Increased Excitotoxicity and Neuroinflammatory Markers In

Home , Excitotoxicity

Molecular Psychiatry (2010) 15, 384–392 & 2010 Nature Publishing Group All rights reserved 1359-4184/10 $32.00 www.nature.com/mp ORIGINAL ARTICLE Increased excitotoxicity and neuroinflammatory markers in postmortem frontal cortex from bipolar disorder patients JS Rao1,3, GJ Harry2, SI Rapoport1 and HW Kim1,3 1Brain Physiology and Metabolism Section, NIA, NIH Bethesda, Bethesda, MD, USA and 2Laboratory of Molecular Toxicology, National Institute of Environmental Health Science, NIH, NC, USA

Reports of cognitive decline, symptom worsening and brain atrophy in bipolar disorder (BD) suggest that the disease progresses over time. The worsening neuropathology may involve excitotoxicity and neuroinflammation. We determined protein and mRNA levels of excitotoxicity and neuroinflammatory markers in postmortem frontal cortex from 10 BD patients and 10 age-matched controls. The brain tissue was matched for age, postmortem interval and pH. The results indicated statistically significant lower protein and mRNA levels of the N-methyl-D- aspartate receptors, NR-1 and NR-3A, but significantly higher protein and mRNA levels of interleukin (IL)-1b, the IL-1 receptor (IL-1R), myeloid differentiation factor 88, nuclear factor- kappa B subunits, and astroglial and microglial markers (glial fibrillary acidic protein, inducible nitric oxide synthase, c-fos and CD11b) in postmortem frontal cortex from BD compared with control subjects. There was no significant difference in mRNA levels of tumor necrosis factor alpha or neuronal nitric oxide synthase in the same region. These data show the presence of excitotoxicity and neuroinflammation in BD frontal cortex, with particular activation of the IL-R cascade. The changes may account for reported evidence of disease progression in BD and be a target for future therapy. Molecular Psychiatry (2010) 15, 384–392; doi:10.1038/mp.2009.47; published online 2 June 2009 Keywords: bipolar disorder; IL-1beta; NMDA receptors; excitotoxicity; inflammation; postmortem brain

Introduction secretory phospholipase A2, and cyclooxygenase-2, were found elevated in postmortem frontal cortex Bipolar disorder (BD) is a severe psychiatric disease from BD patients compared with control subjects.14 characterized by repeated manic and depressive A number of markers have been shown to be altered episodes. Reports of symptom worsening, cognitive by neuroinflammation and excitotoxicity. For exam- decline and progressive brain atrophy suggest that the ple, chronic administration of subconvulsive doses of disease is progressive and has a neurodegenerative N-methyl-D-aspartate (NMDA) decreased rat brain component. Progression is consistent with reports of expression of the NMDA receptor (NR) subunits, structural, metabolic and signaling abnormalities in NR-1 and NR-3A,11 and upregulated mRNA and 1–6 postmortem brain from BD patients, and with activity of cPLA2 and of one of its transcription evidence of excitotoxicity and neuroinflammation in factors, activator protein-2 (AP-2).12 Chronic NMDA BD.7–10 Animal studies have shown that both pro- also upregulated rat brain protein and mRNA levels of cesses are associated with increased levels of pro- neuroinflammatory markers, interleukin-1 beta (IL- inflammatory cytokines, reactive oxygen radicals and 1b), tumor necrosis factor-alpha (TNFa), glial fibril- nitric oxide in brain, and may be accompanied by lary acidic protein (GFAP) and inducible nitric oxide upregulation of brain arachidonic acid signaling synthase (iNOS).12 markers.11–13 Expression levels of enzymes involved Binding of IL-1 to the IL-1 receptor (IL-1R) recruits in arachidonic acid metabolism, including arachido- IL-R-associated kinase to the receptor complex nate-selective cytoplasmic phospholipase A2 (cPLA2), through its association with the IL-1R accessory protein13 and with adaptor protein myeloid differ- 15,16 Correspondence: Dr JS Rao, Brain Physiology and Metabolism entiation factor (MyD88). Upon recruitment, IL-R- Section, National Institute on Aging, National Institutes of Health, associated kinase is highly phosphorylated and Buld 9, Room 1S-126, 9000 Rockville Pike, Bethesda, MD 20892, subsequently dissociates from the receptor complex USA. to interact with TNF receptor-associated factor 6,17 E-mail: [email protected] 3These authors contributed equally to this work. which in turn is involved in activation of nuclear I Received 16 March 2009; accepted 13 April 2009; published kappa kinase and nuclear factor-kappa B (NF-kB) 18 online 2 June 2009 activation (Figure 1). TNFa can regulate cPLA2 and Excitotoxicity and neuroinflammatory markers in BD JS Rao et al 385 Schematic model for activation of IL-1 receptor cascade studied the frontal cortex because earlier studies IL-1 indicated structural, metabolic and signaling abnormalities in this area in BD patients.1–6,14

Materials and methods IL-1 R Postmortem brain samples IL-1 R AcP The study was approved by the Institutional Review Board of McLean Hospital, and by the Office of IRAK Human Subjects Research of NIH (#4380). Frozen MyD88 postmortem human frontal cortex from 10 BD patients and 10 age-matched control subjects were provided TRAF6 AP-2 by the Harvard Brain Tissue Resource Center (McLean TAB2 TAK1 Hospital, Belmont, MA, USA) under Public Health Service grant number R24MH068855. Mean age, PMI and pH of the frozen brain samples (measured by the NIK method of Harrison et al.,25) did not differ signifi- Activation cantly between the BD and control groups: age (years, NF-κB activation control: 43±3.5 vs BD: 49±7.2), PMI (hours, control: 27±1.5 vs BD: 21±3.0) and brain pH (control: 6.6±0.16 vs BD: 6.7±0.09). The BD patients were

GFAP, sPLA2 exposed to various psychotropic medications, as 26 iNOS, COX-2 shown earlier.

