mechanisms of coxsackievirus b5 mediatedβ-cell death depend on the multiplicity of infection

Journal of Medical Virology 72:586–596 (2004)

Mechanisms of Coxsackievirus B5 Mediated b-CellDeath Depend on the Multiplicity of Infection

Suvi Rasilainen,1,2* Petri Ylipaasto,3 Merja Roivainen,3 Risto Lapatto,2 Tapani Hovi,3

and Timo Otonkoski1,2

1Biomedicum Helsinki, Program of Developmental and Reproductive Biology, University of Helsinki, Helsinki, Finland2Hospital for Children and Adolescents, University of Helsinki, Helsinki, Finland3Enterovirus Laboratory, National Public Health Institute, Helsinki, Finland

Coxsackievirus infections may trigger and accel-erate pancreatic b-cell death, leading to type Idiabetes. Unrestricted coxsackievirus B5 replica-tion in cultured b-cells inoculated with highmultiplicityleadstorapidlyticcelldeath.Evidencefrom other virus-host cell systems indicates thathostcellresponsestoinfectionmaydependonthemultiplicityofinfection(MOI).Thus,theaimofthisstudy was to compare the mechanisms of b-celldeath during high versus low multiplicity of cox-sackievirus B5 infection. Cultures of highly differ-entiated mouse insulinoma cells and primaryadult human islets were infected with coxsack-ievirusB5atmultiplicitiesof>1,000or<0.5TCID50

per cell. The results of nuclear morphology andviability stainings, TUNEL staining and electro-phoretic DNA fragmentation analysis showedhigh multiplicity infection to predominantlyinduce necrosis and transient apoptosis. In lowmultiplicity culture, however, necrosis was onlymoderately induced and apoptosis increasedsteadily with time. This was best demonstratedby a tenfold higher apoptosis/necrosis ratio thanafter high multiplicity inoculation. Expression ofg-glutamyl cysteine synthetase increased in bothinfective cultures but the level of intracellularglutathione permanently depleted only at highmultiplicityandrecoveredfullyat lowmultiplicity.Thus, apoptosis represents an important mech-anism of b-cell death after low multiplicity ofcoxsackievirus B5 infection. This process is asso-ciated with maintenance of a physiological intra-cellular glutathione profile differing dramaticallyfrom the high multiplicity infection during whichnecrosisdominatesandintracellular thiolbalancedeteriorates. These data suggest that the patternand mechanisms of coxsackievirus B5 infectioninduced b-cell death depend on theMOI. J.Med.Virol. 72:586–596, 2004. � 2004Wiley-Liss, Inc.

KEY WORDS: enterovirus; diabetes; patho-genesis; islet; apoptosis

INTRODUCTION

Type 1 diabetes mellitus is a consequence of selective,gradual autoimmune destruction of the insulin prod-ucing b-cells in the islets of Langerhans. Although, T-cell-mediated autoimmune mechanisms are consideredthe major effectors of b-cell destruction, direct coxsack-ievirus infection in the pancreatic islets has beendocumented in rare cases of acute type I diabetes[Yoon et al., 1979]. However, prospective epidemiologi-cal studies suggest that enterovirus infections mightmore often be involved in launching and acceleratingthis process [Szopa et al., 1993; Andreoletti et al., 1997;Hiltunen et al., 1997; Nairn et al., 1999; Hyoty andTaylor, 2002; Sadeharju et al., 2003; Salminen et al.,2003]. Recent surveys have shown that several enter-ovirus serotypes are capable of infecting b-cells andshould thus be considered potentially diabetogenic[Roivainen et al., 1998, 2002].

One group of enteroviruses, coxsackieviruses B, hasbeen studied extensively and all six serotypes areconsidered to be associated with the pathogenesis oftype Idiabetes [Chehadehetal., 2000;HyotyandTaylor,2002; Yin et al., 2002]. Infections due to these virusesmay result in fulminant pancreatic damage with cyto-lytic b-cell death and development of overt diabetes[Yoon et al., 1979]. However,most of these infections arelikely to remain subclinical. Besides the strain andserotype of the virus, the aggressiveness of the infectionand the pattern of the resulting b-cell death may alsodepend on the amount of infective viral particles.

Grant sponsor: Academy of Finland; Grant sponsor: SigridJuselius Foundation; Grant sponsor: The Juvenile DiabetesResearch Foundation; Grant sponsor: International (JDRFI) andthe European Union; Grant number: BMH4-CT98-3952; Grantsponsor: Helsinki University Research Fund; Grant sponsor:Ahokas Foundation.

*Correspondence to: Suvi Rasilainen, MD, Biomedicum, roomC504b, PO Box 63 (Haartmaninkatu 8), FIN-00014 University ofHelsinki, Helsinki, Finland. E-mail: [email protected]

Accepted 28 October 2003

DOI 10.1002/jmv.20043

Published online in Wiley InterScience(www.interscience.wiley.com)

� 2004 WILEY-LISS, INC.

We have shown previously a productive coxsackie-virus B infection of high multiplicity to damage humanislet cells first through nuclear pyknosis, which finallyresults in lytic death.Moderate induction of apoptosis isalso evident [Roivainen et al., 2000]. Restriction of viralreplication in poliovirus (another enterovirus) infectedHeLa-cells shifts themode of cell death from pyknosis toapoptosis [Agol et al., 1998]. We have observed a similarphenomenon in primary porcine islets during a coxsack-ievirusB5 infection [Rasilainenetal., inpress]. Further-more, Theiler’s murine encephalomyelitis viruses(family Picornaviridae) and human immunodeficiencyvirus (HIV) kill susceptible cells lytically but restrictedcells apoptotically [Martin et al., 1994; Jelachich andLipton, 1996].

In view of these observations and considering thatboth in vitro [Roivainen et al., 2001] and in vivo, theamount of virus needed to give rise to a local infection isgenerally very small, the analysis of the mechanisms ofb-cell death as a function of the multiplicity of infection(MOI) becomes very important. In this study, we havecharacterised and compared the patterns of coxsack-ievirus B5 induced cell death mechanisms in insulinproducing MIN6 cells and in adult human isletsinoculatedwith lowversus highmultiplicity. Touncoversome possible pathways involved, the levels of glu-tathione and nitrite were measured.

