10−4), and WNK3-1 primer sets were highly correlated (r=0.561; PϽ.01) expression was significantly correlated with WNK1 across all subjects. (r=0.346; P=.03). In contrast, WNK3-1 expression was negatively correlated with KCC2 (r=−0.356; P=.02). ANALYSIS OF POTENTIAL ANTIPSYCHOTIC NKCC1 expression was positively correlated with WNK1 MEDICATION EFFECTS (r=0.583; PϽ10−7), WNK4 (r=0.425; P=.001), and STK39 (r=0.371; P=.01). The expression ratios for OXSR1, WNK3-1, and WNK3- Ͻ Ͼ 1&2 did not differ (all F2,15 0.92 and all P .42) across IMMUNOLOCALIZATION OF OXSR1 the monkeys with long-term exposure to placebo, haloperidol, or olanzapine (Figure 2). In human DLPFC from subjects with PMIs less than 6 hours, immunoreactivity for OXSR1 was highest in lay- ANALYSIS OF POTENTIAL ers 3 and 5, intermediate in layers 2 and 6, and low in CONFOUNDING FACTORS layers 1 and 4 (Figure 5A and B). OXSR1 immunoreactivity was mainly located in the perinuclear cytoplas- The mean expression ratio in the subjects with schizo- mic compartment of neurons, most of which had the mor- phrenia did not differ for OXSR1 (Figure 3A) (all phology of pyramidal cells. In contrast, OXSR1 labeling FϽ1.941 and all P Ն .17, uncorrected) or WNK3-1 was severely reduced with PMIs longer than 12 hours (Figure 3B) (all FՅ3.99 and all PϾ.05, uncorrected) as (Figure 5C). This finding was confirmed by Western blot- a function of sex; diagnosis of schizoaffective disorder; ting in samples of monkey DLPFC with artificially in- suicide; antidepressant medication, benzodiazepines or duced PMIs. At PMIs of 12 and 24 hours, levels of OXSR1 sodium valproate, or antipsychotic medication use at the protein were only 60% and 40%, respectively, of that de- time of death; or diagnosis of substance abuse/ tected at 0 hours (eFigure 2). Thus, a quantitative study dependence at time of death. of OXSR1 protein levels in subjects with schizophrenia and control subjects could not be conducted. HIERARCHICAL CLUSTERING AND CORRELATIONS LAMINAR AND CELLULAR ASSESSMENTS OF OXSR1 AND WNK3-1&2 mRNA EXPRESSION A 2-way cluster analysis performed on delta-delta CT values across the 42 pairs of matched subjects for all 7 tran- To determine if the greater expression of OXSR1 and scripts with Pearson mean distance analysis resulted in WNK3-1&2 transcripts in the subjects with schizophre- the separation of the transcript expression changes in 2 nia arose from the same cortical layer and neurons, we different clusters (eFigure 1). OXSR1 clustered in a path- used laser microdissection to obtain samples of tissue re-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 25

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/23/2021 A F1,34 = 0.12 F1,33 = 0.06 F1,33 = 0.15 F1,33 = 0.99 F1,33 = 1.94 F1,33 = 1.81 F1,33 = 0.002 0.04 P = .73 P = .81 P = .70 P = .33 P = .17 P = .19 P = .96

0.03

0.02 Mean OXSR1 ER

0.01

31 11 28 14 31 11 25 17 27 15 6 36 29 13 0.00

B F1,34 = 0.05 F1,33 = 2.02 F1,33 = 0.04 F1,33 = 0.53 F1,33 = 0.04 F1,33 = 3.99 F1,33 = 0.001 0.008 P = .81 P = .16 P = .83 P = .47 P = .85 P = .06 P = .98

0.006

0.004 Mean WNK3-1 ER

0.002

31 11 28 14 31 11 25 17 27 15 6 36 29 13 0.000 M F No Yes No Yes No Yes No Yes No Yes No Yes Sex Schizoaffective Suicide Antidepressants Benzodiazapines Antipsychotic Substance Abuse Disorder at Time of Death at Time of Death at Time of Death or Dependancy at Time of Death

Figure 3. Mean (bar) and individual (circle) expression ratio (ER) values for OXSR1 (A) and WNK3 isoform 1 (WNK3-1) (B) are shown for the subjects with schizophrenia grouped by potential confounding factors. Neither sex, diagnosis of schizoaffective disorder, suicide, antidepressant medication use at the time of death, use of benzodiazepines or sodium valproate at the time of death, antipsychotic medication use at the time of death, nor diagnosis of substance abuse or dependence at the time of death significantly affected OXSR1 (A) or WNK3-1 (B) transcript expression. Numbers in bars indicate the number of subjects with schizophrenia in each category.

