JAK2 and STAT3 activation contributes to ... - Wiley Online Library

Journal of Neurochemistry, 2006, 98, 1353–1368

doi:10.1111/j.1471-4159.2006.04051.x

JAK2 and STAT3 activation contributes to neuronal damage following transient focal cerebral ischemia Irawan Satriotomo,*,§ Kellie K. Bowen* and Raghu Vemuganti*, ,à *Department of Neurological Surgery, Neuroscience Training Program and àCardiovascular Research Center, University of Wisconsin, Madison, Wisconsin, USA and §Department of Anatomy and Neurobiology, Faculty of Medicine, Kagawa University, Japan

Abstract Increased levels of interleukin-6 (IL-6) play a role in post-ischemic cerebral inflammation. IL-6 binding to its receptors induces phosphorylation of the receptor associated janus kinases (JAKs), and the down-stream signal transducer and activator of transcription (STAT) family of transcription factors, which amplify the IL-6 signal transduction. We evaluated the functional significance of JAK2 and STAT3 activation in focal ischemia-induced neuronal damage. Transient middle cerebral artery occlusion in adult rats led to increased JAK2 and STAT3 phosphorylation in the ipsilateral cortex and striatum after 6–72 h of reperfusion. Fluorescent immunohistochemistry with cell specific markers (NeuN for neurons, glial fibrillary acidic protein for reactive astrocytes and ED1/OX42 for activated macrophages/microglia) showed that both pJAK2 and pSTAT3 staining is predominantly localized in the

macrophages/microglia in the post-ischemic brain. Intracerebroventricular infusion of rats with AG490 (a JAK2 phosphorylation inhibitor) prevented the post-ischemic JAK2 and STAT3 phosphorylation and significantly decreased the infarct volume, number of apoptotic cells and neurological deficits, compared to vehicle control. Furthermore, intracerebral injection of siRNA specific for STAT3 led to curtailed STAT3 mRNA expression and phosphorylation, decreased infarct volume, fewer apoptotic cells and improved neurological function following transient middle cerebral artery occlusion. These studies show that JAK2-STAT3 activation plays a role in post-ischemic brain damage. Keywords: infarction, inflammation, interluekin-6, stroke, suppressor of cytokine signaling, transcription factor. J. Neurochem. (2006) 98, 1353–1368.

Transient focal cerebral ischemia is associated with a robust inflammatory reaction that is known to contribute to tissue damage. Pro-inflammatory cytokines like interleukin (IL)-1, IL-6 and tumor necrosis factor-a (TNF-a) formed in excess during the acute phase after focal ischemia are crucial initiators of the post-ischemic inflammatory neuronal damage (Iadecola and Alexander 2001; Stoll et al. 2002). The effects of these cytokines will be propagated and amplified by the janus kinase (JAK) and signal transducer and activator of transcription (STAT) signaling pathways (O’Shea et al. 2002). Binding of cytokines to their receptors induces transphosphorylation of the receptor-associated JAKs, which in turn leads to phosphorylation of the down-stream STAT family of transcription factors. Phosphorylated STATs dimerize and translocate into nucleus where their binding to conserved genomic regulatory sequences controls the expression of a wide array of genes (Darnell 2005). JAK exists as three isoforms (JAK1, JAK2 and JAK3) and STAT as seven (STAT1, STAT2, STAT3, STAT4, STAT5A,

STAT5B and STAT6). Of these, JAK2 and STAT3 are considered the most-ancient and most-conserved isoforms. Knockout mice that lack JAK2 or STAT3 die in the

Received March 3, 2006; revised manuscript received March 29, 2006; accepted March 29, 2006. Address correspondence and reprint requests to Raghu Vemuganti, PhD, Department of Neurological Surgery, University of Wisconsin, K4/ 8 (Mail stop code CSC-8660), 600 Highland Ave, Madison WI 53792, 608-263-4055, USA. E-mail: [email protected] Abbreviations used: DMSO, dimethylsulfoxide; ECA, external carotid artery; GFAP, glial fibrillary acidic protein; IL, interleukin; JAK, janus kinase; MCAO, middle cerebral artery occlusion; NeuN, neuronal nuclear protein; pJAK, phosphorylated JAK; pSTAT, phosphorylated STAT, suppressor of cytokine signaling; PTEN, phosphatase and tension homolog deleted on chromosome 10; RNAi, RNA interference; siRNA, small interference RNA; SOD, superoxide dismutase; STAT, signal transducer and activator of transcription; TBS, Tris-buffered saline; TBST, TBS with Triton-X 100; TNF, tumor necrosis factor; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling.

2006 The Authors Journal Compilation 2006 International Society for Neurochemistry, J. Neurochem. (2006) 98, 1353–1368

1353

1354 I. Satriotomo et al.

embryonic stage, whereas those that lack the other JAK and STAT isoforms are viable (O’Shea et al. 2002). Previous studies showed that STAT1 and STAT3 phosphorylation will be up-regulated following focal cerebral ischemia in rodents (Planas et al. 1996; Justicia et al. 2000; Suzuki et al. 2001). Takagi et al. (2002) showed that increased phosphorylation of STAT1 in neurons contributes to brain injury following focal ischemia; STAT1 knockout mice show smaller infarcts. However, the functional significance of STAT3 activation to ischemic brain damage was not evaluated before. In the present work, we studied the cellular localization of JAK2 and STAT3 phosphorylation as a time of reperfusion following transient focal ischemia in adult rat brain. Furthermore, the effect of inhibiting JAK2 phosphorylation, and thereby STAT3 phosphorylation, using a JAK2 inhibitor AG490 (Gorina et al. 2005; Qiu et al. 2005) on focal ischemia-induced infarction and neuronal damage was evaluated. As other isoforms of STAT (including STAT1, which is known to be activated after focal ischemia) are also down-stream to JAK2, we studied the effect of specifically preventing STAT3 expression and thus phosphorylation using a small interference RNA (siRNA) specific for STAT3 on post-ischemic brain damage.

Materials and methods Animals The rats used in these studies were cared for in accordance with the Guide for Care and Use of Laboratory Animals, US Department of Health and Human Services Publication Number 86–23 (revised). All the surgical procedures were approved by The Research Animal Resources and Care Committee of the University of WisconsinMadison. In all, 161 adult male spontaneous hypertensive rats (280– 300 g; Charles River, Wilmington, MA, USA) were used in these studies. Of the 161 rats, 124 were subjected to transient middle cerebral artery occlusion (MCAO) and the remaining underwent sham surgery. Of the 124 rats subjected to transient MCAO, 82 rats were treated with AG490 (n ¼ 18), 3% dimethylsulfoxide (DMSO) (n ¼ 18), STAT3 siRNA pool (n ¼ 24) and control siRNA pool (n ¼ 22). Of the 37 sham-operated rats, 24 were treated with AG490 (n ¼ 4), 3% DMSO (n ¼ 4), STAT3 siRNA pool (n ¼ 8) and control siRNA pool (n ¼ 8). Induction of focal ischemia Transient MCAO was conducted as described earlier (Dhodda et al. 2004; Vemuganti et al. 2004; Vemuganti 2005). In brief, a rat was anesthetized with halothane, placed in a stereotaxic frame fitted with a nose cone with 2% halothane anesthesia. A craniotomy (4 mm in diameter, 2–4 mm lateral and 1–2 mm caudal to bregma) was performed with extreme care over the MCA territory using a trephine. The dura was left intact and a laser Doppler flow-meter probe (model PD-434; Vasamedics, LLC, St. Paul, MN, USA) was placed on the surface of the ipsilateral cortex and fixed to the periosteum with a 4–0 silk suture. The probe was connected to a laser flowmeter device (Laserflow blood perfusion monitor BPM