MATERIALS AND METHODS

Cells, Media and Reagents

MIN6 mouse insulinoma cells provided by Dr.Jun-ichi Miyazaki, University of Osaka [Miyazakiet al., 1990]were cultured inDulbecco’smodifiedEagle’smedium (D-MEM) supplemented with 20 mM N-(2Hydroxyethyl)piperazine-N¢-(2-ethanesulfonic acid)(HEPES) pH 7.40 (Sigma, Taufkirchen, Germany),2 mM L-glutamine (Life Technologies, Tiby, Sweden),100 U/ml penicillin, 0.1 mg/ml streptomycin (LifeTechnologies) and 50 mM beta-mercaptoethanol (Gibco,Paisley, Scotland). Tissue culture treated 96-well and 6-well plates (NUNC, Roskilde, Denmark), 5-ml petridishes (Greiner Labortechnic, Vantaa, Finland) and 8-chamber tissue culture slides (Becton Dickinson, Mey-lan Cedex, France) were used for cell culture. Adulthuman islets were isolated and purified at the Scandi-navian Center for islet transplantation in Uppsala,Sweden, sent to Helsinki as free floating islets andcultured inHam’s F-10 supplementedwith 1mMbovineserumalbumin, 20mMHEPES, 100U/ml penicillin and0.1 mg/ml streptomycin. Non-adherent 10-ml petridishes (Sterilin, Staffordshire, UK) were used for isletculture before any experimental procedure. After treat-ments, non-adherent 6- and 96-well plates (BectonDickinson, Heidelberg, Germany) were used. TheFaulkner prototype strain of coxsackievirus B5 (familyPicornaviridae, genus Enterovirus) was obtained fromthe American Type Culture Collection (Manassas, VA)and passaged in green monkey kidney (GMK) cells, acontinuous cell line of green monkey kidney origin. The

identity of the virus was confirmed using a plaqueneutralization assay with type-specific antiserum.

Replication of Viruses

MIN6 cells were detached and allowed to adhereovernight to the appropriate culture carriages. BothMIN6 cells and adult human islets were infected withpreparations of either high (MOI>1,000) or low(MOI<0.5) multiplicity of coxsackievirus B5. Virusadsorption for 60 min at 368C was followed by removalof the inoculum virus, two washes with Hank’s balanc-ed salt solution (HBSS, Life Technologies, Taby,Sweden) supplemented with 100 U/ml penicillin and0.1 mg/ml streptomycin, 20 mM HEPES pH 7.40 and5 mMNaHCO3 (Merck,Whitehouse Station, NJ) and theaddition of cell-preparation-specific culture medium.The infection was carried on for 7–14 days in CO2-atmosphere at 368Cwith culturemedium changed twicea week.

For infectivity measurements, samples of MIN6 cellsandhumanisletswerecollected immediatelyafter (0hr),1, 2 and 3 days post infection in 200 and 100 ml volumesrespectively. All samples were collected before the firstculture medium refreshment. Samples were frozen andthawed three times to release the virus, clarified by lowspeed centrifugation and assayed for total infectivityusing endpoint dilutions in microwell cultures of GMKcells.

Immunocytochemistry for Enterovirus Antigen

Samples of MIN6 cells were collected after 1, 2 and 3days of infection. Cytocentrifuge preparations wereprepared, fixed with methanol (15 min at 48C), rinsedthree times with PBS and stored in PBS at 48C untilstained. First, the samples were rinsed once withPBSþ 0.1% BSA and exposed to PALTAVETGAT-antiserum (enterovirus specific polyclonal rabbit anti-serum 1/400 in PBSþ 0.1% BSAþ0.05% tween20) for60 min at 368C [Hovi and Roivainen, 1993]. Next, thesamples were rinsed three times with PBS and incu-bated with anti-rabbit FITC conjugate (Jackson Immu-noresearch Lab, West Grove, PA, diluted in PBSþ0.1%BSAþ0.05% tween20) for 30 min at 368C (1/250).Finally, the samples were rinsed four times with PBS,twice with sterile aqua and analysed by confocalmicroscopy (Leica TCS NT). A sample of non-infectedcells at every timepoint was used as a negative control.

Nuclear Double Staining With Hoechst33342 and Ethidium Homodimer-1

The nuclear dyes, 5 mg/ml Hoechst 33342 (HO,Sigma, St. Louis, MO) and 2 mM ethidium homodimer-1 (EthD-1,Molecular Probes, Leiden, TheNetherlands),were added to the culture medium of MIN6 cells on the8-chamber slides and into a suspension of human isletscollected into eppendorf tubes in 200ml volume.The cellswere exposed to the dyes for 30 min at þ368C. There-after, cytocentrifuge preparations were prepared from

Apoptosis of Coxsackievirus B5 Infected b-Cells 587

the human islet suspensions while MIN6 cells werewashed further with HBSS (supplemented as describedabove) and in both cases fixed with 3% paraformalde-hyde (PFA) for 30 min at room temperature (RT).Finally, all sampleswere examinedunder afluorescencemicroscopewith ultraviolet excitation at 340–380 nm (adual DAPI/PI filter). The number of viable, apoptotic,pyknotic and necrotic cells were scored as described pre-viously [Hoorens et al., 1996]. Briefly, HO freely passescell membranes and stains nuclear DNA blue. EthD-1is impermeable to intact membranes and only entersnecrotic and late-phase apoptotic cells and stains DNAred. Viable, pyknotic and necrotic cells are identified byintact nuclei with homogenous blue (HO), bright con-densed blue (HO) or homogenous red (HOþEthD-1)fluorescence respectively.Apoptotic cells are detectedbytheir fragmented nuclei that exhibit either blue (HO) orred (HOþEthD-1) fluorescence depending on the stageof the process.

