of May 7, 2018.This information is current as Activation

Bκ and NF-βGlycogen Synthase Kinase-3Cell Apoptosis via Ceramide-RegulatedAntibodies Cause NO-Mediated Endothelial

Dengue Virus Nonstructural Protein 1−Anti

Robert Anderson and Yee-Shin LinWei, Mei-Chun Chen, Trai-Ming Yeh, Hsiao-Sheng Liu, Chia-Ling Chen, Chiou-Feng Lin, Shu-Wen Wan, Li-Shiung

http://www.jimmunol.org/content/191/4/1744doi: 10.4049/jimmunol.1201976July 2013;

2013; 191:1744-1752; Prepublished online 12J Immunol 

MaterialSupplementary

6.DC1http://www.jimmunol.org/content/suppl/2013/07/12/jimmunol.120197

Referenceshttp://www.jimmunol.org/content/191/4/1744.full#ref-list-1

, 10 of which you can access for free at: cites 67 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2013 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Anti–Dengue Virus Nonstructural Protein 1 AntibodiesCause NO-Mediated Endothelial Cell Apoptosis viaCeramide-Regulated Glycogen Synthase Kinase-3band NF-kB Activation

Chia-Ling Chen,*,† Chiou-Feng Lin,*,†,‡ Shu-Wen Wan,*,† Li-Shiung Wei,* Mei-Chun Chen,*

Trai-Ming Yeh,†,x Hsiao-Sheng Liu,*,† Robert Anderson,*,†,{,‖,# and Yee-Shin Lin*,†

Immunopathogenetic mechanisms of dengue virus (DENV) infection are involved in hemorrhagic syndrome resulting from throm-

bocytopenia, coagulopathy, and vasculopathy. We have proposed a mechanism of molecular mimicry in which Abs against DENV

nonstructural protein 1 (NS1) cross-react with human endothelial cells and cause NF-kB–regulated immune activation and NO-

mediated apoptosis. However, the signaling pathway leading to NF-kB activation after the binding of anti-DENV NS1 Abs to en-

dothelial cells is unresolved. In this study, we found that anti-DENV NS1 Abs caused the formation of lipid raftlike structures, and

that disrupting lipid raft formation by methyl-b-cyclodextrin decreased NO production and apoptosis. Treatment with anti-DENV

NS1 Abs elevated ceramide generation in lipid rafts. Pharmacological inhibition of acid sphingomyelinase (aSMase) decreased anti-

DENV NS1 Ab-mediated ceramide and NO production, as well as apoptosis. Exogenous ceramide treatment induced biogenesis of

inducible NO synthase (iNOS)/NO and apoptosis through an NF-kB–regulated manner. Furthermore, activation of glycogen synthase

kinase-3b (GSK-3b) was required for ceramide-induced NF-kB activation and iNOS expression. Notably, anti-DENVNS1 Abs caused

GSK-3b–mediated NF-kB activation and iNOS expression, which were regulated by aSMase.Moreover, pharmacological inhibition of

GSK-3b reduced hepatic endothelial cell apoptosis in mice passively administered anti-DENV NS1 Abs. These results suggest that

anti-DENV NS1 Abs bind to the endothelial cell membrane and cause NO production and apoptosis via a mechanism involving the

aSMase/ceramide/GSK-3b/NF-kB/iNOS/NO signaling pathway. The Journal of Immunology, 2013, 191: 1744–1752.

Pathogenic autoantibodies are involved in several autoim-mune diseases and cancers (1). In particular, anti–endo-thelial cell Abs (AECAs) have been shown, at least par-

tially, to be involved in vasculitis and endothelial cell dysfunction

in patients with lupus, systemic sclerosis, thrombocytopenic pur-pura, and with microbial infections, such as dengue virus (DENV),hepatitis C virus, CMV, and group A Streptococcus (2–4). Thevascular dysfunction induced by AECAs can be mediated throughdirect targeting to cell-surface autoantigens followed by stimu-lating or blocking their intracellular signaling, or indirectly link-ing to immune cells, such as neutrophils, monocytes, and NK cellsto cause complement- and Ab-dependent cytotoxicity, as wellas vessel inflammation (4–6). The mechanisms of autoantibodygeneration after autoantigen exposure are important issues to ex-plore. Also, the molecular basis for AECA-mediated pathogeniceffects remains unclear.Infection of DENV causes dengue fever, dengue hemorrhagic

fever (DHF), and dengue shock syndrome (DSS), as clinicallycharacterized by the onset of fever, hemorrhagic tendency, throm-bocytopenia, and some evidence of plasma leakage (7, 8). For thedevelopment of severe DHF/DSS, both viral and immunopath-ogenetic factors are involved (7, 9–14). Ab-dependent enhance-ment (ADE) of DENV infection has long been recognized as animportant risk factor for severe dengue disease. Extrinsic ADEincreases viral load by increasing virus cell attachment, whereasintrinsic ADE may bias the immune response toward Th2 by in-ducing IL-10 production (15, 16). Furthermore, exuberant immunecell and endothelial cell activation resulting in cytokine and che-mokine production, coagulant, anticoagulant, and complementactivation are typically found in DHF/DSS patients (10, 17–19).DENV infection–induced autoimmunity is also evident by thegeneration of autoantibodies against platelets and endothelial cellsin dengue patients (20–23) and in mouse models (24, 25). Notably,the levels of autoantibodies are higher in DHF/DSS than in dengue

*Department of Microbiology and Immunology, National Cheng Kung UniversityMedical College, Tainan 701, Taiwan; †Center of Infectious Disease and SignalingResearch, National Cheng Kung University Medical College, Tainan 701, Taiwan;‡Institute of Clinical Medicine, National Cheng Kung University Medical College,Tainan 701, Taiwan; xDepartment of Medical Laboratory Science and Biotechnology,National Cheng Kung University Medical College, Tainan 701, Taiwan; {Departmentof Microbiology and Immunology, Dalhousie University, Halifax, Nova Scotia B3H4R2, Canada; ‖Department of Pediatrics, Dalhousie University, Halifax, Nova ScotiaB3H 4R2, Canada; and #Canadian Center for Vaccinology, Dalhousie University,Halifax, Nova Scotia B3H 4R2, Canada

Received for publication July 17, 2012. Accepted for publication June 6, 2013.

This work was supported by the National Science Council, Taiwan (Grants NSC100-2321-B006-004, NSC100-2325-B006-007, and NSC97-2320-B-006-004-MY3), theMultidisciplinary Center of Excellence for Clinical Trial and Research, Departmentof Health, Taiwan (Grant DOH101-TD-B-111-002), and the Center of InfectiousDisease and Signaling Research, Aim for the Top University Project, National ChengKung University, Taiwan.

Address correspondence and reprint requests to Prof. Yee-Shin Lin, Department ofMicrobiology and Immunology, National Cheng Kung University Medical College,1 University Road, Tainan, Taiwan 701. E-mail address: yslin1@mail.ncku.edu.tw

The online version of this article contains supplemental material.

Abbreviations used in this article: ADE, Ab-dependent enhancement; AECA, anti–endothelial cell Ab; aSMase, acid sphingomyelinase; CHL, chlorpromazine hydro-chloride; DENV, dengue virus; DHF, dengue hemorrhagic fever; DSS, dengue shocksyndrome; GSK-3b, glycogen synthase kinase-3b; HMEC-1, human microvascularendothelial cell line 1; iNOS, inducible NO synthase; LiCl, lithium chloride; MbCD,methyl-b-cyclodextrin; NS1, nonstructural protein 1; nSMase, neutral sphingomye-linase; PCNA, proliferating cell nuclear Ag; PDTC, pyrrolidine dithiocarbamate; PI,propidium iodide.

