Next-generation sequencing (NGS) in the identification of encephalitis-causing viruses: Unexpected detection of human herpesvirus 1 while searching for RNA pathogens

ContentslistsavailableatScienceDirect

Journal

of

Virological

Methods

jou rn al h om ep ag e :w w w . e l s e v i e r . c o m / l o c a t e / j v i r o m e t

Short

communication

Next-generation

sequencing

(NGS)

in

the

identification

of

encephalitis-causing

viruses:

Unexpected

detection

of

human

herpesvirus

1

while

searching

for

RNA

pathogens

Karol

Perlejewski

,

Marta

Popiel

,

Tomasz

Laskus

,

Shota

Nakamura

,

Daisuke

Motooka

,

Tomasz

Stokowy

,

Dariusz

Lipowski

,

Agnieszka

Pollak

,

Urszula

Lechowicz

,

Kamila

Caraballo

Cortés

,

Adam

St ˛epie ´n

,

Marek

Radkowski

,

Iwona

Bukowska-O´sko

a_{DepartmentofImmunopathologyofInfectiousandParasiticDiseases,WarsawMedicalUniversity,3CPawinskiegoStreet,02-106Warsaw,Poland} b_{PostgraduateSchoolofMolecularMedicine,61 ˙ZwirkiiWiguryStreet,02-091Warsaw,Poland}

c_{DepartmentofInfectionMetagenomics,GenomeInformationResearchCenter,ResearchInstituteforMicrobialDiseases,OsakaUniversity3-1Yamadaoka,} Suita-City,Osaka,Japan

d_{DepartmentofClinicalScience,UniversityofBergen,5021Bergen,Norway} e_{MunicipalHospitalforInfectiousDiseases,37WolskaStreet,01-201Warsaw,Poland}

f_{DepartmentofGenetics,InstituteofPhysiologyandPathologyofHearing,Mochnackiego10,02-042Warsaw,Poland} g_{DepartmentofNeurology,MilitaryInstituteofMedicine,128SzaserówStreet,04-141Warsaw,Poland}

Articlehistory:

Received18May2015 Receivedinrevisedform 24September2015 Accepted24September2015 Availableonline28September2015

Keywords:

Encephalitis

Next-generationsequencing Unknownetiology

a

b

s

t

r

a

c

t

Background:Encephalitisisasevereneurologicalsyndromeusuallycausedbyviruses.Despitesignificant progressindiagnostictechniques,thecausativeagentremainsunidentifiedinthemajorityofcases.The aimofthepresentstudywastotestanalternativeapproachforthedetectionofputativepathogensin encephalitisusingnext-generationsequencing(NGS).

Methods:RNAwas extractedfrom cerebrospinalfluid(CSF)from a60-year-oldmale patient with encephalitisandsubjectedtoisothermallinearnucleicacidamplification(Ribo-SPIA,NuGen)followed bynext-generationsequencingusingMiSeq(Illumina)systemandmetagenomicsdataanalysis. Results:Thesequencingrunyielded1,578,856readsoveralland2579readsmatchedhumanherpesvirus I(HHV-1)genome;thepresenceofthispathogeninCSFwasconfirmedbyspecificPCR.Insubsequent experimentswe foundthattheDNAseI treatment,whileloweringthebackgroundofhost-derived sequences,loweredthenumberofdetectableHHV-1sequencesbyafactorof4.Furthermore,wefound thattheroutineextractionoftotalRNAbytheChomczynskimethodcouldbeusedforidentificationof bothDNAandRNApathogensintypicalclinicalsettings,asitresultsinretentionofasignificantamount ofDNA.

Conclusion:Insummary,itseemsthatNGSprecededbynucleicacidamplificationcouldsupplement currentlyuseddiagnosticmethodsinencephalitis.

©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Abbreviations:NGS,next-generationsequencing;CSF,cerebrospinalfluid; HHV-1,humanherpesvirus1.

∗_{Correspondingauthor.}

E-mailaddresses:[email protected]

(K.Perlejewski),[email protected](M.Popiel),[email protected]

(T.Laskus),[email protected](S.Nakamura),

[email protected](D.Motooka),[email protected]

(T.Stokowy),[email protected](D.Lipowski),

[email protected](A.Pollak),[email protected]

(U.Lechowicz),[email protected](K.CaraballoCortés),[email protected]

(A.St ˛epie ´n),[email protected](M.Radkowski),

[email protected](I.Bukowska-O´sko).