stricted to layer 3 and samples of 50 pyramidal cells in- the expression levels of OXSR1 and WNK3-1&2 mRNA dividually cut from layer 3. In the layer 3 tissue samples, were significantly correlated (r=0.56; P=.001) across all the expression levels of OXSR1 and WNK3-1&2 mRNA 20 subjects, confirming that both transcripts are ex- were significantly correlated (r=0.88; PϽ.001) across all pressed in layer 3 pyramidal cells and suggesting that their 20 subjects, and the within-pair expression differences expression is coregulated. In addition, both transcripts for OXSR1 and WNK3-1&2 mRNA levels were also sig- showed greater expression in the subjects with schizo- nificantly correlated (r=0.92; PϽ.001). Consistent with phrenia (mean percentage increases: OXSR1, ϩ12.0%; these findings, in the samples of layer 3 pyramidal cells, WNK3-1&2, ϩ12.4%), although because of the smaller

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 26

9.5 3.5

9.0 3.0

8.5 2.5 WNK1 WNK3-1

8.0 2.0

7.5 1.5

7.0 1.0 4.0 4.5 5.0 5.5 6.0 4.5 5.0 5.5 6.0 6.5 OXSR1 OXSR1 C D 4.0 11.0

3.5 10.5

3.0 10.0

2.5 9.5 WNK1 WNK4

2.0 9.0

1.5 8.5

1.0 8.0 2.5 3.0 3.54.0 4.5 5.0 2.5 3.0 3.54.0 4.5 5.0 NKCC1 WNK1

Figure 4. Four most significant Pearson correlations (PϽ10−3) between delta cycle thresholds for 84 subjects. WNK3-1 indicates WNK3 isoform 1.

sample size (n=10 pairs) these findings did not achieve increased expression of OXSR1 and WNK3 are likely to statistical significance (t=−1.007; P=.15 and t=−0.682; lead to shifts in the activity of the chloride transporters P=.25, respectively). Together, these findings suggest an that could substantially alter the nature of cortical GABA upregulation of both transcripts in layer 3 pyramidal cells neurotransmission in schizophrenia. Indeed, the find- in the illness. ing that both transcripts are upregulated in the same layer 3 pyramidal neurons suggests that altered GABA signal- COMMENT ing may be particularly prominent in these cells. The findings for OXSR1 and WNK3 are striking for several other In the DLPFC of subjects with schizophrenia, mRNA ex- reasons. First, the higher levels of these transcripts in sub- pression levels for the chloride transporters NKCC1 and jects with schizophrenia contrast with the more com- KCC2 were not significantly altered. Although it is pos- monly observed lower levels of GABA-related tran- 32 sible that our measures of mRNA in total gray matter ob- scripts in the illness. Second, the elevated transcript levels scured schizophrenia-associated expression differences in schizophrenia indicate that the findings are not at- in neuronal subpopulation(s) in a different layer, we pre- tributable to poorer RNA quality in these subjects, con- viously found that neither NKCC1 nor KCC2 mRNA sistent with the excellent brain pH and RIN measures in showed a preferential laminar pattern of expression in all subjects used in this study and the absence of group control human subjects.43 differences in these variables. Third, relative to the In contrast, the mRNA expression levels of 2 regula- matched healthy comparison subjects, almost all of the tory kinases, OXSR1 and WNK3, were markedly and con- 42 subjects with schizophrenia had higher levels of both sistently higher in the DLPFC of subjects with schizo- OXSR1 and WNK3 mRNA. The apparent consistency of phrenia. Although an increase in protein kinase activity these findings suggests that they are more likely to be re- may not always follow an increase in the corresponding lated to a conserved downstream aspect of the disease mRNA transcript, previous reports have shown that in- process (eg, impaired GABA neurotransmission) than to creasing WNK3 mRNA levels or silencing OXSR1 ex- reflect the etiological complexity of schizophrenia.8 pression is associated with alterations in NKCC1 trans- Fourth, consistent with the findings that these tran- port activity.28,44 Thus, our results suggest that the scripts are not altered in monkeys exposed to antipsy-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 27