Copyright� 2013 by TheAmericanAssociation of Immunologists, Inc. 0022-1767/13/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1201976

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

fever patients (20, 22, 23). For AECAs in dengue autoimmunity,DENV nonstructural protein 1 (NS1) Abs in patient sera account,at least in part, for the endothelial cell cross-reactivity and apopto-sis induction (22). In vitro studies demonstrated that anti-DENVNS1 Abs induce endothelial cells to undergo apoptosis (26) and alsoregulate inflammatory activation (27). Moreover, in vivo studiesrevealed pathogenic effects of anti-DENV NS1 Abs on hepatitisthrough endothelial cell injury (28). These findings suggest thatanti-DENV NS1 Abs may act as AECAs, which promote DENV-induced autoimmune responses (4). In conclusion, molecularmimicry-based autoimmune mechanisms initiated by the immu-nogenicity of DENV NS1 protein are likely involved in denguedisease pathogenesis (29, 30).Anti-DENV NS1 Abs cause inducible NO synthase (iNOS)

expression and NO production in endothelial cells, which lead toNO-mediated cell apoptosis (4, 26). Anti-DENV NS1 Abs alsoinduce NF-kB–regulated inflammatory activation, including cy-tokine production and adhesion molecule expression (27). How-ever, the upstream signaling leading to these effects after anti-DENV NS1 binding to the cell surface is not well characterized.In this study, we aimed to elucidate the relevant signaling path-way(s) in particular, the role of lipid rafts and ceramide. The for-mation of lipid rafts, the membrane microdomains that are enrichedin cholesterol and glycosphingolipids, play an important role in cel-lular signaling in response to various stimuli (31–35). Ceramide,which is enriched in lipid rafts, acts as a second messenger of mul-tiple extracellular stimuli in cell survival, inflammation, and apo-ptotic cell death (36–41). We investigated the formation of raftstructure and the generation of cellular ceramide after binding ofanti-DENV NS1 Abs to endothelial cells. We determined criticalroles for lipid raft formation and ceramide generation in anti-DENV

NS1 Ab-induced glycogen synthase kinase-3b (GSK-3b)–regulatedNF-kB activation followed by NF-kB–regulated iNOS/NO bio-genesis in endothelial cells.

Materials and MethodsCell cultures and reagents

The human microvascular endothelial cell line 1 (HMEC-1) was obtainedfrom the Centers for Disease Control and Prevention (42) and passaged inculture plates containing endothelial cell growth medium M200 (CascadeBiologics, Portland, OR) supplemented with 2% FBS, 1 mg/ml hydro-cortisone, 10 ng/ml epidermal growth factor, 3 ng/ml basic fibroblastgrowth factor, 10 mg/ml heparin, and antibiotics. C2-ceramide (BioMol,Plymouth Meeting, PA), pyrrolidine dithiocarbamate (PDTC), BAY 11-7085 (Sigma-Aldrich, St. Louis, MO), and SB415286 (TOCRIS, Ellisville,MO) were dissolved in DMSO. Lithium chloride (LiCl), methyl-b-cy-clodextrin (MbCD), and chlorpromazine hydrochloride (CHL) (Sigma-Aldrich) were dissolved in serum-free medium.

Mice

C3H/HeN breeder mice were obtained from Charles River BreedingLaboratories (Kanagawa, Japan) and maintained on standard laboratoryfood and water in the Laboratory Animal Center of National Cheng KungUniversity Medical College. Their male 8-wk-old progeny were used forthe experiments. The animal use protocol was reviewed and approved bythe Institutional Animal Care and Use Committee. An in vivo experimentfor examining the pathogenic effects was performed according to previousstudies (28). Mice were injected i.v. with control IgG (500 mg), anti-DENVNS1 (500 mg), control IgG plus LiCl (20 mg/kg), or anti-DENV NS1 plusLiCl for 48 h.

Preparation of recombinant DENV NS1 and anti-DENV NS1 Abs

The recombinant protein expression and purification followed the proce-dures described previously for DENV2 NS1 preparation (26). For Abgeneration, BALB/c mice were i.p. immunized with PBS or 25 mg purified

FIGURE 1. Lipid raft formation is required for anti-

DENV NS1 Ab–induced NO production and apoptosis.

(A) Confocal microscopic detection of FITC-labeled anti-

DENV NS1 Abs (green) on HMEC-1 cells at both 4˚C

and 37˚C, and clustering with Texas Red–labeled cav-

eolin-1 (red) in raftlike structure (yellow, see Merge) at

37˚C, but not 4˚C. Scale bar is shown. (B) Fluorescence

tracings from (A) were analyzed for the colocalization

coefficients of anti-DENV NS1 and caveolin-1 by green

and red fluorescence intensity using Image-Pro Plus

version 6.0 software. (C) In the presence or absence of 2

mM lipid raft inhibitor MbCD, HMEC-1 cells were

treated with 10 mg control IgG, anti-DENV NS1, or anti-

pan cadherin at 37˚C for 1 h. The formation of lipid raft

structure in cell membranes was determined by affinity

dye staining using FITC-conjugated cholera toxin B, and

the cross-binding of control IgG, anti-DENV NS1, and

anti-pan cadherin was stained using Texas Red–conju-

gated anti-mouse and anti-rabbit Abs, followed by con-

focal microscopic analysis. Scale bar is shown. (D) The

fluorescence tracing from (C) was analyzed for the

colocalization coefficients of anti-DENV NS1 Abs and

cholera toxin B by red and green fluorescence intensity

using Image-Pro Plus version 6.0 software. (E) In the

presence or absence of 2 mM MbCD, the generation of

NO by 10 mg anti-DENV NS1 stimulation in HMEC-1

cells was determined by flow cytometry. The time-kinetic

responses of NO production by anti-DENV NS1 are

shown as the means 6 SD of triplicate cultures. (F)

TUNEL assay was performed to detect anti-DENV NS1

Ab–induced cell apoptosis followed by cytometric anal-

ysis. The percentages of apoptotic cells are shown as the

means6 SD of triplicate cultures. *p, 0.05 as compared

with the groups without MbCD treatment.

The Journal of Immunology 1745

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

recombinant DENV NS1 protein emulsified in CFA (Sigma-Aldrich) fol-lowed by four times in IFA. The IgG fractions from both PBS-immunizedcontrol and NS1-immunized mouse sera were purified using a proteinG-Sepharose affinity chromatography column (Amersham PharmaciaBiotech, Piscataway, NJ). All procedures for control IgG and anti-DENVNS1 purification were the same and performed simultaneously. The pre-parations were subjected to testing for endotoxin contamination using aLimulus amebocyte lysate assay (Pyrotell; Associates of Cape Cod, Falmouth,MA); the endotoxin concentrations of anti-DENV NS1 and control IgGwere both ,0.03 endotoxin unit/ml.