403 A; TSI Inc., St. Paul, MN, USA) for continuous monitoring of regional cerebral blood flow. The left femoral artery was cannulated for continuous monitoring of arterial blood pressure and to obtain the measurements of pH, Pao2, Paco2, hemoglobin and blood glucose concentration (i-STAT; Sensor Devices, Waukesha, WI). The rectal temperature was controlled at 37.0 ± 0.5C during surgery with a feedback-regulated heating pad. After a midline skin incision, the left external carotid artery (ECA) was exposed, and its branches were coagulated. A 3–0 surgical monofilament nylon suture, blunted at the end, was introduced into the ECA lumen and gently advanced to the internal carotid artery until regional cerebral blood flow was reduced to 10–16% of the baseline (recorded by laser Doppler flowmeter). After a 1-h occlusion, the suture was withdrawn to restore the blood flow (confirmed by laser Doppler). After suturing the wound, the rat was allowed to recover from anesthesia and returned to the cage with ad libitum access to food and water. During the MCAO, PaO2 (100–200 mmHg) and PaCO2 (30–40 mmHg) were maintained at physiological levels. Intracerebroventricular infusion of AG490 The JAK2 phosphorylation inhibitor AG490 was purchased from Biomol Research Laboratories (Plymouth Meeting, PA, USA). AG 490 (250 lM in 3% DMSO) or vehicle (3% DMSO) was continuously infused into the lateral ventricles of cohorts of rats as described earlier (Dempsey et al. 2003; Vemuganti et al. 2004). The drugs were filled into osmotic minipumps (Alzet model 1003D) which pump at a rate of 1 lL/h (Alza Co., Palo Alto, CA, USA). Each pump was connected to an Alzet brain infusion stainless steel cannula by peristaltic tubing and primed overnight at 37C to ensure immediate delivery after implantation. The cannula was stereotaxically implanted into the lateral ventricle [bregma; 0.8 mm posterior, ) 4.8 mm dorsoventral, ) 1.5 mm lateral; on the basis of the rat brain atlas of Paxinos and Watson (1998)] and secured to the skull with dental cement. The pump was placed in the skin fold on the neck of the rat. The cannula and pump implantation was conducted under halothane anesthesia. Intracerebral injection of siRNA To selectively prevent STAT3 phosphorylation, we employed the RNA interference (RNAi) technology. A double-stranded RNA (small interference RNA; siRNA) was employed to degrade STAT3 mRNA and thus to limit the available STAT3 protein for phosphorylation after focal ischemia. The siRNA experiments were designed and conducted as described earlier (Baker-Herman et al. 2004). The siRNAs directed against the STAT3 mRNA (NM_012747) consisted of four pooled sequences with symmetrical 3¢-UU overhangs and 5¢-phosphorylated antisense strand overhangs (siGENOME SMARTpool Reagents; Dharmacon, Lafayette, CO, USA). The sequences were designed and validated by using the SMARTselection Design Algorithm (Dharmacon) and were submitted to a BLAST search to avoid the possible targeting of other homologous genes. The sequences of the four STAT3 siRNAs represented in the pool are as follows: 5¢-CCA ACG ACC UGC AGC AAU AUU-3¢, 5¢-GGU CAA AUU UCC UGA GUU GUU-3¢, 5¢-GGG CAU CAA UCC UGU GGU AUU-3¢ and 5¢-GGA UGU CGC UGC CCU CAG AUU-3¢. To account for the nonsequence-specific effects, siControl Non-Targeting Pool from Dharmacon was used. This also comprised a pool of four non-targeting siRNAs with comparable GC content to that of the functional siRNA but lacks identity with known gene targets which was confirmed by


Neuronal damage after JAK2 and STAT3 activation

BLAST analysis to have at least four mismatches with all known human, mouse and rat genes. Dharmacon confirmed the lack of effect of these control siRNAs on the expression of various mRNAs using microarray analysis. The sequences of the four individual control siRNAs mixed and used as the pool are as follows: 5¢-AUG AAC GUG AAU UGC UCA A-3¢, 5¢-UAA GGC UAU GAA GAG AUA C3¢, 5¢-AUG UAU UGG CCU GUA UUA G-3¢ and 5¢-UAG CGA CUA AAC ACA UCA A-3¢. STAT3 siRNAs or the control siRNAs were suspended in siRNA universal buffer (Dharmacon) to yield a concentration of 50 lM, and 17 lL siRNAwas combined with 3 lL of oligofectamine and incubated at room temperature for 15 min. The STAT3 or the control siRNAs were slowly injected using a Hamilton syringe into the cerebral cortex (0.2 mm from Bregma, 5 mm lateralmidline, 3 mm deep). After 1 h, cohorts of rats were subjected to either transient MCAO or sham operation. Neurological evaluation Neurological deficits were evaluated on a 6-point scale (Zea Longa et al. 1989; Vemuganti et al. 2001, 2004) before transient MCAO and after 1 day and 3 days of reperfusion by an investigator blinded to the study groups. A score of 0 suggests no neurological deficit (normal), 1 suggests mild neurological deficit (failure to extend right forepaw fully), 2 suggests moderate neurological deficit (circling to the right), 3 suggests severe neurological deficit (falling to the right), and 4 suggests very severe neurological deficit (the rat did not walk spontaneously and had a depressed level of consciousness).

1355

lottesville, VA, USA) antibodies, washed in TBS-T (3 · 5 min), incubated in biotinylated goat anti-rabbit antibody (1 : 1000; Vector Laboratories, Burlingame, CA, USA) for 1 h. The sections were washed in TBS and incubated for 30 min in normal goat serum. Conjugation with avidin-biotin complex (1 : 100; Vecstatin Elite ABC kit, Vector Laboratories) was followed by visualization with 3,3¢-diaminobenzidine-hydrogen peroxidase (Vector Laboratories). The sections were dehydrated, cleared and mounted in Permount. Sections incubated without primary or secondary antibodies served as negative controls. For fluorescent double staining, parallel sets of sections from each brain were washed in TBS (3 · 5 min), incubated in normal goat serum for 30 min and incubated with the above polyclonal pSTAT3 or pJAK2 antibodies together with monoclonal antibodies against neuron-specific nuclear protein (NeuN; neuronal marker; 1 : 200; Chemicon), glial fibrillary acidic protein (GFAP; astroglial marker; 1 : 1000; Chemicon), and ED1 and Cd11b (activated microglia/macrophage markers; 1 : 500 each; Serotec, Oxford, UK). The sections were washed in TBS and incubated for 1 h with secondary antibodies (goat anti-rabbit 495 and goat antimouse 488; 1 : 200 each; Alexa Flour antibodies; Molecular Probes, Eugene, OR, USA). The slides were washed and coverslipped with Vectashield mounting medium (Vector Laboratories) and analyzed with a fluorescent microscope (Olympus, Osaka, Japan). Four sections were used from each rat for each antibody staining.

Estimation of ischemic brain injury Ischemic infarct volume was measured as described earlier (Vemuganti et al. 2001, 2004; Dhodda et al. 2004). The rats were killed after either 1 day or 3 days of reperfusion. Brains were removed, postfixed and sectioned serially (coronal; 40 lm thick at an interval of 360 lm), stained with thionine, and scanned using the NIH Image Program (written by Wayne Rasband, the program can be downloaded at no cost from http://rsb.info.nih.gov/ij). The volume of the ischemic lesion was computed by the numeric integration of data from 8 to 10 serial sections with respect to the sectional interval. To account for the cerebral edema and differential shrinkage resulting from tissue processing, the injury volumes were corrected using the Swanson formula: corrected injury volume ¼ contralateral hemisphere volume – (ipsilateral hemisphere volume – measured injury volume) (Swanson et al. 1990).

TUNEL assay In situ detection of apoptotic cells was performed using terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) using the In Situ Cell Death Detection Kit (Roche Molecular Biochemicals, Indianapolis, IN, USA) according to the manufacturer’s instructions. Briefly, the sections were permeabilized with 1% Proteinase K (in 50 mM Tris/5 mM EDTA buffer) for 15 min, rinsed with phosphate-buffered saline (PBS) and incubated in TUNEL reaction mixture for 1 h at 37C. Samples were rinsed in PBS (3 · 5 min), mounted in antifade solution (Molecular Probes) and analyzed with the fluorescence microscope. The apoptotic cells (FITC green-stained) were counted in the peri-infarct area of ipsilateral cortex and striatum of each rat (three fields were counted in each case at · 300 magnification). The cell number was expressed as cells per mm2. We immunostained a parallel set of sections from each brain first with NeuN antibodies and then subjected to TUNEL staining procedure.