Encephalitisisasevereneurologicalsyndrome,whichmaybe causedbyvirusesaswellasbacteria,fungiandparasites(Chaudhuri and Kennedy, 2002; Glaser et al., 2006; Solomon et al., 2007). Diagnosticproceduresincludeserologicaltestingforspecific anti-bodiesinbloodandcerebrospinalfluid(CSF),detectionofcausative pathogensbymolecularmethodsinCSFandblood,andoccasionally evenpathogenculture(Solomonetal.,2007).Correctdiagnosisis facilitatedbyepidemiologicaldataandclinicalsymptoms,routine CSFanalysisandneuroimagingstudies(ChaudhuriandKennedy, 2002; Solomon etal., 2007).Surprisingly, ina largeproportion http://dx.doi.org/10.1016/j.jviromet.2015.09.010

0166-0934/©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4. 0/).

ofcasestheexactetiologyofencephalitisremainsunknown.For example,TheCalifornianEncephalitisProject,whichincluded1570 patients,failedtoidentifycauseofencephalitisin63%ofpatients (Glaser et al., 2006), while in a French study the etiology of encephalitis remainedunknownin 80% of cases (Mailles et al., 2007).However,identificationofcausativeagentsiscrucialforthe correctdiagnosisandmaybenecessaryforthetimelyintroduction ofspecifictreatment(Studahletal.,2013).

Thehighrateof caseswithunclearetiology ispartly dueto thelimitedsensitivityandspecificityofcurrentlyused diagnos-ticassays,butalsotothesuboptimaltimeofsamplecollection, sinceviralpathogensmaybedetectableonlytransiently(Glaser et al., 2006; Studahl et al., 2013). Moreover, a large and still expandingnumberofpathogensreportedtobecapableofcausing encephalitismakesspecificdiagnosischallengingduetothesheer scopeofrequiredtesting(Kennedy,2005).Arecentlydeveloped next-generation sequencing (NGS) technique offers a potential breakthroughinthediagnosticprocessofencephalitisby eliminat-ingtheneedfortestingofindividualpathogens.High-throughput sequencinghasalreadyestablishedafirmfootholdintheareaof microbiologyasithasbeensuccessfullyemployedforpathogen detection(including novelmicroorganisms), analysisof genetic evolutionanddiversity,aswellasforthedeterminationofdrug resistance(Radfordetal.,2012).Theaimofthepresentstudywasto determinethevalueofNGSindetectingputativecausativeagent(s) inCSFinencephalitis.

For this purposeCSF from60-year-old male withsuspected encephalitis,admitted tothehospital witha 1-weekhistory of fever,drowsinessandsevereheadacheswasused.Thesesymptoms wereprecededbyamildcold.Atthetimeofadmissionthepatient wascomatosewithpositivemenigealsignsandhistemperature was38.5◦C. Analysisof CSF revealed480leukocytes/mm3 _(90%

lymphocytes),proteinlevelof3.21g/l(normalvalue:0.12–0.60g/l), andglucosevalueof2.32mmol/l(normalvalue:2.20–4.40mmol/l). Theperipheralbloodleukocytecountwas6900/mm3_with66.7%

neutrophilsand22.2%lymphocytes.SerumandCSFsamples col-lectedatadmissionwere negativefor thepresenceof IgM and IgGantibodiestohumanherpesvirus(NovaLisaHSV1/HSV2/HSV 1+2IgG/IgMELISA;NovaTecImmundiagnosticaGmBH, Dietzen-bach,Germany).However,bothanti-HHV-1IgMandanti-HHV-1 IgGweredetectedinsubsequentCSFsamplescollectedat1.5,3and 5months.Magneticresonanceimaging(MRI)atadmissionshowed hyperintensityintherighttemporallobe.Patientwastreatedwith Acyclovir(ZoviraxIV)for15daysandfullyrecoveredwithno resid-ualneurologicalsequelae.PCRforHSV-1/2wasnotperformedat admission,asitwasnotapartofroutinediagnosticsinPolandat thattime,andouranalysiswasdoneretrospectively.

TotalRNAwasisolatedfrom1mlof CSFcollectedat admis-sionusingthemodifiedChomczynskimethod(TRIZOLLSReagent; LifeTechnologies,Carlsbad,USA)andsplitintotwoaliquots,one ofwhich (sampleSDNase)wastreated withDNaseI (DNA-free DNA Removal Kit; Invitrogen, Waltham, USA) at a concentra-tionof2U/␮lfor20minin 37◦C ina finalvolumeof100␮lto reducetheamountofDNApresent.Theotheraliquotremained DNaseuntreated(sampleS1).Controlsamplesincluded HCV-RNA-positiveserum sampleadjusted tocontain100viralcopies per reaction(sampleC-HCV)andthreeHHV-1-negativeCSFsamples (samplesC1–C3;Table1).