Viability Staining by Calceinand Ethidium Homodimer-1

The viability of human islets was studied by acommercial live/dead cell assay kit (L-3224, MolecularProbes, Inc.) according to the manufacturers instruc-tions. Briefly, the assay is based on fluorescent doublestaining with calcein and EthD-1, the former of whichstains viable cells green due to their esterase activityand the latter, as previously mentioned, invadesdamaged nuclei and stains them red. The harvestedislets were exposed to the labelling solutions for 30 min(in RT, in the dark) after which preparations of cell-concentrates were analysed by confocal microscopy(Leica TCS NT).

TUNELþ Insulin Staining

Samples of adult human islets were collected after4 and 7 days of infection in 200-ml volume of whichcytocentrifuge preparations were prepared. After dry-ing for 5 min, the slides were fixed in 3% PFA for 30 minin RT followed by three rinses with PBS. Slides werestored in PBS at þ48C until stained. Prior to staining,the preparations were permeabilised with 0.1% citricacid and 0.1%Triton X-100 (Sigma, on ice for 2min) andwashed twice with steril aqua. To detect apoptosis, thepreparations were stained using the terminal dideox-ynucleotidetransferase (Tdt)-mediated dig-ddUTP nickend labelling (TUNEL) procedure [Roivainen et al.,2000]. Briefly, the preparations were preincubated withone time TdT buffer (Roche Diagnostics, Indianapolis,IN) for 10 min followed by DNA-nick end-labelling byTdt (RocheDiagnostics) for 60min at 378C. The sampleswere next exposed to 2% blocking reagent to preventnon-specific binding and then treated with anti-dig Fabfragments (Roche Diagnostics) for 60 min in 378C.The apoptotic nuclei were visualised in a 10–15 minexposure by a peroxidase dye, NBT/BCIP (RocheDiagnostics), followed by termination of the reactionwith 1 mM EDTA.

For detection of cell-specific apoptosis, the procedurewas continued with immunocytochemistry for insulinbeginning with blocking in 3% goat serum (Zymed,South SanFrancisco, CA) for 15min at RT and exposureto insulin antibody (1/500 guinea-pig anti-porcineinsulin antibody, DAKO Carpinteria, CA, in 3% goatserum) overnight at þ48C. After two rinses with PBS,the samples were incubated for 30 min at RT withbiotinylated-goat anti-rabbit IgG (1/200 in PBS, Zymed,San Francisco, CA), rinsed again twice with PBS andincubated for 30 min with peroxidase conjugatedstreptavidin (1/200 in PBS, Zymed). Finally, the insulinsignal was developed with 3-amino-9-ethylcarbazoleAEC substrate (up to 10 min) and rinsed with PBS anddistilled water. Light counterstaining was carried outwith Meyer’s haematoxylin.

DNA Fragmentation Analysis

Samples of MIN6 cells and human islets werecollected after 4 days of infection to harvest DNA. Acommercial kit was used according to the manufac-turer’s instructions (Apoptotic DNA Ladder Kit, Boeh-ringer Mannheim, Germany). Extracted DNA samplesunderwent treatment with RNase A (500 mg/ml).Apoptotic human testisDNA served as a positive controland a 100-bp DNA marker was used for sizing linearfragments. The DNA content was measured spectro-photometrically and 2 mg of DNA per sample was pre-cipitated and resuspended in 10 ml of sterile water. Themethod used for apoptotic DNA ladder analysis hasbeen described previously by Pentikainen et al. [2000].Briefly, the samples were 30-end labelled with digoxi-genin-dideoxy-UTP and fractionated through a 2%SeAKem Agarose gel followed by denaturing andneutralising procedures. The gel was then blotted ontoa positively charged nylon hybridization transfer mem-brane (Amersham, Freiburg, Germany) overnight. Nextday, the DNA was UV-crosslinked onto the membraneand treated with 1% Blocking Reagent (Roche Diagnos-tics) for 30 min at RT followed by anti-digoxigenin anti-body exposure (Roche Diagnostics, 1/4,000) for 30 minat RT. After two rinses with washing buffer, thebound antibody was detected with CSPD (BoehringerMannheim) yielding a chemiluminescent signal. Filmexposures of 20 min to 2 hr were sufficient.

Measurement of Glutathione

To analyse the levels of glutathione, samples ofculture medium and cells were collected separately at0 hr, 1, 2, 3, 5, 7, 10 and 14 days after infection. First,100 ml of culture medium was collected from a 96-welland treated with 5% sulfosalicylic acid to abolish infec-tivity. The remaining culture medium was thenremoved followed by one rinse with PBS and additionof 100 ml new PBS with 0.1% D-penicillamine (Sigma,Steinheim, Germany, to standardise the further analy-sis) into which the cells were detached. The cells weretransferred into Eppendorf tubes, treated with 5%sulfosalicylic acid for 10–15 min at RT and stored at

588 Rasilainen et al.

�708C together with the culture medium samples. Allexperiments were done in duplicate.

Thiol concentrations were measured fluorometricallyusinghigh-performance liquid chromatography (HPLC)withpenicillamineas the internal standard [Ahola etal.,1999]. In brief, free sulfhydryls of thiols were derivedusingmonobromobimane to formfluorescent complexes.Thiols were separated with HPLC and detected fluor-ometrically. Additional samples were first treated withdithiothreitol to reduce disulfides for the calculation oftotal glutathione concentration. NAD and NADH levelswere used to proportion the thiol concentrations to cellnumber.

Northern Blot for Gamma-GlutamylCysteine Synthetase

RNA was extracted from MIN6 cell samples collectedafter 1 and 4 days of infection by a commercial kit(RNAeasy, Qiagen, Hilden, Germany) according to themanufacturer’s instructions. Routine DNAse treatmentwith phenyl, chloroform and isoamylalcohol was per-formed and 8 mg of DNAse treated total RNA per samplewas fractionated through agarose/formaldehyde gels.After capillary transfer onto nylon membranes(Hybond-N, Amersham Pharmacia Biotech, Bucking-hampshire, England), the blots were hybridised bystandard methodology [Sambrook et al., 1989] with32P-labelled (DuPont, Wilmington, DE) complementaryRNA probe corresponding to nt 874–1106 of thepublished sequence [Lu et al., 1992]. After final rinses,the membranes were exposed to autoradiography films(Kodak, Rochester, NY), stripped and reprobed with arandom primed mouse 32P-labelled 18S DNA probe(Ambion) to control RNA loading.