Immunohistochemistry The cellular localization of pJAK2 and pSTAT3 was evaluated in brain sections from cohorts of rats killed after 6 h, 24 h or 72 h reperfusion after transient MCAO (n ¼ 5/group). Five shamoperated rats served as control. The rats were perfused transcardially with 4% buffered paraformaldehyde and the brains were sectioned (coronal; 40 lm thick). The sections were rinsed in 0.1 M Trisbuffered saline with 0.1% Triton-X100 (TBS-T; 3 · 5 min), incubated in 1% H2O2 for 30 min and washed in TBS-T (3 · 5 min). The sections were incubated in 5% normal goat serum for 1 h and overnight in polyclonal STAT3 (1 : 500; Cell Signaling Technologies, Beverly, MA, USA), polyclonal JAK2 (1 : 500; Chemicon, Temecula, CA, USA), polyclonal pSTAT3 (pTyr705) (1 : 400; Cell Signaling Technologies) or polyclonal pJAK2 (pTyr1007/1008) (1 : 500; Upstate Cell Signaling Solutions, Char-

Real-time PCR analysis of STAT3 mRNA We verified the efficacy of the STAT3 siRNA to knockdown postischemic STAT3 mRNA expression using real-time PCR. Fifteen rats subjected to transient MCAO with no injections were sacrificed at various reperfusion times (five each at 6 h, 24 h and 72 h) to establish the time course of STAT3 mRNA expression after ischemia. A cohort of rats injected with STAT3 siRNA or control siRNA (n ¼ 4/group) were subjected to transient MCAO and reperfusion for 24 h. Sham-operated rats injected with STAT3 siRNA or control siRNA or not injected with any agent (n ¼ 4/ group) served as controls. Total RNA was extracted from the ipsilateral cortex of each rat using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). 1 lg total RNA from each sample was reverse transcribed with oligo(dT)15 and random hexamer primers using M-MuLV reverse transcriptase (Life Technologies, Rockville, MD,



USA). Real-time PCR was conducted as described earlier (Dhodda et al. 2004; Vemuganti et al. 2004; Vemuganti and Dempsey 2005). In brief, 10 ng of cDNA and gene-specific primers were added to SYBR Green PCR Master Mix (SYBR Green I Dye, AmpliTaq DNA polymerase, dNTPs with dUTP and optimal buffer components; Applied Biosystems, Foster City, CA, USA) and subjected to PCR amplification (1 cycle at 50C for 2 min, 1 cycle at 95C for 10 min, and 40 cycles at 95C for 15 s and 60C for 1 min) in a TaqMan 5700 Sequence Detection System (Applied Biosystems). PCR reactions were conducted in duplicate. The amplified transcripts were quantified with the comparative CT method using 18S rRNA and glyceraldehyde-3-phosphate dehydrogenase as the internal control (http://docs.appliedbiosystems.com/pebiodocs/ 04303859.pdf). In previous studies, we observed no significant change in the expression of these housekeeping transcripts in rat CNS following acute insults like transient MCAO, traumatic brain injury and spinal cord injury (Song et al. 2001; Vemuganti et al. 2002, 2003). The following primer sequences (5¢)3¢) were designed using the PRIMER EXPRESS software (Applied Biosystems) based on the GenBank accession numbers given in parentheses: STAT3 (NM_012747): GCG ATA GCT TCC CCA TGG A and CAC CAG AGT GGC GTG TGA CT; 18S rRNA (M11188): CGC CGC TAG AGG TGA AAT TCT and CGA ACC TCC GAC TTT CGT TCT and glyceraldehyde-3-phosphate dehydrogenase (NM_017008): AGG GCT GGC CTA AAT GAT ACC and CAC CGA CCT TCA CCA TCT TGT. These primer sequences were validated in our previous papers (Song et al. 2001; Vemuganti et al. 2003, 2004; Dhodda et al. 2004; Naylor et al. 2005). Western blotting The effect of AG490 and STAT3 siRNA on pJAK2 and/or pSTAT3 protein levels after focal ischemia was tested by western blotting as described earlier (Vemuganti et al. 2001, 2004; Dhodda et al. 2004). Ipsilateral cortex from the rats treated with AG490, 3% DMSO, STAT3 siRNA and control siRNA, and subjected to transient MCAO/24 h reperfusion (n ¼ 3/group) was homogenized in ice cold lysis buffer (20 mM Na2HPO4, 50 mM NaF, 10 mM Na4P2O7, 150 mM NaCl, 2% Triton, 0.5% Na-deoxycholate and 1 mM Na3VO4) containing protease inhibitors (1 mM PMSF, 10 lg/mL leupeptin, 10 lg/mL pepstatin, 10 lg/mL aprotinin and 5 mM EDTA). The homogenate was centrifuged at 15 000 g for 10 min at 4C, and the protein content was estimated in the supernatant. The equivalent of 30 lg protein from each sample was electrophoresed on Bio-Rad Criterion precast polyacrylamide gels, transferred to polyvinylidene difluoride membranes and probed with polyclonal pJAK2 (1 : 500; Chemicon) and pSTAT3 (1 : 1000; Cell Signaling Technology) antibodies followed by HRP-conjugated goat antirabbit IgG (1 : 1000). The blots were stripped and reprobed with monoclonal anti-b-tubulin antibody (Sigma Chemical Co., St. Louis, MO, USA). The protein bands recognized by antibodies were detected by enhanced chemiluminescence according to the manufacturer’s instructions (Pierce, Rockford, IL, USA). Blots were prepared twice for each sample. Statistical analysis The data are expressed as mean ± SD. Comparisons among groups were performed by one-way ANOVA followed by Tucky-Kramer multiple comparisons post-test.

Results

Transient focal ischemia induced JAK2 and STAT3 phosphorylation Immunohistochemical analysis showed very little pJAK2 staining in the cortex and striatum of the sham-operated rats (Fig. 1a, b). Following transient MCAO, pJAK2 immunostained cell numbers started to increase rapidly in the ipsilateral cortex and striatum in the ischemic core and the peri-infarct areas. The pJAK2 immunopositive cells increased in number after 6–72 h reperfusion in both the cortex and striatum of the ischemic animals (Figs 1c–n). The sham-operated rats showed neither pJAK2 (Fig. 2a and c) nor ED1 (Figs 2B.C) immunostaining. Fluorescent double staining using ED1 and OX42/Cd11b as markers confirmed that most of the pJAK2 immunostained cells were of monocyte/macrophage lineage at both 24 h (Figs 2d–f) and 72 h (Figs 2g–l) reperfusion after transient MCAO. Neither NeuN positive neurons (Figs 2m–o) nor GFAP-positive astrocytes (Figs 2p–r) showed any pJAK2 immunoreactivity following transient MCAO and 24 h reperfusion. The contralateral cortex or striatum of the rats subjected to transient MCAO showed no appreciable increase in pJAK2 staining over sham-operated rats at any reperfusion time (data not shown). The pSTAT3 immunoreactivity was very faint in the cortex (Fig. 3a) and striatum (Fig. 3b) of the sham-operated rats. After 6 h reperfusion following transient MCAO, the ipsilateral striatum showed pSTAT3 immunoreactive cells in the ischemic core (Fig. 3c) and the peri-infarct area (Fig. 3d). The core (Fig. 3e) and peri-infarct area (Fig. 3f) of the ipsilateral cortex also showed some pSTAT3 immunoreactive cells after 6 h of reperfusion. After 24 h reperfusion, many pSTAT3 immunoreactive cells appeared in the striatal core (Fig. 3g), striatal peri-infarct area (Fig. 3h), cortical core (Fig. 3i) and the cortical peri-infarct area (Fig. 3j). After 72 h reperfusion, all these areas on the ipsilateral half of the brain showed pSTAT3 immunoreactive cells (Figs 3k–n), but fewer in number than those observed after 24 h reperfusion. Fluorescent immunostaining showed neither pSTAT3 (Fig. 4a and c) nor ED-1 (Fig. 4b, c) immunostaining in sham-operated rat brain. Similar to pJAK2, activated microglia/macrophages (ED1 and OX42/ Cd11b positive) are the predominant cell type that showed pSTAT3 immunostaining after 24 h and 72 h reperfusion following transient MCAO (Figs 4d–l). Unlike pJAK2 immunostaining, pSTAT3 immunostaining was also observed in some neurons (NeuN positive; Figs 4m–o) and astroglia (GFAP positive; Figs 4p–r) after 24 h reperfusion. The contralateral cortex or striatum of the rats subjected to transient MCAO showed no appreciable increase in pSTAT3 staining over sham-operated rats at any reperfusion time (data not shown).



(a)

(b)

(o)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

1357

sham

6h

24h

72h

I

Striatum

II

Fig. 1 Transient MCAO increased pJAK2 immunostaining in rat brain. The pJAK2 staining is very faint in the striatum (a) and the cortex (b) of sham-operated (SH) rats. The pJAK2 immunostaining increased significantly as a function of time in all the four areas (I, II, III and IV) studied at 6 h (c–f), 24 h (g–j) and 72 h (k–n) reperfusion following transient MCAO compared to sham. Panel o indicates the ischemic core (I) and periinfarct area (II) in the ipsilateral striatum and the ischemic core (III) and

III

Cortex

IV

peri-infarct area (IV) in the ipsilateral cortex in rats subjected to transient MCAO. Scale bar is 100 lm. The histogram at the bottom indicates the number of pJAK2 immunopositive cells/high power field. The bars represent mean ± SD of five rats/group. For each rat, cells were counted in three ·100 fields by an evaluator blinded to study groups. *p < 0.05 compared to sham by one-way ANOVA followed by Tukey-Kramer multiple comparisons post-test.