ThequantityofRNA extractedfromCSF wassmall andwas belowthesensitivitylimitsofDeNovixDS-11Spectrophotometer (DeNovix,Wilmington,USA;sensitivity≥2ng/␮l)andQubit Fluo-rimeter(ThermoScientific,Waltham,USA;sensitivity≥250pg/␮l). Both CSF samples (S1 and SDNAse) and all control sam-plesunderwent reversetranscription followedby single-primer isothermalamplification(Ribo-SPIA),usingcommerciallyavailable kit(OvationRNA-SeqV2 system(NuGEN,SanCarlos, USA).The

protocolcloselyfollowedmanufacturer’srecommendations.The resultingamplifieddscDNAwaspurifiedusingAgencourtAMPure XPbeads(BeckmanCoulter,Pasadena,USA).

Librariesfor NGSwerepreparedfrom1ngofdscDNAusing NexteraXTKit(Illumina,SanDiego,USA)followingmanufacturer’s protocol. First, cDNA was fragmented using transposon-based method followed by addition of indexes by PCR. The resulting PCRproductswerepurifiedwith1.8volumesofAMPureXPbeads (BeckmanCoulter,Pasadena,USA).Thequalityandaveragelength ofsequencelibraryforeachsamplewasassessedusingBioanalyzer (AgilentTechnologies,SantaClara,USA)andeithertheDNA1000or DNAHSkit,dependingonsampleconcentration.Theindexed sam-pleswerepooledequimolarilyandsequencedonIlluminaMiSeq (150nt,paired-endreads)incaseofsamplesS1,SDNase,C-HCVand onIlluminaHiSeq(101nt,paired-endreads);incaseofsamples C1–C3.

Raw reads were trimmed by the following procedures: (i) removingadaptorsequencesusingcutadapt-1.2.1(Martin,2011), (ii)removingartifactsequencesusingfastxartifactsfilter(http:// hannonlab.cshl.edu/fastxtoolkit/index.html,2015),and(iii) trim-ming bases with quality below Q20 (phred quality score) from 3 end of each read and removing reads shorter than 50bp using fastqqualitytrimmer (http://hannonlab.cshl.edu/ fastxtoolkit/index.html, 2015). Trimmed sequences were then mappedontothehumanreferencesequence(hg19)withthe map-pingsoftwareStampy(LunterandGoodson,2011).Theunmapped sequencesweresubjectedtosimilaritysearchusingtheprogram blastn(Altschuletal.,1990)againstthewholeunfilteredNCBI-nt databasewithe-valuecutoffof1e−5.Thetaxonomicinformationof eachsequencewasassignedandtherelativeabundanceofdetected organismswassummarizedbytextminingofblastnoutputfiles usingBioRuby(Gotoetal.,2010)scripts.

TheinitialsequencingrunincludedsamplesS1,SDNase,C-HCV andyielded4,529,322readsaltogether.Aftertrimming,therewere 1,190,455readsforsampleS1(noDNasetreatment)and1,030,284 readsfor sampleSDNase (treatedwithDNaseI). Sequencingof HepatitisCVirus(HCV)positivecontrol(sampleC-HCV)provided 1,184,251reads(Table1).Humansequenceswerethemost abun-dant(43.913–99.506%ofallsequences;Table1).Viralsequences represented0.07–0.246%ofreadsandthemostfrequentamong themcorrespondedtoHHV-1(99.8%ofallviralsequencesinsample S1and86.6%insampleSDNase).InthecontrolHCVRNA-positive sample99.8%ofviralreadswerepositivelyalignedtoHCVgenome (Table2).Fig.1showsthecoverageplotofdetectedHHV-1and HCVsequenceswithrespecttoreferencegenomes.Importantly,the DNaseItreatment,whileloweringthebackgroundofhost-derived sequencesfrom98.2%to43.9%,atthesametimeloweredthe num-berofdetectableHHV-1sequencesbyafactorof4(from2579to 629sequences).

ControlCSFsamplesC1–C3wereanalyzedin aseparaterun yieldingatotalof107,219,264reads.Aftertrimming,therewere 37,516,857readsforsampleC1,35,638,335readsforsampleC2 and34,064,072readsforsampleC3.Themetagenomicapproach didnotdetectanyHSV-1sequencesinanyofthesethreesamples. TheresultsofNGSsequencingandthemostfrequentlyidentified species/familiesarepresentedinTables1and2.

