RESEARCH Open AccessThe histone deacetylase inhibitor suberoylanilidehydroxamic acid attenuates human astrocyteneurotoxicity induced by interferon-γSadayuki Hashioka1,3*, Andis Klegeris2 and Patrick L McGeer1AbstractBackgrounds: Increasing evidence shows that the histone deacetylase inhibitor suberoylanilide hydroxamic acid(SAHA) possesses potent anti-inflammatory and immunomodulatory properties. It is tempting to evaluate thepotential of SAHA as a therapeutic agent in various neuroinflammatory and neurodegenerative disorders.Methods: We examined the effects of SAHA on interferon (IFN)-γ-induced neurotoxicity of human astrocytes andon IFN-γ-induced phosphorylation of signal transducer and activator of transcription (STAT) 3 in human astrocytes.We also studied the effects of SAHA on the astrocytic production of two representative IFN-γ-inducibleinflammatory molecules, namely IFN-γ-inducible T cell α chemoattractant (I-TAC) and intercellular adhesionmolecule-1 (ICAM-1).Results: SAHA significantly attenuated the toxicity of astrocytes activated by IFN-γ towards SH-SY5Y humanneuronal cells. In the IFN-γ-activated astrocytes, SAHA reduced the STAT3 phosphorylation. SAHA also inhibited theIFN-γ-induced astrocytic production of I-TAC, but not ICAM-1. These results indicate that SAHA suppressesIFN-γ-induced neurotoxicity of human astrocytes through inhibition of the STAT3 signaling pathway.Conclusion: Due to its anti-neurotoxic and anti-inflammatory properties, SAHA appears to have the therapeutic orpreventive potential for a wide range of neuroinflammatory disorders associated with activated astrocytes.Keywords: HDAC inhibitor, SAHA, STAT3, I-TAC, Astrocytes, Neuroinflammation, Neurodegenerative diseasesBackgroundSuberoylanilide hydroxamic acid (SAHA; also known asvorinostat, ChemBank ID 468) is the first histone deace-tylase (HDAC) inhibitor approved by the United StatesFood and Drug Administration. It was licensed in 2006for the treatment of cutaneous T-cell lymphoma (CTCL)[1]. HDAC inhibitors promote the acetylation of his-tones, which are generally associated with transcriptionalactivation. HDAC inhibitors also increase the acetylationstatus and modulate the activity of a wide range of non-histone proteins. Included are inflammatory transcriptionfactors, such as nuclear factor-κB and signal transducerand activator of transcription (STAT) 3 [1,2]. While vari-ous HDAC inhibitors have been studied and developedfor cancer therapy due to their anti-proliferative effects,increasing evidence shows that SAHA, at lower and non-cytotoxic concentrations, exhibits potent anti-inflammatoryand immunomodulatory activities in vitro [3-6] and in vivo[4,7]. Furthermore, animal studies indicate that SAHAcould ameliorate inflammatory bowel disease [3], hepatitis[4], lupus nephritis [5,6], graft versus host disease [7] andrheumatoid arthritis [8].A broad spectrum of neurodegenerative diseases, in-cluding Alzheimer disease (AD), Huntington disease(HD), Parkinson disease and multiple sclerosis, can beconsidered as chronic inflammatory disorders of thecentral nervous system (CNS) [9-11]. Chronic inflamma-tion associated with neuronal damage caused by cerebralischemia [12] and spinal cord injury [13] could beincluded. Chronic activation of astrocytes is believed to* Correspondence: hashioka@f2.dion.ne.jp1Kinsmen Laboratory of Neurological Research, Department of Psychiatry, theUniversity of British Columbia, 2255 Wesbrook Mall, Vancouver, BC V6T 1Z3,Canada3Department of Neuropsychiatry, Graduate School of Medical Sciences,Kyushu University, Maidasi 3-1-1, Higashi-ku, Fukuoka 812-8582, JapanFull list of author information is available at the end of the articleJOURNAL OF NEUROINFLAMMATION© 2012 Hashioka et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Hashioka et al. Journal of Neuroinflammation 2012, 9:113http://www.jneuroinflammation.com/content/9/1/113play an important role in the progression of neuroin-flammation, which includes causing damage to the sur-rounding neurons [10]. The STAT3 signaling pathwayhas been shown to mediate the neurotoxic secretion ofhuman astrocytes induced by interferon (IFN)-γ [14].The facts mentioned above motivated us to examinethe effects of SAHA on IFN-γ-induced neurotoxicityand STAT3 activation of human astrocytes. The purposewas to evaluate the potential of SAHA as a therapeuticagent in various neuroinflammatory and neurodegenera-tive disorders. In order to confirm the anti-inflammatoryproperties of SAHA, we also studied the effects ofSAHA on the astrocytic production of two representativeIFN-γ-inducible inflammatory molecules, IFN-γ-inducibleT cell α chemoattractant (I-TAC) and intercellular adhe-sion molecule-1 (ICAM-1).MethodsChemicals and reagentsSAHA was purchased from BioVision (MountainView,CA, USA). Human recombinant IFN-γ was purchasedfrom PeproTech (Rocky Hill, NJ, USA). 