(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

(p)

(q)

(r)

Fig. 2 Activated microglial and macrophages are the major cell type that showed pJAK2 after transient MCAO. Neither pJAK2 (a) nor activated microglia/macrophage marker ED1 (b) immunostaining was seen in the sham-operated rat cerebral cortex (c). After 24 h reperfusion, several pJAK2 immunostained cells appeared throughout the ischemic area (d) and most of them were ED1 positive activated microglia and macrophages (e and f). The pJAK2/ED1 double labeled cells further increased in number after 72 h reperfusion (g–i). After 72 h reperfusion,

many pJAK2 positive cells also stained for a second microglia/macrophage marker OX42 (j–l). The pJAK2 immunostaining was not present in neurons and astrocytes after transient MCAO. After 24 h reperfusion, the pJAK2 immunopositive cells (m and p) observed in the cerebral cortex were not positive for either the mature neuronal marker NeuN (n and o) or the astroglial marker GFAP (Q and R). The insets in panels d–r show the pJAK2, ED1, OX42, NeuN and GFAP positive cells at a higher magnification. The scale bar in c is 100 lm.

AG490 blocked post-ischemic JAK2 and STAT3 phosphorylation To study the functional significance of JAK2 and the downsteam STAT3 activation, we treated ischemic rats with AG490,

a potent JAK2 phosphorylation inhibitor. Cohorts of rats were given continuous intracerebroventricular infusions of either AG490 or 3% DMSO (vehicle for AG490) starting 1 day before the induction of transient MCAO. After 1 day of



(a)

(b)

(o)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

1359

sham

6h

24h

72h

I

Striatum

II

Fig. 3 Transient MCAO increased pSTAT3 immunostaining in rat brain. The pSTAT3 staining was very faint in the striatum (a) and cortex (b) of sham-operated rats. The pSTAT3 immunostained cell number increased significantly in all four areas (I, II, III and IV) studied at 6 h (c–f), 24 h (g–j) and 72 h (k–n) of reperfusion following transient MCAO. Panel o indicates the ischemic core (I) and peri-infarct area (II) in the ipsilateral striatum and the ischemic core (III) and peri-infarct

III

Cortex

IV

area (IV) in the ipsilateral cortex of rats subjected to transient MCAO. The scale bar is 100 lm. The histogram at the bottom indicates the number of pSTAT3 immunopositive cells/high power field. The bars represent mean ± SD of five rats/group. For each rat, cells were counted in three ·100 fields by an evaluator blinded to study groups. *p < 0.05 compared to sham by one-way ANOVA followed by TukeyKramer multiple comparisons post-test.



(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

(p)

(q)

(r)

Fig. 4 The pSTAT3 was mainly localized in the activated microglial, macrophages and endothelial cells in the post-ischemic brain. Neither pSTAT3 (a) nor activated microglia/macrophage marker ED1 (b) immunostaining was observed in the sham-operated rat cerebral cortex (c). After 24 h reperfusion after transient MCAO, many endothelial cells and the blood cells showed pSTAT3 immunostaining in the ischemic area (d) and many of these cells were observed to be positive for the microglia/macrophage marker ED1 (e–f). At this time point, pJAK2/ED1 double positive cells were seen in both the cerebral tissue as well as in the blood vessels (f). After 72 h of reperfusion, the pSTAT3 (g) and the ED1 (h) positive cell number increased further in the brain parenchyma

(i). The pSTAT3 positive cells seen in the ipsilateral cortex after 72 h reperfusion also stained positive for the other microglia/macrophage marker OX42/Cd11b (j–l). Increased post-ischemic pSTAT3 immunostaining was also seen in neurons and astrocytes. At 24 h reperfusion, many pSTAT3 immunopositive cells observed in the ischemic cortex (m) were positive for NeuN (n and o). After 24 h reperfusion following transient MCAO, pSTAT3 immunostaining (p and r) was also observed in some GFAP positive astrocytes (q and r). The insets in panels d–r show the pSTAT3, ED1, OX42/Cd11b, NeuN and GFAP positive cells at a higher magnification. The scale bar in c is 100 lm.

reperfusion, the pJAK2 immunopositive cells were significantly lower in the ipsilateral cortex (Fig. 5c) and striatum (Fig. 5d) of the AG490/MCAO group compared to the

ipsilateral cortex (Fig. 5a) and striatum (Fig. 5b) of the DMSO/MCAO group. After 24 h reperfusion following transient MCAO, the pSTAT3 immunoreactivity was also



(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

Fig. 5 Infusion of JAK2 phosphorylation inhibitor AG490 into lateral ventricles prevented post-ischemic JAK2 and STAT3 phosphorylation. In the DMSO/MCAO group, cortex (a) and striatum (b) showed several pJAK2 immunopositive cells which were significantly less in number in the cortex (c) and striatum (d) of the AG490/MCAO group. The DMSO/ MCAO group also showed many pSTAT3 immunopositive cells in the cortex (e) and striatum (f) which were significantly less in the cortex (g) and striatum (h) of the AG490/MCAO group. The scale bar is 100 lm. Western blot analysis showed that the pJAK2 and pSTAT3 protein levels were significantly lower in the MCAO/AG490 group compared to the MCAO/DMSO group. MCAO/STAT3 siRNA group showed significantly decreased pSTAT3, but not pJAK2 protein level compared to the MCAO/control siRNA group (i). Similar results were observed with three rats per group.

decreased significantly in the ipsilateral cortex (Fig. 5g) and striatum (Fig. 5h) of the AG490/MCAO group compared to the cortex (Fig. 5e) and striatum (Fig. 5f) of the DMSO/ MCAO group. Western blot analysis also showed significantly reduced protein levels of pJAK2 and pSTAT3 after 1 day of

1361

reperfusion following transient MCAO in the ipsilateral cortex of rats treated with AG490 compared to vehicle (Fig. 5i). AG490 decreased post-ischemic infarction, neuronal death and neurological dysfunction In spontaneous hypertensive rats, 1 h transient MCAO and 24 h reperfusion resulted in an infarct on the ipsilateral side (total infarct volume: 217 ± 35 mm3) of the brain encompassing the cerebral cortex (178 ± 31 mm3) and the striatum (35 ± 5 mm3) (Table 1). Infusion of 3% DMSO had no significant effect on the total or cortical or striatal infarct volume (Table 1). Treating rats with AG490 led to a significant decrease in the cortical (by 40%, p < 0.05), striatal (by 29%, p < 0.05) and total (38%, p < 0.05) infarct volume after transient MCAO, compared with the vehicle-treated control (Table 1). Thionine-stained serial sections from representative rats show a significantly smaller infarct in the MCAO/AG490 group than in the MCAO/DMSO group (Fig. 6a). Compared to the MCAO/DMSO group, the MCAO/ AG490 group showed a significant reduction in the number of apoptotic cells (TUNEL positive) in the peri-infarct cortex (by 65%, p < 0.05) and peri-infarct striatum (by 45%, p < 0.05) (Table 2). Many of these apoptotic cells were observed to be NeuN positive neurons in the cortex of the MCAO/DMSO group (Fig. 6b, left panel) and they were significantly fewer in number in the MCAO/AG490 group (Fig. 6b, right panel). In both groups, there were some TUNEL positive cells which were NeuN negative. This is understandable as astrocytes and possibly some of the invading macrophages and neutrophils will also undergo apoptosis following transient focal ischemia. Compared to the normal cytoarchitecture of the cortical cellular layers observed in the sham/DMSO group (Fig. 6c, left panel), MCAO/DMSO group showed severe neuronal loss in all the six cortical layers, with noticeable near total loss of layer V pyramidal neurons (Fig. 6c, middle panel). Rats subjected to MCAO without any drug treatment also showed comparable neuronal damage (data not shown). The MCAO/ AG490 group showed a significantly higher neuronal density in all the cortical layers (Fig. 6c, right panel) compared to MCAO/DMSO group. The neurological deficits analyzed at 24 h reperfusion were moderate to severe in both the MCAO/no drug group (median neuroscore was 2.44) and MCAO/DMSO group (median neuroscore was 2.57) (Table 1). The MCAO/AG490 group showed only mild neurological deficits with a significantly higher median neuroscore of 1.34 compared to the other two groups (p < 0.05) (Table 1). STAT3 siRNA treatment decreased post-ischemic STAT3 phosphorylation As pJAK2 can induce phosphorylation of STAT3 as well as other STAT isoforms, the neuroprotection afforded by AG490 could not be assigned exclusively to the prevention of STAT3 phosphorylation. This is particularly important as



