Circuitry-based expression profiles in GABA cells of the trisynaptic pathway in schizophrenics versus bipolars

Francine M. Benesa,b,c,1, Benjamin Lima, David Matzilevicha,c, Sivan Subburajua,c, and John P. Walsha

aProgram in Structural and Molecular Neuroscience, McLean Hospital, Belmont, MA 02478; and bProgram in Neuroscience and cDepartment of Psychiatry, Harvard Medical School, Boston, MA 02215

Communicated by Erminio Costa, University of Illinois, Chicago, IL, October 15, 2008 (received for review April 15, 2008) Significant reductions in GABAergic cell numbers and/or activity have tional microscopic criteria (7), the interneurons in these layers been demonstrated in the of subjects with schizophre- showed a distinctive cresyl violet staining of their cytoplasm. nia and bipolar disorder. To understand how different subpopula- Glial cells, however, typically showed no such staining, appearing tions of interneurons are regulated, laser microdissection and gene as isolated nuclei suspended within the neuropil. This funda- expression profiling have been used to ‘‘deconstruct’’ the trisynaptic mental difference between the two cell types suggests that the pathway, so that subtypes of GABA cells could be defined by their concentrations of RNA in interneurons of SO and SR is much location in various layers of CA3/2 and CA1. The results suggest that higher than in glial cells. Additionally, preliminary assessments, the cellular endophenotypes for SZ and BD may be determined by using in situ hybridization demonstrated that antisense RNA multiple factors that include unique susceptibility for the associated with a broad array of genes (i.e., GAD67, GAD65, respective disorders and altered integration among hippocampal GluR5, GluR6, KCNJ3, KCJN6, HCN3, HCN4, HDAC1, GABA cells with extrinsic and intrinsic afferent fiber systems. The DAXX, PAX5, Runx2) was found over neuronal cell bodies, extensive and intricate data that has come from this study has whereas no grains were observed over glial cell profiles. provided insights into how a complex circuit, like the trisynaptic As shown in Table S1, the BDs showed different ratios for pathway, may be regulated in human hippocampus in both health laterality (R/L) and gender (M/F) when compared with the CON and disease. and SZ groups that were well-matched with one another. An earlier study from our group did not find that laterality or gender influ- ͉ ͉ ͉ ͉ GAD67 potassium ion transport synaptic transmission kainate enced the nature of changes (8). Additionally, the nicotinic expression of mRNA for BDNF and trkB also shows no relationship to these two variables (9). Although an influence of these potential ene expression plays a central role in the regulation of confounds on gene expression cannot be definitively ruled out, it Gneural circuitry involved in cognitive behavior. Identifying does seems unlikely that these variables could account for the molecular mechanisms within neurons of complex circuits pre- unusually robust differences noted in the BD group. sents one of the foremost challenges to understanding the human The quality of amplified antisense RNA used was evaluated using brain. In the past 20 years, postmortem studies of 18S/28S ratios, 3Ј/5Ј ratios and the percentage of present calls (SZ) and bipolar disorder (BD) have provided evidence for a generated with dChip 1.0 (see ref. 6). The percentage of present dysfunction of GABAergic neurons in frontal cortices and calls showed good correspondence between the SZ and BD groups. hippocampus (1). It is well known that GABAergic interneurons When the changes in expression between the groups were com- provide potent inhibitory modulation of principle neurons (2) pared, the number of genes satisfying the inclusionary criterion of and are critical for the regulation of feed-forward inhibition (3) P Յ 0.05 (CON vs. SZ or BD) for the post hoc analysis of GenMapp and oscillatory rhythms (4, 5). A network of genes involved in the biopathways and/or clusters resulted in large numbers being de- regulation of glutamate decarboxylase 67 (GAD67), a key marker tected in SO of both CA1 and CA3/2 of the BD and SZ groups for the GABA cell phenotype (6), shows changes in expression (Table 1). In the SR and SP of both sectors, very few genes satisfied in SZ that are different from those seen in BD, suggesting that the inclusionary criterion in both groups, particularly the BDs. The there may be unique molecular endophenotypes for each disor- number of genes with P Յ 0.05 level did not covary with the der. To learn more about the molecular regulation of hippocam- percentage of present calls (Table 1). For example, in the SP of pal GABA cells in SZ and BD, a combination of laser micro- CA3/2 of BDs and SZs the number of genes at the P Յ 0.05 level dissection (LMD) and gene expression profiling has been used was 181 and 244, respectively, whereas the percentage of present to ‘‘deconstruct’’ the trisynaptic pathway into subtypes of GABA calls in the same samples of all three groups ranged from 30–36%. neurons defined by their location and connectivity. Several clusters of genes have been examined across a broad array of QRT-PCR Validation. As shown in Fig. S2 and predicted from the cellular functions that include transduction, signaling, metabo- microarray data, the KCNJ3 gene (G dependent inwardly lism, translation, transcription and cell cycle regulation. These rectifying potassium channel; GIRK) (10) and the hyperpolariza- clusters have been separately analyzed in various layers and tion-activated Ih (HCN3 and 4) (11) channels showed significantly sectors with a preponderance of GABA cells. To our knowledge, this is the first demonstration that the regulation of gene expression in GABA cells varies not only according to diagnosis, Author contributions: F.M.B. designed research; F.M.B., D.M., and J.P.W. performed re- but also to location within a complex circuit. search; F.M.B. contributed new reagents/analytic tools; F.M.B., B.L., and J.P.W. analyzed data; and F.M.B. and S.S. wrote the paper. Results The authors declare no conflict of interest. Samples of the stratum oriens (SO), pyramidale (SP), and Freely available online through the PNAS open access option. radiatum (SR) of sectors CA3/2 and CA1 (Fig. S1) were obtained 1To whom correspondence should be addressed. E-mail: [email protected]. for the normal control (CON), SZ and BD groups (Table S1). This article contains supporting information online at www.pnas.org/cgi/content/full/