Immunostaining

Mouse anti-DENV NS1 or control IgG were incubated with cells at 4˚C or37˚C for 1 h. Cells were washed briefly in PBS, fixed with 1% formal-dehyde in PBS at room temperature for 10 min, and washed again withPBS followed by FITC-conjugated or Alexa 594–conjugated goat anti-mouse IgG (Invitrogen, Camarillo, CA). For the detection of lipid raftformation, specific Abs against raft-associated protein caveolin-1 (SantaCruz Biotechnology, Santa Cruz, CA), FITC-conjugated cholera toxinsubunit B, and Alexa 594–conjugated cholera toxin subunit B (Invitrogen)were used followed by confocal microscopy analysis (Leica TCS SPII,Nussloch, Germany; Olympus FV-1000, Tokyo, Japan). For the detectionof acid sphingomyelinase (aSMase) and ceramide generation, fixed or

nonfixed HMEC-1 cells were incubated respectively with specific Absagainst aSMase (Santa Cruz Biotechnology) and ceramide (GlycoTechProduktions and Handels, Kukels, Germany) at 4˚C or 37˚C for 1 h fol-lowed by FITC-conjugated goat anti-rabbit IgG or anti-mouse IgM (Cal-biochem, San Diego, CA) staining, and analyzed by confocal mi-croscopy or flow cytometry (FACSCalibur; BD Biosciences, San Jose,CA). Polyclonal rabbit anti-pan cadherin IgG (Abcam, Cambridge, MA) wasused for plasma membrane binding analysis, followed by Alexa 594–con-jugated anti-rabbit Ab staining. For the detection of ceramide generation,cells were analyzed using flow cytometry with excitation set at 488 nm. Theemission was detected with the FL-1 channel followed by CellQuest Pro4.0.2 software (BD Biosciences) analysis, and quantification was doneusing WinMDI 2.8 software (The Scripps Research Institute, La Jolla,CA). The ceramide-positive cells were gated and compared with untreated(mock) cells. For the measurement of NF-kB nuclear translocation, rabbitanti–NF-kB p65 (Cell Signaling Technology, Beverly, MA) was usedfollowed by FITC-conjugated goat anti-rabbit IgG staining and analyzedby fluorescence microscopy. DAPI (Calbiochem) was used for nuclearstaining. All immunostaining studies were performed in at least two in-dependent experiments. ImageJ software (National Institutes of Health,Bethesda, MD) and Image-Pro Plus version 6.0 software (Media Cyber-netics, Bethesda, MD) were used for image quantification analysis.

NO assay

The production of NO was detected using the ApoAlert NO/Annexin VDual Sensor Kit (Clontech, Palo Alto, CA) according to the manufacturer’sinstructions. In brief, cells were incubated with 5 mM NO Sensor Dye for30 min and then treated with anti-DENV NS1 or ceramide for indicatedperiods of time. Cells were washed and resuspended in PBS, then analyzedusing flow cytometry with excitation set at 488 nm. The emission wasdetected with the FL-1 channel. Samples were analyzed using CellQuestPro 4.0.2 software (BD Biosciences), and quantification was done usingWinMDI 2.8 software (The Scripps Research Institute). The NO+ cellswere gated and compared with untreated (mock) cells. The NO productionwas also detected using Griess reagent. In brief, 80 ml of the culture su-pernatant was reacted with 20 ml enzyme cofactor and nitrate reductase

FIGURE 2. Anti-DENV NS1 Ab treatment induces ceramide genera-

tion. (A) HMEC-1 cells were treated with 10 mg control IgG or anti-DENV

NS1 at 37˚C for 1 h. After treatment, the colocalization of lipid rafts and

ceramide generation in nonfixed HMEC-1 cells was determined using

Alexa 594–conjugated cholera toxin B (red) and anti-ceramide Abs (green)

followed by confocal microscopic analysis. Scale bar is shown. (B) The

fluorescence tracing from (A) was analyzed for the colocalization coef-

ficients of ceramide and cholera toxin B by green and red fluorescence

intensity using Image-Pro Plus version 6.0 software. (C) HMEC-1 cells

were treated with or without 10 mg control IgG or anti-DENV NS1 at 37˚C

for 1 h and then immunostained with anti-ceramide Abs after formalde-

hyde fixation. The generation of ceramide was determined by flow cyto-

metric analysis. Data are shown as the means 6 SD of triplicate cultures.

***p , 0.001 as compared with the control IgG–treated group.

FIGURE 3. Accumulation of aSMase in anti-DENV NS1 Ab–induced

ceramide generation, NO production, and apoptosis. (A) HMEC-1 cells

were treated with 10 mg control IgG or anti-DENV NS1 Ab at 37˚C for

1 h. The distribution of aSMase was determined by immunostaining using

specific Abs and analyzed by confocal microscopy. (B) In the presence or

absence of 1 mM of aSMase inhibitor CHL, the generation of ceramide

was measured using specific Ab staining followed by flow cytometric

analysis. Data are shown as the means 6 SD of triplicate cultures. *p ,0.05, **p , 0.01. (C) In the presence or absence of 1 mM CHL, the time-

kinetic responses of NO production after 10 mg anti-DENV NS1 Ab

treatment were detected by flow cytometry. (D) HMEC-1 cell apoptosis

was determined by TUNEL assay followed by cytometric analysis. Data

are shown as the means6 SD of triplicate cultures. *p, 0.05 as compared

with the groups without CHL treatment.

1746 ANTI-DENV NS1 MEDIATES CERAMIDE SIGNALING

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

mixture (Cayman Chemical Company, Ann Arbor, MI) for 30 min atroom temperature followed by adding 100 ml Griess reagent (1% sul-fanilamide, 0.1% naphthylethylenediamine dihydrochloride, and 2.5%H3PO4) and incubation for 10 min at room temperature. The relativeconcentration of nitrite was measured using a microplate reader (Emaxmicroplate reader; Molecular Devices) at 540 nm.

Cell apoptosis analysis

Cells were fixed with 70% ethanol in PBS for TUNEL reaction using theApoAlert DNA Fragmentation Assay Kit (Clontech) according to the manu-facturer’s instructions, or for propidium iodide (PI; Sigma-Aldrich) staining,and then analyzed by flow cytometry with excitation set at 488 nm. Theemission was detected with the FL-1 channel for TUNEL assay or the FL-2channel for PI assay. Samples were analyzed using CellQuest Pro 4.0.2 soft-ware (BD Biosciences), and quantification was done using WinMDI 2.8software (The Scripps Research Institute). The apoptotic cells were gated andcompared with untreated (mock) cells. To detect apoptotic cells in formalin-fixed sections of liver tissues, we performed TUNEL staining (Clontech)according to the manufacturer’s instructions. Apoptotic cells were analyzedbyfluorescencemicroscopy.DAPI (Calbiochem)wasused for nuclear staining.Tomeasurecell viability,weused theWST-8cell proliferationassaykit (CaymanChemical Company) according to the manufacturer’s instructions.

Immunoblotting analysis

Total cell lysates were extracted using a Triton X-100–based lysis buffercontaining 1% Triton X-100, 150 mM NaCl, 5 mM EDTA, 10 mM Tris(pH 7.5), 5 mM NaN3, 10 mM NaF, 10 mM sodium pyrophosphate, pro-

tease inhibitor mixture I, and protein phosphatase inhibitor mixture I (Sigma-Aldrich). Proteins were resolved using SDS-PAGE and then transferred toa polyvinylidene fluoride membrane (Millipore Corporation, Billerica, MA).After blocking, blots were developed with a series of Abs as indicated. Absspecific for GSK-3b, phospho–GSK-3b (S9), iNOS, and NF-kB were pur-chased from Cell Signaling Technology. Mouse Abs specific for prolifer-ating cell nuclear Ag (PCNA; ZYMED) and GAPDH (Santa Cruz Bio-technology) were used for internal controls. Finally, blots were hybridizedwith HRP-conjugated goat anti-rabbit IgG or anti-mouse IgG (Cell Sig-naling Technology) and developed using an ECLWestern blot detectionkit (Millipore Corporation) according to the manufacturer’s instructions.All immunoblotting studies were performed in at least two independentexperiments. The band intensity was measured using ImageJ software.