Table 1 Transient MCAO-induced infarct volumes and neurological deficits in rats treated with AG490, DMSO, STAT3 siRNA and control siRNA Infarct volume; mm3

1 day of reperfusion None 3% DMSO AG490 Control si RNA STAT3 si RNA 3 days of reperfusion 3% DMSO AG490 Control si RNA STAT3 si RNA

n

Cortex

12 10 10 10 12

178 188 114 181 115

± ± ± ± ±

5 5 5 5

201 5126 204 143

± ± ± ±

Striatum

Total

31 18 17a,b 21 19a,c

35 38 26 37 27

± ± ± ± ±

5 5 5a,b 6 5a,c

217 229 142 223 151

± ± ± ± ±

33 16a,b 32 20a,c

39 29 38 28

± ± ± ±

4 4a,b 4 5a,c

242 157 243 170

± ± ± ±

Median neuroscore

Neurological deficit

35 27 23a,b 35 22a,c

2.44 2.57 1.34a,b 2.69 1.25a,c

Moderate-severe Moderate-severe Mild Moderate-severe Mild

37 20a,b 41 31a,c

2.73 1.48a,b 2.81 1.53a,c

Moderate-severe Mild ModerateMild

Values are mean ± SD. MCA occlusion was 1 h in all cases. Ischemic infarct volumes and neuroscores were measured at either 1 day or 3 days of reperfusion. The scale for neurological scoring is as follows: 0, no neurological deficit; 1, mild neurological deficit; 2, moderate neurological deficit; 3, severe neurological deficit and 4, very severe neurological deficit. ap < 0.05 compared to none control group, bp < 0.05 compared to the 3% DMSO (vehicle for AG490) infused group, cp < 0.05 compared to the control siRNA group by one-way ANOVA followed by Tukey-Kramer multiple comparisons post-test.

increased STAT1 phosphorylation was shown to mediate post-ischemic neuronal damage (West et al. 2004). As isoform-specific STAT phosphorylation inhibitors are not currently available, we employed the RNAi technique to specifically prevent STAT3 expression and thus phosphorylation. Rather than degrading STAT3 mRNA per se, our goal is to impair STAT3 mRNA translation sufficiently to minimize new STAT3 protein synthesis so that the STAT3 protein available for phosphorylation will be limited in the post-ischemic brain. To assure the effectiveness of the siRNA to knockdown STAT3 protein, we used a pool of four STAT3-specific sequences. The pSTAT3 positive cells were negligible in number in normal rat brain, but were significantly elevated in both the ipsilateral cortex and striatum at 24 h reperfusion following transient MCAO. Treating rats with control siRNA pool had no significant effect on the post-ischemic pSTAT3 immunoreactivity in either the cortex (Fig. 7a) or the striatum (Fig. 7b), whereas STAT3 siRNA treatment abrogated the post-ischemic pSTAT3 immunoreactive cell number by 78 ± 9% in both ipsilateral cortex (Fig. 7c) and striatum (Fig. 7d) compared to control siRNA injected rat cortex and striatum. Western blotting also showed significantly decreased pSTAT3 protein levels in the MCAO/STAT3 siRNA group compared to the MCAO/ control siRNA group (Fig. 5i). Treating ischemic rats with STAT3 siRNA or the control siRNA had no effect on pJAK2 protein level as observed by western blotting (Fig. 5i). Effect of STAT3 siRNA on post-ischemic STAT3 mRNA levels Real-time PCR analysis showed increased STAT3 mRNA levels in the rats subjected to transient MCAO. STAT3

mRNA expression increased significantly in the ipsilateral cortex at 6 h (by 3.2 ± 0.7 fold, p < 0.05), 24 h (5.1 ± 1.2 fold, p < 0.05) and 72 h (by 2.9 ± 0.5 fold, p < 0.05) of reperfusion compared to sham control (Fig. 7e). The MCAO/ control siRNA group sacrificed after 24 h reperfusion also showed a 4.7 ± 0.7 fold increase (p < 0.05) over the sham/ control siRNA group (Fig. 7e). In contrast, the MCAO/ STAT3 siRNA group sacrificed after 24 h reperfusion showed no significant increase in STAT3 mRNA levels over the sham/STAT3 siRNA group (Fig. 7e). The two injected sham groups (control siRNA and STAT3 siRNA) showed no significant differences in the STAT3 mRNA levels compared to uninjected sham group (Fig. 7e). None of the groups showed any change in the expression of the housekeeping transcripts 18S rRNA and glyceraldehyde-3-phosphate dehydrogenase (data not shown). STAT3 siRNA decreased infarct volume, neuronal loss, apoptosis and neurological dysfunction following transient MCAO Injecting rats with STAT3 siRNA decreased the cortical (by 36%, p < 0.05), striatal (by 25%, p < 0.05) and total (by 33%, p < 0.05) infarct volume compared to control siRNA (Table 1). Figure 8a shows thionine-stained serial coronal sections from the representative rats from MCAO/STAT3 siRNA and MCAO/control siRNA groups. No significant difference was observed in the infarct volume between MCAO/no drug and MCAO/control siRNA groups (Table 1). The number of apoptotic cells was significantly reduced in the MCAO/STAT3 siRNA group compared to the MCAO/control siRNA group in the ipsilateral peri-infarct cortex (by 64%, p < 0.05) and the peri-infarct striatum (by



1363

Fig. 6 AG490 treatment decreased infarction, neuronal damage and apoptosis following transient MCAO. Panel a shows serial coronal sections from the brains of rats from MCAO/DMSO (vehicle) and MCAO/AG490 (JAK2 phosphorylation inhibitor) groups. AG490 treatment resulted in a significantly smaller infarct (outlined with a black line) compared to vehicle control (a). The values on the left in panel a indicate the distance in mm from Bregma. AG490 treatment also prevented the post-ischemic cortical neuronal loss. The peri-infarct cortex of MCAO/DMSO group showed several TUNEL+ apoptotic cells at (b, left panel) which were significantly less in the MCAO/AG490 group (b, right panel). Similar result was observed in the peri-infarct striatum. Many TUNEL+ cells in the MCAO/DMSO group were also positive for the neuronal marker NeuN. The TUNEL+ cell number in the cortex and striatum of different groups is given in Table 2. Panel c

shows the microscopic images of cerebral cortex of rats from the sham/DMSO (left panel), MCAO/DMSO (middle panel) and MCAO/ AG490 (right panel) groups. No evidence of neuronal damage was observed in the cortical layers I–VI of the sham/DMSO group. The neuronal density in all the cortical layers decreased significantly in the MCAO/DMSO group; and the MCAO/AG490 group showed significant neuroprotection over the MCAO/DMSO group. The insets show the magnified portions outlined by black boxes in the panels in c. The images showed in this figure were obtained from representative rats from each group. Similar results were observed in the other rats in each group. The mean cortical, striatal and the total infarct volumes are given in Table 1. The rats were subjected to 1 h transient MCAO and 24 h reperfusion.

49%, p < 0.05) (Table 2). A representative section from the penumbral cortex of a rat from the MCAO/control siRNA group (Fig. 8b, left panel) showed that many of the TUNEL positive cells were NeuN positive neurons; there were significantly fewer of these cells in a section from a rat from the MCAO/STAT3 siRNA group (Fig. 8b, right panel). Similar to the AG490 study, some TUNEL positive cells in both groups (control and STAT3 siRNA-treated) were

negative for NeuN. The sham/control siRNA group showed no significant neuronal damage (Fig. 8c, left panel). After 24 h reperfusion, the MCAO/control siRNA group showed severe neuronal loss in all the cortical layers, including an almost total loss of layer V pyramidal neurons (Fig. 8c, middle panel). In contrast, the MCAO/STAT3 siRNA group showed significant neuroprotection with a higher cell density in all the cortical layers (Fig. 8c, right panel) compared to the



Table 2 TUNEL positive cell number in the peri-infarct area following transient MCAO in the rats treated with AG490, DMSO, STAT3 siRNA and control siRNA

(a)

(b)

(c)

(d)

TUNEL+ cells (/mm2)

3% DMSO AG490 Control si RNA STAT3 si RNA

n

Cortex

Striatum

10 10 10 12

221 89 237 84

255 161 286 156

± ± ± ±

49 15* 45 16*

± ± ± ±

35 27* 55 34*

Values represent mean ± SD. In each animal, the TUNEL+ cells were counted microscopically in three fields at ·300. MCA occlusion was 1 h in all cases. TUNEL+ cells were counted at 1 day of reperfusion. *p < 0.05 compared to the respective control group (one-way ANOVA followed by Tukey-Kramer multiple comparisons post-test).