Figure 1 Representation of interleukin-1 receptor (IL-1R) Preparation of membrane, nuclear and cytoplasmic cascade activation by IL-1b. Activation of the type I IL-1R by extracts b IL-1 leads to recruitment of interleukin receptor-associated Membrane, cytoplasmic and nuclear extracts were kinase (IRAK) to the receptor complex through its associa- 11 tion with the IL-1R accessory protein and an adaptor prepared from frozen tissue samples as described. protein, myeloid differentiation factor 88 (MyD88). On Briefly, the tissue was homogenized in 20 mM Tris association, IRAK becomes highly phosphorylated and HCl (pH 7.5) and 0.2 mM EDTA buffer containing a subsequently dissociates from the receptor complex to cocktail of protease inhibitors (Roche, Indianapolis, interact with tumor necrosis factor receptor-associated IN, USA). The suspension was centrifuged at factor 6 (TRAF6), which in turn is involved in nuclear I 100,000 g for 1 h at 4 1C and the supernatant (S1), kappa kinase (NIK) and nuclear factor-kappa B (NF-kB) containing mostly cytosolic constituents, was re- activation. Activator protein-2 (AP-2) transcription factor is moved. The pellet was re-suspended in the earlier also activated by IL-1b. GFAP, glial fibrillary acidic protein; mentioned buffer containing 0.1% Triton-X 100, and sPLA2, secretory phospholipase A2; iNOS, inducible nitric incubated for 1 h at 4 1C. This mixture was centrifuged oxide synthase; COX-2, cyclooxygenase-2. at 100 000 g for 1 h at 4 1C. The resulting supernatant (S2), containing membrane constituents, was re- moved. Protein concentrations of cytosolic and cyclooxygenase-2 expression in a mitogen-activated membrane extracts were determined using a Bio-Rad protein kinase-dependent manner in various cell protein reagent (Bio-Rad, Hercules, CA, USA). Each types.19–21 Activation of IL-1R by IL-1 triggers activa- membrane and cytosolic fraction was characterized tion of its cascade, increasing expression of using cadherin and tubulin antibodies. 13,15,16 16 cPLA2, secretory phospholipase A2, cyclo- Nuclear extracts were prepared from frozen tissue as oxygenase-2,22 GFAP19–21 and iNOS through the described earlier.27 The nuclear fraction was character- activation of transcription factors, AP-2 or NF-kB, in ized by using lamin B antibody. Protein concentrations mouse astrocytes and other cell types. In response to of cytoplasmic and nuclear extracts were determined an excitotoxic insult, iNOS23 and the immediate early using Bio-Rad Protein Reagent (Bio-Rad). gene, c-fos, are expressed and considered as specific biomarkers of excitotoxicity.24 Western blot analysis To further clarify the possible contributions of Protein (50 mg) from the membrane, cytoplasmic and neuroinflammation and excitotoxicity in BD, in this nuclear extracts was separated on 4–20% SDS- study we measured protein and mRNA levels of polyacrylamide gels (Bio-Rad). After electrophoresis, markers of these processes in the postmortem frontal the proteins were transferred to a nitrocellulose cortex from BD patients and control subjects, matched membrane. Membrane protein blots were incubated for age, postmortem interval (PMI) and pH. Among overnight in Tris-buffered saline solution, containing others, we measured expression of NR subunits, of IL- 5% nonfat dried milk and 0.1% Tween-20, with 1b, IL-1R and MyD88, and markers of activated glia specific primary antibodies (1:200 dilution) for the and astroglia. We chose these markers to focus on NR-1, NR-2A, NR-2B, NR-3A, IL-1R and cadherin possible changes in the IL-1 R cascade (Figure 1). We (Cell Signaling, Beverly, MA, USA). Cytosolic protein