Statistical analysis

Between-group comparisons were performed using Student t test withSigmaPlot version 8.0 and GraphPad Prism version 5.0. The p values ,0.05 were considered significant.

ResultsLipid raft formation is involved in anti-DENV NS1 Ab-inducedNO production and apoptosis

We previously showed that anti-DENV NS1 Abs induced endo-thelial cell apoptosis, which was mediated through the productionof NO, upregulation of p53 and Bax, downregulation of Bcl-2 and

FIGURE 4. Exogenous ceramide treatment induces NF-kB–mediated iNOS expression, NO production, and apoptosis. (A) HMEC-1 cells were treated

with 50 mM C2-ceramide for various time periods as indicated. The expression of iNOS was determined by Western blot analysis. GAPDH was an internal

control. The ratios of iNOS and GAPDH are shown. (B) The generation of NO after 50 mM C2-ceramide treatment was determined by flow cytometry. The

dose- and time-kinetic responses of NO production induced by C2-ceramide are shown as the means 6 SD of triplicate cultures. (C) Endothelial cell

apoptosis induced by C2-ceramide stimulation for different dosages and time periods was measured using PI staining followed by flow cytometric analysis.

The percentages of apoptotic cells are shown as the means 6 SD of triplicate cultures. Cells treated with DMSO were used for negative control. *p , 0.05,

**p , 0.01 as compared with untreated group. (D) HMEC-1 cells were treated with 50 mM C2-ceramide for 2 h; NF-kB (green) nuclear translocation was

measured using a specific Ab and then FITC-labeled secondary Ab staining followed by fluorescence microscopic observation. Nucleus was stained by

DAPI (blue). Scale bar is shown. NF-kB nuclear translocation was quantified using ImageJ software, and the percentages of nuclear translocation are shown

as the means 6 SD of triplicate cultures. ***p , 0.001 as compared with the untreated group (right panel). (E) After 50 mM C2-ceramide treatment for the

indicated time periods, nuclear fractions were isolated. The expression level of NF-kB was determined. The nuclear protein PCNAwas used as an internal

control. The ratios of NF-kB and PCNA are shown. (F) In the presence or absence of 2.5 mM BAY or 50 mM PDTC, the protein expression of iNOS was

measured after 50 mM C2-ceramide treatment for 4 h. GAPDH was used as an internal control. The ratios of iNOS and GAPDH are shown.

The Journal of Immunology 1747

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Bcl-xL, cytochrome c release, and caspase-3 activation (4, 26, 29).However, the upstream signaling pathways of anti-DENVNS1Ab–induced endothelial cell apoptosis are still unclear. Because anti-DENVNS1Ab–induced apoptosis was initiated after Ab binding tocells, the involvement of biomembrane changes through lipid raftformation was first investigated. By immunostaining and confocalmicroscopic observation, the formation of membrane raftlike struc-tures by anti-DENV NS1 clustering with endothelial surface mole-cules was shown by colocalization with caveolin-1 protein (Fig.1A). The colocalization coefficients of anti-DENV NS1 Abs andcaveolin-1 were analyzed (Fig. 1B) and quantified (SupplementalFig. 1A, 1B). To confirm the formation of raftlike structures afteranti-DENVNS1 stimulation of endothelial cells, we performed cellstaining with cholera toxin B, a high-affinity protein for lipid rafts.Results showed that bindingof anti-DENVNS1Abs toHMEC-1cellscould induce lipid raft formation, which was inhibited by pretreat-ment with the lipid raft–specific inhibitor MbCD (Fig. 1C). Thecolocalization coefficients of anti-DENVNS1 Abs and cholera toxinBwere also analyzed (Fig. 1D) and quantified (Supplemental Fig. 1C,1D). The binding of anti-pan cadherin Abs to HMEC-1 could notinduce raft formation, which indicated that anti-DENV NS1 specifi-cally regulated lipid raft formation.In our previous studies, we found that anti-DENV NS1 Abs

induced NO-mediated endothelial cell apoptosis (26). To furtherstudy the effects of raft formation on anti-DENV NS1 Ab–inducedendothelial cell injury, we measured the NO production in thepresence of MbCD. Pretreatment with MbCD inhibited anti-DENV NS1 Ab–induced NO production (Fig. 1E) and cell apo-ptosis (Fig. 1F). These results suggest the involvement of lipidrafts in anti-DENV NS1 Ab–induced cell cytotoxicity.

Anti-DENV NS1 treatment induces ceramide generation

Lipid raft formation has been observed in response to multiplestimuli (31–33, 43–45). Generally, lipid rafts play an important rolein the activation of signal transduction after extracellular stimula-tion. Moreover, the generation of ceramide, an intracellular lipidwith signaling functions, has been demonstrated in lipid rafts (43,46, 47). To further investigate the involvement of lipid raft forma-tion in anti-DENV NS1 signaling, the accumulation of ceramidewas investigated. Using confocal microscopic analysis, we foundceramide-rich raft formation after anti-DENV NS1 treatment innonfixed HMEC-1 cells (Fig. 2A). The colocalization coefficientsof ceramide and cholera toxin B were also analyzed (Fig. 2B) andquantified (Supplemental Fig. 2). Anti-DENV NS1 Ab–inducedceramide generation was further demonstrated using flow cyto-metric analysis as compared with control IgG treatment (Fig. 2C).

aSMase regulates anti-DENV NS1 Ab–induced ceramidegeneration, NO production, and apoptosis

Ceramide can be generated from sphingomyelin hydrolysis byaSMase (41, 43, 46, 47). Interestingly, after anti-DENV NS1stimulation, the accumulation of aSMase could be detected byconfocal microscopic observation (Fig. 3A). The quantificationanalysis showed significant increase in cells with aSMase con-densation after anti-DENV NS1 Ab stimulation as compared withcontrol IgG treatment (Supplemental Fig. 3). Anti-DENV NS1Ab-induced ceramide generation was inhibited by pretreatmentwith CHL, an inhibitor of aSMase (Fig. 3B). Moreover, the pro-duction of NO (Fig. 3C) and cell apoptosis (Fig. 3D) by anti-DENVNS1 stimulation were also inhibited with CHL pretreatment. Anti-DENV NS1, but not control IgG, caused NO production and celldeath in HMEC-1 cells. Treatment with MbCD or CHL did notcause an effect in control IgG group (Supplemental Fig. 4). Forcompleteness,we also examined the role of neutral sphingomyelinase

(nSMase) in these processes. The role of nSMase in the forma-tion of ceramide-rich platforms is less clear than that of aSMase(41). The accumulation of nSMase on plasma membrane was alsodetectable after anti-DENV NS1 stimulation; however, the nSMaseinhibitor sphingolactone-24 (Sph-24) could not block anti-DENVNS1 Ab–induced ceramide generation (data not shown). Theseresults demonstrate that the activation of aSMase may accountfor ceramide generation, and trigger the subsequent NO productionand apoptosis in anti-DENV NS1 Ab–stimulated endothelial cells.