(e)

MCAO/control siRNA group. The neurological deficits analyzed after 24 h reperfusion were mild in the MCAO/ STAT3 siRNA group compared to moderate to severe deficits observed in the MCAO/control siRNA group (the median neuroscores were 1.25 for the STAT3 siRNA group and 2.69 for the control siRNA group) (Table 1). Neuroprotection after 3 days of reperfusion in AG490 and STAT3 siRNA treated rats In adult spontaneous hypertensive rats, the infarct evolves rapidly in the first 24 h of reperfusion following transient MCAO. Between day 1 and day 3 of reperfusion, the size of infarct increases by < 10%. As described above, we observed significantly smaller infarcts in the rats treated with the JAK2 phosphorylation inhibitor AG490 and STAT3 siRNA than in their respective controls after 1 day of reperfusion. To confirm that preventing JAK-STAT3 activation is not merely delaying the development and maturation of infarct, we tested the effect of AG490 and STAT3 siRNA in a cohort of rats subjected to 3 days of reperfusion following transient MCAO. Both the groups showed significantly decreased cortical, striatal and total infarct volumes and improved neuroscores compared to the DMSO and control siRNA treated 3-day vehicle control groups (Table 1). The percent decreases for both the infarct volumes and the neuroscores were similar between the 1-day and 3-day AG490 and STAT3 siRNA groups compared to the respective control groups (Table 1). AG490 and STAT3 siRNA had no effect on physiological parameters The physiological parameters (mean arterial blood pressure, pH, PaO2, PaCO2, hemoglobin and blood glucose) were not significantly different in the five groups of rats at different periods (before, during, and after MCAO) (Table S1). No significant differences were observed between the five groups of rats (no treatment, DMSO, AG490, control siRNA and STAT3 siRNA treated) in the regional cerebral blood

Fig. 7 Intracerebral administration of STAT3 siRNA prevented postischemic STAT3 phosphorylation and STAT3 mRNA expression. The ipsilateral hemisphere of control siRNA pool injected rats subjected to transient MCAO showed several pSTAT3 positive cells (a is cortex and b is striatum). The pSTAT3 immunostaining was abrogated significantly in both the cortex (c) and striatum (d) of the STAT3 siRNA pool injected rats subjected to transient MCAO. All rats were subjected to a 24 h reperfusion following a 1-h transient MCAO. The scale bar is 100 lm. Realtime PCR analysis showed that transient MCAO significantly enhanced the expression of STAT3 mRNA between 6 h and 72 h of reperfusion (e). Treating rats with STAT3 siRNA prevented the focal ischemia-induced STAT3 mRNA expression. Control siRNA treatment had no effect on the STAT3 mRNA levels. Bars represent mean ± SD (four to five rats/ group). Statistics: *p < 0.05 compared to the respective sham group and p < 0.05 compared to the 24 h MCAO/no treatment group.

flow monitored before, during and after the MCAO (Fig. S1). Discussion

The results of this study can be briefly summarized as follows: • up-regulation of JAK2 and STAT3 phosphorylation following transient MCAO



(a)

1365

(b)

(c)

Fig. 8 STAT3 knockdown decreased infarction, neuronal damage and apoptosis following transient MCAO. Panel a shows coronal brain sections of rats from the MCAO/control siRNA and MCAO/STAT3 siRNA groups. STAT3 knockdown resulted in a smaller infarct (outlined with a black line) compared to control (b). The values on the left in panel a indicate mm from Bregma. STAT3 siRNA also prevented the post-ischemic cortical neuronal loss. The peri-infarct cortex of MCAO/ control siRNA group showed several TUNEL+ apoptotic cells (a, left panel) which were significantly lower in the MCAO/STAT3 siRNA group. Similar result was observed in the peri-infarct striatum. Many TUNEL+ cells in the MCAO/control siRNA group were also positive for the neuronal marker NeuN. The TUNEL+ cell number in the cortex and striatum of different groups is given in Table 2. Panel a shows the

microscopic images of cerebral cortex of rats from the sham/control siRNA (left panel), MCAO/control siRNA (middle panel) and MCAO/ STAT3 siRNA (left panel) groups. No evidence of neuronal damage was observed in the cortical layers I–VI of the sham/control siRNA group. The neuronal density in all the cortical layers decreased significantly in the MCAO/control siRNA group and the MCAO/STAT3 siRNA group showed significant neuroprotection over the MCAO/ control siRNA group. The insets show the magnified portions outlined by black boxes in the panels in c. The images showed in this figure were obtained from representative rats from each group. Similar results were observed in the other rats in each group. The mean cortical, striatal and the total infarct volumes are given in Table 1. The rats were subjected to 1 h transient MCAO and 24 h reperfusion.

• activated microglia/macrophages as the main cell type that show JAK2 and STAT3 phosphorylation in the ischemic brain • significantly decreased infarction, apoptosis and neurological dysfunction by preventing STAT3 phosphorylation with either a JAK2 inhibitor or STAT3 siRNA.

cytokines like IL-6, mediates in vivo ischemic neuronal damage. Increased cerebral STAT1 and STAT3 phosphorylation following transient focal ischemia was reported previously (Planas et al. 1996; Justicia et al. 2000; Suzuki et al. 2001). Although STAT1 activation was shown to promote ischemic neuronal damage (Takagi et al. 2002), the functional significance of STAT3 activation for ischemic outcome was not evaluated before. As JAK2 is up-stream to STAT3, we

This is the first study to provide evidence that increased phosphorylation of STAT3, a down-stream effector of gp130



assumed that blocking JAK2 phosphorylation prevents STAT3 phosphorylation, leading to decreased downstream biological consequences of STAT3 activation. To test this hypothesis, we infused rats with JAK phosphorylation inhibitor AG490, which abrogated both JAK2 and STAT3 phosphorylation in the post-ischemic brain. AG490 infusion also resulted in smaller infarcts, a decreased number of apoptotic cells, and improved neurological function compared to vehicle-infused controls. This suggests that prevention of ischemic-induced JAK2 and the down-stream STAT3 phosphorylation is neuroprotective. However, in addition to STAT3, pJAK2 can also phosphorylate STAT1, and a previous study showed that AG490 prevents phosphorylation of both STAT1 and STAT3 in the preconditioned heart (Bolli et al. 2003). Furthermore, Takagi et al. (2002) showed that focal ischemia in STAT1 knockout mice (which are incapable of showing increased STAT1 phosphorylation) induces less brain damage. Hence, the possibility that AG490 neuroprotected by preventing phosphorylation of both STAT1 and STAT3 can not be ruled out. The induction of ischemia in STAT3 knockout mice might decipher the specific role of post-ischemic STAT3 activation, but such studies are not feasible as STAT3 knockout mice die at the embryonic stage (O’Shea et al. 2002). Hence, to specifically prevent STAT3 phosphorylation, we used the RNAi technique, which is a sequence-specific post-transcriptional gene silencing mechanism triggered by double stranded siRNAs that target homologous gene transcripts for degradation, allowing efficient and high-throughput gene function analysis (Ichim et al. 2004). Using specific siRNA sequences, recent studies showed efficient knockdown of a2A-adrenergic receptors, glutamate receptor subunits (GluR2a and GluR2d), epidermal growth factor transgene, dopamine transporter and the tetraspanin CD81 gene in young and adult rodent brain (Shishkina et al. 2004; Thakker et al. 2004; Akaneya et al. 2005; Bahi et al. 2005; Thakker et al. 2005). The promise of siRNA was also shown in animal models of CNS injury. Using siRNA, polyglutamine-induced neurodegeneration caused by mutant ataxin-1 was inhibited (Xia et al. 2004), mutant huntingtin-induced motor abnormalities were reduced (Harper et al. 2005), and mutant SOD1-induced spinal cord damage was prevented in the mouse models (Ralph et al. 2005). The siRNA-induced knockdown of the tumor suppressor PTEN during global ischemia in brain (Ning et al. 2004) and knock-down of BDNF following hypoxia in spinal cord (Baker-Herman et al. 2004) was shown to induce neuroprotection in adult rats. In the present study, using a pool of four siRNAs directed to different sites on the STAT3 mRNA, we could achieve 88% decrease in STAT3 mRNA and 78% decrease in STAT3 phosphorylation in the post-ischemic rat brain compared to control siRNA pool treated rats. Prevention of the post-ischemic STAT3 phosphorylation led to significant neuroprotection and a decrease in neurological deficits.