Molecular Psychiatry Excitotoxicity and neuroinflammatory markers in BD JS Rao et al 386 blots were incubated similarly, but with primary and immersion-fixed with 4% paraformaldehyde for antibodies (1:200 dilution) for IL-1b, MyD88, GFAP, 18 h and cyroprotected in 30% sucrose. The cryostat iNOS, nNOS (neuronal nitric oxide synthase) and sections were air dried for 60 min, rinsed in 1 Â tubulin (Cell Signaling). Nuclear NF-kB p50 and NF- automation buffer (Biomedia Corp., Foster City, CA, kB p65 protein levels were determined using specific USA). Sections were treated with 0.3% hydrogen (1:200) primary antibodies (Cell Signaling). Mem- peroxide to quench endogenous peroxidase activity, brane, cytoplasmic and nuclear protein blots were and then incubated for 1 h at room temperature with a incubated with appropriate HRP-conjugated second- monoclonal human leukocyte antigen-D-related anti- ary antibodies (Bio-Rad) and visualized using a body (1:750; MBL, Woburn, MA, USA). Sections were chemiluminescence reaction (Amersham, Piscataway, rinsed, incubated for 30 min at room temperature with NJ, USA) on X-Ray film (Kodak, Rochester, NY, USA). secondary antibody (Vector Elite Kit, Burlingame, Optical densities of immunoblot bands were mea- CA) and detected with DAB (diamino-benzidine) sured using Alpha Innotech Software (Alpha Inno- chromagen (Dako, Carpinteria, CA, USA). An anti- tech, San Leandro, CA, USA) and were normalized to body to GFAP was used to detect astrocytes. b-actin (Sigma, St. Louis, MO, USA). Experiments Staining was conducted on an IHC Omni-UltraMap were carried out in duplicate for the 10 control and 10 HRP (Ventana Medical Systems, Tucson, AZ, USA). BD brain samples. Mean density values were ex- Sections were incubated with anti-GFAP (1:3500; pressed as percent of control. Dako) for 32 min at room temperature, rinsed, incubated with OmniMap anti-Rb HRP for 16 min, and Total RNA isolation and real-time reverse transcriptase counter-stained with modified Harris hematoxylin. PCR Images were collected using an Aperio Scanscope T2 Total RNA from brain and lipid-rich tissue was Scanner (Aperio Technologies, Vista, CA, USA) and isolated using an RNeasy mini kit (Qiagen, Valencia, viewed using an Aperio Imagescope v. 6.25.0.1117. CA, USA). The RNA integrity number was measured Images were rated by two independent scorers using a Bioanalyzer (Agilent 2100 bioanalyzer, Santa blind to subject classification. Samples were Clara, CA, USA). The RNA integrity number values rank-ordered on the basis of the level of the glial were 6.9±0.4 for the control samples and BD responses and then clustered on the basis of the 7.15±0.5 (mean±s.e.m.) for the BD samples. The subject classification. cDNA (complementary DNA) was prepared from total RNA using a high capacity cDNA Archive kit (Applied Biosystems, Foster City, CA, USA). Messen- Statistical Analysis ger RNA levels of NR-1, NR-2A, NR-2B, NR-3A, IL-1b, ± IL-1R, MyD88, TNFa GFAP, iNOS, nNOS, CD11b, Data are expressed as mean s.e.m. Statistical sig- p50, p65 and c-fos were measured by quantitative nificance of means was calculated using a two-tailed reverse transcriptase PCR, using an ABI PRISM 7000 unpaired t-test at P < 0.05. Pearson correlations were sequence detection system (Applied Biosystems). made between age, PMI and pH and mRNA levels of Specific primers and probes for NR-1, NR-2A, NR- NR-1, NR-2A, NR-2B, NR-3A, IL-1, IL-1R, Myd88, 2B, NR-3A, IL-1b, IL-1R, MyD88, GFAP, iNOS, nNOS, GFAP, iNOS, NF-kB65, NF-kB50 and CD11B in CD11b, p50, p65 and c-fos were purchased from postmortem frontal cortex from control subjects and TaqMan gene expression assays (Applied Biosystems) BD patients, separately. Statistical significance was and consisted of a 20 Â mix of unlabeled PCR primers set at P < 0.05. and TaqMan minor groove binder probe (FAM (6-carboxy-fluorescein) dye labeled). The fold-change in gene expression was determined by the DDC T Results method.27 Data were expressed as the relative level of the target gene in the BD tissue normalized to the Decreased protein and mRNA levels of NR-1 and endogenous control (b-globulin) and relative to the NR-3A in frontal cortex from BD patients 28 control (calibrator), as described earlier. All experi- The mean protein levels of NR-1, NR-2A, NR-2B and ments were carried out twice in triplicates with 10–10 NR-3A in postmortem BD brain were compared with independent samples, and the data were reported as the matched control brain. NR-1 and NR-3A protein relative expression. levels were decreased significantly by 42% (P < 0.05) and 41% (P < 0.01), respectively, in BD brain when Immunohistochemistry compared with control brain (Figures 2a and 2d), Frozen tissue, held on a dry ice bed, was sectioned in whereas the mean NR-2A and NR-2B protein levels a cryostat. The cut surface was immediately refrozen. did not differ significantly between the groups One section was used for biochemical or molecular (Figures 2b and 2c). The mRNA levels of NR-1 and analysis, whereas the corresponding section was used NR-3A were significantly decreased in BD compared for immunohistochemistry. with control brain by 0.42-fold (P < 0.01) and 0.46-fold For immunohistochemistry, frozen sections were (P < 0.01), respectively (Figures 2e and 2h), whereas warmed to À20 1C from À80 1C storage conditions. NR-2A and NR-2B mRNA levels were not signifi- Each section was gently raised to room temperature cantly different (Figures 2f and 2g).

Molecular Psychiatry Excitotoxicity and neuroinflammatory markers in BD JS Rao et al 387

125 120 120 120 100 100 100 100 75 80 80 80 * *** 60 *** 60 50 60 ** 40 40 40 25 % of control % of control % of control 20 20 20 % of control 0 0 0 0 Control BD Control BD Control BD Control BD

NR-1 103 KD NR-2A 163 KD NR-2B 163 KD NR-3 103 KD β-actin 45 KD β-actin 45 KD β-actin 45 KD β-actin 45 KD

NR-1 mRNA NR-2A mRNA NR-2B mRNA NR-3A mRNA 1.2 1.2 1.5 1.8 1.5 1.0 0.9 1.2 1.2 0.8 0.9 0.6 ** 0.9 0.6 ** 0.6 0.6 0.4 0.3 0.3 0.3 0.2 gene expression gene expression gene expression 0.0 0.0 0.0 0.0 in gene expression Relative fold change in Relative fold change Relative fold change in Relative fold change in Control BD Control BD Control BD Control BD Figure 2 Mean NMDA receptor (NR)-1 (a), NR-2A (b), NR-2B (c) and NR-3A (d) protein (with representative immunoblots) levels as percent of control levels in frontal cortex from control (n = 10) and bipolar disorder (BD) (n = 10) subjects. Data are optical densities relative to that of b-actin. Mean mRNA as percent of control of NR-1 (e), NR-2A (f), NR-2B (g) and NR-3A (h) in frontal cortex from control (n = 10) and BD (n = 10) subjects measured using reverse transcriptase PCR. Data are normalized to the endogenous control (b-globulin) and expressed relative to the control (calibrator), using the DDCT method. Mean±s.e.m., *P < 0.05, **P < 0.01.

IL-1 IL-1R MyD88

200 200 200 * 150 * 150 150 * 100 100 100

50 50 50 % of control % of control % of control 0 0 0 Control BD Control BD Control BD

IL-1 beta 32 KD IL-1R 44 KD MyD88 35 KD β-actin 45 KD β-actin 45 KD β-actin 45 KD

IL-1 mRNA IL-1R mRNA MyD88 mRNA

4 ** 6 *** 6 ** 5 3 5 4 4 2 3 3 2 1 2 1 1 0 0 in gene expression 0 in gene expression in gene expression Relative fold change Relative fold change Control BD Relative fold change Control BD Control BD Figure 3 Mean interleukin (IL)-1b (a), IL-1R (IL-1R) (b) and myeloid differentiation factor 88 (MyD88) (c) protein (with representative immunoblots) as percent of control in frontal cortex, from control (n = 10) and bipolar disorder (BD) (n = 10) subjects. Data are optical densities relative to that of b-actin. IL-1b (d), IL-1R (e) and MyD88 (f) mRNA levels in the frontal cortex from controls (n = 10) and BD patients (n = 10), measured using reverse transcriptase PCR. Data are normalized to the endogenous control (b-globulin) and expressed relative to the control (calibrator), using the DDCT method. Mean±s.e.m., *P < 0.05, **P < 0.01, ***P < 0.001.