Figure 5. A-C, Brightfield photomicrographs showing Nissl staining (A) and laminar distribution pattern of OXSR1 immunoreactivity in human dorsolateral prefrontal cortex area 9 for a subject with a short postmortem interval (5.5 hours) (B) and a long postmortem interval (19.3 hours) (C) (scale bar=300 µm). D, The insert shows a magnification of OXSR1 immunoreactivity in layer 3 (scale bar=30 µm). WM indicates white matter.

chotic medications long-term, the conservation of these transcript expression level was significantly altered in sub- alterations across almost all subjects examined strongly jects with schizophrenia, despite suggestions of overex- suggests that they are not attributable to other factors (eg, pression (ϩ7%; P=.36 and ϩ12.6%; P=.17, respec- substance use, suicide, mood symptoms) that were pres- tively), indicating that the overexpression of the OXSR1 ent in some of the subjects with schizophrenia, an inter- transcript could be primarily responsible for an increase pretation confirmed by the direct assessment of these fac- in NKCC1 transporter activity. tors (Figure 3). WNK3 is a potent activator of NKCC128,29 and they A conserved OXSR1 binding motif present on NKCC117 are colocalized in neurons.28 Furthermore, WNK3 pos- has also been detected on WNK1 and WNK445 and the ac- sesses an OXSR1/STK39 consensus binding motif,45 sug- tivation of NKCC1 by OXSR1 is regulated through up- gesting that an increase in WNK3 expression in schizo- stream interactions with WNK1 and WNK4.23,25-27,46 Con- phrenia could lead to increased NKCC1 transporter sistent with these observations, we found strong activity through OXSR1. Consistent with this hypoth- correlations across all subjects between OXSR1 and WNK1 esis, we found very strong correlations (r=0.744; PϽ10−7) (r=0.478; PϽ.01) and between WNK1 and WNK4 between OXSR1 and WNK3 mRNA levels across all sub- (r=0.516; PϽ.001) mRNA levels. The expression differ- jects as well as within-pair expression differences in layer ences between subjects with schizophrenia and control sub- 3 samples (r=0.92; PϽ.001), suggesting that their ef- jects for WNK1 and WNK4 kinases clustered together, con- fects on chloride transporter activity are synergistic. sistent with the idea that WNK1 and WNK4 regulate each WNK3 can inhibit KCC2 activity regardless of cellu- other’s activity.47 Nevertheless, neither WNK1 nor WNK4 lar tonicity,28,29 suggesting that the elevated WNK3 ex-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 28

GABA GABA

NKCC1 KCC2 GABAA NKCC1 KCC2 GABAA

P P

OXSR1 WNK3 OXSR1 WNK3

Figure 6. Putative interaction model between NKCC1 and KCC2 and OXSR1 and WNK3 isoform 1 (WNK3-1) kinases. In both panels, the orange bar represents the cell membrane, with the extracellular domain above and the intracellular domain below the bar. The size and orientation of the green arrows indicate the

magnitude and direction of chloride ion flow mediated by NKCC1, KCC2, and ␥-aminobutyric acidA (GABAA) receptor chloride channels. A, In normal adult neurons, intracellular chloride concentration is low because of low levels of NKCC1 and high levels of KCC2. The binding of GABA to GABAA receptors triggers chloride entry (chloride flow is represented by green arrows) and hyperpolarization. B, In schizophrenia, increased OXSR1 and WNK3-1 kinase levels lead to increased phosphorylation (black P) and consequently increased NKCC1 activity and decreased KCC2 activity, producing a greater intracellular chloride

concentration. Thus, when GABAA receptors are activated, chloride influx is reduced (or perhaps reversed) and the nature of GABA neurotransmission is altered.