Although normal levels of STAT3 activation are essential for cellular functions like growth and survival (Bowman et al. 2000), the present study shows that excess STAT3 activation after focal ischemia is detrimental to brain. This is understandable as IL-6 is known to be formed in excess after focal ischemia and its pro-inflammatory actions are mediated by a rapid activation of JAK/STAT signaling (Campbell 2005). In fact, STAT3 was first identified as a transcription factor that mediates the induction of many genes involved in acute inflammatory responses down-stream to IL-6 activation (Akira et al. 1994; Zhong et al. 1994). Once IL-6 binds to its receptors, it activates JAK2 and STAT3 phosphorylation and the pSTAT3 dimerizes and translocates into nucleus, where its binding to DNA increases cytokine gene expression to generate more interleukins including IL-6, IL-1b and TNF-a. This vicious cycle leads to sustained inflammation unless controlled. The pSTAT3 also stimulates the expression of suppressor of cytokine signaling (SOCS) family of proteins, which act as the intracellular negative feedback regulators of JAK and STAT phosphorylation to switch off or dampen the cytokine signal transduction (Rakesh and Agrawal 2005). Previous studies from our laboratory showed significant induction of SOCS3 in rat brain following focal ischemia and exacerbation of ischemic neuronal damage by prevention of SOCS3 expression using antisense oligonucleotides (Vemuganti et al. 2002). However, the endogenous SOCS3 induction might only partially prevent the IL-6-JAKSTAT inflammatory signaling. Preliminary studies from our laboratory showed significant prevention of post-ischemic infarction and decreased STAT3 phosphorylation in rat brain following adenovirus-mediated SOCS3 overexpression (Satriotomo et al. 2005; Vemuganti et al. 2005). The present studies extend these findings and indicate a neurotoxic role of excessive activation of STAT3 after focal cerebral ischemia. In addition to IL-6, other factors like leukemia inhibitory factor, ciliary neurotrophic factor, oncostatin-M and cardiotropin-1 can also modulate the JAK-STAT signal transduction system by acting via gp130 cytokine receptor (Suzuki et al. 2001). Hence, the currently observed increased JAK2-STAT3 phosphorylation after focal ischemia might not exclusively restricted to stimulation by IL-6. Focal ischemia-induced STAT3 phosphorylation was previously reported to be localized in various cell types including microglia/macrophages, astrocytes and neurons (Planas et al. 1996; Justicia et al. 2000; Suzuki et al. 2001). Using fluorescent double immunostaining of pSTAT3 with NeuN (neuronal marker), GFAP (astroglial marker), and ED1 and OX42 (microglia/macrophage markers), we observed increased pSTAT3 in all these cell types following focal ischemia confirming the previous studies. However, activated microglia and macrophages were observed to be the predominant cell type that shows pSTAT3 after transient MCAO. Increased JAK2 phosphorylation following transient MCAO was observed only in microglia and macrophages,



indicating that pJAK2 might be up-stream to microglia/ macrophage pSTAT3 in the post-ischemic brain. A previous study showed increased neuronal and astrocytic JAK1 phosphorylation following transient focal ischemia (Justicia et al. 2000), indicating that pJAK1 might be up-stream to neuronal and astroglial pSTAT3. This suggests a fine cellular compartmentation of JAK-STAT signaling system in the CNS. The functional significance of the compartmentation of ischemia-induced JAK-STAT activation is not clear at present. The 5¢-upstream promoter of the GFAP contains several cis-acting elements (Miura et al. 1990; Brenner and Messing 1996) that include potential STAT3 binding sites (Nakashima et al. 1999; Takizawa et al. 2001), which regulate GFAP expression. In CNS, the JAK/STAT pathway is known to play an important role in normal astrocyte differentiation during development and abnormal astrocytic responses after CNS injury (Sun et al. 2001; Xia et al. 2002). Increased phosphorylation of STAT3 in the GFAP+ astroglia observed in the post-ischemic brain might be one of the promoters of the astrogliosis known to occur after focal ischemia. On the other hand, previous studies showed that up-regulation of STAT3 phosphorylation in neurons play a significant neuroprotective role following conditional spinal axonal injury (Qiu et al. 2005). Erythropoietin-induced cultured retinal ganglion cell outgrowth capacity was shown to be mediated via increased STAT3 phosphorylation (Kretz et al. 2005). Thus, increased neuronal pSTAT3 might be an attempt by the neurons to promote their survival by inducing neuroprotective SOCS3 expression which was previously reported to be localized in the neurons of the post-ischemic brain (Vemuganti et al. 2002). However, results of the present study show that JAK2/STAT3 activated in the microglia/macrophages might be neurotoxic, as JAK2 inhibitor AG490 and STAT3 siRNA induced significant neuroprotection and improved neurological recovery following transient focal cerebral ischemia. In conclusion, STAT3 activation in different cellular subtypes may serve different functions. However, STAT3 activation to an inappropriately high level after a CNS insult like focal ischemia might be detrimental. Acknowledgements Funded by grants to R. Vemuganti (NIH RO1 NS044173, NIH RO1 NS049448, AHA Grant-in-Aid 0350164 N). The authors wish to thank Dr R. J. Dempsey (University of Wisconsin-Madison, USA) and Dr Y. Takeuchi (Kagawa University, Japan) for valuable suggestions.

Supplementary material The following supplementary material is available for this article online:

1367

Fig. S1. Regional cerebral blood flow during and after transient MCAO. The regional cerebral blood flow was not significantly different between the groups of rats treated with AG490, 3% DMSO (vehicle for AG490), STAT3 siRNA or control siRNA at any period (before MCAO, during MCAO and the first 30 min of reperfusion). Table S1. Physiological parameters during transient MCAO. This material is available as part of the online article from http:// www.blackwell-synergy.com

References Akaneya Y., Jiang B. and Tsumoto T. (2005) RNAi-induced gene silencing by local electroporation in targeting brain region. J. Neurophysiol. 93, 594–602. Akira S., Nishio Y., Inoue M., Wang X. J., Wei S., Matsusaka T., Yoshida K., Sudo T., Naruto M. and Kishimoto T. (1994) Molecular cloning of APRF, a novel IFN-stimulated gene factor 3 p91-related transcription factor involved in the gp130-mediated signaling pathway. Cell 77, 63–71. Bahi A., Boyer F., Kolira M. and Dreyer J. L. (2005) In vivo gene silencing of CD81 by lentiviral expression of small interference RNAs suppresses cocaine-induced behaviour. J. Neurochem. 92, 1243–1255. Baker-Herman T. L., Fuller D. D., Bavis R. W., Zabka A. G., Golder F. J., Doperalski N. J., Johnson R. A., Watters J. J. and Mitchell G. S. (2004) BDNF is necessary and sufficient for spinal respiratory plasticity following intermittent hypoxia. Nature Neurosci. 7, 48–55. Bolli R., Dawn B. and Xuan Y. T. (2003) Role of the JAK-STAT pathway in protection against myocardial ischemia/reperfusion injury. Trends Cardiovasc. Med. 13, 72–79. Bowman T., Garcia R., Turkson J. and Jove R. (2000) STATs in oncogenesis. Oncogene 19, 2474–2488. Brenner M. and Messing A. (1996) GFAP transgenic mice. Methods 10, 351–364. Campbell I. L. (2005) Cytokine-mediated inflammation, tumorigenesis, and disease-associated JAK/STAT/SOCS signaling circuits in the CNS. Brain Res. Rev. 48, 166–177. Darnell J. E. (2005) Validating Stat3 in cancer therapy. Nature Med. 11, 595–596. Dempsey R. J., Sailor K. A., Bowen K. K., Tureyen K. and Vemuganti R. (2003) Stroke-induced progenitor cell proliferation in adult spontaneously hypertensive rat brain: effect of exogenous IGF-1 and GDNF. J. Neurochem. 87, 586–597. Dhodda V. K., Sailor K. A., Bowen K. K. and Vemuganti R. (2004) Putative endogenous mediators of preconditioning-induced ischemic tolerance in rat brain identified by genomic and proteomic analysis. J. Neurochem. 89, 73–89. Gorina R., Petegnief V., Chamorro A. and Planas A. M. (2005) AG490 prevents cell death after exposure of rat astrocytes to hydrogen peroxide or proinflammatory cytokines: involvement of the Jak2/ STAT pathway. J. Neurochem. 92, 505–518. Harper S. Q., Staber P. D., He X., Eliason S. L., Martins I. H., Mao Q., Yang L., Kotin R. M., Paulson H. L. and Davidson B. L. (2005) RNA interference improves motor and neuropathological abnormalities in a Huntington’s disease mouse model. Proc. Natl. Acad. Sci. U S A 102, 5820–5825. Iadecola C. and Alexander M. (2001) Cerebral ischemia and inflammation. Curr. Opin. Neurol. 14, 89–94. Ichim T. E., Li M., Qian H., Popov I. A., Rycerz K., Zheng X., White D., Zhong R. and Min W. P. (2004) RNA interference: a potent tool for gene-specific therapeutics. Am. J. Transplant. 4, 1227–1236. Justicia C., Gabriel C. and Planas A. M. (2000) Activation of the JAK/ STAT pathway following transient focal cerebral ischemia: signaling through Jak1 and Stat3 in astrocytes. Glia 30, 253–270.