Exogenous ceramide treatment induces NF-kB–mediated iNOSexpression, NO production, and apoptosis

Because the generation of ceramide seems to play a role in anti-DENV NS1 Ab–induced endothelial cell injury, we treated cellsdirectly with ceramide. Results showed that ceramide treatmentcould induce iNOS protein expression (Fig. 4A), NO production(Fig. 4B), and cell apoptosis (Fig. 4C). NF-kB can be involved inthe cell death pathway mediated by ceramide (48, 49). Our pre-vious study also showed that anti-DENV NS1 Ab–induced en-dothelial cell activation was regulated by NF-kB activation (27).We therefore investigated the activation of NF-kB by ceramidestimulation. Results showed that ceramide caused NF-kB nucleartranslocation in HMEC-1 cells within 2 h (Fig. 4D, 4E). Ceramide-

FIGURE 5. GSK-3b activation regulates ceramide-induced NF-kB ac-

tivation and iNOS expression. (A) HMEC-1 cells were treated with DMSO

or 50 mM C2-ceramide for various time periods as indicated. The ex-

pression of phospho–GSK-3b at S9 (p-GSK-3b-S9), total GSK-3b, and

GAPDH were determined by specific Abs. The ratios of phospho–GSK-3b

and GSK-3b are shown. (B) In the presence or absence of 20 mM

SB415286 or 20 mM LiCl, nuclear fractions were isolated from HMEC-1

cells after 50 mM C2-ceramide treatment for 2 h followed by NF-kB de-

tection. PCNAwas used as an internal control marker of nuclear fractions.

The ratios of NF-kB and PCNA are shown. (C) In the presence or absence

of 20 mM SB415286 or 20 mM LiCl, HMEC-1 cells were treated with

50 mM C2-ceramide for the indicated time periods. The protein expression

of iNOS was measured. The expression of GAPDH was used as an internal

control. The ratios of iNOS and GAPDH are shown.

1748 ANTI-DENV NS1 MEDIATES CERAMIDE SIGNALING

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

induced iNOS protein expression was inhibited by pretreatment withNF-kB inhibitors, PDTC and BAY 11-7085 (Fig. 4F). These resultsdemonstrate that ceramide induces NF-kB activation followed byiNOS/NO biogenesis.

The activation of GSK-3b is involved in regulation ofceramide-induced NF-kB activation and iNOS expression

We previously showed ceramide-induced apoptosis via GSK-3bactivation in various types of cells (50). We further showed thatceramide induced dephosphorylation of GSK-3b at S9, which is theinhibitory phosphorylation site of GSK-3b, in HMEC-1 cells (Fig.5A). We therefore investigated the possible involvement of GSK-3bin regulating ceramide-mediated NF-kB activation and iNOS ex-pression. Pretreatment with GSK-3b inhibitors, SB415286 andLiCl, inhibited ceramide-induced NF-kB nuclear translocation(Fig. 5B). Furthermore, ceramide-induced iNOS expression wasalso blocked by GSK-3b inhibition (Fig. 5C).

Anti-DENV NS1 Ab induces GSK-3b–mediated NF-kBactivation

Because ceramide induces GSK-3b–regulated NF-kB activation andiNOS expression in HMEC-1 cells, we further investigated the in-volvement of GSK-3b in anti-DENV NS1 stimulation. Resultsshowed the dephosphorylation of GSK-3b at S9 after anti-DENVNS1Ab treatment (Fig. 6A). Pretreatment with SB415286 inhibitedanti-DENVNS1Ab–induced NF-kB nuclear translocation (Fig. 6B).Anti-DENV NS1 Ab–induced expression of iNOS was also inhibitedby GSK-3b inhibition (Fig. 6C). We further tested whether ceramidegeneration is involved in anti-DENV NS1 Ab–induced GSK-3b ac-tivation. Results showed that anti-DENV NS1 Ab–induced dephos-phorylation of GSK-3b at S9 was blocked by CHL pretreatment(Fig. 6D). These results suggest that anti-DENV NS1 Ab inducesGSK-3b–mediatedNF-kBexpression through ceramidegeneration.

LiCl inhibits anti-DENV NS1 Ab–induced hepatic endothelialcell apoptosis

We further confirmed the involvement of GSK-3b activation inanti-DENV NS1 Ab–induced endothelial cell injury using a pas-

sively immunized mouse model as described previously (28).After i.v. administration with anti-DENV NS1 Abs, mouse liverdamage as determined by serum levels of aspartate aminotrans-ferase and alanine aminotransferase was higher than the controlIgG group (28) (data not shown). We thus verified whether anti-DENV NS1 Ab–induced endothelial cell apoptosis in vivo wasinhibited by GSK-3b inactivation. TUNEL results showed thatLiCl treatment reduced anti-DENV NS1 Ab–induced hepatic en-dothelial cell apoptosis (Fig. 7A), which was further quantifiedand shown in Fig. 7B. These results suggest that GSK-3b playsa crucial role in regulating anti-DENV NS1 Ab–mediated endo-thelial cell injury (Fig. 8).

DiscussionDengue has become a globally emerging problem in public healthwith high incidence and extensive geographic distribution (13–15,51, 52). Although several candidate vaccines are under clinical trials(8, 53, 54), as yet no effective dengue vaccines are available. Acomplicating factor in dengue vaccine design is ADE, by which Absagainst viral surface proteins can amplify infection. Viral non-structural proteins, including NS1, are potential dengue vaccinecandidates to avoid ADE. However, anti-DENV NS1 Abs cross-react with endothelial cells by a mechanism of molecular mimicryand cause their dysfunction (29, 30). The cross-reactive Ab levelsthat bind to endothelial cells correlate with the disease severityand the disease phase (22). These results are consistent witha dengue pathogenetic model in which cross-reactive Abs play acontributory role. The generation of anti-DENV NS1 Abs maytherefore need to be considered in dengue vaccine development.In our previous studies, we explored the pathogenic effects,

including apoptosis and inflammation (26–28), induced by anti-DENV NS1 Abs. In this study, we further elucidate the molecularsignaling of the endothelial cell response and apoptosis caused byanti-DENV NS1 Abs. As summarized in Fig. 8, we found that thebinding of anti-DENV NS1 Abs to endothelial cells can efficientlycause lipid raft formation followed by aSMase-regulated ceramidegeneration. The sequential process includes GSK-3b–regulatedNF-kB activation followed by NF-kB–regulated iNOS/NO bio-

FIGURE 6. Ceramide-mediated GSK-3b activation

is involved in anti-DENV NS1 Ab-induced NF-kB ac-

tivation and iNOS expression. (A) HMEC-1 cells were

treated with or without 10mg control IgG or anti-DENV

NS1 for the indicated time periods, the expression of

phospho–GSK-3b at S9 (p-GSK-3b-S9) and total

GSK-3b was determined. GAPDH was used as an in-

ternal control. The ratios of phospho-GSK-3b and

GSK-3b are shown. (B) In the presence or absence of

20 mM SB415286, HMEC-1 cells were treated with or

without 10 mg control IgG or anti-DENV NS1 for 3 h.

NF-kB (green) nuclear translocation was measured us-

ing a specific Ab and then FITC-labeled secondary Ab

staining followed by fluorescence microscopic obser-

vation. Nuclei were stained with DAPI (blue). Scale bar

is shown. (C) In the presence or absence of 20 mM

SB415286 or 20 mM LiCl, the protein expression of

iNOS in HMEC-1 cells was measured after 10 mg

control IgG or anti-DENV NS1 Ab treatment for 8 h.

GAPDH was used as an internal control. The ratios of

iNOS and GAPDH are shown. (D) In the presence or

absence of 1mMCHL,HMEC-1 cells were treated with

or without 10 mg control IgG or anti-DENV NS1 for 2

h. The expression of phospho–GSK-3b at S9 (p-GSK-

3b-S9) and total GSK-3b was determined by specific

Abs. GAPDHwas used as an internal control. The ratios

of phospho–GSK-3b and GSK-3b are shown.