The SO and SR samples were collected by LMD from a location 0810153105/DCSupplemental. NEUROSCIENCE where there were very few glial cells. Consistent with conven- © 2008 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0810153105 PNAS ͉ December 30, 2008 ͉ vol. 105 ͉ no. 52 ͉ 20935–20940 Downloaded by guest on October 1, 2021 Table 1. Numbers of genes showing significant changes and genes with predominantly decreased expression in BDs (blue). In percentages of present calls SZs, however, the genes in the same clusters were both increased Number of genes Percentage of and decreased. Group Sector Layer (P Յ 0.05) present calls Mapping Genes Associated with Synaptic Transmission and Potassium BD CA1 Oriens 4,062.0 42.3 Ion Transport to the Trisynaptic Pathway. To evaluate the potential SZ 1,028.0 33.0 significance of these various changes for the activity of the trisyn- CON 30.5 aptic pathway, the gene expression profile changes for various functional categories were mapped onto schematic diagrams of BD CA3/2 Oriens 2,979.0 36.4 GABA cells representing the various layers and sectors studied. To SZ 1,875.0 27.3 focus primarily on GABA cells, the layers reported herein are those CON 32.4 in which interneurons are the predominant neuronal cell type and include the SO and SR of CA3/2 and CA1. SP is not reported below BD CA1 Radiatum 173.0 34.5 because it contains both pyramidal cells and GABA cells in a ratio SZ 539.0 33.1 of Ϸ10:1. If GAD67 expression was significantly decreased, partic- CON 35.0 ularly if other signaling and/or metabolic pathways also showed an overall decrease in expression, a cell was given a blue background BD CA3/2 Radiatum 339.0 27.8 coloration (Fig. S4). If, however, GAD67 expression was unchanged SZ 15.0 21.9 and/or other signaling and metabolic pathways were markedly CON 22.6 up-regulated, then the cell was given a red background to indicate that its neuronal activity may be significantly increased. Because of BD CA1 Pyramidale 220.0 35.7 the importance of synaptic transmission for the up- and down- SZ 133.0 30.5 stream regulation of the trisynaptic pathway, this functional cluster CON 30.3 is given primary emphasis. Some of the genes related to potassium ion transport were differentially regulated in SZs and BDs and, BD CA3/2 Pyramidale 181.0 30.0 because of their importance for understanding changes in synaptic SZ 244.0 36.2 mechanisms, are also included in the results described below: CON 34.5 SO of CA3/2. Schizophrenia. As shown in Fig. S4A for genes related to The number of genes with P Յ 0.05 refers to comparison of bipolars (BDs), and schizophrenics (SZs) with normal controls (CONs) using a 2-way ANOVA. synaptic transmission, GAD67 (GAD1) and GAD65 (GAD2) ex- The percent of present calls were obtained with dCHIP 1.3. pression showed significant decreases in expression in the SO of CA3/2. There was also an increase in the rho1 subunit of the GABAA receptor (GABRR1). The rho1 subunit has been associ- increased expression in SZs, but decreased expression in BDs. The ated with GABA-to-GABA ionotropic inputs to cerebellar Pur- GRIA1 (ionotropic AMPA1 receptor) gene, that is thought to be kinje cells (14) and this finding is consistent with a previously a susceptibility gene for schizophrenia (12), showed significant reported increase of specific GABAA binding activity on interneu- negative fold changes in this group, but not in BDs. These various rons in CA3/2 in SZs (15). The expression of another gene involved changes all occurred in the direction observed with microarray- in the degradation of GABA, ALDH5A1 gene, i.e., succinate based gene expression profiling. semialdehyde dehydrogenase (16), was also significantly decreased at this and could potentially be associated with a compensa- Analyses of Functional Biopathways and Clusters. The composite tory increase of GABA concentrations. Several other genes in the synaptic transmission cluster also showed significant changes in SZs probability, Pc, was calculated (13) for 48 different functional gene clusters, including those associated with synaptic function, signaling, (refer to Fig. S3A). Those genes that showed decreased expression ϭϪ ϭ metabolism, transcription/translation and other miscellaneous cat- included the GRIA1 [fold change (FC) 1.77, P 0.008], GRIA3 (FC ϭϪ1.47, P ϭ 0.05), GRIK1 (the egories (Table S2). For each biopathway/cluster in SO of CA3/2 for ϭϪ ϭ the CONs vs. BDs, the number of genes meeting the P Յ 0.05 subunit, GluR5; FC 1.35, P 0.05) and the metabotropic (mGluR) subunits GRM3 (FC ϭϪ1.91, P ϭ criterion (Ni), the total number of genes in the cluster (Nt) and the ϭϪ ϭ overall percentage change were calculated for each layer and sector 0.001) and GRM5 (FC 2.02, P 0.004). In contrast, some glutamate receptor subunits were up-regulated, including the of both groups. The most significant P values were observed in the c GRIK2 (GluR6; FC ϭ 1.5, P ϭ 0.04), GRIK 3 (GluR7; FC ϭ 1.6, SO of CA3/2 and CA1 of the BD and SZ groups (Table S2). In the P ϭ 0.01) and GRIN2A (NR2A; FC ϭ 1.46, P ϭ 0.04). The SP and SR of both sectors, very few biopathways/clusters had a Pc Յ Ϫ10 synaptosomal 25-kDa protein (SNAP25) showed a very robust 10 . Some examples of gene clusters that showed very prom- decrease in expression (FC ϭϪ2.24, P ϭ 0.006). For the voltage- inent changes in SO of CA3/2 and CA1 in BDs are neurogenesis (Pc gated potassium ion transport cluster (Fig. S3B), the G protein ϭ Ϫ109 Ϫ117 ϭ Ϫ81 10 and 10 , respectively), cell cycle regulation (Pc 10 dependent, inwardly rectifying potassium (KCNJ3; FC ϭ 1.51 P ϭ Ϫ92 ϭ Ϫ87 and 10 , respectively) and synaptic transmission (Pc 10 and 0.015) and hyperpolarization-activated cationic (HCN3; FC ϭ 1.68; Ϫ91 10 , respectively). In the SZs, similar clusters showed robust P ϭ 0.001) channels were also significantly up-regulated. changes, except that the magnitude of the Pc values was not quite Bipolar disorder. As shown in Fig. S3A and S4C, BDs, like the SZs, as robust as those seen in BDs, generally because the number of also showed a significant decrease in the expression of GAD67 and genes with P Յ 0.05 was smaller (Table S2). The Pc values for Ϫ Ϫ GAD65 in SO of CA3/2. For the receptors, neurogenesis (10 77 and 10 36, respectively), cell cycle regulation glutamate and dopamine subtypes generally showed changes. Al- Ϫ Ϫ Ϫ (10 67 and 10 22, respectively) and for synaptic transmission (10 54 though the expression of the GRIK1 kainate receptor subunit Ϫ and 10 22, respectively) generally showed a smaller negative expo- (GluR5) was also significantly decreased, there were no other nent. The genes associated with the same biopathway/cluster of the changes in AMPA or NMDA receptors. In contrast, the expression same layer/sector in the SZs vs. BDs were typically quite different of muscarinic (CHRM5; FC ϭϪ1.88, P ϭ 0.02) and nicotinic in terms of the specific genes and their direction of change (Fig. S3). cholinergic receptor polypeptides alpha 3 (CHRNA3; FC ϭϪ1.34, For example, in the SO of CA3/2, the synaptic transmission and P ϭ 0.03), alpha 7 (CHRNA7; FC ϭϪ1.58, P ϭ 0.04), delta potassium ion transport clusters, respectively, contained many (CHRND; FC ϭϪ1.23, P ϭ 0.04) and epsilon (CHRNE; FC ϭ