Increased protein and mRNA levels of IL-1b, IL-1R and levels of the inflammatory cytokines, IL-1b MyD88 in frontal cortex from BD patients (41%; P < 0.05), IL-R (41%;P < 0.05) and MyD88 As illustrated in Figure 3, compared with control (38%; P < 0.05) in frontal cortex from BD patients subjects, there were significantly elevated protein (Figures 3a–3c). Significant increases also were seen

Molecular Psychiatry Excitotoxicity and neuroinflammatory markers in BD JS Rao et al 388 in mRNA levels for IL-1b (2.2-fold; P < 0.01), IL-R Increased nuclear protein and mRNA levels of p50 and (3.6-fold; P < 0.001) and MyD88 (3.6-fold; P < 0.01) p65 in frontal cortex from BD patients (Figures 3d–3f). There were significant increases in nuclear protein and mRNA levels of p50 (45% and 0.8-fold, respectively) (Figures 4a–c; P < 0.05) and of p65 (62%

p50 protein p65 protein and 1.4-fold, respectively) (P < 0.01, P < 0.001, Figures 4b–4d) in the BD brain. 200 200 ** * 150 150 Increased cell markers of astrocytes and microglia in 100 100 frontal cortex from BD patients 50 50 A significant increase was observed in the mean % of control % of control 0 0 protein levels of GFAP (46%; P < 0.05) and of iNOS Control BD Control BD (37%; P < 0.05) in BD frontal cortex when compared with control frontal cortex (Figures 5a and 5b). The p50 50 KD p65 65KD increased protein levels were accompanied by sig- β β-actin 45KD -actin 45 KD nificant increases in mRNA levels of GFAP (0.7-fold; P < 0.01) and iNOS (1.7-fold; P < 0.01) (Figures 5d and p50 mRNA p65 mRNA 5e). However, there was no significant difference in the protein or mRNA level of nNOS (Figures 5c and 2.5 * 4 2.0 3 *** 5f). c-fos and CD11b protein and mRNA levels were 1.5 significantly higher in BD brain when compared with 2 1.0 control brain (Figures 6a–6d). 0.5 1 There was no significant difference in TNFa mRNA 0.0 0 level in the frontal cortex from BD patients (Figure in gene expression in gene expression Relative fold change Control BD Relative Fold change Control BD 6e). The elevation in astrocyte and microglia markers Figure 4 Mean nuclear factor-kappa B (NF-kB) p50 (a) and was supported by immunohistochemical staining for NF-kB p65 (b) protein levels (with representative immuno- both GFAP and HLD-A (Figure 6f). In control tissue, blots) in frontal cortex from control (n = 10) and bipolar GFAP þ astrocytes displayed fine fibrous processes, disorder (BD) (n = 10) subjects. Bar graphs are ratios of whereas in the BD tissue, hypertrophic GFAP þ optical densities of NF-kB p50 or NF-kB p65 to that of b- astrocytes were detected. In BD tissue, furthermore, actin, expressed as percent of control. NF-kB p50 (c) and human leukocyte antigen-D-related antibody staining k NF- B p65 (d) mRNA in postmortem frontal cortex from the detected increased staining for process-bearing micro- control (n = 10) and BD (n = 10) subjects, measured using reverse transcriptase PCR. Data are levels of NF-kB p50 and glia displaying a thickening of processes. NF-kB p65 in BD normalized to the endogenous control (b-globulin) and relative to the control (calibrator), using Correlation data with brain variables

the DDCT method. Mean±s.e.m., *P < 0.05, **P < 0.01, Pearson correlations between mRNA levels in control ***P < 0.001. and BD brains with respect to PMI, age and pH, were

200 150 * 120 * 150 100 100 80 100 60 50 50 40 % of control % of control % of control 20 0 0 0 Control BD Control BD Control BD GFAP (55 KD) iNOS (131 KD) nNOS (160 KD) β β β-actin (45 KD) -actin (45 KD) -actin (45KD)

GFAP mRNA iNOS mRNA nNOS mRNA 2.5 4.8 1.2 4.0 2.0 ** 0.9 3.2 ** 1.5 2.4 0.6 1.0 1.6 0.3 0.5 0.8 in gene expression in gene expression in gene expression Relative fold change Relative fold change 0.0 0.0 Relative fold change 0.0 Control BD Control BD Control BD Figure 5 Mean glial fibrillary acidic protein (GFAP) (a), inducible nitric oxide synthase (iNOS) (b) and neuronal nitric oxide synthase (nNOS) (c) protein (with representative immunoblots) in control (n = 10) and bipolar disorder (BD) (n = 10) frontal cortex. Data are optical densities of GFAP, iNOS and nNOS proteins to b-actin, expressed as percent of control. The mRNA levels of GFAP (d), iNOS (e) and iNOS (f) in postmortem control (n = 10) and BD (n = 10) frontal cortex, measured using reverse transcriptase PCR. Data are levels of GFAP, iNOS and nNOS in the BD patients normalized to the endogenous control

(b-globulin) and relative to control level (calibrator), using the DDCT method. Mean±s.e.m., *P < 0.05, **P < 0.01.