The Journal of Immunology 1749

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

genesis and apoptosis. These findings further clarify the patho-genic effects of AECAs in dengue disease pathogenesis caused byanti-DENV NS1 Abs.Lipid rafts have been shown to play important roles in cell

responses to multiple extracellular stimuli, including cell–cellcontact, cytokine and chemokine stimulation, drug therapy, and Ab-binding reactions (32, 33, 44, 45). For example, anti-CD95 (55, 56)and anti-phospholipid Abs (57) mediate protein clustering in theplasma membrane followed by lipid raft formation. This processgenerally triggers intracellular signal transduction and cellular ac-tivation, including apoptosis and inflammation. In this study, ourfindings show an essential role of lipid raft formation in anti-DENVNS1 Ab–induced signaling of NO production and apoptosis.Ceramide is enriched in lipid rafts and acts as the second

messenger in response to multiple stimuli (37, 41, 43, 46, 47). Wefound that ceramide is also increased in anti-DENV NS1 Ab–treated endothelial cells and that aSMase is required. Similar toCD95 signaling via aSMase activation and ceramide-rich lipid raftformation leading to cell death (55), anti-DENV NS1 stimulationalso induces apoptotic signaling through aSMase/ceramide. Inaddition to aSMase, ceramide is also generated by de novo bio-synthesis and by nSMase-mediated hydrolysis of sphingomyelin(37, 58). However, inhibiting ceramide synthase and nSMase did

not block anti-DENV NS1 Ab–induced ceramide generation (datanot shown).We further showed that treatment with exogenous ceramide

caused NO production and apoptosis in endothelial cells. Ceramide-regulated NO production has been reported in vascular smoothmuscle cells, astrocytes, microglias, macrophages, and C6 rat gliomacells after TNF-a and LPS/IFN-g stimulation (59, 60). Expressionof iNOS is required for NO production under inflammatory stim-ulation. We previously demonstrated that anti-DENV NS1 Abscould induce NF-kB activation and iNOS expression in endothelialcells (26, 27); however, the molecular mechanism upstream of NF-kB activation remains unclear. In this study, we demonstrated thatlipid raft formation and aSMase-mediated ceramide generation playimportant roles in NF-kB–regulated iNOS/NO biogenesis after anti-DENV NS1 Ab stimulation.In this study, we also found GSK-3b may act as an upstream

regulator for NF-kB activation in anti-DENV NS1 Ab–induced

ceramide signaling. As demonstrated in an in vitro model, both

exogenous ceramide and anti-DENV NS1 Ab treatment caused

GSK-3b activation followed by NF-kB activation and iNOS ex-

pression. Although several critical transcription factors, including

NF-kB, are regulated by GSK-3b, the effects of GSK-3b on NF-

kB in regulating apoptosis are complex (61–63). GSK-3b–facili-

tated NF-kB activation and NO biosynthesis were demonstrated in

this study and in a previous report (64). In anti-DENV NS1 Ab–

stimulated endothelial cells, aSMase effectively regulated GSK-

3b activation, suggesting an essential role for ceramide in this

pathway. The proapoptotic effect of ceramide through a GSK-3b–

regulated manner has been previously demonstrated (50). To

further demonstrate the pathogenic role of GSK-3b, we investi-

gated an in vivo model of anti-DENV NS1–induced endothelium

injury in liver using GSK-3b inhibitor LiCl. The results showed

a GSK-3b–regulated endothelial cell apoptosis in mice passively

administrated anti-DENV NS1 Abs. We have therefore demon-

strated the crucial regulatory role of GSK-3b in anti-DENV NS1

Ab–stimulated endothelial cell apoptosis not only in vitro but also

in vivo. However, LiCl could not significantly decrease the serum

FIGURE 7. Anti-DENV NS1 Ab–induced apoptosis of liver endothelial

cells is reduced by LiCl treatment in mice. (A) Liver sections were ob-

tained from control IgG–, anti-DENV NS1–, control IgG plus LiCl–, and

anti-DENV NS1 plus LiCl–treated mice (n = 3), and cell apoptosis

(arrows) was detected using a TUNEL assay (original magnification 320

or 3100). Scale bar is shown. (B) The TUNEL+ endothelial cells were

quantified from three mouse liver sections with six random fields for a total

of 18 fields. Each field may contain one to two vessels. ***p , 0.001.

FIGURE 8. Anti-DENV NS1 Abs induce endothelial cell NO produc-

tion and apoptosis via raft formation and ceramide generation, and GSK-

3b and NF-kB activation. The binding of anti-DENV NS1 to endothelial

cells can trigger aSMase accumulation in the cell membrane. Meanwhile,

the activation of aSMase accelerates ceramide generation and causes

ceramide-rich lipid raft formation. Ceramide generation prompts down-

stream signals of GSK-3b activation and further regulates NF-kB nuclear

translocation. The activation of NF-kB is required for iNOS expression.

Taken together, anti-DENV NS1 Ab-induced NO production and apoptosis

are regulated via ceramide-activated GSK-3b and NF-kB activation.

1750 ANTI-DENV NS1 MEDIATES CERAMIDE SIGNALING

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

levels of aspartate aminotransferase and alanine aminotransferaseinduced by anti-DENV NS1 Abs (data not shown). This is prob-ably due to the fact that, in addition to cell apoptosis, other factorssuch as complement activation induced by anti-DENV NS1 Absmay also be involved in vivo.The symptoms of DHF/DSS are multifaceted, including throm-

bocytopenia, coagulopathy, and vasculopathy, which are causedby viral effects and immunopathological host responses (7–15,51). Currently, no dengue vaccine is available, although DENVNS1 has been proposed as a possible candidate. Given the patho-logical effects reported for anti-NS1 Abs, the underlying mecha-nisms need to be characterized and the relevant epitopes identified.Considerable progress toward these goals has been achieved (30,65–68), raising the hope that safer NS1 candidate vaccines maybe developed with modified or deleted harmful epitopes such asthose involved in autoimmune responses.

DisclosuresThe authors have no financial conflicts of interest.

References1. Lleo, A., P. Invernizzi, B. Gao, M. Podda, and M. E. Gershwin. 2010. Definition

of human autoimmunity—autoantibodies versus autoimmune disease. Auto-immun. Rev. 9: A259–A266.

2. Savage, C. O., and S. P. Cooke. 1993. The role of the endothelium in systemicvasculitis. J. Autoimmun. 6: 237–249.

3. Salojin, K. V., M. Le Tonqueze, E. L. Nassovov, M. T. Blouch, A. A. Baranov,A. Saraux, L. Guillevin, J. N. Fiessinger, J. C. Piette, and P. Youinou. 1996. Anti-endothelial cell antibodies in patients with various forms of vasculitis. Clin. Exp.Rheumatol. 14: 163–169.

4. Lin, Y. S., C. F. Lin, H. Y. Lei, H. S. Liu, T. M. Yeh, S. H. Chen, and C. C. Liu.2004. Antibody-mediated endothelial cell damage via nitric oxide. Curr. Pharm.Des. 10: 213–221.

5. Tesfamariam, B., and A. F. DeFelice. 2007. Endothelial injury in the initiationand progression of vascular disorders. Vascul. Pharmacol. 46: 229–237.

6. Savage, C. O. 2009. Vascular biology and vasculitis. APMIS Suppl. 117: 37–40.7. Kurane, I. 2007. Dengue hemorrhagic fever with special emphasis on immu-

nopathogenesis. Comp. Immunol. Microbiol. Infect. Dis. 30: 329–340.8. Murphy, B. R., and S. S. Whitehead. 2011. Immune response to dengue virus and

prospects for a vaccine. Annu. Rev. Immunol. 29: 587–619.9. Pang, T., M. J. Cardosa, and M. G. Guzman. 2007. Of cascades and perfect

storms: the immunopathogenesis of dengue haemorrhagic fever-dengue shocksyndrome (DHF/DSS). Immunol. Cell Biol. 85: 43–45.