3-(4,5-dimethylthiazol-2-yl) 2,5-diphenyl tetrazolium bromide(MTT) and dimethyl sulfoxide (DMSO) were obtainedfrom Sigma-Aldrich (St. Louis, MO, USA). SAHA wasinitially dissolved in DMSO. The final concentration ofDMSO in tissue culture medium was less than 0.001%.At this concentration, DMSO had no effect on cellviability.Cell culturesThe human astrocytic U-373 MG cell line was obtainedfrom the American Type Culture Collection (ATCC,Manassas, VA, USA). The human neuroblastoma SH-SY5Y cell line was a gift from Dr. Robert Ross. Thesecells were grown in Dulbecco’s modified Eagle medium(DMEM), nutrient mixture F12 Ham (DMEM-F12) sup-plemented with 10% fetal bovine serum (FBS) and peni-cillin (200 U/ml)/streptomycin (200 μg/ml) (all fromInvitrogen Canada, Burlington, ON, Canada). Both celllines were used without initial differentiation.Human astrocytes were obtained from epilepticpatients undergoing temporal lobe surgery. The speci-mens were from normal tissue overlying the epilepticfoci. The use of human brain materials was approved bythe Clinical Research Ethics Board for Human Subjectsof the University of British Columbia. Astrocytes wereisolated as described previously [15,16]. They weregrown in DMEM-F12 supplemented with 10% FBS andpenicillin/streptomycin. The cells were cultured forthree to four weeks. Purity of the astrocyte cultures wasestimated by immunostaining with an antibody againstthe astrocytic marker glial fibrillary acidic protein(GFAP, from Dako, Z334, Carpinteria, CA, USA). Underour culture conditions, more than 99% cells were posi-tive for GFAP.Cytotoxicity of human astrocytes and U-373 MG cellstowards SH-SY5Y cellsHuman astrocytes or astrocytic U-373 MG cells wereseeded into 24-well plates at a concentration of 2 × 105cells/ml in 0.8 ml of DMEM-F12 medium containing 5%FBS. The cells were treated with various drugs for 1 hprior to the addition of activating stimulant (50 U/ml ofIFN-γ). The cells in the control group were incubatedwith medium only. After 24 h incubation of U-373 MGcells or 48 h incubation of astrocytes at 37°C, 0.4 ml ofcell-free supernatants were transferred to each well con-taining SH-SY5Y cells. At this time point, viability of U-373 MG or astrocytes was measured by the MTT assay.SH-SY5Y cells had been plated 24 h earlier at a concen-tration of 2 × 105 cells/ml in 0.4 ml of DMEM-F12medium containing 5% FBS. After 72 h incubation at 37°C, evaluation of surviving SH-SY5Y cells was performedby the MTT assay. The neuronal culture media weresampled for lactate dehydrogenase (LDH) to determineits release from dead cells. To establish that SAHA atthe concentration, which showed anti-neurotoxic effects,did not neutralize neurotoxins in the supernatants, thefollowing procedures were used. Supernatants fromastrocytes treated with IFN-γ for 48 h in the absence ofthe drug were collected first. One μM of SAHA wasadded into the supernatants just before applying them tothe SH-SY5Y cells. After 72 h incubation at 37°C, theSH-SY5Y cell viability was measured by the MTT assay.LDH release from dead cells was also measured.MTT assayMTT reduction was measured as described previously[17]. Briefly, the MTT reagent was added to cell culturesto reach a final concentration of 0.5 mg/ml. Following1 h incubation at 37°C, the dark crystals formed weredissolved by adding to the wells an equal volume of so-dium dodecyl sulfate/N, N-dimethylformamide (SDS/DMF) extraction buffer (20% SDS, 50% DMF, pH 4.7).Subsequently, plates were placed overnight at 37°C inorder to dissolve aggregates of lysed cells. Optical dens-ity (OD) was measured at 570 nm. Viable cell valueswere expressed as a percentage of the value obtainedfrom cells incubated in fresh medium only. The residualvalue for 0% cell survival was determined by lysing thecells with 1% Triton X-100.LDH assayLDH activity in supernatants was measured as describedpreviously [17]. Briefly, 100 μl of cell culture superna-tants were transferred into the wells of 96-well plates,followed by the addition of 15 μl of lactate solutionHashioka et al. Journal of Neuroinflammation 2012, 9:113 Page 2 of 8http://www.jneuroinflammation.com/content/9/1/113(36 mg/ml in phosphate-buffered saline (PBS)) and 15 μlof p-iodonitrotetrazolium violet solution (2 mg/ml inPBS). The enzymatic reaction was started by theaddition of 15 μl of NAD+/diaphorase solution (3 mg/mlNAD+; 2.3 mg solid/ml diaphorase). OD was measuredat 490 nm. The amount of LDH that had been releasedwas expressed as a fraction of the value obtained incomparative wells where the remaining cells were com-pletely lysed by 1% Triton X-100.Analysis of cellular morphologyIn order to analyze the morphological changes of SH-SY5Y cells, the cultures were observed with an invertedphase-contrast microscope (Axiovert 200, Carl Zeiss,Oberkochen, Germany) and photographed with a digitalcamera (Retiga 1300, Qimaging, Surrey, BC, Canada)72 h after transfer of supernatants from astrocytes. 