Nucleotide Analysis

Similar samples as above, except that no penicilla-mine was added, were used to determine the concentra-tion of purine nucleotides in culture medium and incells. The measurements were performed at day 0–3only, to avoid the difficulties in interpreting the datacausedby culturemediumrefreshmentat day3.For thisexamination, the method described by Stocchi et al.[1987] was used. NAD or NADH was again used tonormalise the data.

Nitrite Measurements

To examine the role of nitric oxide, the level of nitritewasmeasured from culturemediumof non-infected, lowMOI and high MOI coxsackievirus B5 infected MIN6-cell cultures after 0 hr, 1, 2, 3, 5 and 7 days of infection.The method is slightly modified from that describedpreviously by Green et al. [1982]. Briefly, culturemedium samples were collected and divided intotriplicates (volume, 150 ml per tube) followed by incuba-tion with 1/10 volume of 10% sulphanilamide in 50%phosphoric acid and 1% naphtyl ethylenediaminedihydrochloride for 2min at 608C.Nitrite wasmeasuredphotometrically at 550 nm with a Victor 2 1420 multi-

label counter (Wallac, Turku, Finland) with sodiumnitrate as standard.

The DNA contents of cells corresponding to theculture medium samples used for nitrite analyses werequantified to proportion the nitrate data to cell number.Briefly, after culture medium collection, cells wererinsed once with PBS, detached into 300 ml of redistilledwater and homogenised ultrasonically (10 sec persample). To quantify DNA, a sample of the homogenatewas analysed in duplicate by a fluorometric methodbased on diaminobenzoic acid-induced fluorescence[Hinegardner, 1971].

Statistical Analysis

The statistical significance of differences betweenmultiple groups was calculated with StatView 5.0software for theMacintosh (Abacus Concepts, Berkeley,CA), using one-way ANOVA followed by Fischer’sprotected least significant difference (PLSD) test taking95% level as the limit of significance.

RESULTS

Infectivity

In MIN6 cells, the coxsackievirus B5 titres reachedhigh levels already on the first day following high MOIinoculation. As expected, after lowMOI inoculation, thetitre increased at a significantly slower pace (P<0.0001at day 1). This same phenomenon could be observed bythe immunofluoresence staining with enterovirus spe-cific antiserum (Fig. 1A). However, under both condi-tions, a similar maximal titre was obtained after 3 daysof infection, which provedMIN6 cells highly susceptibleto coxsackievirus B5 infection (Fig. 1B). In our earlierstudies, it was observed that coxsackievirus B5 repli-cates well in human islets [Roivainen et al., 2000].

Cell Viability and Apoptosis

Nuclear double staining (HOþEthD-1) was used tomeasure viability and screen the mechanism of celldeath inMIN6 cells. Individual stainingwas carried outat day 1, 2, 3 and 6 post infection. High MOI culture ofcoxsackievirus B5 led dominantly to steadily increasingnecrosis (49.5�6% of the total cell count at day 6) whilein low MOI culture, necrosis increased slowly (7.0�1%at day 6) and reached a steady state after 3 days.Within3 days both conditions stimulated nuclear fragmenta-tion at a similar level (3.8�0.6% and 3.8�1.4% for highand low MOI respectively), however, the trend increas-ing in low and decreasing in high MOI culture. In linewith this, the ratio of apoptotic tonecrotic cells increasedin low and decreased in high MOI culture showing atenfold difference on day 6 (0.5�0.03 vs. 0.05�0.01,Figs. 2 and 3).

In human isletsHOþEthD-1 stainingwas performedafter 4 and 7 days of virus inoculation. Already after4 days, high MOI coxsackievirus B5 culture showed agrowing number of necrotic cells within the islets.Instead, in low MOI culture several apoptotic cellsemerged at themargins of the islets. By day 7, the scarce


islets still remaining in the highMOI culture were filledwith necrotic cells and their integrity was severelydamaged. However, the low MOI culture still showedpreserved islet morphology. The viability of humanislets was also analysed by EthD-1þCalcein staining.After 4 days of highMOI culture islets showed rupturedmorphologywithmainly single cells present. Instead, inlow MOI culture the islet structure was well preserved.By day 7mild morphological disintegration was evidentalso in islets cultured at low MOI (Fig. 4).

TUNELþ Insulin Staining

To verify MOI-dependent induction of cell-specificapoptosis and the histological integrity of human islets,a double staining for TUNEL and insulin was carriedout. On day 4, the sample of high MOI cultured isletsalready showed somewhat disturbed morphology and afew apoptotic b-cells were detected. By day 7, normalislet morphology was lost and mostly damaged singlecells and some scattered islets were detected. Instead,during lowMOI culture, the islet integrity was intact at

day 4 and well-shaped islets were still present at day 7.Apoptotic b-cells could be detected in the peripheralparts of the islets with increasing frequency (Fig. 4).

DNA Fragmentation Analysis

At day 4, the DNA from both MIN6 cells and humanislets inoculated with low MOI coxsackievirus B5showed a classical ladder-pattern characterising DNAfragmentation.WeakerDNA ladderingwas also evidentin the sample of highMOI cultured islets as expected bythe data gathered from nuclear stainings (not shown).Only faint apoptotic patterning was found in the un-infected controls (Fig. 5).

Evaluation of Oxidative Stress: Statusof Intracellular Glutathione and Expression

of Gamma-Glutamyl CysteineSynthetase in MIN6 Cells

Intracellular glutathione concentrations decreasedfor 5 days reflecting the oxidative stress caused by cell

Fig. 1. A: Immunocytochemistry for enterovirus antigen inMIN6 cells infected with coxsackievirus B5.Uninfected control cells were used as a negative control. B: Coxsackievirus B5 replication in MIN6 cellsbased on total infectivitymeasurements fromendpoint dilutions; TCID50, tissue culture infectiousdose 50.Results are presented as mean�SEM from four to six individual experiments.


culture. The decrease, however, was more evident invirus infected cells suggesting additional oxidativestress caused by infection. After 4 days, adaptive pro-cesses were able to replenish the glutathione poolobserved by increased expression of gamma-glutamylcysteine synthetase (the rate-limiting enzyme of glu-tathione synthesis) in Northern blot analysis (Fig. 6A).During the second week of infection, the glutathionelevels collapsed inhighMOI culturewhereas in lowMOIculture they recovered and were stimulated even morethan under uninfected control circumstances (Fig. 6B).This result is presented as the level of total intracellularglutathione, since the level of oxidised glutathione(GSSG) was negligible compared to the total count.