The presence of HHV-1 in CSF of the analyzed patient was confirmed by specific PCR (Fig. 2) performed as follows. Human herpesvirus 1 (HHV-1) DNA was isolated from cere-brospinal fluid (200␮l) using QIAamp DNA Mini Kit (Qiagen, Venlo,Netherlands),thenelutedin5␮lofwater(Water Molec-ular Biology Reagen; Sigma–Aldrich,Saint Louis, USA) and the HHV specific PCR was performed using primers specific for bothHHV-1andHHV-2(5-TTGAAGCGGTCGGCGGCGTA-3and5 -GCATCGTCGAGGAGGTGGA-3) resulting in 148bp amplification product(54836nt–54983nt;[GenBank:NC001806.1]). Extracted

Table1

Next-generationsequencing(NGS)ofCSFfrompatientwithHHV-1encephalitisandfromHHV-1-negativecontrols.SequenceswerecomparedtotheNCBI-ntdatabase.

Age/gender 60/male 56/female 20/female 43/female

Diagnosis HHV-1encephalitis Encephalitisof

unknown etiology Encephalitisof unknown etiology Multiple sclerosis Sample S1 SDNase C1 C2 C3 C-HCV Totalreads 1,578,856 1,427,356 37,516,857 35,638,335 34,064,072 1,523,110

Readsaftertrimming 1,190,455 1,030,284 375,168,57 35,638,335 34,064,072 1,184,251

Human 1,169,244(98.218%) 452,428(43.913%) 36,061,952 (96.122%) 35,374,962 (99.261%) 32,616,258 (95.750%) 1,178,397(99.506%) Viral 2583(0.217%) 726(0.070%) 52,226 (0.139%) 71,493 (0.201%) 140,890 (0.414%) 2916(0.246%) Bacterial 5112(0.429%) 242,633(23.550%) 830,811 (2.215%) 120,107 (0.337%) 668,362 (1.962%) 1142(0.096%) Fungal 251(0.021%) 2591(0.251%) 45,550 (0.121%) 3190(0.009%) 78,112 (0.229%) 32(0.003%) Protozoan 22(0.002%) 795(0.077%) 6386(0.017%) 1100(0.003%) 16,298 (0.048%) 13(0.001%) Othera _{12,018(1.010%) 315,615(30.634%) 428,866} (1.143%) 36,106 (0.104%) 468,183 (1.374%) 1015(0.086%) Nomatch 1225(0.103%) 15,496(1.504%) 91,066 (0.243%) 31,377 (0.088%) 75,969 (0.223%) 736(0.062%)

S1,noDNaseItreatment;SDNase,treatedwithDNaseI;C,HCV-RNApositiveserum. SamplesC1–C3werenegativeforHHV-1byPCR.

a_{Sequencesrelatedtoplants,plantviruses,syntheticconstructs.}

Table2

Themostfrequentlyidentifiedspecies/familiesincerebrospinalfluidfromapatientwithHHV-1encephalitisandcontrols.

Viruses Bacteriaa _Fungia _Protozoa

S1 Human herpesvirus1 (2579) InfluenzaC virus(3) HepatitisC virus(1) Moraxellaceae(871) Micrococcaceae(554) Corynebacteriaceae(321) Propionibacteriaceae(204) Gordoniaceae(206) Pyrenophorateres(151) Pyrenophoratritici-repentis(29) Malasseziasympodialis(14) Malasseziaglobosa(10) Debaryomyceshansenii(7) – SDNase Human herpesvirus1 (629) InfluenzaC virus(55) HepatitisC virus(39) Moraxellaceae(37,130) Staphylococcaceae(35,753) Micrococcaceae(48,011) Corynebacteriaceae(19,993) Neisseriaceae(8240) Unculturedfungus(519) UnculturedCryptococcus(206) UnculturedRhodotorula(180) Malasseziasympodialis(142) Malasseziaglobosa(147) Albugolaibachii(441) Besnoitiabesnoiti(89) Trypanoplasmaborreli(83) Ichthyophthiriusmultifiliis (75) C1 – Corynebacteriaceae(2,75,825) Staphylococcaceae(93,611) Micrococcaceae(73,191) Moraxellaceae(51,378) Sphingobacteriaceae(29,855) Unculturedfungus(10,062) Pseudogymnoascusdestructans (1199) Melampsoralarici-populina (995) Leucosporidiumsp.AY30(787) Unculturedsoilfungus(786)