Molecular Psychiatry Excitotoxicity and neuroinflammatory markers in BD JS Rao et al 389 c-fos Protein c-fos mRNA TNFα mRNA

200 2.5 * 1.4 ** 1.2 2.0 150 1.0 1.5 0.8 100 1.0 0.6 0.4 50 0.5 % of control 0.2 in gene expression Relative fold change

0 expression gene in 0.0 0.0 Relative fold change Control BD Control BD Control BD

c-fos (62 KD) β-actin (45 KD)

CONTROL BD CD11b Protein CD11b mRNA

GFAP 200 2.5 ** Astrocytes 150 2.0 * 1.5 100 1.0 50

% of control % 0.5 0 HLA-DR in gene expression 0.0 Control BD change Relative fold Control BD Microglia

CD11b (128KD) β-actin (45 KD) Figure 6 Mean c-fos (a) and CD11B (c) protein (with representative immunoblots) in control (n = 10) and bipolar disorder (BD) (n = 10) frontal cortex. Data are optical densities of c-fos and CD11B proteins to b-actin, expressed as percent of control. The mRNA levels of c-fos (b), CD11B (d) and tumor necrosis factor (TNF)a (e) in postmortem control (n = 10) and BD (n = 10) frontal cortex, measured using reverse transcriptase PCR. Data are levels of c-fos, CD11B and TNFa in the BD patients normalized to the endogenous control (b-globulin) and relative to control level (calibrator), using the DDCT method. Mean±s.e.m., *P < 0.05, ** P < 0.01. Lower right (f): representative histology of control and BD frontal cortex. Microglial activation was characterized by human leukocyte antigen-D-related antibody and visualized by DAB. Astrocytes were detected by glial fibrillary acidic protein antibody and stained on an IHC Omni-Ultra MAP HRP. Sections were stained with Harris hematoxylin. Scale bar, 25 mm. Arrows indicate hypertrophic astrocytes and activated microglia in BD tissue.

Table 1 Pearson correlation coefficients relating brain mRNA levels to subject age, PMI and brain pH

Controls NR-1 NR-2AB NR-2B NR-3A IL-1b IL-R Myd88 GFAP iNOS NF-kB p65 NF-kB p50 CD11B

Age (year) P 0.64 0.55 0.17 0.19 0.06 0.77 0.15 0.72 0.28 0.32 0.34 0.49 r2 0.02 0.04 0.21 0.19 0.37 0.01 0.23 0.01 0.14 0.11 0.11 0.06 PMI (hours) P 0.50 0.23 0.25 0.44 0.15 0.09 0.07 0.38 0.21 0.39 0.09 0.25 r2 0.05 0.17 0.15 0.07 0.24 0.31 0.34 0.09 0.18 0.09 0.30 0.16 pH P 0.12 0.64 0.69 0.72 0.39 0.31 0.82 0.45 0.09 0.29 0.09 0.50 r2 0.26 0.02 0.02 0.01 0.09 0.12 0.00 0.07 0.31 0.13 0.30 0.05 BD Age (year) P 0.53 0.89 0.85 0.97 0.17 0.90 0.81 0.15 0.20 0.19 0.81 0.64 r2 0.05 0.00 0.00 0.00 0.21 0.00 0.00 0.02 0.16 0.19 0.00 0.02 PMI (hours) P 0.49 0.73 0.89 0.56 0.71 0.77 0.30 0.80 0.90 0.19 0.30 0.09 r2 0.06 0.01 0.00 0.04 0.01 0.01 0.12 0.00 0.00 0.19 0.12 0.00 pH P 0.93 0.26 0.37 0.29 0.26 0.22 0.34 0.76 0.32 0.14 0.34 0.40 r2 0.00 0.15 0.09 0.13 0.15 0.17 0.11 0.01 0.12 0.24 0.11 0.08

Abbreviations: BD, bipolar disorder; GFAP, glial fibrillary acidic protein; IL, interleukin; IL-R, IL receptor; iNOS, inducible nitric oxide synthase; Myd88, myeloid differentiation factor 88; NF-kB, nuclear factor-kappa B; NR, NMDA receptor; PMI, postmortem interval. all statistically insignificant (P > 0.05) (Table 1). Discussion Mean values of the three parameters did not differ significantly between the patient and the control This study showed a significant increase in excito- groups. toxicity and neuroinflammatory markers in BD frontal

Molecular Psychiatry Excitotoxicity and neuroinflammatory markers in BD JS Rao et al 390 cortex when compared with control frontal cortex. NR-2B) to form NRs with reduced activity and Ca2 þ Decreased mean protein and mRNA levels of the NR1 influx,38,39 whereas mice lacking the NR-3A subunit and NR-3A subunits were accompanied by increased have increased NR activity.40 Increased NMDA func- levels of markers of excitotoxicity, c-fos and iNOS tion could increase arachidonic acid signaling.14 Rats mRNA. In addition, protein and mRNA levels of IL- given a daily subconvulsive dose of NMDA for 3 1b, IL-1R, MyD88 and NF-kB subunits (p50 and p65) weeks showed reduced brain levels of NR-1 and NR- in the same region were increased significantly. There 3A, increased arachidonic acid turnover in phospho-