pression in subjects with schizophrenia could be asso- may be consistent with preliminary reports of decreased ciated with lower KCC2 activity. Consistent with this levels of KCC2 mRNA in the hippocampus of patients interpretation, KCC2 and WNK3 expression levels were with schizophrenia.53,54 This combined shift in trans- inversely correlated (r=−0.356; P=.02) across all 84 sub- porter activities would be expected to increase intracel- jects. Additionally, in our data set, both KCC2 and WNK3 lular chloride levels, reducing or reversing the net in- clustered together, suggesting that indeed both KCC2 and ward flow of chloride ions when GABAA receptors are WNK3 belong to the same regulatory pathway. Consis- activated, and thus altering the nature of synaptic GABA tent with this hypothesis, KCC2 and WNK3 share the neurotransmission. same developmental trajectory.28,48 Early in development, the ratio of NKCC1 to KCC2 Because kinase-dead WNK3 activates KCC2,28,29 KCC2 is high, resulting in high intracellular chloride content inhibition triggered by the increase in WNK3 could be and a depolarizing excitatory response on GABAA recep- indirect, through the potential inhibition of a yet to be tor activation. Later, as the expression of NKCC1 de- identified phosphatase or the activation of another ki- clines and KCC2 increases, the level of intracellular chlo- 55 nase. Coimmunoprecipitation studies have shown that ride decreases and GABAA receptor activation triggers WNK3 can interact with WNK4 via their carboxy ter- chloride influx, hyperpolarization, and inhibition. In- mini49 and it has been suggested that WNK4 could act terestingly, an increase in OXSR1 and WNK3 kinase ac- as a regulator of WNK3 activity and therefore KCC2/ tivity could promote a change of transporter activities (ie, NKCC1 activity.50 In our data set, WNK4 gene expres- increased NKCC1 to KCC2 ratio) and a shift in the na- sion did display an upregulated trend (ϩ12.6%; P=.17) ture of GABA neurotransmission to be more similar to but its expression level in all 84 subjects was correlated that observed earlier in development. Specifically, our find- with NKCC1 and not with WNK3. This suggests that ings suggest that schizophrenia is associated with a ten- NKCC1 and KCC2 regulation by WNK3 may happen dency for GABA neurotransmission to be less strongly through different regulatory pathways. Finally, the strong inhibitory in the DLPFC, at least in the neurons that have Pearson correlation between OXSR1 and WNK3 (0.744; elevated levels of OXSR1 and WNK3 expression. PϽ.01) suggests that the regulation of both NKCC1 and A provisional model summarizing these potential in- KCC2 activities could be linked through the interaction teractions between NKCC1 and KCC2 chloride trans- of these 2 kinases. porters and OXSR1 and WNK3 kinases is shown in Figure 6. In normal adult neurons (Figure 6A), NKCC1 SYNERGISTIC EFFECTS OF ALTERED KINASE levels are low and KCC2 levels are high resulting in a low LEVELS ON CHLORIDE TRANSPORT level of intracellular chloride. GABA binding to GABAA IN SCHIZOPHRENIA receptors triggers an inward flow of chloride leading to hyperpolarization. In schizophrenia (Figure 6B), greater The combined and coordinated greater expression of levels of OXSR1 and WNK3 result in greater phosphory- OXSR1 and WNK3 in schizophrenia would be expected lation of both chloride transporters, and a consequent in- to increase both NKCC1 and KCC2 phosphorylation lev- crease in NKCC1 and decrease in KCC2 activities, lead- els. Because NKCC1 and KCC2 are activated through ing to a higher level of intracellular chloride. This change phosphorylation and dephosphorylation, respec- in intracellular chloride concentration has the potential tively,24,51,52 our findings suggest that schizophrenia is as- to reduce the hyperpolarizing effect triggered by GABA sociated with both increased NKCC1 activity and de- binding and therefore to substantially change the na- creased KCC2 activity. Interestingly, the latter finding ture of GABA neurotransmission.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 29