20936 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0810153105 Benes et al. Downloaded by guest on October 1, 2021 Ϫ1.51, P ϭ 0.03) were all decreased in SO of CA3/2 in BDs. As in were analyzed according to low versus high dose treatment with SZs, mRNA for the synaptic protein SNAP25 (FC ϭϪ2.98, P ϭ antipsychotic medications (APDs). These evaluations were per- 0.002) was also decreased; in BDs expression of synapsin II (FC ϭ formed for all sectors and layers reported in this study; however, Ϫ1.56, P ϭ 0.03) was also reduced. With one exception, the only those for sector CA3/2 are described below (Fig. S8). The expression of all genes associated with potassium ion transport was expression profiles on high dose neuroleptics (i.e., those receiving decreased in the BDs, including KCNJ6 or GIRK (FC ϭϪ2.31, P ϭ 500 CPZ mg-equivalents/day or higher; 4 cases; average ϭ 775 mg) 0.003) and HCN4 or Ih (FC ϭϪ1.38, P ϭ 0.04) (see Fig. S3B). were compared with those receiving a low dose (i.e., Ͻ500 CPZ mg-equivalents per day; 3 cases; average ϭ 312 mg-equivalents per SO of CA1. Schizophrenia. The expression of GAD 65 showed a day). The number of genes showing significant changes in the high significant decrease (6) (Figs. S4B and S5). Other genes related to dose group was 37.5% lower (n ϭ 35) when compared with the low synaptic transmission, however, did not show prominent changes in dose group (n ϭ 56 highlighted in red and blue). A similar pattern this group (Fig. S5 Left). Additionally, mRNA for the NR1 subunit was observed in other functional clusters, such as those related to (GRIN1) of the NMDA receptor was significantly decreased, as cell cycle regulation and neurogenesis (data not shown) where the were those for two subunits of the nicotinic cholinergic receptor differences were 36 and 38%, respectively, suggesting that APDs (CHRND and CHRNA2), GRM4 or the mGluR receptor and may have a generalized tendency to suppress the expression of norepinephrine transporter (SLC6A2). Unlike SO in sector CA3/2, genes representing a wide variety of cell functions that were both potassium ion transport in the KCNJ and HCN classes did not show increased or decreased. Some genes showed a significant change in significant changes in SZs at this locus. However, other genes in this the low dose group, but not the high dose group (n ϭ 40; 58%). A cluster, including several voltage-gated potassium channels (e.g., smaller proportion of genes showed significant changes only in the KCNS3, KCND1, KCNE1, KCNA2 and KCNAB3) showed sig- high dose group (n ϭ 16; 23%). For example, SNAP-25 mRNA was nificant decreases in expression (Fig. S6). significantly decreased in SZs receiving high dose APDs, but not in Bipolar disorder. GAD67 and GAD65 showed no expression changes those receiving low dose treatment. In contrast, one rodent study (6), although the overall pattern of expression among the various reported that chronic APD administration results in an increase of functional clusters of genes was otherwise uniquely up-regulated in SNAP-25 expression in the CA3 region (17). Other genes, however, this layer/sector (Fig. S5 Right). The GRIA 1,2 and 3 receptor genes may not be influenced by APD exposure, because similar changes (AMPA1,2 and 3) all showed 2-fold increases of expression in SO were observed in both the low and high dose groups (n ϭ 12; 17%). of CA1; the FC for GRIA 1 was 2.10 (P ϭ 0.04). Other genes These genes included, but were not limited to GAD 65, GAD 67, showing an up-regulation at this locus included the glutamate GRM3, GRM5 and SLC1A1 (a neuronal high affinity glutamate (SLC1A3) and GABA (SLC6A1) transporters and transporter). ␤ subunit. Genes that showed decreased expression notably in- Only 3 subjects (two SZs and one BD) were treated with APDs cluded several nicotinic cholinergic receptor subunits (CHRNA4, alone, but no other mood stabilizing agents. All of the remaining 11 CHRNB3, CHRNA2, CHRNA1, and CHRNG). Additionally, the SZ and BD cases received APDs in combination with lithium ␣ GluR5 subunit of the kainate receptor (GRIK 1), the glycine 2 carbonate and/or other mood stabilizing agents, suggesting that the (GLRA2), mGLUR4, NR1 NMDA subunit, dopamine D1 differences in gene expression vs. CONs observed in the respective (DRD1), and D3 (DRD3) receptors and norepinephrine trans- groups were probably not related to medication effects. porter (SLC6A2 and -3) all showed decreased expression. For potassium ion transport in the SO of CA1 in BDs (Fig. S6), KCNJ6 Discussion and HCN4 showed no changes, except for one inwardly rectifying channel (KCNJ15) that was significantly decreased. Several other The findings reported herein demonstrate that GABAergic voltage-gated potassium channels showed increased (KCNA5 and interneurons at specific loci along the trisynaptic pathway show KCNA3) or decreased (KCNS3, KCND1, KCNAB1 and -3, unique expression profiles that vary according to layer, sector KCNE1, and KCNA2) expression. and diagnosis (Fig. 1A and 1B). As reported in ref. 6, both the Many genes associated with metabolic pathways were strikingly SZs and BDs showed significant decreases in GAD67 expression up-regulated in the SO of CA1 of BDs (Fig. S4D). As shown in Fig. in layers containing a preponderance of GABA cells, particularly S7, notable examples included glycolysis and gluconeogenesis, those located in sector CA3/2. Although the number of cases in Krebs tricarboxylic acid cycle and the electron transport chain each group was small, the fold changes and probabilities that where most genes showed robust increases in expression. Genes were detected were quite robust, indicating that the variance in associated with translation, ribosomal regulation, t-RNA synthesis, the data for each group was low. As shown in Fig. 1, changes in proteosomal degradation, TGF-␤ and Wnt signaling and cell cycle expression profiles noted in these cells of SZs and BDs may regulation also showed a preponderance of up-regulated genes in potentially limit or enhance their ability to provide inhibitory the SO of CA1 in BDs (data not shown). modulation to pyramidal neurons. However, substantial differ- ences in expression were also found for many functional gene SR of CA3/2 and CA1. Schizophrenia. The expression of GAD67 was clusters (Fig. 1), suggesting that a reduced expression of GAD67 significantly decreased in SZ of CA3/2 (6); however, none of the in SO of CA3/2 occurs within the context of complex molecular functional gene clusters shown showed any significant expression changes that are unique to the two disorders. Additionally, the changes in this group (Table S2). Very few functional gene clusters findings presented here suggest that GABA cell regulation may attained significance in the SR; however, the Pc for calcium channel be related to a complex interaction of layer and sector with regulation, voltage-gated ion channels, electron transport and neu- psychiatric diagnosis and exposure to psychotropic medications. rogenesis exceeded 10Ϫ10. Within the synaptic transmission cluster (data not shown), a small number of genes, including SNAP25, the Medication Effects. Psychotropic effects may potentially increase the potassium voltage-gated KCNQ2 channel, 4-aminobutyrate ami- intersubject variability of gene expression profiling within the BD notransferase (ABAT) and dynamin I were all up-regulated. and/or SZ groups. We noted prominent qualitative differences Bipolar disorder. GAD67 showed no change in expression in SR of between these two groups, but it is difficult to specifically relate CA3/2 of BDs. In CA1, none of the functional gene clusters at this them to one or another class of drugs to which they were exposed locus in BDs showed significant differences (Table S2). (refer to Fig. S9). For example, APDs and mood stabilizers did not appear to influence GAD67 expression. In an earlier study, GAD65-