was a significant increase in GFAP expression, and lipids, increased protein and mRNA levels of cPLA2 also in the level of CD11b mRNA (a marker of and secretory phospholipase A2, increased AP-2 DNA astrocyte and microglial activation) in postmortem binding activity and increased AP-2a and AP-2b frontal cortex from BD patients. Increases in protein protein in the frontal cortex.14 and mRNA levels were associated with increased staining for GFAP and human leukocyte antigen-D- Involvement of neuroinflammation in BD related antibody in the same brain region from BD Interleukin 1 is a major cytokine responsible for patients. However, there was no significant difference inducing a number of proteins associated with in nNOS or TNFa expression in postmortem frontal inflammation.41 Many of these responses are induced cortex from BD patients compared with control by the rapid activation of the transcription factor, NF- subjects. In sum, these results are consistent with a kB, after signal transduction caused by IL-1b binding marked activation of the IL-1R cascade, which is to the type I IL-1R.42 Similarly, this study shows an characteristic of both systemic and local insults.29,30 increased protein and mRNA levels of IL-1b, IL-1R and MyD88 in BD brain, which may be responsible for Involvement of excitotoxicity in BD the observed upregulated NF-kB transcription factor. Studies have shown that an excitotoxic insult caused The increase in NF-kB may also be responsible for by chronic NMDA administration decreased expres- increased expression of iNOS and GFAP, markers for sion of NR subunits and increased levels of neuroin- astrocytes. The observed increase in GFAP expression flammatory markers in rat brain.11,12 Similarly, we in this study is not in line with other observations on observed decreased expression of NR-1 and NR-3 BD brain,43,44 but this discrepancy may be because of subunits in the postmortem frontal cortex from BD regional differences or drug exposure. An earlier patients. Excitotoxicity23 in this study was indicated clinical study did indicate an increase in serum TNFa by the increased protein and mRNA levels of iNOS level in BD patients.45 However, in this study we did without a significant change in nNOS expression in not see a similar change in postmortem frontal cortex the BD brain. Although this observation is not from BD patients. consistent with an earlier study31 that reported an The observed increases in neuroinflammatory mar- increase in nNOS in postmortem hippocampal region kers in the BD brain14 may have been induced, in part, from BD patients, the difference may represent by underlying process of excitotoxicity. Thus, chronic differences in the response between the different NMDA administration to rats upregulated neuroin- brain regions. Further, characterization of the BD flammatory markers, and cerebroventricular infusion frontal cortex revealed a significant increase in c-fos of high lipopolysaccharide concentrations for weeks expression, a marker of excitotoxicity,24,32,33 suggest- to months induced IL-1b, TNFa and amyloid pre- ing the presence of excitotoxicity in BD. cursor protein, leading to degeneration of hippocam- Consistent with our findings regarding the presence pal CA3 pyramidal neurons.46 Cytokines associated

of excitotoxicity markers, earlier studies indicated an with neuroinflammation can activate both cPLA2 and elevated brain glutamate/glutamine ratio in children secretory phospholipase A2 at astrocytic cytokine and adults with BD.8 Postmortem brains of receptors.47–50 BD patients displayed decreased levels of the NR None of the mRNA levels in either the BD or control subunits, NR-1, NR-2A and NR-3A.9,10 The mRNA brains was correlated significantly with PMI, brain levels of NR-1, NR-2A and NR-3A also were found to pH or subject age, and mean values of these be decreased in the postmortem BD brain, as were parameters did not differ significantly between the concentrations of the NR-associated postsynaptic two groups. Nevertheless, the BD patients had been proteins, PSD-95 and SAP102.34 Finally, gene variants exposed to a variety of drugs not experienced by the of NR-1 and NR-2 have been linked to the dis- control subjects, which may have confounded the ease.10,35,36 These studies suggest that an increase in results. Owing to their selective exposure, our find- glutamate transmission is associated with decreased ings may be related to differences in drug exposure, expression of NR subunits. rather than being specific to the BD trait. Thus, future The observed differences in expression of the studies should examine arachidonic cascade markers NMDA subunits in the BD brain indicate increased in brains from patients with schizophrenia (to use a changes in NR activity. The NR stimulation by roughly comparable drug exposure as a control) or glutamate or by NMDA decreases NR1 expression, with unipolar (primary major) depression, or with which we have observed earlier.14,37 Furthermore, Alzheimer’s disease (to test for disease specificity).51 in vitro studies indicate that the NR3A subunit co- In conclusion, markers of the excitotoxicity and assembles with other subunits (NR-1, NR-2A or neuroinflammation are significantly upregulated in