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/23/2021 The affected cortical neurons are likely to include py- Published Online: September 6, 2010. doi:10.1001/ ramidal cells, given the immunocytochemical evidence archgenpsychiatry.2010.114 that OXSR1 is preferentially localized in pyramidal neu- Correspondence: David A. Lewis, MD, Department of Psy- rons in layers 3 and 5 and the correlated expression of chiatry, University of Pittsburgh, 3811 O’Hara St, W1651 OXSR1 and WNK3-1&2 mRNA in layer 3 pyramidal neu- BST, Pittsburgh, PA 15213 ([email protected]). rons. Effective synchronization of networks of pyrami- Author Contributions: Dr Lewis had full access to all of dal neurons at the gamma band frequencies associated the data in the study and takes responsibility for the in- with working memory requires strongly hyperpolariz- tegrity of the data and the accuracy of the data analysis. ing inputs to a network of pyramidal cells so that the post- Financial Disclosure: Dr Lewis currently receives inves- synaptic pyramidal neurons then escape from inhibi- tigator-initiated research support from the Bristol- tion simultaneously and fire subsequent action potentials Myers Squibb Foundation, Bristol-Myers Squibb, Cu- in unison, producing synchronized activity.56-58 Thus, a ridium Ltd, and Pfizer and in 2007 to 2009 served as a shift to increased NKCC1 and decreased KCC2 activity consultant in the areas of target identification and vali- in pyramidal neurons in schizophrenia could alter the dation and new compound development to AstraZen- reversal potential of the GABAA receptor current to be eca, BioLine Rx, Bristol-Myers Squibb, Hoffmann-La closer to the resting membrane potential, reducing the Roche, Eli Lilly, Merck, Neurogen, and SK Life Science. strong inhibition required for gamma oscillations,6 and Funding/Support: This work was supported by grants thus contributing to the alterations in the prefrontal MH-084053 and MH-043784 from the National Insti- gamma oscillations associated with impaired working tutes of Health. memory in schizophrenia.4 Online-Only Material: The eAppendix, eTables, and eFig- ures are available at http://www.archgenpsychiatry OTHER POTENTIAL EFFECTS OF ALTERED .com. OXSR1 AND WNK3 EXPRESSION Additional Contributions: Mary Brady, BA, assisted with IN SCHIZOPHRENIA the graphics and Kelly Rogers, MS, and Anthony Cipri- ano, BS, assisted with the immunocytochemistry runs. The increased expression of OXSR1 and WNK3 could also We express our deep gratitude to the families of the brain contribute to other circuitry disturbances in the DLPFC. donors who made this research possible. For example, vesicular glutamate transport appears to be highly dependent on the chloride content outside of synaptic vesicles.59 Thus, the change in intracellular chlo- REFERENCES ride concentration predicted by our findings could also alter vesicular glutamate transporter activity and con- 1. Elveva˚g B, Goldberg TE. Cognitive impairment in schizophrenia is the core of the disorder. Crit Rev Neurobiol. 2000;14(1):1-21. tribute to alterations in glutamate transmission and syn- 2. Levy R, Goldman-Rakic PS. Segregation of working memory functions within 60 aptic plasticity reported in schizophrenia. the dorsolateral prefrontal cortex. Exp Brain Res. 2000;133(1):23-32. Recent work also indicates that KCC2, through its in- 3. Uhlhaas PJ, Haenschel C, Nikolic´ D, Singer W. The role of oscillations and syn- teraction with the dendritic cytoskeleton, regulates spine chrony in cortical networks and their putative relevance for the pathophysiology maturation.61 Thus, a decrease in KCC2 activity, second- of schizophrenia. Schizophr Bull. 2008;34(5):927-943. 4. Cho RY, Konecky RO, Carter CS. Impairments in frontal cortical gamma syn- ary to increased OXSR1 and WNK3 expression, could con- chrony and cognitive control in schizophrenia. Proc Natl Acad Sci U S A. 2006; tribute to impaired spine maturation and thus to lower 103(52):19878-19883. dendritic spine density in the DLPFC.30 Consistent with 5. Gonzalez-Burgos G, Lewis DA. GABA neurons and the mechanisms of network this interpretation, we observed higher OXSR1 and WNK3 oscillations: implications for understanding cortical dysfunction in schizophrenia. Schizophr Bull. 2008;34(5):944-961. expression in layer 3 pyramidal cells where the spine defi- 6. Bartos M, Vida I, Jonas P. Synaptic mechanisms of synchronized gamma oscil- 62 cit in schizophrenia is most pronounced. lations in inhibitory interneuron networks. Nat Rev Neurosci. 2007;8(1):45-56. 7. Lewis DA, Hashimoto T, Volk DW. Cortical inhibitory neurons and schizophrenia. Nat Rev Neurosci. 2005;6(4):312-324. CONCLUSIONS 8. Lewis DA, Sweet RA. Schizophrenia from a neural circuitry perspective: advanc- ing toward rational pharmacological therapies. J Clin Invest. 2009;119(4):706- In summary, we report substantial increases in the ex- 716. pression of the regulatory kinases OXSR1 and WNK3 in 9. Payne JA, Stevenson TJ, Donaldson LF. Molecular characterization of a putative K-Cl cotransporter in rat brain: a neuronal-specific isoform. J Biol Chem. 1996; the DLPFC that appear to be highly conserved across sub- 271(27):16245-16252. jects with schizophrenia. This combination is likely to 10. Plotkin MD, Kaplan MR, Peterson LN, Gullans SR, Hebert SC, Delpire E. Expres- alter the relative activities of the chloride transporters sion of the Na(ϩ)-K(ϩ)-2Cl- cotransporter BSC2 in the nervous system. Am J NKCC1 and KCC2, producing a higher than normal in- Physiol. 1997;272(1, pt 1):C173-C183. 11. Kanaka C, Ohno K, Okabe A, Kuriyama K, Itoh T, Fukuda A, Sato K. The differ- tracellular chloride concentration, such that activation ential expression patterns of messenger RNAs encoding K-Cl cotransporters of GABAA receptors results in less chloride influx and a (KCC1,2) and Na-K-2Cl cotransporter (NKCC1) in the rat nervous system. smaller increase in hyperpolarization, or depending on Neuroscience. 2001;104(4):933-946. the chloride reversal potential, could even lead to an out- 12. Lauf PK, Adragna NC. K-Cl cotransport: properties and molecular mechanism. flow of chloride ions, and some degree of membrane de- Cell Physiol Biochem. 2000;10(5-6):341-354. 13. Lauf PK, Bauer J, Adragna NC, Fujise H, Zade-Oppen AM, Ryu KH, Delpire E. polarization of the postsynaptic cell. Erythrocyte K-Cl cotransport: properties and regulation. Am J Physiol. 1992; 263(5, pt 1):C917-C932. 14. Flatman PW. Regulation of Na-K-2Cl cotransport by phosphorylation and protein- Submitted for Publication: January 22, 2010; final re- protein interactions. Biochim Biophys Acta. 2002;1566(1-2):140-151. vision received June 29, 2010; accepted July 4, 2010. 15. Payne JA, Rivera C, Voipio J, Kaila K. Cation-chloride co-transporters in neuro-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 30