40xand 20x objectives were used.Western blot analysisTotal protein was extracted from subconfluent humanastrocyte cultures in 10 cm culture dishes. Astrocyteswere incubated with or without SAHA for 1 h followedby incubation with 50 U/ml of IFN-γ for a further 30minutes. Astrocytes in the control group were incubatedwith medium only. The cells were washed twice withPBS and then fixed with 10% trichloroacetic acid for 30minutes at 4°C. Subsequently, the cells were scraped andlysed in ice-cold RIPA buffer (50 mM Tris–HCl (pH 8.0),150 mM NaCl, 1% deoxycholic acid, 1% TritonX100,0.1% SDS) supplemented with complete protease inhibi-tor cocktail (Roche Diagnostics, Mannheim, Germany).The lysed cells were sonicated and then centrifuged at13,000 g for 5 minutes at 4°C and the supernatants werecollected. Two μg of protein were subjected to SDS-polyacrylamide gel electrophoresis using an 8% acryl-amide gel at 120 V for 70 minutes. The protein wastransferred to a PVDF membrane at 70 V for 2 h. Themembrane was blocked with 5% skim milk plus 3% bo-vine serum albumin (BSA) in PBS at room temperature(RT) for 1 h. Subsequently, the membrane was incubatedwith specific rabbit antibodies against phospho-Tyr701-STAT1 (1:2,000), total STAT1 (1:1,000), phospho-Tyr705-STAT3 (1:2,000) or total STAT3 (1:1,000) at 4°C over-night and then treated with horseradish peroxidase-conjugated anti-rabbit IgG antibody (1:2,000) at RT for1 h. All antibodies used for immunoblotting were pur-chased from Cell Signaling Technology (Danvers, MA,USA). Blots were developed by the chemiluminescentECL system (Amersham, GE Healthcare, Buckingham-shire, UK). The band intensity was quantified by densi-tometry using the NIH Image analysis software version1.63 (NIH, Bethesda, MD, USA). Individual expressionlevel of phosphorylated STAT1 or STAT3 was normal-ized to the corresponding level of total protein.Measurement of I-TAC production: enzyme-linkedimmunosorbent assay (ELISA)Human astrocytes were seeded into 48-well plates at aconcentration of 2 x 105 cells/ml in 0.4 ml of DMEM-F12 medium containing 5% FBS. The cells were incu-bated in the presence or absence of SAHA for 1 h priorto the addition of activating stimulant (50 U/ml of IFN-γ). Astrocytes in the control group were incubated withmedium only. After 48 h incubation at 37°C, 100 μl ofcell-free supernatants were assayed for I-TAC accumula-tion. The concentrations of I-TAC were measured withan ELISA development kit supplied by PeproTech. Theassay was carried out according to the protocol suppliedby the manufacturer.Measurement of ICAM-1 expressionHuman astrocytes were seeded into 48-well plates at aconcentration of 2 x 105 cells/ml in 0.4 ml of DMEM-F12 medium containing 5% FBS. The cells were incu-bated in the presence or absence of SAHA for 1 h priorto the addition of activating stimulant (50 U/ml of IFN-γ). Astrocytes in the control group were incubated withmedium only. After 48 h incubation at 37°C, the cellswere fixed in 4% paraformaldehyde at 4°C for 5 minutesand then incubated with PBS containing 0.1% Triton X-100 at RT for 5 minutes. After blocking with 5% BSA inPBS for 1 h at RT, the cells were incubated with mono-clonal anti-ICAM-1 antibody (1:1,000; MU326-UC,1 H4, Biogenex, San Ramon, CA, USA) at RT for 2 h fol-lowed by incubation with alkaline phosphatase-conjugated goat anti-mouse IgG (1:3,000; Sigma-Aldrich) at RT for 2 h. After washing with PBS, they were incu-bated with 1 mg/ml of phosphate substrate (Sigma-Aldrich) in 0.1 M diethanolamine buffer (pH 9.8) at RTfor 1 h. Subsequently, OD was measured at 405 nm.StatisticsAll values are expressed as the means ± standard error ofmean (S.E.M.). Comparisons were made with a one-wayanalysis of variance (ANOVA) followed by the post hocTukey-Kramer test using StatView 5.0 software (SASInstitute Inc., Cary, USA). The significance was estab-lished at a level of P <0.05.ResultsEffects of SAHA on IFN-γ-induced neurotoxicity of humanastrocytes and astrocytoma cellsWe first investigated the effects of SAHA on IFN-γ-induced neurotoxicity of human astrocytic U-373 MGcells. The MTT assay revealed that SAHA did not affectthe U-373 MG cell viability in the 0.1 to 1 μM rangeHashioka et al. Journal of Neuroinflammation 2012, 9:113 Page 3 of 8http://www.jneuroinflammation.com/content/9/1/113(Figure 1A). U-373 MG cells caused significant toxicity to-wards SH-SY5Y cells after 24 h incubation with 50 U/mlof IFN-γ as shown by both the MTT (Figure 1B) andLDH assays (Figure 1C). Pretreatment of U-373 MG cellswith 1 μM of SAHA for 1 h significantly prevented theIFN-γ-induced neurotoxicity according to the MTT assay(Figure 1B). The LDH assay also showed significant reduc-tion of the IFN-γ-induced neurotoxicity by SAHA at 0.3and 1 μM (Figure 1C). In our preliminary studies, we con-firmed that 50 U/ml of IFN-γ when added directly to SH-SY5Y cells had no effect on their viability according to theMTTassay (data not shown).We further established the SAHA neuroprotection byusing primary human astrocytes. The MTT assaydemonstrated that SAHA did not affect the viability ofhuman astrocytes in the 0.1 to 1 μM range (Figure 2A).Human astrocytes caused significant toxicity towardsSH-SY5Y cells after 48 h incubation with 50 U/ml ofIFN-γ (Figure 2B, C). Similar to the results with U-373MG cells, 1 μM of SAHA