Molecular Psychiatry Excitotoxicity and neuroinflammatory markers in BD JS Rao et al 391 postmortem frontal cortex from BD patients. Their 8 Hashimoto K, Sawa A, Iyo M. Increased levels of glutamate in upregulation might result in cell death, and account brains from patients with mood disorders. Biol Psychiatry 2007; for the brain atrophy and cognitive decline that have 62: 1310–1316. 9 Mueller HT, Meador-Woodruff JH. NR3A NMDA receptor subunit been reported in BD patients. Our observations suggest mRNA expression in schizophrenia, depression and bipolar that agents that attenuate excitotoxicity or neuroin- disorder. Schizophr Res 2004; 71: 361–370. flammation could prove beneficial for treating BD. 10 Mundo E, Tharmalingham S, Neves-Pereira M, Dalton EJ, Chronic administration to rats of lithium, valproic Macciardi F, Parikh SV et al. Evidence that the N-methyl-D- acid, carbamazepine or lamotrigine, mood stabilizers aspartate subunit 1 receptor gene (GRIN1) confers susceptibility to bipolar disorder. Mol Psychiatry 2003; 8: 241–245. that have been approved for treating BD, was reported 11 Rao JS, Ertley RN, Rapoport SI, Bazinet RP, Lee HJ. Chronic NMDA to decrease the markers of brain arachidonic acid administration to rats up-regulates frontal cortex cytosolic phos- metabolism that had been shown to be upregulated in pholipase A2 and its transcription factor, activator protein-2. rat models of excitotoxicity and neuroinflammation, J Neurochem 2007; 102: 1918–1927. as well as in the BD brain.12,14 These markers include 12 Chang YC, Kim HW, Rapoport SI, Rao JS. Chronic NMDA administration increases neuroinflammatory markers in rat frontal cPLA2 and cyclooxygenase-2, and some of their cortex: cross-talk between excitotoxicity and neuroinflammation. transcription factors, AP-2 and NF-kB. Some of the Neurochem Res 2008; 33: 2318–2323. mood stabilizers also prevented NMDA- and lipopo- 13 Huang J, Gao X, Li S, Cao Z. Recruitment of IRAK to the interleukin lysaccharide-induced increases in arachidonic 1 receptor complex requires interleukin 1 receptor accessory 12,52–55 protein. Proc Natl Acad Sci USA 1997; 94: 12829–12832. acid markers in rat brain. This study 14 Rao JS, Kim H-W, Lee HJ, Rapoport SI. Up-regulated arachidonic supports the presence of excitotoxicity and neuroin- acid cascade enzymes and their transcription factors in post- flammation in BD. mortem frontal cortex from bipolar disorder patients. Society for Neuroscience; 2007 Sandi ego 2007 p 707 705/Z704. 15 Cao Z, Henzel WJ, Gao X. IRAK: a kinase associated with the interleukin-1 receptor. Science 1996; 271: 1128–1131. Conflict of interest 16 Wesche H, Henzel WJ, Shillinglaw W, Li S, Cao Z. MyD88: an adapter that recruits IRAK to the IL-1 receptor complex. Immunity The authors declare no conflict of interest. 1997; 7: 837–847. 17 Cao Z, Xiong J, Takeuchi M, Kurama T, Goeddel DV. TRAF6 is a signal transducer for interleukin-1. Nature 1996; 383: 443–446. Acknowledgments 18 Ninomiya-Tsuji J, Kishimoto K, Hiyama A, Inoue J, Cao Z, Matsumoto K. The kinase TAK1 can activate the NIK-I kappaB We thank the Harvard Brain Bank, Boston, MA for as well as the MAP kinase cascade in the IL-1 signalling pathway. Nature 1999; 398: 252–256. providing the postmortem brain samples under PHS 19 Hoeck WG, Ramesha CS, Chang DJ, Fan N, Heller RA. Cytoplasmic grant number R24MH068855. This research was phospholipase A2 activity and gene expression are stimulated by entirely supported by the Intramural Research Pro- tumor necrosis factor: dexamethasone blocks the induced synth- grams of the National Institute on Aging and the esis. Proc Natl Acad Sci USA 1993; 90: 4475–4479. 20 Spriggs DR, Sherman ML, Imamura K, Mohri M, Rodriguez C, National Institute of Environmental Health Sciences, Robbins G et al. Phospholipase A2 activation and autoinduction of National Institutes of Health Bethesda, MD 20892. We tumor necrosis factor gene expression by tumor necrosis factor. thank the National Cancer Institute (NCI) and the Cancer Res 1990; 50: 7101–7107. Center for Cancer Research (CCR) Fellows Editorial 21 Jupp OJ, Vandenabeele P, MacEwan DJ. Distinct regulation of Board for proofreading the manuscript. cytosolic phospholipase A2 phosphorylation, translocation, pro- teolysis and activation by tumour necrosis factor-receptor sub- types. Biochem J 2003; 374: 453–461. 22 Bauer MK, Lieb K, Schulze-Osthoff K, Berger M, Gebicke-Haerter References PJ, Bauer J et al. Expression and regulation of cyclooxygenase-2 in rat microglia. Eur J Biochem 1997; 243: 726–731. 1 Lopez-Larson MP, DelBello MP, Zimmerman ME, Schwiers ML, 23 Acarin L, Peluffo H, Gonzalez B, Castellano B. Expression of Strakowski SM. Regional prefrontal gray and white inducible nitric oxide synthase and cyclooxygenase-2 after matter abnormalities in bipolar disorder. Biol Psychiatry 2002; excitotoxic damage to the immature rat brain. J Neurosci Res 52: 93–100. 2002; 68: 745–754. 2 Lyoo IK, Kim MJ, Stoll AL, Demopulos CM, Parow AM, Dager SR 24 Rogers A, Schmuck G, Scholz G, Griffiths R, Meredith C, et al. Frontal lobe gray matter density decreases in bipolar I Schousboe A et al. Improvements in an in-vitro assay for disorder. Biol Psychiatry 2004; 55: 648–651. excitotoxicity by measurement of early gene (c-fos mRNA) levels. 3 Rajkowska G. Cell pathology in bipolar disorder. Bipolar Disord Arch Toxicol 2005; 79: 129–139. 2002; 4: 105–116. 25 Harrison PJ, Heath PR, Eastwood SL, Burnet PW, McDonald B, 4 Buchsbaum MS, Wu J, DeLisi LE, Holcomb H, Kessler R, Johnson J Pearson RC. The relative importance of premortem acidosis and et al. Frontal cortex and basal ganglia metabolic rates assessed by postmortem interval for human brain gene expression studies: positron emission tomography with [18F]2-deoxyglucose in selective mRNA vulnerability and comparison with their encoded affective illness. J Affect Disord 1986; 10: 137–152. proteins. Neurosci Lett 1995; 200: 151–154. 5 Rubinsztein JS, Fletcher PC, Rogers RD, Ho LW, Aigbirhio FI, 26 Rao JS, Rapoport SI, Kim H-W. Decreased GRK3 but not GRK2 Paykel ES et al. Decision-making in mania: a PET study. Brain expression in frontal cortex from bipolar disorder patients. Int J 2001; 124: 2550–2563. Neuropharmacol 2009; 29: 1–10. 6 Suhara T, Nakayama K, Inoue O, Fukuda H, Shimizu M, Mori A 27 Rao JS, Ertley RN, Lee HJ, Rapoport SI, Bazinet RP. Chronic et al. D1 dopamine receptor binding in mood disorders measured fluoxetine upregulates activity, protein and mRNA levels of by positron emission tomography. Psychopharmacology (Berl) cytosolic phospholipase A2 in rat frontal cortex. Pharmacoge- 1992; 106: 14–18. nomics J 2006; 6: 413–420. 7 Kim YK, Jung HG, Myint AM, Kim H, Park SH. Imbalance between 28 Livak KJ, Schmittgen TD. Analysis of relative gene expression data pro-inflammatory and anti-inflammatory cytokines in bipolar using real-time quantitative PCR and the 2(-Delta Delta C(T)) disorder. J Affect Disord 2007; 104: 91–95. method. Methods 2001; 25: 402–408.