Psychotropic Medications. To evaluate the effects of medications, containing terminals showed APD dose-related increases, suggest- NEUROSCIENCE the GenMapp clusters/pathways related to synaptic transmission ing that these drugs may contribute to compensatory sprouting of

Benes et al. PNAS ͉ December 30, 2008 ͉ vol. 105 ͉ no. 52 ͉ 20937 Downloaded by guest on October 1, 2021 A ainerhpozihcS Pareltnorf xC lanihrotnE xC Pre forantPaht C12/3A AC AlymA ym ggaad la MS

Vageoltlo t -GteCa dh g G1AD e GAAG DD11 Vageolttlo -GteCa ddh h R g AMPAR NMDAR 2 AMPAR NMDAR e GBAAGABA DgrdeeD gra d + + + AMPAR NMDAR AMPAR K RgeReg GBAA Dgrde a K Chanlsn e RAMPAR NMDAR Graun le eC ll Leayr a GBAA Degra d K Chanlen s 2 C +

GPCRs a Glyolsyisyols c yis. Glyols c yis. C GPCRs CT A Ribmo soosese GPCRs TraTranslaattino TCCT A RRiisbbmo sosomes

G Prot Sig Translatio n G Prot ETC G Prot DAN aD mage ETC KainateR APOPOPP R espoesponnssee KainateR DAN aD mage KainateR KainateR RS Sig

Sig RS AAPOPPOP R Respespoose nse MAPK Sig MAPK Pot.r D eg MAPK Pot.r D eg OSx..x tsser TFG β WtnWnt OSx..x tTrGFse s β Wtn Sig NneN rue goru go enelsisislslC eC Cyylcelc e ueN rogeeeg nsen sisis Clelle CcyyC cle Sig C AhRc GBAAGABA-RRA D A-R 5 HTTH5 -R a AhRc GBAGABA-RA D -R5R 5 HT-R ll C c MFosisy besr i i a u ll m c VVo ii u o C m ll h tt h V a o C g R ll e e tt h e a -- g G g R G e e e a A - g tt G e MMP d a A P Glyolsc yis. tt C e MPM A d h R P Glyols c yis. C A N TCA h R K Glyolsc yis. MD TCA RiRibsob soso mmo es + Ribmo sose N C K M h A + M Translatin o Translatio n a R D TCA Risbeo osm C n h A n K a R Translati no e nan ETCCE T ll a ECT s n ii K n e a a ll ETC G tt s ii n DAN aD mage e PS n DAN aD mage P PS R a C G tt APOP Response e R P R DAN aD mage s DAN aD mage C R APOP Response G s APOPRespose n P G r Pot.r D eg o P Por t.e D g t r S o i tt OS.x tsrTse GFβ Wtn g S OS.x tsrTes GFβ Wtn ii Pot.r D eg g M A Nngorue eiess eC llCylce P M Neurogenesis ClCyle lc e K OS.x tsrTse GFβ Wtn A P S i K GBAA -RAchRAch-RARN D A-RTR5 H5 HT-R g S GBAA -R Ahc - NR AD -R5 HT-R gorueN eines sl eC lCycel ii g GBGBAA A-R AhAhRc R AD -RR5 5 HT-R

VageVagtlo e-GteCa dh . GD1A G E AMPAR + RAMPAR GBAA Dgrde a K+CannelCannh el 2 GAD1,2 . + GD1,2A Vagetlo -GteCa dh g a e reffahcS slaretalloC C R AMPAR + AMPAR GBAAAG BA DgrdDge rad K ChaChanln een + 2 Glyolsyis c a C Ribsoso me NMDAR GycGlycoysisl NMDAR TAC Translatin o Ribsoso me GPCRs OS TAC Traationsn l NMDAR NMDAR ETC GPCRs KainateR DAN aD mage KainateR