significantly decreased theIFN-γ-induced neurotoxicity of human astrocytes(Figure 2B, C). To establish that SAHA acts directly onastrocytes and to rule out the possibility that it neutra-lizes neurotoxins, we collected supernatants from astro-cytes that had been stimulated with IFN-γ for 48 hwithout any drug treatment. We then added 1 μM ofSAHA into the supernatants just before applying themto SH-SY5Y cells. Addition of 1 μM SAHA did not affectthe SH-SY5Y cell viability compared with supernatantswithout such additions (Figure 2B, C, right bars), sug-gesting that SAHA does not act by neutralizing neuro-toxins following their secretion into the supernatants.The morphology of SH-SY5Y cells incubated in super-natants of human astrocytes was also analyzed. Thesupernatants of astrocytes stimulated with IFN-γ causedsignificant changes in cellular morphology (Figure 2E).The majority of the cells showed bright and circularlyshrunk cytoplasm in contrast to the typical healthymorphology presented in the control group (Figure 2D).This change was considerably attenuated by pretreat-ment with 1 μM SAHA (Figure 2F). These observationswere in line with the results obtained by both the MTTand LDH assay (Figure 2B, C).Effects of SAHA on IFN-γ-induced phosphorylation ofSTAT3 in human astrocytesOur recent studies have indicated that STAT3 signaling,but not STAT1 signaling, mediates IFN-γ-inducedneurotoxicity of human astrocytes [14]. Therefore, weinvestigated the effects of SAHA on the IFN-γ-inducedphosphorylation of Tyr701-STAT1 and Tyr705-STAT3 inhuman astrocytes. Treatment of astrocytes with 50 U/mlof IFN-γ for 30 minutes phosphorylated both Tyr701-STAT1 (Figure 3A) and Tyr705-STAT3 (Figure 3C).Densitometry revealed that 1 h pretreatment with 1 μMof SAHA significantly inhibited the STAT3 phosphoryl-ation (Figure 3D), while the drug did not affect theSTAT1 phosphorylation (Figure 3B). These results sug-gest that SAHA reduces IFN-γ-induced neurotoxicity ofhuman astrocytes via inhibition of STAT3 phosphorylation.Effects of SAHA on IFN-γ-induced I-TAC production andICAM-1 expression by human astrocytesWe finally examined the effect of SAHA on productionof the inflammatory chemokine I-TAC and on expres-sion of the inflammatory adhesion molecule ICAM-1 byhuman astrocytes stimulated with IFN-γ. Incubation ofastrocytes with 50 U/ml of IFN-γ for 48 h significantlyA CB0255075100125% live cellsControl - 0.1 0.3 1SAHA (μ M)Control - 0.1 0.3 1SAHA (μ M)Control - 0.1 0.3 1SAHA (μ M)0255075100125% live cells*05101520% lysed cells**Figure 1 Effects of SAHA on human astrocytic U-373 MG cell viability and their IFN-γ-induced toxicity toward SH-SY5Y humanneuronal cells. U-373 MG cells were incubated with or without SAHA at the concentrations indicated for 1 h before stimulation with IFN-γ (50U/ml). Control U-373 MG cells were incubated with medium only. After 24 h incubation, the cell-free supernatants of U-373 MG cells werecollected and the viability of U-373 MG was measured by the MTT assay (A). The collected supernatants were transferred to each well containingSH-SY5Y cells. After 72 h incubation, the SH-SY5Y cell viability was assessed by the MTT (B) and the LDH (C) assays. Data (means± S.E.M.) areexpressed as the percent of live cells, where the 100% value is obtained from either non-stimulated astrocytes in the control group (A) or SH-SY5Y cells incubated with fresh medium only (B), or the percent of lysed cells, where the 100% value is obtained from SH-SY5Y cells lysed by 1%Triton X-100 (C). *Significantly different from IFN-γ stimulation only. n = 6 to 7.Hashioka et al. Journal of Neuroinflammation 2012, 9:113 Page 4 of 8http://www.jneuroinflammation.com/content/9/1/113increased the I-TAC production (Figure 4A) and ICAM-1 expression (Figure 4B). SAHA significantly reducedthe IFN-γ-induced I-TAC production in aconcentration-dependent manner (Figure 4A). SAHA, inthe same concentration range, did not suppress the IFN-γ-induced ICAM-1 expression (Figure 4B).DiscussionThere were three major findings in the present study.First, SAHA significantly reduced the IFN-γ-inducedneurotoxicity of human astrocytes and U-373 MG cellsat non-cytotoxic concentrations. Second, SAHA inhib-ited the phosphorylation of Tyr705-STAT3 in humanastrocytes stimulated with IFN-γ. Third, SAHA signifi-cantly suppressed the I-TAC production, but not ICAM-1 expression, by IFN-γ-activated human astrocytes.The inhibitory effect of SAHA on human astrocyteneurotoxicity is compatible with emerging data fromseveral in vitro studies using various stimulated immunecells, which show anti-inflammatory properties ofSAHA. Specifically, treatment of SAHA has been shownto down-regulate the cellular production of inflamma-tory mediators, such as tumor necrosis factor (TNF)-α,interleukin (IL)-1β, IL-6, IL-12, IFN-γ and nitric oxide(NO), which are all potentially neurotoxic [3-6,18].Moreover, the anti-inflammatory activities of SAHAhave also been established in vivo. Administration ofSAHA has been demonstrated to reduce serum levels