Purine Nucleotide ReleaseInto Culture Medium

To evaluate further the damage to the cellular energymetabolism in MIN6 cells, the concentrations of hypox-anthine and xanthine in culture medium were mea-sured. The release of these purine catabolism products

increased day by day, and on day 3 the virus-infectedcells released more than non-infected cells. No differ-ences between high and lowMOI cultures were detected(data not shown).

Levels of Medium Nitrite

Independent of the culture condition (control orinfected), the trend of nitrite concentration in theculture medium of MIN6 cells was increasing linearlyin a time-dependent fashion. This pattern did notchange when analysed as a ratio to DNA. Thus, novirus- or multiplication-dependent differences could bedetected, suggesting that nitric oxide was not criticallyinvolved in the induction of apoptosis (data not shown).

DISCUSSION

Both apoptotic and necrotic cell deaths have beenrecognised as potential virus-induced end results, andtheir differential emergence might also relate to theprogressionof infection.Astraditionallydefined,aggres-sive necrosis leads rapidly to host cells lytic death, while

Fig. 2. MIN6-cell viability andmechanisms of coxsackievirus B5 induced death based on nuclear doublestaining with Hoechst dye and ethidium homodimer-1. Results are presented as the mean�SEM fromthree to five individual experiments. *P<0.05, **P<0.005, ***P<0.0001.


apoptosis results in controlled elimination of the dyingcellswith no inflammatory reactions following. Further-more, the apoptotic regulatory machinery could beexploited by the virus and adjusted in favour of theprogression of infection [Teodoro andBranton, 1997]. Asmentioned earlier, our previous study in human isletsshows an unrestricted coxsackievirus B5 infection ofhigh multiplicity to result at its early stage dominantlyin pyknosis and only moderately in increased apoptosis[Roivainen et al., 2000]. The present study similarlyshows necrosis to dominate the death of cells inoculatedwith high MOI coxsackievirus B5 and apoptosis to betransiently increased during the early days of infection.However, following inoculation with low MOI of cox-sackievirus B5, the number of apoptotic cells increasestime-dependently and finally reaches a level similar tothe transient peak during high MOI culture at day 2.Necrosis is only induced moderately in comparison tohigh MOI conditions. Thus, during low MOI infection,the dynamics of both apoptosis and necrosis differ dram-atically from high MOI. Taken together, these observa-tions point towards a more prominent role for apoptosisas a contributor to beta cell death during a slowly

progressing infection. Previous studies on picornavirusinfections also acknowledge the apoptotic potential. Asan example, the extensively studied model of coxsack-ievirus B3 induced myocarditis is associated withdevelopment of persistence and apoptotic myocardialdeath [Carthy et al., 1998;Huber et al., 1999]. Poliovirusinfection or its individual proteases alone have beenreported to be capable of inducing apoptotic deathin various cell types [Tolskaya et al., 1995; Girardet al., 1999; Barco et al., 2000; Goldstaub et al., 2000;Lopez-Guerrero et al., 2000]. These data together withthe previous results from models of restricted virusinfection [Martin et al., 1994; Jelachich and Lipton,1996] do appreciate the link of apoptosis and slowlyprogressing enterovirus infection.

The observed phenomenon, that cells infected withlow multiplicity maintain a higher viability also afterthe point (3 days) of reaching a viral titre equal to thehigh multiplicity infection, may have several explana-tions. First, the virus might somehow convert and losesome of its cytolytic properties. Second, during the firstdays of infection the scarce infected cells might producefactors (such as interferon) inhibiting the virus from

Fig. 3. Illustrations of Hoechst dyeþ ethidium homodimer-1 staining. Hoechst dye (HO) passes intactplasma membranes freely and enters all cells staining DNA blue. Ethidium homodimer-1 (EthD-1) onlypenetrates cells with damagedmembranes stainingDNA red. Thus, viable nuclei are detected in blue (HO)and necrotic nuclei in red (EthD-1). Objective magnification 20�. Some of the apoptotic cells are indicatedwith arrows; CVB-5, coxsackievirus B5.


entering and infecting other cells. Interferons, known toreduce viral multiplication and thus function in endo-genous antiviral defence, may inhibit fulminant lytichost cell death and instead mediate the developmentof viral persistence in human islets [Chehadeh et al.,2000]. On the other hand, interferons are also reportedto be able to sensitise infected cells to apoptosis [Tanakaet al., 1998]. We carried out an interferon a/b assay butfailed to detect any stimulation of their production (datanot shown). This might be due to an interruption of theinterferon cascade, ameans known to be used by severaldifferent viruses in order to counteract the host [Katze,1995; Ronco et al., 1998; Komatsu et al., 2000; van Peschet al., 2001]. Also, if only small amounts of interferonwere produced, it could be totally bound to cells andthuswouldnot bedetectable in theusedbiological assay,which only detects secreted, extracellular interferon.

Nitric oxide associates with the cytokine-mediatedimmune response against coxsackievirus infection[Zaragoza et al., 1998; Robinson et al., 1999; Saura

et al., 1999; Gluck et al., 2000; Flodstrom et al., 2001],but additionally it provokesapoptosis inmanycell types.In our experimental setting though, no role for nitricoxide could be demonstrated as a contributor to virus-induced b-cell death. Nitric oxide neither contributed tothe dsRNA-mediated apoptosis in rat islets nor to thehuman b-cell death by a productive coxsackievirus Binfection [Roivainen et al., 2000] and seems to mediateonly cytokine-induced metabolic alterations and necro-tic death but not apoptosis in b-cells [Delaney et al.,1997; Liu et al., 2000, 2002].