Nannochloropsisoceanica (1549) Albugolaibachii(401) Amphifilidaesp.H-1(373) Babesiadivergens(371) Exocolpodaaugustini(324) C2 – Burkholderiaceae(40,597) Sphingobacteriaceae(30,089) Streptococcaceae(5702) Bradyrhizobiaceae(4649) Enterobacteriaceae(4546) Unculturedfungus(551) Geotrichumklebahnii(250) Malasseziaglobosa(225) Dipodascustetrasporeus(132) Mortierellaalpina(82) Nannochloropsisoceanica (3204) Plasmodiumovale(1201) Albugolaibachii(964) Babesiadivergens(549) Plasmodiuminui(312) C3 – Moraxellaceae(94,524) Enterobacteriaceae(67,339) Corynebacteriaceae(59,747) Micrococcaceae(48,846) Bradyrhizobiaceae(47,559) Unculturedfungus(12,841) Triposporiumcycadicola(4346) Mollisinauncinata(3997) Pseudogymnoascusdestructans (2790) Melampsoralarici-populina (2465) Nannochloropsisoceanica (2468) Albugolaibachii(2274) Plasmodiuminui(2075) Eunotiasp.ECR-2014 (1780) Plasmodiumvinckei(588) C-HCV HepatitisC virus(2905) InfluenzaC virus(3) Moraxellaceae(197) Micrococcaceae(166) Propionibacteriaceae(67) Streptococcaceae(66) Comamonadaceae(62) Debaryomyceshansenii(19) Staninwardiasuttonii(2) Malasseziasympodialis(2) Leptosphaeriamaculans(2) Entylomacalendulae(2) –

S1,noDNaseItreatment;SDNase,treatedwithDNaseI;C-HCV,HCV-RNApositiveserum;C1andC2,CSFofpatientswithencephalitisofunknownetiology;C3,CSFofpatient withmultiplesclerosis.

SamplesC1–C3werenegativeforHHV-1byPCR.

Thenumbersofsequencesrepresentingeachspecies/familyareshowninbrackets.

Fig.1.CoverageplotofHHV-1andHCVgenomes.SequencingreadsfromsamplesS1andSDNaseweremappedtothereferencegenomeofHHV-1(GenBankaccession numberAB618031)andcontrolsampleC-HCVwasmappedtoHCVgenome(GenBankaccessionnumberEF407475).

Fig.2.PCRdetectionofhumanherpesvirustype1(HHV-1).Lane:1,CSFsample drawnatadmission;2,positivecontrol(400targetcopiesofHHV-1DNAfrom HHV-1infectedVerocells3,negativecontrolconsistingofwater;M,GeneRuler50bpDNA Ladder(Feremntas,ThermoScientific,Waltham,USA;50bp).Expectedproductsize was148bp.

DNA(38␮l)wasmixedwith5␮lof10×buffer(LifeTechnologies, Carlsbad,USA), 2_␮lof PCR primers(10_␮M),2_␮lof 50_×dNTP (10mM),5␮lof25mMMgCl2(LifeTechnologies,Carlsbad,USA)

and2.5UofTaqPolymerase(LifeTechnologies,Carlsbad,USA)in

finalvolumeof50␮l.Aftertheinitial5mindenaturationin94◦C,50 cyclesof94◦Cfor30sand68◦Cfor30swererunusing thermocy-cler9700AppliedBiosystem(AppliedBiosystem,LifeTechnologies, Carlsbad,USA).PCRproductswereanalyzedbyelectrophoresisin 2%agarose.DetectionlimitofourPCRwas≥44viralcopiesper reac-tion.PCRsensitivitywasdeterminedusingserialdilutionsofDNA extractedfromHHV-1infectedVerocellculture(VirusIsolationand ReferenceServiceLaboratory,NewYorkStateHealthDepartment). InthepresentstudyweunexpectedlydetectedHHV-1,whichis aDNAvirus,whileanalyzingCSFsamplesubjectedtoRNA extrac-tionbythemostwidelyusedChmoczynskimethodfollowedby RNAamplification.Thisfindingcouldhavebeentheresultofviral mRNAdetection,oralternatively,a substantialquantityofDNA couldhavebeenretainedafterRNAextraction,thusallowingfor thedetectionofbothRNAandDNAsequences.Toverifythelatter possibility,serial10-folddilutionsofplasmidtemplate (pCR2.1-TOPO,Invitrogen, Waltham, USA) werespikedinto CSF sample fromanuninfectedpatientandsubjectedtoparallelRNAandDNA extractions(TRIZOLLSReagent,LifeTechnologiesandQIAampDNA MiniKit,Qiagen,Venlo,Netherlands).Geneticmaterialextracted fromeachsamplewassubjectedtoreal-timePCR(10min dena-turationin95◦Cfollowedby55cyclesof95◦Cfor10sand50◦C for 15s) using LightCycler FastStart DNA Master SYBR Green I (RocheMolecularDiagnostics,Basel,Switzerland)andprimers(5 -GTAAAACGACGGCCAG-3;5-CAGGAAACAGCTATGAC-3)resulting in 202bp amplificationproduct. As seenin Fig. 3, RNA extrac-tionretainedasignificantamountofDNAwhichwassufficientfor