Molecular Psychiatry Excitotoxicity and neuroinflammatory markers in BD JS Rao et al 392 29 Rothwell NJ, Luheshi G. Pharmacology of interleukin-1 actions in 43 Webster MJ, O0Grady J, Kleinman JE, Weickert CS. Glial fibrillary the brain. Adv Pharmacol 1994; 25: 1–20. acidic protein mRNA levels in the cingulate cortex of individuals 30 Benveniste EN. Cytokines: influence on glial cell gene expression with depression, bipolar disorder and schizophrenia. Neu- and function.. In Blalock JE (ed). Neuroimmunoendocrinology. roscience 2005; 133: 453–461. Karger, Basel, 1997, pp 31–75. 44 Fatemi SH, Laurence JA, Araghi-Niknam M, Stary JM, Schulz SC, 31 Oliveira RM, Guimaraes FS, Deakin JF. Expression of neuronal Lee S et al. Glial fibrillary acidic protein is reduced in cerebellum nitric oxide synthase in the hippocampal formation in affective of subjects with major depression, but not schizophrenia. disorders. Braz J Med Biol Res 2008; 41: 333–341. Schizophr Res 2004; 69: 317–323. 32 Shan Y, Carlock LR, Walker PD. NMDA receptor overstimulation 45 Brietzke E, Kapczinski F. TNF-alpha as a molecular target in triggers a prolonged wave of immediate early gene expression: bipolar disorder. Prog Neuropsychopharmacol Biol Psychiatry relationship to excitotoxicity. Exp Neurol 1997; 144: 406–415. 2008; 32: 1355–1361. 33 Rogers A, Schmuck G, Scholz G, Williams DC. c-fos mRNA 46 Hauss-Wegrzyniak B, Dobrzanski P, Stoehr JD, Wenk GL. Chronic expression in rat cortical neurons during glutamate-mediated neuroinflammation in rats reproduces components of the neuro- excitotoxicity. Toxicol Sci 2004; 82: 562–569. biology of Alzheimer0s disease. Brain Res 1998; 780: 294–303. 34 McCullumsmith RE, Kristiansen LV, Beneyto M, Scarr E, Dean B, 47 Tong W, Shah D, Xu J, Diehl JA, Hans A, Hannink M et al. Meador-Woodruff JH. Decreased NR1, NR2A, and SAP102 tran- Involvement of lipid mediators on cytokine signaling and script expression in the hippocampus in bipolar disorder. Brain induction of secretory phospholipase A2 in immortalized astro- Res 2007; 1127: 108–118. cytes (DITNC). J Mol Neurosci 1999; 12: 89–99. 35 Itokawa M, Yamada K, Iwayama-Shigeno Y, Ishitsuka Y, Detera- 48 Sun GY, Xu J, Jensen MD, Simonyi A. Phospholipase A2 in the Wadleigh S, Yoshikawa T. Genetic analysis of a functional GRIN2A central nervous system: implications for neurodegenerative dis- promoter (GT)n repeat in bipolar disorder pedigrees in humans. eases. J Lipid Res 2004; 45: 205–213. Neurosci Lett 2003; 345: 53–56. 49 Luschen S, Adam D, Ussat S, Kreder D, Schneider-Brachert W, 36 Martucci L, Wong AH, De Luca V, Likhodi O, Wong GW, King N Kronke M et al. Activation of ERK1/2 and cPLA(2) by the p55 TNF et al. N-methyl-D-aspartate receptor NR2B subunit gene GRIN2B in receptor occurs independently of FAN. Biochem Biophys Res schizophrenia and bipolar disorder: polymorphisms and mRNA Commun 2000; 274: 506–512. levels. Schizophr Res 2006; 84: 214–221. 50 Dinarello CA. The IL-1 family and inflammatory diseases. Clin Exp 37 Gascon S, Deogracias R, Sobrado M, Roda JM, Renart J, Rheumatol 2002; 20: S1–13. Rodriguez-Pena A et al. Transcription of the NR1 subunit of 51 Benes FM. Searching for unique endophenotypes for schizophre- the N-methyl-D-aspartate receptor is down-regulated by excito- nia and bipolar disorder within neural circuits and their molecular toxic stimulation and cerebral ischemia. J Biol Chem 2005; 280: regulatory mechanisms. Schizophr Bull 2007; 33: 932–936. 35018–35027. 52 Basselin M, Chang L, Bell JM, Rapoport SI. Chronic lithium 38 Ciabarra AM, Sullivan JM, Gahn LG, Pecht G, Heinemann S, chloride administration attenuates brain NMDA receptor-initiated Sevarino KA. Cloning and characterization of chi-1: a devel- signaling via arachidonic acid in unanesthetized rats. Neuropsy- opmentally regulated member of a novel class of the ionotropic chopharmacology 2006; 31: 1659–1674. glutamate receptor family. J Neurosci 1995; 15: 6498–6508. 53 Basselin M, Chang L, Chen M, Bell JM, Rapoport SI. Chronic 39 Sucher NJ, Akbarian S, Chi CL, Leclerc CL, Awobuluyi M, Deitcher administration of valproic acid reduces brain NMDA signaling via DL et al. Developmental and regional expression pattern of a novel arachidonic acid in unanesthetized rats. Neurochem Res 2008;33: NMDA receptor-like subunit (NMDAR-L) in the rodent brain. J 2229–2240. Neurosci 1995; 15: 6509–6520. 54 Basselin M, Villacreses NE, Chen M, Bell JM, Rapoport SI. Chronic 40 Das S, Sasaki YF, Rothe T, Premkumar LS, Takasu M, Crandall JE carbamazepine administration reduces N-methyl-D-aspartate re- et al. Increased NMDA current and spine density in mice lacking ceptor-initiated signaling via arachidonic acid in rat brain. Biol the NMDA receptor subunit NR3A. Nature 1998; 393: 377–381. Psychiatry 2007; 62: 934–943. 41 Dinarello CA. The biological properties of interleukin 1. Eur 55 Basselin M, Villacreses NE, Lee HJ, Bell JM, Rapoport SI. Chronic Cytokine Netw 1994; 5: 517–531. lithium administration attenuates up-regulated brain arachidonic 42 Siebenlist U, Franzoso G, Brown K. Structure, regulation and acid metabolism in a rat model of neuroinflammation. J Neuro- function of NF-kappa B. Annu Rev Cell Biol 1994; 10: 405–455. chem 2007; 102: 761–772.

Molecular Psychiatry