ofpro-inflammatory cytokines, including TNF-α, IL-1β andIFN-γ, in mice injected with lipopolysaccharide (LPS) [4]or mice transplanted with allogenic bone marrow [7,18].Therefore, SAHA appears to have therapeutic or pre-ventive potential for a wide range of neuroinflammatoryand neurodegenerative disorders. In fact, recent preclin-ical studies have shown that SAHA administration res-cues cognitive deficits in the APPswe/PS1dE9 transgenicmouse model of AD [19] and that SAHA administrationimproves motor impairments in the R6/2 transgenicmouse model of HD [20]. SAHA has been also demon-strated to decrease ischemic injury in the mouse brainsubjected to middle cerebral artery occlusion [21].Reduction of IFN-γ-induced STAT3 phosphorylationin human astrocytes by SAHA is consistent with recentin vitro studies, which showed that SAHA treatmentdecreased STAT3 phosphorylation in human CTCLHuT78 cells transfected with a STAT3-specific reporterconstruct [22] and in murine splenocytes stimulatedwith LPS [18]. Furthermore, HDAC inhibitors other thanSAHA, such as trichostatin A (TSA) [23] and AR-42[24], have also been reported to suppress STAT3A BSAHA (μ M)0255075100125% live cellsControl - 0.1 0.3 1SAHA (μ M)Control - 0.1 0.3 1*0255075100% live cellssupernatant +SAHA(1μ M) SAHA (μ M)Control - 0.1 0.3 1 supernatant +SAHA(1μ M)C051015202530% lysed cells*D E FFigure 2 Effects of SAHA on viability of human primary astrocytes and their IFN-γ-induced toxicity toward SH-SY5Y cells. Humanastrocytes were incubated with or without SAHA at the concentrations indicated for 1 h before stimulation with IFN-γ (50 U/ml) for 48 h (A-F), orSAHA was added directly to cell-free supernatants from IFN-γ-stimulated astrocytes (B, C, right bars). Astrocytes in the control group wereincubated with medium only. After 48 h incubation, cell-free supernatants were collected from astrocyte cultures and the astrocyte viability wasmeasured by the MTT assay (A). The collected supernatants were transferred to cultures of SH-SY5Y cells and their viability assessed by the MTT(B) and LDH (C) assays 72 h later. To establish that SAHA does not neutralize neurotoxins in stimulated supernatants, 1 μM of SAHA was addedinto the supernatants from astrocytes stimulated with IFN-γ (50 U/ml) for 48 h just before applying such supernatants to SH-SY5Y cell cultures.After 72 h incubation, the SH-SY5Y cell viability was assessed by the MTT assay (B, right bar) or the LDH assay (C, right bar). A phase-contrastmicroscopy with 20x and 40x (see inserts) objectives was used to analyze morphology of SH-SY5Y cells (D-F). D, control; E, IFN-γ alone; F, 1 μMSAHA+ IFN-γ. Every scale bar indicates 100 μm (D-F). Data (mean± S.E.M.) are expressed as percent of live cells, where the 100% value is obtainedfrom either non-stimulated astrocytes in the control group (A) or SH-SY5Y cells incubated with fresh medium only (B), or percent of lysed cells,where the 100% value is obtained from cells lysed by 1% Triton X-100 (C). *Significantly different from IFN-γ stimulation only. n = 6.Hashioka et al. Journal of Neuroinflammation 2012, 9:113 Page 5 of 8http://www.jneuroinflammation.com/content/9/1/113phosphorylation in various cancer cells. On the otherhand, it was reported that SAHA did not affect proteinlevels of phosphorylated STAT3 in HuT78 cells [25,26].We currently have no clear rationale for the discrepancyand consider that the effects of HDAC inhibition on theintracellular STAT3 phosphorylation remain inconclu-sive. Nevertheless, histone acetylation induced by HDACinhibitors may reduce STAT3 phosphorylation by up-regulating expression of suppressors of cytokine signaling(SOCS) 1 and SOCS3, which are negative regulators of theFigure 4 Effects of SAHA on I-TAC production and ICAM-1 expression by IFN-γ-activated human astrocytes. Human astrocytes werepretreated with or without SAHA at the concentrations indicated for 1 h before stimulation with IFN-γ (50 U/ml). Astrocytes in the control groupwere incubated with medium only. After 48 h incubation, I-TAC concentrations were determined in the cell-free supernatants of astrocytes (A)and the astrocytic expression of ICAM-1 was measured (B). *Significantly different from IFN-γ stimulation only. #Significantly different fromunstimulated control. n = 7. O.D., optical density units.A B 1.5SAHA (μM)Control - 0.1 1 1pSTAT1 96KDa0.5STAT1 96KDa pSTAT1/STAT1pSTAT3/STAT30Control - 0.1 1C D 0.5SAHA (μM) 0.4Control - 0.1 1pSTAT3 92KDa0.30.2 *STAT3 92KDa 0.10Control - 0.1 1Figure 3 Effects of SAHA on IFN-γ-induced phosphorylation of STAT1 and STAT3 in human astrocytes. Human astrocytes were incubatedwith or without SAHA at the concentrations indicated for 1 h. The cells were subsequently stimulated with IFN-γ for 30 minutes. Astrocytes in thecontrol group were incubated with medium only. Cell lysates were separated by 8% SDS-PAGE and immunoblotted for phospho-Tyr701-STAT1(pSTAT1) and total STAT1 (A) or phospho-Tyr705-STAT3 (pSTAT3) and total STAT3 (C). The density ratios of phosphorylated to total protein areshown as