Fig.3.Real-timePCRdetectionofserial10-folddilutionsofplasmidtemplate(pCR2.1-TOPO;Invitrogen,Waltham,USA)spikedintoCSFfromanuninfectedpatient.Samples wereeitherextractedforDNA(D),orRNA(R).Samples1,2,3,4contained103_,104_,105_and106_{plasmidcopiesperreaction,respectively.}

real-timePCRdetectionof≥103_{plasmidcopies.Surprisingly,for}

awiderangeofplasmidconcentrationstested(103_,104_,105_and

106_{perreaction)theRNAextractionresultedinonly1.6–3.1fold}

sensitivitydecreasewhencomparedtoDNAextraction.

Determiningtheetiologyofencephalitisremainschallenging inclinicalpractice,partlydue tothesheer numberofpotential pathogens.Sofarover100viruseswerereportedtobecapable ofcausingencephalitis,andthisnumberisstillgrowing(Debiasi andTyler,2004).Theresultsofsomerecentstudiesindicatethat high-throughputsequencingcouldbesuccessfullyusedtoidentify causativepathogensinpatientswithcentralnervoussystem infec-tions.Forexample,Chanetal.(2014)identifiedmeaslesvirusand HHV-1infrozenbraintissuesamplesusingIlluminaHiSeq2000, whereasTanleetal.(2013)usedFLXgenomesequencer(454Life Science,Roche)andWilsonetal.(2014)usedIlluminaMiSeqto detectCyclovirusand LeptospirasanatarosainCSF,respectively. Moreover,NGSbasedapproachesenableidentificationofallagents inasinglerun,regardlesswhethertheirsequenceisknown,ornot (Bigetal.,2009;Dupuisetal.,2011;Glaseretal.,2006;Quanetal., 2010).

Ourapproach,whichemployedNuGenRibo-SPIAfollowedby NGSandmetagenomicdataanalysis,ledtotheidentificationof HHV-1inCSFfromapatientwithacuteencephalitisatatimewhen specificantibodiescouldnot yetbedetected.Subsequently,the etiologyofencephalitiswasconfirmedbyspecificPCRaswellas HHV-1seroconversioninthefollowupCSFsamples.

Host-derived geneticmaterial is commonlypresent in clini-calsamplesandpresentsamajorchallengeforNGSanalysis(Lim etal.,2013).Acommonlyusedmethodtoreducehumangenetic backgroundisDNaseItreatmentofthesamples(Allanderetal., 2001).Whilewefoundthatthisapproachwouldresultinmore thantwofolddecreaseinthenumberofhost-derivedsequences,it significantly(30–50fold)increasedthenumberofbacterialand “other” sequences probably becauseof diminishedcompetition fromhumanDNA.Notunexpectedly,theuseofDNaseIresulted alsointhereductionofHHV-1sequences.

Interestingly,samplesprovidedasignificantnumberofbacterial readsandsequencesclassifiedasother(plants,plantviruses).This problemwaspreviouslyreportedforotherNGSstudies(Bzhalava etal.,2012;Salteretal.,2014).Bacterialcontaminationisoften foundinexistinghuman-derivedRNAsequencedatasetsandwill remainanimportantissueforfutureNGSdiagnosticapplication (Laurenceetal.,2014;Strongetal.,2014).ThepresenceofHCV

sequencesinanalyzedCSFsampleswasmostlikelytheresultof indexmis-assigninIlluminasequencing(Kircheretal.,2012)and thesecondindependentrunwiththreeHHV-1negativecontrolCSF samplesrevealedneitherHCVnorHHV-1sequences.Intriguing, samplesfromthefirstruncontainedasmallnumberofinfluenzaC sequences,whichraisesthequestionoftheirorigin.However,the analysiswasconductedonthepeakoffluseason(endofFebruary) anditiscurrentlyalmostimpossibletocompletelyeliminate envi-ronmentalcontaminantsduringlibrarypreparation processand mis-assignedreadsfromothersamples.