10. Rothman, A. L. 2011. Immunity to dengue virus: a tale of original antigenic sinand tropical cytokine storms. Nat. Rev. Immunol. 11: 532–543.

11. Lei, H. Y., T. M. Yeh, H. S. Liu, Y. S. Lin, S. H. Chen, and C. C. Liu. 2001.Immunopathogenesis of dengue virus infection. J. Biomed. Sci. 8: 377–388.

12. Clyde, K., J. L. Kyle, and E. Harris. 2006. Recent advances in deciphering viraland host determinants of dengue virus replication and pathogenesis. J. Virol. 80:11418–11431.

13. Green, S., and A. Rothman. 2006. Immunopathological mechanisms in dengueand dengue hemorrhagic fever. Curr. Opin. Infect. Dis. 19: 429–436.

14. Guzman, M. G., S. B. Halstead, H. Artsob, P. Buchy, J. Farrar, D. J. Gubler,E. Hunsperger, A. Kroeger, H. S. Margolis, E. Martınez, et al. 2010. Dengue:a continuing global threat. Nat. Rev. Microbiol. 8(12 Suppl.): S7–S16.

15. Halstead, S. B. 2007. Dengue. Lancet 370: 1644–1652.16. Halstead, S. B., S. Mahalingam, M. A. Marovich, S. Ubol, and D. M. Mosser.

2010. Intrinsic antibody-dependent enhancement of microbial infection inmacrophages: disease regulation by immune complexes. Lancet Infect. Dis. 10:712–722.

17. Nguyen, T. H., H. Y. Lei, T. L. Nguyen, Y. S. Lin, K. J. Huang, B. L. Le,C. F. Lin, T. M. Yeh, Q. H. Do, T. Q. Vu, et al. 2004. Dengue hemorrhagic feverin infants: a study of clinical and cytokine profiles. J. Infect. Dis. 189: 221–232.

18. Sosothikul, D., P. Seksarn, S. Pongsewalak, U. Thisyakorn, and J. Lusher. 2007.Activation of endothelial cells, coagulation and fibrinolysis in children withDengue virus infection. Thromb. Haemost. 97: 627–634.

19. Hatch, S., T. P. Endy, S. Thomas, A. Mathew, J. Potts, P. Pazoles, D. H. Libraty,R. Gibbons, and A. L. Rothman. 2011. Intracellular cytokine production bydengue virus-specific T cells correlates with subclinical secondary infection. J.Infect. Dis. 203: 1282–1291.

20. Lin, C. F., H. Y. Lei, C. C. Liu, H. S. Liu, T. M. Yeh, S. T. Wang, T. I. Yang,F. C. Sheu, C. F. Kuo, and Y. S. Lin. 2001. Generation of IgM anti-plateletautoantibody in dengue patients. J. Med. Virol. 63: 143–149.

21. Oishi, K., S. Inoue, M. T. Cinco, E. M. Dimaano, M. T. Alera, J. A. Alfon,F. Abanes, D. J. Cruz, R. R. Matias, H. Matsuura, et al. 2003. Correlation be-tween increased platelet-associated IgG and thrombocytopenia in secondarydengue virus infections. J. Med. Virol. 71: 259–264.

22. Lin, C. F., H. Y. Lei, A. L. Shiau, C. C. Liu, H. S. Liu, T. M. Yeh, S. H. Chen, andY. S. Lin. 2003. Antibodies from dengue patient sera cross-react with endothelialcells and induce damage. J. Med. Virol. 69: 82–90.

23. Saito, M., K. Oishi, S. Inoue, E. M. Dimaano, M. T. Alera, A. M. Robles,B. D. Estrella, Jr., A. Kumatori, K. Moji, M. T. Alonzo, et al. 2004. Associationof increased platelet-associated immunoglobulins with thrombocytopenia andthe severity of disease in secondary dengue virus infections. Clin. Exp. Immunol.138: 299–303.

24. Huang, K. J., S. Y. Li, S. C. Chen, H. S. Liu, Y. S. Lin, T. M. Yeh, C. C. Liu, andH. Y. Lei. 2000. Manifestation of thrombocytopenia in dengue-2-virus-infectedmice. J. Gen. Virol. 81: 2177–2182.

25. Sun, D. S., C. C. King, H. S. Huang, Y. L. Shih, C. C. Lee, W. J. Tsai, C. C. Yu,and H. H. Chang. 2007. Antiplatelet autoantibodies elicited by dengue virus non-structural protein 1 cause thrombocytopenia and mortality in mice. J. Thromb.Haemost. 5: 2291–2299.

26. Lin, C. F., H. Y. Lei, A. L. Shiau, H. S. Liu, T. M. Yeh, S. H. Chen, C. C. Liu,S. C. Chiu, and Y. S. Lin. 2002. Endothelial cell apoptosis induced by antibodiesagainst dengue virus nonstructural protein 1 via production of nitric oxide. J.Immunol. 169: 657–664.

27. Lin, C. F., S. C. Chiu, Y. L. Hsiao, S. W. Wan, H. Y. Lei, A. L. Shiau, H. S. Liu,T. M. Yeh, S. H. Chen, C. C. Liu, and Y. S. Lin. 2005. Expression of cytokine,chemokine, and adhesion molecules during endothelial cell activation inducedby antibodies against dengue virus nonstructural protein 1. J. Immunol. 174:395–403.

28. Lin, C. F., S. W. Wan, M. C. Chen, S. C. Lin, C. C. Cheng, S. C. Chiu,Y. L. Hsiao, H. Y. Lei, H. S. Liu, T. M. Yeh, and Y. S. Lin. 2008. Liver injurycaused by antibodies against dengue virus nonstructural protein 1 in a murinemodel. Lab. Invest. 88: 1079–1089.

29. Lin, C. F., S. W. Wan, H. J. Cheng, H. Y. Lei, and Y. S. Lin. 2006. Autoimmunepathogenesis in dengue virus infection. Viral Immunol. 19: 127–132.

30. Lin, Y. S., T. M. Yeh, C. F. Lin, S. W. Wan, Y. C. Chuang, T. K. Hsu, H. S. Liu,C. C. Liu, R. Anderson, and H. Y. Lei. 2011. Molecular mimicry between virusand host and its implications for dengue disease pathogenesis. Exp. Biol. Med.(Maywood) 236: 515–523.

31. Harder, T. 2003. Formation of functional cell membrane domains: the interplayof lipid- and protein-mediated interactions. Philos. Trans. R. Soc. Lond. B Biol.Sci. 358: 863–868.

32. Simons, K., and D. Toomre. 2000. Lipid rafts and signal transduction. Nat. Rev.Mol. Cell Biol. 1: 31–39.

33. Hoekstra, D., O. Maier, J. M. van der Wouden, T. A. Slimane, and S. C. vanIJzendoorn. 2003. Membrane dynamics and cell polarity: the role of sphingo-lipids. J. Lipid Res. 44: 869–877.

34. Li, P. L., Y. Zhang, and F. Yi. 2007. Lipid raft redox signaling platforms inendothelial dysfunction. Antioxid. Redox Signal. 9: 1457–1470.

35. Milhas, D., C. J. Clarke, and Y. A. Hannun. 2010. Sphingomyelin metabolism atthe plasma membrane: implications for bioactive sphingolipids. FEBS Lett. 584:1887–1894.

36. Pettus, B. J., C. E. Chalfant, and Y. A. Hannun. 2002. Ceramide in apoptosis: anoverview and current perspectives. Biochim. Biophys. Acta 1585: 114–125.