Virus infections may trigger the production of oxida-tive molecules in the host [Schwarz, 1996]. They mayactivate phagocytes to produce reactive oxygen species[Peterhans et al., 1987] and pro-oxidative cytokines(TNF-a, IL-1b), which potentiate further both oxidativeand viral damage through impaired mitochondrialfunction and increased multiplication of the virus re-spectively [Schulze-Osthoff et al., 1992; Polla et al.,1996]. Infections may also decrease concentrations of

Fig. 4. Illustrations of the pattern andmechanisms of coxsackievirus B5 induced human islet cell deathanalysed by double staining with TUNELþ insulin immunocytochemistry (A,D, at 4 and 7 days postinfection respectively), staining by Hoechst dyeþEthD-1 (B) and CalceinþEthD-1(C). Objectivemagnification 40� (A,B,D), 10� (C). Some of the apoptotic cells are indicated with arrows. CVB-5,coxsackievirus B5.


several ROS scavengers and thus weaken the host’santi-oxidative defence [Hennet et al., 1992; Allard et al.,1998]. We found the concentration of glutathione, themajor intracellular anti-oxidant, to decrease steadily forthe first 5 days in both virus-exposed cultures. This wasassociated with a relatively more intensive expressionof the rate-limiting glutathione synthesising enzyme inthe highMOI culture. This differencemay simply reflectthe intensity of cell damage and the need for repairbetween low and high MOI cultures. However, duringthe second week of infection, high MOI culture resultedin depletion of the thiol levels while they recovered fullyduring lowMOI culture. This data implies to a potentialrole for the intracellular anti-oxidative machinery indefence against the coxsackievirus B5 induced b-celldamage. The recovering thiol profile and the probablymilder oxidative burst associated with low multiplicityinfection indicate a moderate, physiological type ofdamage, which has previously been linked to apoptoticcell death [Lennon et al., 1991; Dypbukt et al., 1994;Hampton and Orrenius, 1997; Chandra et al., 2000]. Asimilar time-dependent thiol profile has previously beendocumented during coxsackievirus B3 induced myocar-ditis (V. Kyto, unpublished observations) in the protec-tion of which glutathione peroxidase activity has been

demonstrated relevant [Beck et al., 1998]. Furthermore,this model showed the transient decrease in GSH-levelto coincide with the peak in cardiomyocyte apoptosis(V. Kyto, unpublished data). It should be emphasisedthat these overall changes in thiol metabolism reflectthe situation in thewhole culture, and theremaywell besubpopulations of cells behaving differently. Further-more, at late time points, MIN6-cell cultures inoculatedwith high MOI of coxsackievirus B5 show extensivenecrotic damage and cell loss, which might have aninfluence on further statistical analyses.

In conclusion, these results demonstrate an accentu-ated role for apoptotic death as a result of low multi-plicity coxsackievirus B5 infection in b-cells. The morephysiological glutathione profile associated with lowmultiplicity infection reflects moderate viral damage,which may lead to the induction of programmed b-cell

Fig. 5. DNAfragmentationanalysisafter4daysof infection.A:Adulthuman islets, (B) MIN6 cells. Co, uninfected control; CVB-5, coxsack-ievirus B5.

Fig. 6. A: Northern blot for gamma-glutamyl cysteine synthetaseand its densitometric analysis (MIN6 cells); Co, uninfected control;CVB-5, coxsackievirusB5).B: Intracellular total glutathione content inMIN6 cells. Results are presented as a ratio to NADH. The resultsrepresent the mean�SEM from three individual experiments.


death. Thus, the pattern and mechanisms of coxsack-ievirus B5 infection in b-cells seem to depend on theinitial amount of infective virus.

ACKNOWLEDGMENTS

We thank the technicians Ritva Lofman and SariLinden for assistance in the thiol analyses.

REFERENCES

Agol VI, Belov GA, Bienz K, Egger D, Kolesnikova MS, Raikhlin NT,RomanovaLI, SmirnovaEA, TolskayaEA. 1998. Two types of deathof poliovirus-infected cells: Caspase involvement in the apoptosisbut not cytopathic effect. Virology 252:343–353.

Ahola T, Fellman V, Laaksonen R, Laitila J, Lapatto R, Neuvonen PJ,Raivio KO. 1999. Pharmacokinetics of intravenous N-acetylcys-teine in pre-term new-born infants. Eur J Clin Pharmacol 55:645–650.

Allard JP, Aghdassi E, Chau J, Salit I, Walmsley S. 1998. Oxidativestress and plasma antioxidant micronutrients in humans with HIVinfection. Am J Clin Nutr 67:143–147.

Andreoletti L, Hober D, Becquart P, Belaich S, Copin MC, Lambert V,Wattre P. 1997. Experimental CVB3-induced chronic myocarditisin twomurine strains: Evidence of interrelationships between virusreplication and myocardial damage in persistent cardiac infection.J Med Virol 52:206–214.

Barco A, Feduchi E, Carrasco L. 2000. Poliovirus protease 3C(pro) killscells by apoptosis. Virology 266:352–360.

BeckMA, Esworthy RS, Ho YS, Chu FF. 1998. Glutathione peroxidaseprotects mice from viral-induced myocarditis. FASEB J 12:1143–1149.

Carthy CM, Granville DJ, Watson KA, Anderson DR,Wilson JE, YangD, Hunt DW, McManus BM. 1998. Caspase activation and specificcleavage of substrates after coxsackievirus B3-induced cytopathiceffect in HeLa cells. J Virol 72:7669–7675.

Chandra J, Samali A, Orrenius S. 2000. Triggering and modulation ofapoptosis by oxidative stress. Free Radic Biol Med 29:323–333.

Chehadeh W, Weill J, Vantyghem MC, Alm G, Lefebvre J, Wattre P,Hober D. 2000. Increased level of interferon-alpha in blood ofpatients with insulin-dependent diabetes mellitus: Relationshipwith coxsackievirus B infection. J Infect Dis 181:1929–1939.