mean± S.E.M. of three independent experiments (B, D). *Significantly different from IFN-γ stimulation only.Hashioka et al. Journal of Neuroinflammation 2012, 9:113 Page 6 of 8http://www.jneuroinflammation.com/content/9/1/113Janus kinase/STAT signaling, as demonstrated by Xiong etal. (2012) [23].The finding that SAHA attenuates both IFN-γ-inducedneurotoxicity and IFN-γ-induced STAT3 phosphoryl-ation of human astrocytes is in line with our recentstudy which demonstrated that IFN-γ-induced neurotox-icity of human astrocytes is mediated, at least in part, bythe STAT3 signaling pathway [14]. Proton pump inhibi-tors [27] and L-type calcium channel blockers [28] havealso been demonstrated to suppress IFN-γ-inducedastrocytic neurotoxicity and STAT3 activation in humanastrocytes. These emerging data further support our hy-pothesis that neuroprotective activity of many reagentsthat reduce IFN-γ-induced neurotoxicity of humanastrocytes is exerted through inhibition of the STAT3signaling pathway in these cells.HDAC inhibitors have been shown to confer neuro-protection in experimental models of various neurode-generative diseases, including HD [29], amyotrophiclateral sclerosis [30] and multiple sclerosis [31], eventhough the exact mechanisms underlying their neuro-protective actions are still elusive. As we demonstratedin this study using SAHA, inhibition of activated astro-cytes by decreasing intracellular STAT3 phosphorylationseems to be one of the mechanisms. Effects of HDACinhibitors on astrocytes have not been studied well.SAHA is shown to inhibit the increased amount ofTNF-α and NO secretion from Abcd1/2-silenced murineastrocytes, which are associated with inflammatoryresponses [32]. TSA is indicated to alleviate 1-methyl-4-phenylpyridinium-induced impairment of glutamate up-take by rat astrocytes [33]. HDAC inhibitors are alsoreported to increase gene expression of the neurotro-phins glial cell line-derived neurotrophic factor andbrain-derived neurotrophic factor in rat astrocytes[34,35]. All these astrocytic events could contribute tothe HDAC inhibitor neuroprotection. The exploration ofthe relationship between HDAC inhibitor-elicited neuro-protection and astrocytic functions affected by HDACinhibitors appears to be still in its infancy.To the best of our knowledge, this is the first study todetermine the effects of SAHA on the cellular produc-tion of I-TAC. Our results showed that SAHA sup-pressed the IFN-γ-induced astrocytic production of I-TAC, a non-ELR CXC chemokine which attracts acti-vated T cells during immune and inflammatoryresponses. This finding is in agreement with a numberof previous studies, which have established that variousHDAC inhibitors exert anti-inflammatory effects viainhibiting levels of chemokines [36-39] as well as pro-inflammatory cytokines [37,40].We observed no influence of SAHA on the IFN-γ-induced astrocytic expression of ICAM-1, which con-trasts the data obtained by Takada et al. (2006) [41]demonstrating that SAHA represses the levels of ICAM-1 expressed by KBM-5 human myeloid cells stimulatedwith TNF-α. Further studies exploring this discrepancyare clearly warranted.SAHA may be suitable for a clinical intervention tar-geting the CNS due to its safety and permeability acrossthe blood–brain barrier (BBB). SAHA is generally welltolerated in clinical trials involving lymphoma patients[1,2] and is reported to cross the BBB and cause bio-logical responses in the mouse brain [20]. The aboveobservations combined with the major findings of thepresent study identify SAHA as an excellent candidatedrug for preclinical testing in a wide range of neuroin-flammatory disorders associated with activated astrocytes.AbbreviationsAD: Alzheimer disease; BBB: Blood–brain barrier; CNS: Central nervous system;CTCL: Cutaneous T-cell lymphoma; ELISA: Enzyme-linked immunosorbentassay; GFAP: Glial fibrillary acidic protein; HD: Huntington disease;HDAC: Histone deacetylase; ICAM-1: Intercellular adhesion molecule-1;IFN: Interferon; I-TAC: IFN-γ-inducible T cell α chemoattractant; IL: Interleukin;LDH: Lactate dehydrogenase; LPS: Lipopolysaccharide; MTT: 3-(4,5-dimethylthiazol-2-yl) 2,5-diphenyl tetrazolium bromide; NO: Nitric oxide;SAHA: Suberoylanilide hydroxamic acid; SOCS: Suppressors of cytokinesignaling; STAT: Signal transducer and activator of transcription; TNF: Tumornecrosis factor; TSA: Trichostatin A.Competing interestsAuthors declare that they have no competing interests.AcknowledgmentsThis research was supported by the Pacific Alzheimer Research Foundation(SH and PLM) and the Jack Brown and Family Alzheimer’s Disease ResearchFoundation (AK). This study was supported in part by Grant-in-Aid forScientific Research (C, #24591721) (SH).Author details1Kinsmen Laboratory of Neurological Research, Department of Psychiatry, theUniversity of British Columbia, 2255 Wesbrook Mall, Vancouver, BC V6T 1Z3,Canada. 2Department of Biology, I.K. Barber School of Arts and Sciences, theUniversity of British Columbia Okanagan, 3333 University Way, Kelowna, BCV1V 1 V7, Canada. 