ETC DADADN Daamagemage APOP Rnpse oes Sig Prot G RnseRne opse ops se KainateR KainateR OS.x tsser ProD t.eg AOPP Sig Prot G OS.x tsser ProD t.eg TFG β Wtn OS.x tsser Pot.r D eg Neoru gein se s ClCle ylce TFTGFβ Wtn AhRc GBAA -R AD 3 5TH -R Smbu iibu uluc Nnue or ggorue eeiessise s ClClCelle Cyylcel AhRcAchR GBA A-R AD -R5RT HTH5 -R Almyym gaad hrotnE ni lanihrotnE la CxCx Parelfro tnorf tn CxCx

SptalNe uiclelc e AHC GBAA GmG tul atmae FdeFoder Fdee deeF ooF F wwardard E xorawr cnEiEati tiitati icx tonata it

ix Basolaa ret l A mygdala rn o F Asuevl

B iB p ralo rosiD d re Parelxorf nt C C12/3A AC lanihrotnE xC Prforne atPtah Alym gaad MS

Graneu l leC lLeray Vagelo tlo t -GteCa dh g GD1A e AMPAR AMPAR VageVoltlo tage-GteCGatedh C g R GD1A e GBGBAA A Dgrde a K+ Chanlsn e 2 K Chanlsn e + + R

2 AMPAR NMDAR

K Rge NMDAR AMPAR a GBAA DgrdDgre ad + CGPCRs Glyolsyolsyis c yis. a

Ribmos ose Glyyols colyiss . CGPCRs NMDAR TAC Traationsn l NMDAR TAC RRiisbmo soosese

G Prot Sig Prot G TranslaTranslattino ETC Sig Prot G DAN aD mageage KainateR AAPOPPOP RRnseespoopse nse KainateR ETC DAN aD mage DAN aD mage KainateR KainateR Pot.r D eg APOPAPOP Responseesponse RS MAPK Sig RS MAPK Sig

β C Pot.Por t. D eg OS.x tsrTeses GF Wtn a ll β eN u gor einess Celcll yC e c OSO .x StstrTese ss GF WtWnt Nneur go e se is Celclel C y ii MFosiys besr u C m Nne gorue goru eie elesis slC CeC le ylcec e a AhRccA hR GBAAAG BA-R D A-RT 5 H5H -R V ll c o C AhRc GBAGABA-RA D -RT 5H5 H -R ii ll u t h m a m R V g e o C e - g ll - t h G a a g R t A e t e e M -- g d G P C a A A tt h e M R d P N C A K Glyols c yis. M h R + D C D A Glcyolsyis. N h K a R M TCA Ri ob sosem + n C D n C Translatio n n K h A h a TCA Risbeo osm lele a R s ii Translation n n n a K ETC G tt e a lele P R s ii ECTTE C n C SPSP a DAN aD mage R G tt e s P R AOPP Rnseopse C G DAN Da mage R P s APAOPPOp P seR oens r Pot.r D eg o G tt S P ii r g o β t OS.x tsrTse GF Wtn Pot.r Deg S M i g A Neu or g ne essi ClCle ycel P K OSx. tsr se T GFβ nW t M A S i P g Nne goru eiess eC llCycel K GBAGABA-RAch-RA D -R5 HT-R S i GBA A-RAch-RA D -R5RT HTH5 -R g

GD1AGAD1 Vagetlo -GteCa dhd h g e AMPAR + RAMPAR Vao tltlo ge-GteCGta edh C GBAAGABA DgDgrde rda K+ Chanlsn e 2 GD1,2GDA 1,2 g + e a AMPAR NMDAR AMPAR NMDAR + R CGPCRs K+ ChaChanlsnln e s 2 Glyols c yis GBAA DgrdeDegrda + reffahcS slaretalloC a Traationsn l CGPCRs Gl cy olyis s. NMDAR TAC Ribsoso me NMDAR Ribsoso me G Prot CT A Traaaationnssn l tion ETC G Prot G Prot Sig KainateR AOPAOP P DAN D aDa mageage KainateR ETCECTD DNA Daaamam ge OS

KainateR Rnseopse Sig KainateR Rnseopse Sig RnsRnsepse oopse e xO StStsser s Pot.Por t.e D egMAPK Sig PA PO

PA PO MAPK Sig Pot.r D eg TFG β Wtn OS.x tsser Neuenueor g neiss Cely l C cle TFG β Wtn AccA hNhNR GBAAGABA-R AD 3,1 5T5HT A2 NnueN r gogorue einels sslC C e l Cylce Su ib cul mu AhNRc GBAA -R AD -R 5TH -R ymA galad Exlanihrotn C Parelxtnorf C

SptSptalNe al N uielc AHC GBGBAA A Gmtul atae Fodee F rndraw itaticxE o ix rn o F Asuevl Boaas latAre lm ygdala

Fig. 1. Schematic diagrams of the trisynaptic pathway showing pyramidal neurons (triangular) and GABA cells (square) in stratum oriens (SO), stratum pyramidal (SP), and stratum radiatum (SR) of sectors CA3/2 and CA1. The perforant path projection from the entorhinal cortex project to the granule cells and these in turn send mossy fiber projections that synapse on the apical dendrites of pyramidal neurons in sector CA3/2. The latter cell sends projects axons into the SO where they travel as Schaffer collaterals that eventually form excitatory synapse with the apical dendrites of pyramidal neurons in sector CA1. Collateral branches of pyramidal neurons in CA3/2 and CA1 form synapses with GABAergic interneurons on the SO. The arrows at the bottom of each diagram show the direction of feed-forward excitation along the trisynaptic pathway and suggest that this may be increased in schizophrenics because of diminished GABAergic tone in both CA3/2 and CA1. In BDs, however, feed-forward excitation may be attenuated at the level of CA1 because of the heightened activity of GABA cells with in SO at this locus.