Delaney CA, Pavlovic D, Hoorens A, Pipeleers DG, Eizirik DL. 1997.Cytokines induce deoxyribonucleic acid strand breaks and apopto-sis in human pancreatic islet cells. Endocrinology 138:2610–2614.

Dypbukt JM, AnkarcronaM, Burkitt M, SjoholmA, StromK,OrreniusS, Nicotera P. 1994. Different prooxidant levels stimulate growth,trigger apoptosis, or produce necrosis of insulin-secreting RINm5Fcells. The role of intracellular polyamines. J Biol Chem 269:30553–30560.

Flodstrom M, Horwitz MS, Maday A, Balakrishna D, Rodriguez E,Sarvetnick N. 2001. A critical role for inducible nitric oxidesynthase in host survival following coxsackievirus B4 infection.Virology 281:205–215.

Girard S, Couderc T, Destombes J, Thiesson D, Delpeyroux F, BlondelB. 1999. Poliovirus induces apoptosis in the mouse central nervoussystem. J Virol 73:6066–6072.

Gluck B, Merkle I, Dornberger G, Stelzner A. 2000. Expression ofinducible nitric oxide synthase in experimental viral myocarditis.Herz 25:255–260.

Goldstaub D, Gradi A, Bercovitch Z, Grosmann Z, Nophar Y, Luria S,Sonenberg N, Kahana C. 2000. Poliovirus 2A protease inducesapoptotic cell death. Mol Cell Biol 20:1271–1277.

Green LC, Wagner DA, Glogowski J, Skipper PL, Wishnok JS,Tannenbaum SR. 1982. Analysis of nitrate, nitrite, and[15N]nitrate in biological fluids. Anal Biochem 126:131–138.

HamptonMB, Orrenius S. 1997. Dual regulation of caspase activity byhydrogen peroxide: Implications for apoptosis. FEBSLett 414:552–556.

Hennet T, Peterhans E, Stocker R. 1992. Alterations in antioxidantdefences in lung and liver of mice infected with influenza A virus.J Gen Virol 73:39–46.

Hiltunen M, Hyoty H, Knip M, Ilonen J, Reijonen H, Vahasalo P,Roivainen M, Lonnrot M, Leinikki P, Hovi T, Akerblom HK. 1997.Islet cell antibody seroconversion in children is temporally

associated with enterovirus infections. Childhood Diabetes inFinland (DiMe) Study Group. J Infect Dis 175:554–560.

Hinegardner RT. 1971. An improved fluorometric assay for DNA. AnalBiochem 39:197–201.

Hoorens A, Van de Casteele M, Kloppel G, Pipeleers D. 1996. Glucosepromotes survival of rat pancreatic beta cells by activatingsynthesis of proteins which suppress a constitutive apoptoticprogram. J Clin Invest 98:1568–1574.

Hovi T, Roivainen M. 1993. Peptide antisera targeted to a conservedsequence in poliovirus capsid VP1 cross-react widely withmembersof the genus Enterovirus. J Clin Microbiol 31:1083–1087.

Huber SA, Budd RC, Rossner K, Newell MK. 1999. Apoptosis incoxsackievirus B3-induced myocarditis and dilated cardiomyopa-thy. Ann NY Acad Sci 887:181–190.

Hyoty H, Taylor KW. 2002. The role of viruses in human diabetes.Diabetologia 45:1353–1361.

Jelachich ML, Lipton HL. 1996. Theiler’s murine encephalomyelitisvirus kills restrictive but not permissive cells by apoptosis. J Virol70:6856–6861.

Katze MG. 1995. Regulation of the interferon-induced PKR: Canviruses cope? Trends Microbiol 3:75–78.

Komatsu T, Takeuchi K, Yokoo J, Tanaka Y, Gotoh B. 2000. Sendaivirus blocks alpha interferon signaling to signal transducers andactivators of transcription. J Virol 74:2477–2480.

Lennon SV, Martin SJ, Cotter TG. 1991. Dose-dependent induction ofapoptosis in human tumour cell lines by widely diverging stimuli.Cell Prolif 24:203–214.

LiuD, PavlovicD, ChenMC, FlodstromM,Sandler S, EizirikDL. 2000.Cytokines induce apoptosis in beta-cells isolated frommice lackingthe inducible isoform of nitric oxide synthase (iNOS�/�). Diabetes49:1116–1122.

Liu D, Cardozo AK, Darville MI, Eizirik DL. 2002. Double-strandedRNA cooperates with interferon-gamma and IL-1 beta to induceboth chemokine expression and nuclear factor-kappa B-dependentapoptosis in pancreatic beta-cells: Potential mechanisms for viral-induced insulitis and beta-cell death in type 1 diabetes mellitus.Endocrinology 143:1225–1234.

Lopez-Guerrero JA, Alonso M, Martin-Belmonte F, Carrasco L. 2000.Poliovirus induces apoptosis in the human U937 promonocytic cellline. Virology 272:250–256.

Lu Y, O’Dowd BF, Orrego H, Israel Y. 1992. Cloning and nucleotidesequence of human liver cDNA encoding for cystathionine gamma-lyase. Biochem Biophys Res Commun 189:749–758.

Martin SJ,Matear PM, VyakarnamA. 1994. HIV-1 infection of humanCD4þ T cells in vitro. Differential induction of apoptosis in thesecells. J Immunol 152:330–342.

Miyazaki J, Araki K, Yamato E, Ikegami H, Asano T, Shibasaki Y,Oka Y, Yamamura K. 1990. Establishment of a pancreatic beta cellline that retains glucose-inducible insulin secretion: Specialreference to expression of glucose transporter isoforms. Endocri-nology 127:126–132.

Nairn C, Galbraith DN, Taylor KW, Clements GB. 1999. Enterovirusvariants in the serum of children at the onset of Type 1 diabetesmellitus. Diabet Med 16:509–513.

Pentikainen V, Erkkila K, Suomalainen L, Parvinen M, Dunkel L.2000. Estradiol acts as a germ cell survival factor in the humantestis in vitro. J Clin Endocrinol Metab 85:2057–2067.