3Department of Neuropsychiatry, Graduate School ofMedical Sciences, Kyushu University, Maidasi 3-1-1, Higashi-ku, Fukuoka812-8582, Japan.Authors’ contributionsSH and PLM participated in the design of the study. SH carried out allexperiments, collected the data and performed the statistical analysis. SHand AK interpreted the data. SH drafted the manuscript. AK and PLM revisedthe manuscript. All authors read and approved the final manuscript.Received: 1 March 2012 Accepted: 30 May 2012Published: 30 May 2012References1. Grant S, Easley C, Kirkpatrick P: Vorinostat. Nat Rev Drug Discov 2007, 6:21–22.2. Lemoine M, Younes A: Histone deacetylase inhibitors in the treatment oflymphoma. Discov Med 2010, 10:462–470.3. Glauben R, Batra A, Fedke I, Zeitz M, Lehr HA, Leoni F, Mascagni P, Fantuzzi G,Dinarello CA, Siegmund B: Histone hyperacetylation is associated withamelioration of experimental colitis in mice. J Immunol 2006, 176:5015–5022.4. Leoni F, Zaliani A, Bertolini G, Porro G, Pagani P, Pozzi P, Dona G, Fossati G,Sozzani S, Azam T, Bufler P, Fantuzzi G, Goncharov I, Kim SH, Pomerantz BJ,Reznikov LL, Siegmund B, Dinarello CA, Mascagni P: The antitumor histonedeacetylase inhibitor suberoylanilide hydroxamic acid exhibitsantiinflammatory properties via suppression of cytokines. Proc Natl AcadSci U S A 2002, 99:2995–3000.Hashioka et al. Journal of Neuroinflammation 2012, 9:113 Page 7 of 8http://www.jneuroinflammation.com/content/9/1/1135. Mishra N, Reilly CM, Brown DR, Ruiz P, Gilkeson GS: Histone deacetylaseinhibitors modulate renal disease in the MRL-lpr/lpr mouse. J Clin Invest2003, 111:539–552.6. Reilly CM, Mishra N, Miller JM, Joshi D, Ruiz P, Richon VM, Marks PA,Gilkeson GS: Modulation of renal disease in MRL/lpr mice bysuberoylanilide hydroxamic acid. J Immunol 2004, 173:4171–4178.7. Reddy P, Maeda Y, Hotary K, Liu C, Reznikov LL, Dinarello CA, Ferrara JL:Histone deacetylase inhibitor suberoylanilide hydroxamic acid reducesacute graft-versus-host disease and preserves graft-versus-leukemiaeffect. Proc Natl Acad Sci U S A 2004, 101:3921–3926.8. Lin HS, Hu CY, Chan HY, Liew YY, Huang HP, Lepescheux L, Bastianelli E,Baron R, Rawadi G, Clement-Lacroix P: Anti-rheumatic activities of histonedeacetylase (HDAC) inhibitors in vivo in collagen-induced arthritis inrodents. Br J Pharmacol 2007, 150:862–872.9. McGeer EG, McGeer PL: Neuroinflammation in Alzheimer’s disease andmild cognitive impairment: a field in its infancy. J Alzheimers Dis 2010,19:355–361.10. McGeer PL, McGeer EG: Glial reactions in Parkinson’s disease. Mov Disord2008, 23:474–483.11. Klegeris A, McGeer PL: Non-steroidal anti-inflammatory drugs (NSAIDs)and other anti-inflammatory agents in the treatment ofneurodegenerative disease. Curr Alzheimer Res 2005, 2:355–365.12. Xia W, Han J, Huang G, Ying W: Inflammation in ischaemic brain injury:current advances and future perspectives. Clin Exp Pharmacol Physiol2010, 37:253–258.13. Fehlings MG, Nguyen DH: Immunoglobulin G: a potential treatment toattenuate neuroinflammation following spinal cord injury. J Clin Immunol2010, 30(Suppl 1):S109–S112.14. Hashioka S, Klegeris A, Qing H, McGeer PL: STAT3 inhibitors attenuateinterferon-gamma-induced neurotoxicity and inflammatory moleculeproduction by human astrocytes. Neurobiol Dis 2011, 41:299–307.15. Hashioka S, Klegeris A, Schwab C, McGeer PL: Interferon-gamma-dependent cytotoxic activation of human astrocytes and astrocytomacells. Neurobiol Aging 2009, 30:1924–1935.16. Hashioka S, Klegeris A, Schwab C, Yu S, McGeer PL: Differential expressionof interferon-gamma receptor on human glial cells in vivo and in vitro. JNeuroimmunol 2010, 225:91–99.17. Klegeris A, McGeer PL: Chymotrypsin-like proteases contribute to humanmonocytic THP-1 cell as well as human microglial neurotoxicity. Glia2005, 51:56–64.18. Leng C, Gries M, Ziegler J, Lokshin A, Mascagni P, Lentzsch S, Mapara MY:Reduction of graft-versus-host disease by histone deacetylase inhibitorsuberonylanilide hydroxamic acid is associated with modulation ofinflammatory cytokine milieu and involves inhibition of STAT1. ExpHematol 2006, 34:776–787.19. Kilgore M, Miller CA, Fass DM, Hennig KM, Haggarty SJ, Sweatt JD,Rumbaugh G: Inhibitors of class 1 histone deacetylases reversecontextual memory deficits in a mouse model of Alzheimer’s disease.Neuropsychopharmacology 2010, 35:870–880.20. Hockly E, Richon VM, Woodman B, Smith DL, Zhou X, Rosa E, Sathasivam K,Ghazi-Noori S, Mahal A, Lowden PA, Steffan JS, Marsh JL, Thompson LM,Lewis CM, Marks PA, Bates GP: Suberoylanilide hydroxamic acid, a histonedeacetylase inhibitor, ameliorates motor deficits in a mouse model ofHuntington’s disease. Proc Natl Acad Sci U S A 2003, 100:2041–2046.21. Faraco G, Pancani T, Formentini L, Mascagni P, Fossati G, Leoni F, Moroni F,Chiarugi A: Pharmacological inhibition of histone deacetylases bysuberoylanilide hydroxamic acid specifically alters gene expression andreduces ischemic injury in the mouse brain. Mol Pharmacol 2006,70:1876–1884.22. Tiffon C, Adams J, van der Fits L, Wen S, Townsend P, Ganesan