GABAergic terminals, particularly in the SO of CA3/2 (18). either SZs or BDs, suggesting that APD exposure might have Consistent with this, rats receiving chronic administrations of increased its expression to normal levels. Most of the SZs and BDs haloperidol showed marked increases in the number of GABA- included in this study were treated with both APDs and mood containing terminals forming axosomatic contacts with pyramidal stabilizers, making the interpretation of the complex expression neurons in the medial prefrontal cortex (19). Moreover, in the patterns difficult. Gene expression profiles in the two groups were current report, mRNA for GAD65 showed no significant changes in fundamentally different, despite overlapping treatments, suggesting

20938 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0810153105 Benes et al. Downloaded by guest on October 1, 2021 that most of these changes are not related to psychotropic treatment it is not surprising that abnormalities in gamma oscillations have alone. Finally, the decrease in the expression of various nicotinic been reported in schizophrenia (46). receptor subunits raises the possibility that smoking cigarettes In BD, a prominent decreased in the expression of nicotinic might have contributed to these changes in SZs (20) and BDs (21). cholinergic receptor subunits was the most prominent change Because BDs do not smoke as heavily as SZs (clinical observation), observed in SO of CA3/2 and CA1. These findings are consistent they might be expected to show less striking changes in nicotinic with an earlier postmortem study reporting that this receptor receptor expression; however, this group showed more subunits activity was reduced in subjects with psychotic disorders (47). A with decreased expression when compared with SZs, suggesting decrease of excitatory cholinergic activity impinging on interneu- that ingestion alone does not account for the expression rons could contribute to GABAergic dysfunction in SO of CA3/2 changes noted for these receptors. and CA1 (48). In BDs, GABA cells within the SO of CA3/2 might be particularly compromised in their ability to provide inhibitory Functional Implications of Gene Expression Changes. An important modulation, because the kainate receptor subunits and Ih channels factor that probably influenced gene expression changes is the also show significantly decreased expression, changes that could unique connectivity found within each layer and sector of the promote a diminished firing of GABA cells (33, 34). In CA1 of BDs, trisynaptic pathway (Fig. 1). The mossy fiber system provides the expression of kainate receptor subunits is also decreased; afferent inputs to pyramidal neurons and GABA cells in the SR of however, dramatic increases in the expression of genes associated CA3 (22); kainate (23) and AMPA receptors both contribute to with a broad range of metabolic signaling, transcriptional and feed-forward inhibition in this circuit (24–26). No changes in the translational clusters are present at this locus in BDs. Together with expression of any of the glutamate receptors were observed in the a normal expression of GAD67 the GABA cells found therein may SR of CA3/2 in SZs or BDs. Pyramidal neurons receiving mossy be hyperactive and capable of exerting higher than normal levels of fiber inputs provide excitatory recurrent collaterals to GABAergic inhibitory modulation on the pyramidal neurons in SO of CA1. This cells; these are mediated by AMPA receptors and generate NMDA- locus represents the final common pathway for the output of mediated LTP responses (27). In SR of CA3/2, AMPA and NMDA excitatory activity from the trisynaptic pathway to its many termi- receptor subunits showed no expression changes in either group, nation sites and its modulation by GABA cells has important suggesting that mossy fiber activity may not be altered in SZ or BD. implications for functional integration within the hippocampus. In addition to SR, GABAergic neurons provide inhibitory mod- The model shown in Fig. 1 A and B postulates that the overall ulation to pyramidal neurons in the SO of CA3/2. In SZs, complex flow of excitatory activity along the trisynaptic pathway may be changes in the expression of AMPA, NMDA and kainate receptor increased in SZ and help explain the paradoxical finding of an subunits were observed in this layer/sector; this occurred to a lesser increased basal cerebral blood flow that was recorded in the degree in BDs. With GABAergic activity decreased at this locus, hippocampus of SZs, using positron emission tomography (49). In these differences in gene expression profiles for glutamate recep- CA1 of BDs, however, information processing by the trisynaptic tors suggest that the integration of glutamatergic inputs with pathway may be fundamentally different from that seen in SZ. interneurons is fundamentally different in the two disorders. Neurons of CA3/2 also receive inputs from subcortical regions, Conclusions such as the hypothalamus, basal forebrain and basolateral amygdala In summary, the current study describes the results of a cross- (2). Cholinergic (28) and GABAergic (29) inputs to GABA cells in sectional analysis of gene expression profiling at key sites along the SO of CA3/2 originate in the septal nuclei, whereas glutamatergic trisynaptic pathway that almost exclusively contain GABAergic projections (30) originate in the basolateral amygdala (BLa) (31). interneurons. These results suggest that the cellular endopheno- Experimental stimulation of the BLa is associated with a reduction types for SZ and BD may be determined by multiple factors that in the number of GABA cells in sector CA3/2, a pattern remarkably include unique susceptibility genes for the respective disorders and similar to that seen in postmortem studies of SZ (1). Accordingly, altered integration among hippocampal neurons with extrinsic and GABA cell dysfunction inferred from our gene expression profiling intrinsic afferent fiber systems. The extensive and intricate data that studies in the SO of CA3/2 in SZs and BDs could involve gluta- has come from this study has provided novel insights into how a matergic inputs from the BLa (32). Kainate receptors may poten- complex circuit, like the trisynaptic pathway may be regulated in tially mediate some of the amygdalar effects on GABA cells and human hippocampus in both health and disease. regulate the expression of GAD67 in interneurons of this layer/ sector (6). The up-regulation of kainate receptor subunits channels Methods in SO of CA3/2 of SZs could be part of a larger mechanism that All methodological details have been described (6) and careful assessments of promotes neuronal excitability along GABA cell dendrites (32) and RNA quality and reproducibility have been reported (50). A brief description increases their firing rate (33, 34). This effect could be potentiated appears below: through an activation of Ih channels (35) that are associated with Subjects. As shown in Table S1, the cohort used in these studies consisted of 7 an attenuation of after-hyperpolarizing currents (34, 36) associated CONs, 7 SZs and 7 BDs matched for age, postmortem interval, pH and 18S/28S with the activation of GABA cells by pyramidal neurons via their ratios and the method for establishing retrospective diagnoses is described recurrent collaterals. The up-regulation of the HCN4 gene is elsewhere (6). consistent with such a mechanism operating in SZ. Tissue and RNA Preparation. Frozen tissue sections were cut (8 ␮m) on a Microm Kainate receptors are also critical for the generation of gamma HM 560 CryoStar cryostat, mounted on LEICA Frame Slides (PET-membrane 1.4 oscillations (37). Interestingly, Ih channels are involved in pace- ␮m) and fixed in Streck Tissue Fixative (STF; Streck Laboratories). As shown in Fig. maker activity (35) and have the potential to reset the phase S1, the sections were lightly stained with 0.1% cresyl violet to visualize the currents that comprise oscillatory rhythms (38). Together with cytoarchitectonic details of the hippocampus and to identify the stratum oriens kainate receptors, Ih channels expressed by ‘‘horizontal,’’ fast- (SO), stratum pyramidale (SP) and stratum radiatum (SR) within sectors CA3/2 and spiking interneurons located in the region of the oriens-alveus (39) CA1. The frame slides were mounted on a LEICA AS LMD apparatus and a pulsed ϫ may contribute to the generation of gamma (40) and theta (41, 42) UV laser was used to ‘‘cut’’ samples (2 mm 1 mm) of the various layers that fell into a lysis/denaturing solution. Total RNA was extracted using the Rneasy Micro oscillations that influence long-range synchronization (43), involv- Kit (Qiagen) and T7-based linear amplification was performed by MessageAmp ing the induction of GABA currents in pyramidal neurons (44). II aRNA amplification kit (Ambion). Subsequently, target labeling was performed Oscillatory rhythms are thought to be important in establishing the with the MessageAMP Biotin Enhanced Kit (Ambion) and generated aRNA with