Kretz A., Happold C. J., Marticke J. K. and Isenmann S. (2005) Erythropoietin promotes regeneration of adult CNS neurons via Jak2/ Stat3 and PI3K/AKT pathway activation. Mol. Cell Neurosci. 29, 569–579. Miura M., Tamura M. and Mikoshiba K. (1990) Cell-specific expression of the mouse glial fibrillary acidic protein gene: identification of cis- and trans-acting promoter lements for astrocytes-specific expression. J. Neurochem. 55, 1180–1188. Nakashima K., Yanagisawa M., Arakawa H., Kimura N., Hisatsune T., Kawabata M., Miyazono K. and Taga T. (1999) Synergistic signaling in fetal brain by STAT3-Smad1 complex bridged by p300. Science 284, 479–482. Naylor M., Bowen K. K., Sailor K. A., Dempsey R. J. and Vemuganti R. (2005) Preconditioning-induced ischemic tolerance stimulates growth factor expression and neurogenesis in adult rat hippocampus. Neurochem. Int. 47, 565–572. Ning K., Pei L., Liao M. et al. (2004) Dual neuroprotective signaling mediated by downregulating two distinct phosphatase activities of PTEN. J. Neurosci. 24, 4052–4060. O’Shea J. J., Gadina M. and Schreiber R. D. (2002) Cytokine signaling in 2002: new surprises in the Jak/Stat pathway. Cell 109, S121– S131. Paxinos G. and Watson C. (1998) The Rat Brain in Stereotaxic Coordinates, 4th edn. Academic Press, New York. Planas A. M., Soriano M. A., Berruezo M., Justicia C., Estrada A., Pitarch S. and Ferrer I. (1996) Induction of Stat3, a signal transducer and transcription factor, in reactive microglia following transient focal cerebral ischaemia. Eur. J. Neurosci. 8, 2612– 2618. Qiu J., Cafferty W. B., McMahon S. B. and Thompson S. W. (2005) Conditioning injury-induced spinal axon regeneration requires signal transducer and activator of transcription 3 activation. J. Neurosci. 25, 1645–1653. Rakesh K. and Agrawal D. K. (2005) Controlling cytokine signaling by constitutive inhibitors. Biochem. Pharmacol. 70, 649–657. Ralph G. S., Radcliffe P. A., Day D. M et al. (2005) Silencing mutant SOD1 using RNAi protects against neurodegeneration and extends survival in an ALS model. Nature Med. 11, 429–433. Satriotomo I., Kalluri H., Bowen K. and Vemuganti R. (2005) Increased JAK and STAT phosphorylation following transient focal cerebral ischemia. J. Neurochem. 94, 96. Shishkina G. T., Kalinina T. S. and Dygalo N. N. (2004) Attenuation of alpha2A-adrenergic receptor expression in neonatal rat brain by RNA interference or antisense oligonucleotide reduced anxiety in adulthood. Neuroscience 129, 521–528. Song G., Cechvala C., Resnick D. K., Dempsey R. J. and Vemuganti R. (2001) GeneChip analysis after acute spinal cord injury in rat. J. Neurochem. 79, 804–815. Stoll G., Jander S. and Schroeter M. (2002) Detrimental and beneficial effects of injury-induced inflammation and cytokine expression in the nervous system. Adv. Exp. Med. Biol. 513, 87–113. Sun Y., Nadal-Vicens M., Misono S., Lin M. Z., Zubiaga A., Hua X., Fan G. and Greenberg M. E. (2001) Neurogenin promotes neurogenesis and inhibits glial differentiation by independent mechanisms. Cell 104, 365–376. Suzuki S., Tanaka K., Nogawa S., Dembo T., Kosakai A. and Fukuuchi Y. (2001) Phosphorylation of signal transducer and activator of transcription-3 (Stat3) after focal cerebral ischemia in rats. Exp. Neurol. 170, 63–71. Swanson R. A., Morton M. T., Tsao-Wu G., Savalos R. A., Davidson C. and Sharp F. R. (1990) A semiautomated method for measuring brain infarct volume. J. Cereb. Blood Flow Metab. 10, 290–293.

Takagi Y., Harada J., Chiarugi A. and MoskowitZ. M. A. (2002) STAT1 is activated in neurons after ischemia and contributes to ischemic brain injury. J. Cereb. Blood Flow Metab. 22, 1311–1318. Takizawa T., Yanagisawa M., Ochiai W., Yasukawa K., Ishiguro T., Nakashima K. and Taga T. (2001) Astrocyte differentiation of fetal neuroepithelial cells involving cardiotrophin-1-induced activation of STAT3. Cytokine 9, 272–279. Thakker D. R., Natt F., Husken D., Maier R., Muller M., van der Putten H., Hoyer D. and Cryan J. F. (2004) Neurochemical and behavioral consequences of widespread gene knockdown in the adult mouse brain by using nonviral RNA interference. Proc. Natl. Acad. Sci. U S A 101, 17 270–17 275. Thakker D. R., Natt F., Husken D., van der Putten H., Maier R., Hoyer D. and Cryan J. F. (2005) siRNA-mediated knockdown of the serotonin transporter in the adult mouse brain. Mol. Psychiat. 10, 7892–7789. Vemuganti R. (2005) Decreased expression of vesicular GABA transporter, but not vesicular glutamate, acetylcholine and monoamine transporters in rat brain following focal ischemia. Neurochem. Int. 47, 136–142. Vemuganti R. and Dempsey R. J. (2005) Carotid atherosclerotic plaques from symptomatic stroke patients share the molecular fingerprints to develop in a neoplastic fashion: a microarray analysis study. Neuroscience 131, 359–374. Vemuganti R., Dogan A., Todd K. G., Bowen K. K., Kim B. T., Rothstein J. D. and Dempsey R. J. (2001) Antisense knockdown of the glial glutamate transporter GLT-1, but not the neuronal glutamate transporter EAAC1, exacerbates transient focal cerebral ischemiainduced neuronal damage in rat brain. J. Neurosci. 21, 1876–1883. Vemuganti R., Bowen K. K., Dhodda V., Song G., Franklin J. and Dempsey R. J. (2002) Microarray analysis of the gene expression changes in rat brain after focal ischemia. J. Neurochem. 83, 1072– 1086. Vemuganti R., Dhodda V., Song G., Bowen K. K. and Dempsey R. J. (2003) Traumatic brain injury-induced acute gene expression changes in rat cerebral cortex identified by GeneChip analysis. J. Neurosci. Res. 71, 208–219. Vemuganti R., Dempsey R. J. and Bowen K. K. (2004) Inhibition of intercellular adhesion molecule-1 protein expression by antisense oligonucleotides is neuroprotective after transient middle cerebral artery occlusion in rat. Stroke 35, 179–184. Vemuganti R., Satriotomo I. and Bowen K. K. (2005) Role of STAT/ SOCS signaling in controlling stroke-induced inflammation and Brain Damage. J. Neurochem. 94, 117. West D. A., Valentim L. M., Lythgoe M. F., Stephanou A., Proctor E., van der Weerd L., Ordidge R. J., Latchman D. S. and Gadian D. G. (2004) MR image-guided investigation of regional signal transducers and activators of transcription-1 activation in a rat model of focal cerebral ischemia. Neuroscience 127, 333–339. Xia H., Mao Q., Paulson H. L. and Davidson B. L. (2002) siRNAmediated gene silencing in vitro and in vivo. Nature Biotechnol. 20, 1006–1010. Xia H., Mao Q., Eliason S. L., Harper S. Q., Martins I. H., Orr H. T., Paulson H. L., Yang L., Kotin R. M. and Davidson B. L. (2004) RNAi suppresses polyglutamine-induced neurodegeneration in amodel of spinocerebellar ataxia. Nature Med. 10, 816–820. Zea Longa E., Weinstein P. R., Carlson S. and Cummins R. (1989) Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke 20, 84–91. Zhong Z., Wen Z. and Darnell J. E. (1994) Stat3: a STAT family member activated by tyrosine phosphorylation in response to epidermal growth factor and interleukin-6. Science 264, 95–98.