Peterhans E, Grob M, Burge T, Zanoni R. 1987. Virus-inducedformation of reactive oxygen intermediates in phagocytic cells.Free Radic Res Commun 3:39–46.

Polla BS, Jacquier-Sarlin MR, Kantengwa S, Mariethoz E, Hennet T,Russo-Marie F,CossarizzaA. 1996. TNFalphaaltersmitochondrialmembrane potential in L929 but not in TNF alpha-resistantL929.12 cells: Relationship with the expression of stress proteins,annexin 1 and superoxide dismutase activity. Free Radic Res 25:125–131.

Rasilainen S, Ylipaasto P, RoivainenM, Bouwens L, Lapatto R, Hovi T,Otonkoski T. Mechanisms of beta cell death during restricted andunrestricted enterovirus infection. J Med Virol 72:451–461.

Robinson NM, Zhang HY, Bevan AL, De Belder AJ, Moncada S,Martin JF, Archard LC. 1999. Induction of myocardial nitricoxide synthase by Coxsackie B3 virus in mice. Eur J Clin Invest29:700–707.

Roivainen M, Knip M, Hyoty H, Kulmala P, Hiltunen M, Vahasalo P,Hovi T, Akerblom HK. 1998. Several different enterovirus ser-otypes canbeassociatedwithprediabetic autoimmune episodes and


onset of overt T1DM. Childhood Diabetes in Finland (DiMe) StudyGroup. J Med Virol 56:74–78.

Roivainen M, Rasilainen S, Ylipaasto P, Nissinen R, Ustinov J,Bouwens L, Eizirik DL, Hovi T, Otonkoski T. 2000. Mechanismsof coxsackievirus-induced damage to human pancreatic beta-cells.J Clin Endocrinol Metab 85:432–440.

Roivainen M, Ylipaasto P, Ustinov J, Hovi T, Otonkoski T. 2001.Screening enteroviruses for beta-cell tropism using foetal porcinebeta-cells. J Gen Virol 82:1909–1916.

Roivainen M, Ylipaasto P, Savolainen C, Galama J, Hovi T, OtonkoskiT. 2002. Functional impairment and killing of human beta cells byenteroviruses: The capacity is shared by a wide range of serotypes,but the extent is a characteristic of individual virus strains.Diabetologia 45:693–702.

Ronco LV, Karpova AY, Vidal M, Howley PM. 1998. Humanpapillomavirus 16 E6 oncoprotein binds to interferon regulatoryfactor-3 and inhibits its transcriptional activity. Genes Dev 12:2061–2072.

Sadeharju K, Hamalainen AM, Knip M, Lonnrot M, Koskela P,Virtanen SM, Ilonen J, Akerblom HK, Hyoty H. 2003. Enterovirusinfections as a risk factor for type I diabetes: Virus analyses in adietary intervention trial. Clin Exp Immunol 132:271–277.

SalminenK, SadeharjuK, LonnrotM,Vahasalo P,Kupila A,KorhonenS, IlonenJ,SimellO,KnipM,HyotyH. 2003.Enterovirus infectionsare associated with the induction of beta-cell autoimmunity in aprospective birth cohort study. J Med Virol 69:91–98.

Sambrook J, Fritsch EF, Maniatis T, editors. 1989. Molecular cloning:A laboratory manual. Cold Spring Harbor, NY: Cold Spring HarborPress.

SauraM, Zaragoza C,McMillan A, Quick RA,Hohenadl C, LowensteinJM, Lowenstein CJ. 1999. An antiviral mechanism of nitric oxide:Inhibition of a viral protease. Immunity 10:21–28.

Schulze-Osthoff K, Bakker AC, Vanhaesebroeck B, Beyaert R, JacobWA, Fiers W. 1992. Cytotoxic activity of tumor necrosis factor ismediated by early damage of mitochondrial functions. Evidence for

the involvement of mitochondrial radical generation. J Biol Chem267:5317–5323.

Schwarz KB. 1996. Oxidative stress during viral infection: A review.Free Radic Biol Med 21:641–649.

Stocchi V, Cucchiarini L, Canestrari F, Piacentini MP, Fornaini G.1987. A very fast ion-pair reversed-phase HPLC method forthe separation of the most significant nucleotides and theirdegradation products in human red blood cells. Anal Biochem167:181–190.

Szopa TM, Titchener PA, Portwood ND, Taylor KW. 1993. Diabetesmellitus due to viruses—Some recent developments. Diabetologia36:687–695.

Tanaka N, Sato M, Lamphier MS, Nozawa H, Oda E, Noguchi S,Schreiber RD, Tsujimoto Y, Taniguchi T. 1998. Type I interferonsare essential mediators of apoptotic death in virally infected cells.Genes Cells 3:29–37.

Teodoro JG, Branton PE. 1997. Regulation of apoptosis by viral geneproducts. J Virol 71:1739–1746.

Tolskaya EA, Romanova LI, Kolesnikova MS, Ivannikova TA,Smirnova EA, Raikhlin NT, Agol VI. 1995. Apoptosis-inducingand apoptosis-preventing functions of poliovirus. J Virol 69:1181–1189.

van Pesch V, van Eyll O, Michiels T. 2001. The leader protein ofTheiler’s virus inhibits immediate-early alpha/beta interferonproduction. J Virol 75:7811–7817.

YinH, Berg AK, TuvemoT, FriskG. 2002. Enterovirus RNA is found inperipheral blood mononuclear cells in a majority of type 1 diabeticchildren at onset. Diabetes 51:1964–1971.

Yoon JW, Austin M, Onodera T, Notkins AL. 1979. Isolation of a virusfrom the pancreas of a child with diabetic ketoacidosis. N Engl JMed 300:1173–1179.

Zaragoza C, Ocampo C, SauraM, LeppoM,Wei XQ, Quick R,MoncadaS, Liew FY, Lowenstein CJ. 1998. The role of inducible nitric oxidesynthase in the host response to Coxsackievirus myocarditis. ProcNatl Acad Sci USA 95:2469–2474.


mechanisms of coxsackievirus b5 mediatedβ-cell death depend on the multiplicity of infection

Documents