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/23/2021 nal communication, development and trauma. Trends Neurosci. 2003;26(4): 38. Hashimoto T, Arion D, Unger T, Maldonado-Avile´s JG, Morris HM, Volk DW, Mir- 199-206. nics K, Lewis DA. Alterations in GABA-related transcriptome in the dorsolateral 16. Ushiro H, Tsutsumi T, Suzuki K, Kayahara T, Nakano K. Molecular cloning and prefrontal cortex of subjects with schizophrenia. Mol Psychiatry. 2008;13(2): characterization of a novel Ste20-related protein kinase enriched in neurons and 147-161. transporting epithelia. Arch Biochem Biophys. 1998;355(2):233-240. 39. Eggan SM, Lewis DA. Immunocytochemical distribution of the cannabinoid CB1 17. Piechotta K, Lu J, Delpire E. Cation chloride cotransporters interact with the stress- receptor in the primate neocortex: a regional and laminar analysis. Cereb Cortex. related kinases Ste20-related proline-alanine-rich kinase (SPAK) and oxidative 2007;17(1):175-191. stress response 1 (OSR1). J Biol Chem. 2002;277(52):50812-50819. 40. Chen W, Yazicioglu M, Cobb MH. Characterization of OSR1, a member of the 18. Piechotta K, Garbarini N, England R, Delpire E. Characterization of the interac- mammalian Ste20p/germinal center kinase subfamily. J Biol Chem. 2004;279 tion of the stress kinase SPAK with the Naϩ-Kϩ-2Cl- cotransporter in the ner- (12):11129-11136. vous system: evidence for a scaffolding role of the kinase. J Biol Chem. 2003; 41. Hashimoto T, Nguyen QL, Rotaru D, Keenan T, Arion D, Beneyto M, Gonzalez- 278(52):52848-52856. Burgos G, Lewis DA. Protracted developmental trajectories of GABAA receptor 19. Delpire E, Gagnon KB. SPAK and OSR1: STE20 kinases involved in the regula- alpha1 and alpha2 subunit expression in primate prefrontal cortex. Biol Psychiatry. tion of ion homoeostasis and volume control in mammalian cells. Biochem J. 2009;65(12):1015-1023. 2008;409(2):321-331. 42. Hayes TL, Cameron JL, Fernstrom JD, Lewis DA. A comparative analysis of the 20. Flemmer AW, Gimenez I, Dowd BF, Darman RB, Forbush B. Activation of the Na- distribution of prosomatostatin-derived peptides in human and monkey neocortex. K-Cl cotransporter NKCC1 detected with a phospho-specific antibody. J Biol Chem. J Comp Neurol. 1991;303(4):584-599. 2002;277(40):37551-37558. 43. Arion D, Unger T, Lewis DA, Mirnics K. Molecular markers distinguishing su- 21. Darman RB, Forbush B. A regulatory locus of phosphorylation in the N terminus pragranular and infragranular layers in the human prefrontal cortex. Eur J Neurosci. of the Na-K-Cl cotransporter, NKCC1. J Biol Chem. 2002;277(40):37542-37550. 2007;25(6):1843-1854. 22. Dowd BF, Forbush B. PASK (proline-alanine-rich STE20-related kinase), a regu- 44. Geng Y, Hoke A, Delpire E. The STE20 kinases SPAK and OSR1 regulate NKCC1 latory kinase of the Na-K-Cl cotransporter (NKCC1). J Biol Chem. 2003;278 function in sensory neurons. J Biol Chem. 2009;139(7):1353-1365. (30):27347-27353. 45. Delpire E, Gagnon KB. Genome-wide analysis of SPAK/OSR1 binding motifs. Physiol 23. Vitari AC, Thastrup J, Rafiqi FH, Deak M, Morrice NA, Karlsson HK, Alessi DR. Genomics. 