A, Hodges E,Vermeer M, Packham G: The histone deacetylase inhibitors vorinostat andromidepsin downmodulate IL-10 expression in cutaneous T-celllymphoma cells. Br J Pharmacol 2011, 162:1590–1602.23. Xiong H, Du W, Zhang YJ, Hong J, Su WY, Tang JT, Wang YC, Lu R, Fang JY:Trichostatin A, a histone deacetylase inhibitor, suppresses JAK2/STAT3signaling via inducing the promoter-associated histone acetylation ofSOCS1 and SOCS3 in human colorectal cancer cells. Mol Carcinog 2012,51:174–184.24. Lin TY, Fenger J, Murahari S, Bear MD, Kulp SK, Wang D, Chen CS, KisseberthWC, London CA: AR-42, a novel HDAC inhibitor, exhibits biologic activityagainst malignant mast cell lines via down-regulation of constitutivelyactivated Kit. Blood 2010, 115:4217–4225.25. Fantin VR, Loboda A, Paweletz CP, Hendrickson RC, Pierce JW, Roth JA, Li L,Gooden F, Korenchuk S, Hou XS, Harrington EA, Randolph S, Reilly JF, WareCM, Kadin ME, Frankel SR, Richon VM: Constitutive activation of signaltransducers and activators of transcription predicts vorinostat resistancein cutaneous T-cell lymphoma. Cancer Res 2008, 68:3785–3794.26. Shao W, Growney JD, Feng Y, O’Connor G, Pu M, Zhu W, Yao YM, Kwon P,Fawell S, Atadja P: Activity of deacetylase inhibitor panobinostat (LBH589)in cutaneous T-cell lymphoma models: Defining molecular mechanismsof resistance. Int J Cancer 2010, 127:2199–2208.27. Hashioka S, Klegeris A, McGeer PL: Proton pump inhibitors reduceinterferon-gamma-induced neurotoxicity and STAT3 phosphorylation ofhuman astrocytes. Glia 2011, 59:833–840.28. Hashioka S, Klegeris A, McGeer PL: Inhibition of human astrocyte andmicroglia neurotoxicity by calcium channel blockers. Neuropharmacology.doi:10.1016/j.neuropharm.2012.05.033. in press.29. Gardian G, Browne SE, Choi DK, Klivenyi P, Gregorio J, Kubilus JK, Ryu H,Langley B, Ratan RR, Ferrante RJ, Beal MF: Neuroprotective effects ofphenylbutyrate in the N171-82Q transgenic mouse model ofHuntington’s disease. J Biol Chem 2005, 280:556–563.30. Petri S, Kiaei M, Kipiani K, Chen J, Calingasan NY, Crow JP, Beal MF: Additiveneuroprotective effects of a histone deacetylase inhibitor and a catalyticantioxidant in a transgenic mouse model of amyotrophic lateralsclerosis. Neurobiol Dis 2006, 22:40–49.31. Camelo S, Iglesias AH, Hwang D, Due B, Ryu H, Smith K, Gray SG, Imitola J,Duran G, Assaf B, Langley B, Khoury SJ, Stephanopoulos G, De Girolami U,Ratan RR, Ferrante RJ, Dangond F: Transcriptional therapy with thehistone deacetylase inhibitor trichostatin A ameliorates experimentalautoimmune encephalomyelitis. J Neuroimmunol 2005, 164:10–21.32. Singh J, Khan M, Singh I: HDAC inhibitor SAHA normalizes the levels ofVLCFAs in human skin fibroblasts from X-ALD patients anddownregulates the expression of proinflammatory cytokines in Abcd1/2-silenced mouse astrocytes. J Lipid Res 2011, 52:2056–2069.33. Wu JY, Niu FN, Huang R, Xu Y: Enhancement of glutamate uptake in 1-methyl-4-phenylpyridinium-treated astrocytes by trichostatin A.Neuroreport 2008, 19:1209–1212.34. Chen PS, Peng GS, Li G, Yang S, Wu X, Wang CC, Wilson B, Lu RB, Gean PW,Chuang DM, Hong JS: Valproate protects dopaminergic neurons inmidbrain neuron/glia cultures by stimulating the release of neurotrophicfactors from astrocytes. Mol Psychiatry 2006, 11:1116–1125.35. Wu X, Chen PS, Dallas S, Wilson B, Block ML, Wang CC, Kinyamu H, Lu N,Gao X, Leng Y, Chuang DM, Zhang W, Lu RB, Hong JS: Histone deacetylaseinhibitors up-regulate astrocyte GDNF and BDNF gene transcription andprotect dopaminergic neurons. Int J Neuropsychopharmacol 2008,11:1123–1134.36. Buglio D, Georgakis GV, Hanabuchi S, Arima K, Khaskhely NM, Liu YJ, YounesA: Vorinostat inhibits STAT6-mediated TH2 cytokine and TARCproduction and induces cell death in Hodgkin lymphoma cell lines.Blood 2008, 112:1424–1433.37. Halili MA, Andrews MR, Sweet MJ, Fairlie DP: Histone deacetylase inhibitorsin inflammatory disease. Curr Top Med Chem 2009, 9:309–319.38. Huang N, Katz JP, Martin DR, Wu GD: Inhibition of IL-8 gene expression inCaco-2 cells by compounds which induce histone hyperacetylation.Cytokine 1997, 9:27–36.39. Marumo T, Hishikawa K, Yoshikawa M, Hirahashi J, Kawachi S, Fujita T:Histone deacetylase modulates the proinflammatory and -fibroticchanges in tubulointerstitial injury. Am J Physiol Renal Physiol 2010, 298:F133–F141.40. Blanchard F, Chipoy C: Histone deacetylase inhibitors: new drugs for thetreatment of inflammatory diseases? Drug Discov Today 2005, 10:197–204.41. Takada Y, Gillenwater A, Ichikawa H, Aggarwal BB: Suberoylanilidehydroxamic acid potentiates apoptosis, inhibits invasion, and abolishesosteoclastogenesis by suppressing nuclear factor-kappaB activation. JBiol Chem 2006, 281:5612–5622.doi:10.1186/1742-2094-9-113Cite this article as: Hashioka et al.: The histone deacetylase inhibitorsuberoylanilide hydroxamic acid attenuates human astrocyteneurotoxicity induced by interferon-γ. Journal of Neuroinflammation 20129:113.Hashioka et al. Journal of Neuroinflammation 2012, 9:113 Page 8 of 8http://www.jneuroinflammation.com/content/9/1/113