cognitive relevance of hippocampal output (45). Given the gene the incorporation of biotinylated nucleotides necessary for signal detection on NEUROSCIENCE expression findings reported here for GABA cells in SO of CA3/2, the Affymetrix Genechip system. Yields of aRNA were determined using a

Benes et al. PNAS ͉ December 30, 2008 ͉ vol. 105 ͉ no. 52 ͉ 20939 Downloaded by guest on October 1, 2021 ␮Quant Microplate Spectrophotometer (Bio-Tek Instruments) and the quality of pathways and/or biologically related clusters of genes. The ␣-level of significance aRNA was determined using a Bioanalyzer 2100 (Agilent Technologies). for individual genes was set at P Յ 0.05. To assess the changes in gene expression for each GenMapp biopathway or cluster, an ad hoc metric, composite proba- Gene Expression Profiling. Biotinylated target RNA from each sample was indi- bility, Pc, based on a combination of probability theory and two separate correc- vidually hybridized to the U133A array (Affymetrix) as described in ref. 6. tions for multiple comparisons was used (13).

GenMapp Biopathway and Cluster Analyses. To identify biologically relevant Quantitative RT-PCR. Five genes were used to validate the microarray findings clusters of interrelated genes, GenMapp algorithms (www.genmapp.org) were with qRT-PCR and were processed as previously described (6), and the data are used to relate the ANOVAs generated with dChip to several different biochemical shown in Fig. S2.