2007;28(2):223-231. Functional interactions of the SPAK/OSR1 kinases with their upstream activator 46. Richardson C, Rafiqi FH, Karlsson HK, Moleleki N, Vandewalle A, Campbell DG, WNK1 and downstream substrate NKCC1. Biochem J. 2006;397(1):223-231. Morrice NA, Alessi DR. Activation of the thiazide-sensitive Naϩ-Cl- cotrans- 24. Gagnon KB, England R, Delpire E. Volume sensitivity of cation-Cl- cotransport- porter by the WNK-regulated kinases SPAK and OSR1. J Cell Sci. 2008;121 ers is modulated by the interaction of two kinases: Ste20-related proline-alanine- (pt 5):675-684. rich kinase and WNK4. Am J Physiol Cell Physiol. 2006;290(1):C134-C142. 47. Yang CL, Zhu X, Wang Z, Subramanya AR, Ellison DH. Mechanisms of WNK1 25. Moriguchi T, Urushiyama S, Hisamoto N, Iemura S, Uchida S, Natsume T, Mat- and WNK4 interaction in the regulation of thiazide-sensitive NaCl cotransport. sumoto K, Shibuya H. WNK1 regulates phosphorylation of cation-chloride- J Clin Invest. 2005;115(5):1379-1387. coupled cotransporters via the STE20-related kinases, SPAK and OSR1. J Biol 48. Lu J, Karadsheh M, Delpire E. Developmental regulation of the neuronal- Chem. 2005;280(52):42685-42693. specific isoform of K-Cl cotransporter KCC2 in postnatal rat brains. J Neurobiol. 26. Anselmo AN, Earnest S, Chen W, Juang YC, Kim SC, Zhao Y, Cobb MH. WNK1 1999;39(4):558-568. and OSR1 regulate the Naϩ,Kϩ, 2Cl- cotransporter in HeLa cells. Proc Natl Acad 49. Yang CL, Zhu X, Ellison DH. The thiazide-sensitive Na-Cl cotransporter is regu- Sci U S A. 2006;103(29):10883-10888. lated by a WNK kinase signaling complex. J Clin Invest. 2007;117(11):3403- 27. Vitari AC, Deak M, Morrice NA, Alessi DR. The WNK1 and WNK4 protein kinases 3411. that are mutated in Gordon’s hypertension syndrome phosphorylate and acti- 50. McCormick JA, Yang CL, Ellison DH. WNK kinases and renal sodium transport vate SPAK and OSR1 protein kinases. Biochem J. 2005;391(pt 1):17-24. in health and disease: an integrated view. Hypertension. 2008;51(3):588-596. 28. Kahle KT, Rinehart J, de Los Heros P, Louvi A, Meade P, Vazquez N, Hebert SC, 51. Kahle KT, Rinehart J, Ring A, Gimenez I, Gamba G, Hebert SC, Lifton RP. WNK Gamba G, Gimenez I, Lifton RP. WNK3 modulates transport of Cl- in and out of protein kinases modulate cellular Cl- flux by altering the phosphorylation state cells: implications for control of cell volume and neuronal excitability. Proc Natl of the Na-K-Cl and K-Cl cotransporters. Physiology (Bethesda). 2006;21:326- Acad Sci U S A. 2005;102(46):16783-16788. 335. 29. de Los Heros P, Kahle KT, Rinehart J, Bobadilla NA, Va´zquez N, San Cristobal P, 52. Russell JM. Sodium-potassium-chloride cotransport. Physiol Rev. 2000;80(1): Mount DB, Lifton RP, Hebert SC, Gamba G. WNK3 bypasses the tonicity require- 211-276. ment for K-Cl cotransporter activation via a phosphatase-dependent pathway. 53. Hyde TM. Cation chloride co-transporters in human brain development and schizo- Proc Natl Acad Sci U S A. 2006;103(6):1976-1981. phrenia: NKCC1 and KCC2 [abstract]. Biol Psychiatry. 2008;63(symposium 7): 30. Glantz LA, Lewis DA. Decreased dendritic spine density on prefrontal cortical py- 47S. ramidal neurons in schizophrenia. Arch Gen Psychiatry. 2000;57(1):65-73. 