Surprisingly,wedetectedHHV-1geneticmaterialinCSFwhich wasextractedforRNAandsimilaroccurrencehasbeenreported by Chan et al. (2014).Modified acid guanidinium thiocyanate-phenol-chloroformRNAextractionisknowntogenerategenomic DNAcontaminants(SiebertandChenchik,1993)andinourstudy significantDNApresenceafterRNAextractionwasconfirmedby quantitativeanalysisofplasmidtemplate.

Inconclusions,itseemsthatNGS,whennecessaryprecededby unspecificnucleicacidamplification,couldsupplementcurrently used diagnostic methods for the determination of encephalitis etiology.However,duetocommonindexmis-assignmentand con-taminationwithreagents-derivednucleicacids,thepresenceofany identifiedpathogenshouldbeverifiedbyPCRorculture.Wealso foundthatextractionoftotalRNAcouldbeusedforthe identifica-tionofbothDNAandRNApathogensinasinglerun.

Acknowledgment

ThisstudywassupportedbyFoundationforPolishScience,grant No.POMOST/2013-7/2andPolishNationalScienceCenter(NCN), grantNo.N/N401/646940.

References

Allander,T.,Emerson,S.U.,Engle,R.E.,Purcell,R.H.,Bukh,J.,2001.Avirusdiscovery methodincorporatingDNasetreatmentanditsapplicationtothe

identificationoftwobovineparvovirusspecies.Proc.Natl.Acad.Sci.U.S.A.98, 11609–11614.

Altschul,S.F.,Gish,W.,Miller,W.,Myers,E.W.,Lipman,D.J.,1990.Basiclocal alignmentsearchtool.J.Mol.Biol.215,403–410.

Big,C.,Reineck,L.A.,Aronoff,D.M.,2009.Viralinfectionsofthecentralnervous system:acase-basedreview.Clin.Med.Res.7,142–146.

Bzhalava,D.,Ekstrom,J.,Lysholm,F.,Hultin,E.,Faust,H.,Persson,B.,Lehtinen,M., deVilliers,E.M.,Dillner,J.,2012.PhylogeneticallydiverseTTvirusviremia amongpregnantwomen.Virology432,427–434.

Chan,B.K.,Wilson,T.,Fischer,K.F.,Kriesel,J.D.,2014.Deepsequencingtoidentify thecausesofviralencephalitis.PLOSONE9,e93993.

Chaudhuri,A.,Kennedy,P.G.,2002.Diagnosisandtreatmentofviralencephalitis. Postgrad.Med.J.78,575–583.

Debiasi,R.L.,Tyler,K.L.,2004.Molecularmethodsfordiagnosisofviral encephalitis.Clin.Microbiol.Rev.17,903–925,tableofcontents.

Dupuis,M.,Hull,R.,Wang,H.,Nattanmai,S.,Glasheen,B.,Fusco,H.,Dzigua,L., Markey,K.,Tavakoli,N.P.,2011.Moleculardetectionofviralcausesof encephalitisandmeningitisinNewYorkState.J.Med.Virol.83,2172–2181.

Glaser,C.A.,Honarmand,S.,Anderson,L.J.,Schnurr,D.P.,Forghani,B.,Cossen,C.K., Schuster,F.L.,Christie,L.J.,Tureen,J.H.,2006.Beyondviruses:clinicalprofiles andetiologiesassociatedwithencephalitis.Clin.Infect.Dis.43,1565– 1577.

Goto,N.,Prins,P.,Nakao,M.,Bonnal,R.,Aerts,J.,Katayama,T.,2010.BioRuby: bioinformaticssoftwarefortheRubyprogramminglanguage.Bioinformatics 26,2617–2619,http://hannonlab.cshl.edu/fastxtoolkit/index.html,2015, FASTX-Toolkit.

Kennedy,P.G.,2005.Viralencephalitis.J.Neurol.252,268–272.

Kircher,M.,Sawyer,S.,Meyer,M.,2012.Doubleindexingovercomesinaccuracies inmultiplexsequencingontheIlluminaplatform.NucleicAcidsRes.40,e3.

Laurence,M.,Hatzis,C.,Brash,D.E.,2014.Commoncontaminantsin next-generationsequencingthathinderdiscoveryoflow-abundance microbes.PLOSONE9,e97876.