37. Lin, C. F., C. L. Chen, and Y. S. Lin. 2006. Ceramide in apoptotic signaling andanticancer therapy. Curr. Med. Chem. 13: 1609–1616.

38. Pandey, S., R. F. Murphy, and D. K. Agrawal. 2007. Recent advances in theimmunobiology of ceramide. Exp. Mol. Pathol. 82: 298–309.

39. Hannun, Y. A., and L. M. Obeid. 2008. Principles of bioactive lipid signalling:lessons from sphingolipids. Nat. Rev. Mol. Cell Biol. 9: 139–150.

40. Kitatani, K., J. Idkowiak-Baldys, and Y. A. Hannun. 2008. The sphingolipidsalvage pathway in ceramide metabolism and signaling. Cell. Signal. 20: 1010–1018.

41. Stancevic, B., and R. Kolesnick. 2010. Ceramide-rich platforms in transmem-brane signaling. FEBS Lett. 584: 1728–1740.

42. Ades, E. W., F. J. Candal, R. A. Swerlick, V. G. George, S. Summers,D. C. Bosse, and T. J. Lawley. 1992. HMEC-1: establishment of an immortalizedhuman microvascular endothelial cell line. J. Invest. Dermatol. 99: 683–690.

43. Gulbins, E. 2003. Regulation of death receptor signaling and apoptosis byceramide. Pharmacol. Res. 47: 393–399.

44. Patra, S. K. 2008. Dissecting lipid raft facilitated cell signaling pathways incancer. Biochim. Biophys. Acta 1785: 182–206.

45. George, K. S., and S. Wu. 2012. Lipid raft: A floating island of death or survival.Toxicol. Appl. Pharmacol. 259: 311–319.

46. Gulbins, E., and R. Kolesnick. 2003. Raft ceramide in molecular medicine.Oncogene 22: 7070–7077.

47. Gulbins, E., S. Dreschers, B. Wilker, and H. Grassme. 2004. Ceramide, mem-brane rafts and infections. J. Mol. Med. 82: 357–363.

48. Manna, S. K., N. K. Sah, and B. B. Aggarwal. 2000. Protein tyrosine kinasep56lck is required for ceramide-induced but not tumor necrosis factor-inducedactivation of NF-kappa B, AP-1, JNK, and apoptosis. J. Biol. Chem. 275: 13297–13306.

49. Gill, J. S., and A. J. Windebank. 2000. Ceramide initiates NFkappaB-mediatedcaspase activation in neuronal apoptosis. Neurobiol. Dis. 7: 448–461.

50. Lin, C. F., C. L. Chen, C. W. Chiang, M. S. Jan, W. C. Huang, and Y. S. Lin.2007. GSK-3beta acts downstream of PP2A and the PI 3-kinase-Akt pathway,and upstream of caspase-2 in ceramide-induced mitochondrial apoptosis. J. CellSci. 120: 2935–2943.

51. Srikiatkhachorn, A., and S. Green. 2010. Markers of dengue disease severity.In Dengue Virus, Current Topics in Microbiology and Immunology 338.A. L. Rothman, ed. Springer-Verlag, Berlin, Heidelberg, p. 67–82.

The Journal of Immunology 1751

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

52. Kyle, J. L., and E. Harris. 2008. Global spread and persistence of dengue. Annu.Rev. Microbiol. 62: 71–92.

53. Swaminathan, S., G. Batra, and N. Khanna. 2010. Dengue vaccines: state of theart. Expert Opin. Ther. Pat. 20: 819–835.

54. Thomas, S. J., and T. P. Endy. 2011. Critical issues in dengue vaccine devel-opment. Curr. Opin. Infect. Dis. 24: 442–450.

55. Grassme, H., A. Jekle, A. Riehle, H. Schwarz, J. Berger, K. Sandhoff,R. Kolesnick, and E. Gulbins. 2001. CD95 signaling via ceramide-rich mem-brane rafts. J. Biol. Chem. 276: 20589–20596.

56. Zhang, Y., X. Li, K. A. Becker, and E. Gulbins. 2009. Ceramide-enrichedmembrane domains—structure and function. Biochim. Biophys. Acta 1788:178–183.

57. Sorice, M., A. Longo, A. Capozzi, T. Garofalo, R. Misasi, C. Alessandri,F. Conti, B. Buttari, R. Rigano, E. Ortona, and G. Valesini. 2007. Anti-beta2-glycoprotein I antibodies induce monocyte release of tumor necrosis factor alphaand tissue factor by signal transduction pathways involving lipid rafts. ArthritisRheum. 56: 2687–2697.

58. Mathias, S., L. A. Pena, and R. N. Kolesnick. 1998. Signal transduction of stressvia ceramide. Biochem. J. 335: 465–480.

59. Katsuyama, K., M. Shichiri, F. Marumo, and Y. Hirata. 1998. Role of nuclearfactor-kappaB activation in cytokine- and sphingomyelinase-stimulated induc-ible nitric oxide synthase gene expression in vascular smooth muscle cells.Endocrinology 139: 4506–4512.

60. Won, J. S., Y. B. Im, M. Khan, A. K. Singh, and I. Singh. 2004. The role ofneutral sphingomyelinase produced ceramide in lipopolysaccharide-mediatedexpression of inducible nitric oxide synthase. J. Neurochem. 88: 583–593.

61. Beurel, E., and R. S. Jope. 2006. The paradoxical pro- and anti-apoptotic actionsof GSK3 in the intrinsic and extrinsic apoptosis signaling pathways. Prog.Neurobiol. 79: 173–189.

62. Dugo, L., M. Collin, and C. Thiemermann. 2007. Glycogen synthase kinase3beta as a target for the therapy of shock and inflammation. Shock 27: 113–123.

63. Beurel, E., S. M. Michalek, and R. S. Jope. 2010. Innate and adaptive immuneresponses regulated by glycogen synthase kinase-3 (GSK3). Trends Immunol. 31:24–31.

64. Kai, J. I., W. C. Huang, C. C. Tsai, W. T. Chang, C. L. Chen, and C. F. Lin. 2010.Glycogen synthase kinase-3b indirectly facilitates interferon-g-induced nuclearfactor-kB activation and nitric oxide biosynthesis. J. Cell. Biochem. 111: 1522–1530.

65. Wan, S. W., C. F. Lin, M. C. Chen, H. Y. Lei, H. S. Liu, T. M. Yeh, C. C. Liu, andY. S. Lin. 2008. C-terminal region of dengue virus nonstructural protein 1 isinvolved in endothelial cell cross-reactivity via molecular mimicry. Am. J. Infect.Dis. 4: 85–91.

66. Cheng, H. J., C. F. Lin, H. Y. Lei, H. S. Liu, T. M. Yeh, Y. H. Luo, and Y. S. Lin.2009. Proteomic analysis of endothelial cell autoantigens recognized by anti-dengue virus nonstructural protein 1 antibodies. Exp. Biol. Med. (Maywood) 234:63–73.

67. Chen, M. C., C. F. Lin, H. Y. Lei, S. C. Lin, H. S. Liu, T. M. Yeh, R. Anderson,and Y. S. Lin. 2009. Deletion of the C-terminal region of dengue virus non-structural protein 1 (NS1) abolishes anti-NS1-mediated platelet dysfunction andbleeding tendency. J. Immunol. 183: 1797–1803.

68. Perng, G. C., H. Y. Lei, Y. S. Lin, and K. Chokephaibulkit. 2011. Denguevaccines: challenge and confrontation. World J. Vaccines 1: 109–130.

1752 ANTI-DENV NS1 MEDIATES CERAMIDE SIGNALING

by guest on May 7, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from