1. Benes FM, Berretta S (2001) GABAergic interneurons: Implications for understanding 24. Frotscher M (1989) Mossy fiber synapses on glutamate decarboxylase-immunoreactive schizophrenia and bipolar disorder. Neuropsychopharmacology 25:1–27. neurons: Evidence for feed-forward inhibition in the CA3 region of the hippocampus. Exp 2. Rosene DL, Van Hoesen GW (1987) The hippocompal formation of the primate brain in Brain Res 75:441–445. Cerebral cortex, Vol. 6: Further Aspects of Cortical Function, Including Hippocampus, eds 25. Hirbec H, et al. (2003) Rapid and differential regulation of AMPA and kainate receptors at Peters A, Jones EG (Plenum, New York) pp 345–456. hippocampal mossy fibre synapses by PICK1 and GRIP. Neuron 37:625–638. 3. Alger BE, Nicoll RA (1982) Feed-forward dendritic inhibition in rat hippocampal pyramidal 26. Lauri SE, et al. (2001) Synaptic activation of a presynaptic kainate receptor facilitates AMPA cells studied in vitro. J Physiol 328:105–123. receptor-mediated synaptic transmission at hippocampal mossy fibre synapses. Neuro- 4. Csicsvari J, Jamieson B, Wise KD, Buzsaki G (2003) Mechanisms of gamma oscillations in the pharmacology 41:907–915. hippocampus of the behaving rat. Neuron 37:311–322. 27. Maccaferri G, McBain CJ (1996) Long-term potentiation in distinct subtypes of hippocam- 5. Freund TF, Buzsaki G (1996) Mechanisms of gamma oscillations in the hippocampus of the pal nonpyramidal neurons. J Neurosci 16:5334–5343. behaving rat. Hippocampus 6:347–470. 28. Dringenberg HC, Vanderwolf CH (1996) Cholinergic activation of the electrocorticogram: 6. Benes FM, Lim B, Matzilevich D, Walsh JP, Subburaju S, Minns M (2007) Regulation of the An amygdaloid activating system. Exp Brain Res 108:285–296. GABA cell phenotype in hippocampus of schizophrenics and bipolars. Proc Natl Acad Sci 29. Toth K, Freund TF, Miles R (1997) Disinhibition of rat hippocampal pyramidal cells by USA 104:10164–10169. GABAergic afferents from the septum. J Physiol 500:463–474. 7. Benes FM, Davidson J, Bird ED Quantitative cytoarchitectural studies of the cerebral cortex 30. Swanson LW, Petrovich GD (1998) What is the amygdala? Trends Neurosci 21:323–331. of schizophrenics.Arch Gen Psychiatry 43:31–35, 1986. 31. Cunningham MG, Bhattacharyya S, Benes FM (2008) Increasing interaction of amygdalar 8. Heckers S, et al. (2002) Differential hippocampal expression of glutamic acid decarboxylase afferents with GABAergic interneurons between birth and adulthood. Cereb Cortex 65 and 67 messenger RNA in bipolar disorder and schizophrenia. Arch Gen Psychiatry 18:1529–1535. 59:521–529. 32. Schaefer AT, et al. (2007) Dendritic voltage-gated Kϩ conductance gradient in pyramidal 9. Webster MJ, Herman MM, Kleinman JE, Shannon Weickert C (2006) BDNF and trkB mRNA neurones of neocortical layer 5B from rats. J Physiol 579:737–752. expression in the hippocampus and temporal cortex during the human lifespan. Gene Expr 33. Lupica CR, Bell JA, Hoffman AF, Watson PL (2001) Contribution of the hyperpolarization- Patterns 6:941–951. activated current (I(h)) to membrane potential and GABA release in hippocampal inter- 10. Luscher C, Jan LY, Stoffel M, Malenka RC, Nicoll RA (1997) G protein-coupled inwardly neurons. J Neurophysiol 86:261–268. rectifying Kϩ channels (GIRKs) mediate postsynaptic but not presynaptic transmitter 34. Lauri SE, et al. (2005) Endogenous activation of kainate receptors regulates glutamate actions in hippocampal neurons. Neuron 19:687–695. release and network activity in the developing hippocampus. J Neurosci 25:4473–4484. 11. Chen X, Johnston D (2005) Constitutively active G-protein-gated inwardly rectifying 35. Robinson RB, Siegelbaum SA (2003) Hyperpolarization-activated cation currents: From Kϩ channels in dendrites of hippocampal CA1 pyramidal neurons. J Neurosci molecules to physiological function. Ann Rev Physiology 65:453–480. 25:3787–3792. 36. Chen S, Wang J, Siegelbaum SA (2001) Properties of hyperpolarization-activated pace- 12. Magri C, et al. (2006) Glutamate AMPA receptor subunit 1 gene (GRIA1) and DSM-IV-TR maker current defined by coassembly of HCN1 and HCN2 subunits and basal modulation schizophrenia: A pilot case-control association study in an Italian sample. Am J Med Genet by cyclic nucleotide. J Gen Physiol 117:491–504. B 141:287–293. 37. Fisahn A, et al. (2004) Distinct roles for the kainate receptor subunits GluR5 and GluR6 in 13. Benes FM, et al. (2004) Acute amygdalar activation induces an upregulation of multiple kainate-induced hippocampal gamma oscillations. J Neurosci 24:9658–9668. monoamine G protein coupled pathways in rat hippocampus. Mol Psychiatry 9:932–945. 38. Yang EJ, Harris AZ, Pettit DL (2007) Synaptic kainate currents reset interneuron firing 14. Harvey VL, Duguid IC, Krasel C, Stephens GJ (2006) Evidence that GABA rho subunits phase. J Physiol (London) 578:259–273. contribute to functional ionotropic GABA receptors in mouse cerebellar Purkinje cells. 39. Maccaferri G, Lacaille JC (2003) Interneuron Diversity series: Hippocampal interneuron classi- J Physiol 577:127–139. fications—making things as simple as possible, not simpler. Trends Neurosci 26:564–571. 15. Benes FM, Vincent SL, Alsterberg G, Bird ED, San Giovanni JP (1992) Increased GABAA 40. Traub RD, et al. (2003) GABA-enhanced collective behavior in neuronal axons underlies receptor binding in superficial layers of cingulate cortex in schizophrenics.J Neurosci persistent gamma-frequency oscillations. Proc Natl Acad Sci USA 100:11047–11052. 12:924–929. 41. Cobb SR, et al. (2003) Activation of Ih is necessary for patterning of mGluR and mAChR induced 16. van der Laan JW, Jacobs AW, Bruinvels J (1980) Effects of branched-chain fatty acids on network activity in the hippocampal CA3 region. Neuropharmacology 44:293–303. GABA-degradation and behavior: Further evidence for a role of GABA in quasi-morphine 42. Dugladze T, et al. (2007) Impaired hippocampal rhythmogenesis in a mouse model of abstinence behavior. Pharmacol Biochem Behav 13:843–849. mesial temporal lobe epilepsy. Proc Natl Acad Sci USA 104:17530–17535. 17. Barr AM, Young CE, Phillips AG, Honer WG (2006) Selective effects of typical antipsychotic 43. Traub RD, Bibbig A, LeBeau FE, Buhl EH, Whittington MA (2004) Cellular mechanisms of drugs on SNAP-25 and synaptophysin in the hippocampal trisynaptic pathway. Int J Neu- neuronal population oscillations in the hippocampus in vitro. Annu Rev Neurosci 27:247–278. ropsychopharmacol 9:457–463. 44. Palva JM, et al. (2000) Fast network oscillations in the newborn rat hippocampus in vitro. 18. Benes FM, Todtenkopf MS (1998) Meta-Analysis of nonpyramidal neuron (NP) loss in layer J Neurosci 20:1170–1178. II in anterior cingulate cortex (ACCx-II) from three studies of postortem schizophrenic 45. Whittington MA, Doheny HC, Traub RD, LeBeau FE, Buhl EH (2001) Differential expression brain. Soc Neurosci Abstr 24:1275. of synaptic and nonsynaptic mechanisms underlying stimulus-induced gamma oscillations 19. Vincent SL, Adamec E, Sorensen I, Benes FM (1994) The effects of chronic haloperidol in vitro. J Neurosci 21:1727–1738. administration on GABA- immunoreactive axon terminals in rat medial prefrontal cortex. 46. Bucci P, Mucci A, Merlotti E, Volpe U, Galderisi S (2007) Induced gamma activity and Synapse 17:26–35. event-related coherence in schizophrenia. Clin EEG Neurosci 38:96–104. 20. Adler LE, et al. (1998) Schizophrenia, sensory gating, and nicotinic receptors. Schizophr 47. Freedman R, Hall M, Adler LE, Leonard S (1995) Evidence in postmortem brain tissue for Bull 24:189–202. decreased numbers of hippocampal nicotinic receptors in schizophrenia. Biol Psychiatry 21. McEvoy JP, Allen TB (2002) The importance of nicotinic receptors in schizo- 38:22–33. phrenia, bipolar disorder and Tourette’s syndrome. Curr Drug Targets CNS Neurol Disord 48. Sotty F, et al. (2003) Distinct electrophysiological properties of glutamatergic, cholinergic 1:433–442. and GABAergic rat septohippocampal neurons: Novel implications for hippocampal rhyth- 22. Acsady L, Kamondi A, Sik A, Freund T, Buzsaki G (1998) GABAergic cells are the micity. J Physiol 551:927–943. major postsynaptic targets of mossy fibers in the rat hippocampus. J Neurosci 49. Heckers S, et al. (1998) Impaired recruitment of the hippocampus during conscious 18:3386–3403. recollection in schizophrenia. Nat Neurosci 1:318–323. 23. Bortolotto ZA, Lauri S, Isaac JT, Collingridge GL (2003) Kainate receptors and the induction 50. Benes FM (2006) Strategies for improving sensitivity of gene expression profiling: Regu- of mossy fibre long-term potentiation. Philos Trans R Soc Lond B 358:657–666. lation of apoptosis in the limbic lobe of schizophrenics and bipolars (Elsevier, Amsterdam).

20940 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0810153105 Benes et al. Downloaded by guest on October 1, 2021