54. Hyde TM. The cation chloride co-transporters KCC2 and NKCC1 and schizophre- 31. Volk DW, Austin MC, Pierri JN, Sampson AR, Lewis DA. Decreased glutamic acid nia [abstract]. Biol Psychiatry. 2009;64(symposium 8):170S. decarboxylase67 messenger RNA expression in a subset of prefrontal cortical 55. Ben-Ari Y, Khazipov R, Leinekugel X, Caillard O, Gaiarsa JL. GABAA, NMDA and gamma-aminobutyric acid neurons in subjects with schizophrenia. Arch Gen AMPA receptors: a developmentally regulated ‘me´nage à trois’. Trends Neurosci. Psychiatry. 2000;57(3):237-245. 1997;20(11):523-529. 32. Hashimoto T, Bazmi HH, Mirnics K, Wu Q, Sampson AR, Lewis DA. Conserved 56. Wang XJ, Buzsa´ki G. Gamma oscillation by synaptic inhibition in a hippocampal regional patterns of GABA-related transcript expression in the neocortex of sub- interneuronal network model. J Neurosci. 1996;16(20):6402-6413. jects with schizophrenia. Am J Psychiatry. 2008;165(4):479-489. 57. Mann EO, Paulsen O. Role of GABAergic inhibition in hippocampal network 33. Mimmack ML, Brooking J, Bahn S. Quantitative polymerase chain reaction: vali- oscillations. Trends Neurosci. 2007;30(7):343-349. dation of microarray results from postmortem brain studies. Biol Psychiatry. 2004; 58. Mann EO, Kohl MM, Paulsen O. Distinct roles of GABA(A) and GABA(B) recep- 55(4):337-345. tors in balancing and terminating persistent cortical activity. J Neurosci. 2009; 34. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, Spele- 29(23):7513-7518. man F. Accurate normalization of real-time quantitative RT-PCR data by geo- 59. Schenck S, Wojcik SM, Brose N, Takamori S. A chloride conductance in VGLUT1 metric averaging of multiple internal control genes. Genome Biol. 2002;3(7): underlies maximal glutamate loading into synaptic vesicles. Nat Neurosci. 2009; H0034. 12(2):156-162. 35. Holden S, Cox J, Raymond FL. Cloning, genomic organization, alternative splic- 60. Stone JM, Morrison PD, Pilowsky LS. Glutamate and dopamine dysregulation ing and expression analysis of the human gene WNK3 (PRKWNK3). Gene. 2004; in schizophrenia—a synthesis and selective review. J Psychopharmacol. 2007; 335:109-119. 21(4):440-452. 36. Lepre J, Rice JJ, Tu Y, Stolovitzky G. Genes@Work: an efficient algorithm for 61. Li H, Khirug S, Cai C, Ludwig A, Blaesse P, Kolikova J, Afzalov R, Coleman SK, pattern discovery and multivariate feature selection in gene expression data. Lauri S, Airaksinen MS, Keinänen K, Khiroug L, Saarma M, Kaila K, Rivera C. Bioinformatics. 2004;20(7):1033-1044. KCC2 interacts with the dendritic cytoskeleton to promote spine development. 37. Dorph-Petersen KA, Pierri JN, Perel JM, Sun Z, Sampson AR, Lewis DA. The in- Neuron. 2007;56(6):1019-1033. fluence of chronic exposure to antipsychotic medications on brain size before 62. Kolluri N, Sun Z, Sampson AR, Lewis DA. Lamina-specific reductions in den- and after tissue fixation: a comparison of haloperidol and olanzapine in ma- dritic spine density in the prefrontal cortex of subjects with schizophrenia. Am J caque monkeys. Neuropsychopharmacology. 2005;30(9):1649-1661. Psychiatry. 2005;162(6):1200-1202.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 68 (NO. 1), JAN 2011 WWW.ARCHGENPSYCHIATRY.COM 31