Lim,Y.W.,Schmieder,R.,Haynes,M.,Willner,D.,Furlan,M.,Youle,M.,Abbott,K., Edwards,R.,Evangelista,J.,Conrad,D.,Rohwer,F.,2013.Metagenomicsand metatranscriptomics:windowsonCF-associatedviralandmicrobial communities.J.Cyst.Fibros.12,154–164.

Lunter,G.,Goodson,M.,2011.Stampy:astatisticalalgorithmforsensitiveandfast mappingofIlluminasequencereads.GenomeRes.21,936–939.

Mailles,A.,Vaillant,V.,Stahl,J.P.,2007.InfectiousencephalitisinFrancefrom2000 to2002:thehospitaldatabaseisavaluablebutlimitedsourceofinformation forepidemiologicalstudies.Med.Mal.Infect.37,95–102.

Martin,M.,2011.Cutadaptremovesadaptersequencesfromhigh-throughput sequencingreads.EMBnetJ.(NorthAmerica),17.

Quan,P.L.,Wagner,T.A.,Briese,T.,Torgerson,T.R.,Hornig,M.,Tashmukhamedova, A.,Firth,C.,Palacios,G.,Baisre-De-Leon,A.,Paddock,C.D.,Hutchison,S.K.,

Egholm,M.,Zaki,S.R.,Goldman,J.E.,Ochs,H.D.,Lipkin,W.I.,2010.Astrovirus encephalitisinboywithX-linkedagammaglobulinemia.Emerg.Infect.Dis.16, 918–925.

Radford,A.D.,Chapman,D.,Dixon,L.,Chantrey,J.,Darby,A.C.,Hall,N.,2012.

Applicationofnext-generationsequencingtechnologiesinvirology.J.Gen. Virol.93,1853–1868.

Salter,S.J.,Cox,M.J.,Turek,E.M.,Calus,S.T.,Cookson,W.O.,Moffatt,M.F.,Turner,P., Parkhill,J.,Loman,N.J.,Walker,A.W.,2014.Reagentandlaboratory

contaminationcancriticallyimpactsequence-basedmicrobiomeanalyses. BMCBiol.12,87.

Siebert,P.D.,Chenchik,A.,1993.Modifiedacidguanidinium

thiocyanate-phenol-chloroformRNAextractionmethodwhichgreatlyreduces DNAcontamination.NucleicAcidsRes.21,2019–2020.

Solomon,T.,Hart,I.J.,Beeching,N.J.,2007.Viralencephalitis:aclinician’sguide. Pract.Neurol.7,288–305.

Strong,M.J.,Xu,G.,Morici,L.,SplinterBon-Durant,S.,Baddoo,M.,Lin,Z.,Fewell,C., Taylor,C.M.,Flemington,E.K.,2014.Microbialcontaminationinnext generationsequencing:implicationsforsequence-basedanalysisofclinical samples.PLoSPathog.10,e1004437.

Studahl,M.,Lindquist,L.,Eriksson,B.M.,Gunther,G.,Bengner,M.,Franzen-Rohl,E., Fohlman,J.,Bergstrom,T.,Aurelius,E.,2013.Acuteviralinfectionsofthe centralnervoussysteminimmunocompetentadults:diagnosisand management.Drugs73,131–158.

Tanle,V.,vanDoorn,H.R.,Nghia,H.D.,Chau,T.T.,Tule,T.P.,deVries,M.,Canuti,M., Deijs,M.,Jebbink,M.F.,Baker,S.,Bryant,J.E.,Tham,N.T.,Boni,N.T.B.K.M.F.,Loi, T.Q.,Phuongle,T.,Verhoeven,J.T.,Crusat,M.,Jeeninga,R.E.,Schultsz,C.,Chau, N.V.,Hien,T.T.,vanderHoek,L.,Farrar,J.,deJong,M.D.,2013.Identificationof anewcyclovirusincerebrospinalfluidofpatientswithacutecentralnervous systeminfections.mBio4,e00231–e313.

Wilson,M.R.,Naccache,S.N.,Samayoa,E.,Biagtan,M.,Bashir,H.,Yu,G.,Salamat, S.M.,Somasekar,S.,Federman,S.,Miller,S.,Sokolic,R.,Garabedian,E.,Candotti, F.,Buckley,R.H.,Reed,K.D.,Meyer,T.L.,Seroogy,C.M.,Galloway,R.,Henderson, S.L.,Gern,J.E.,DeRisi,J.L.,Chiu,C.Y.,2014.Actionablediagnosisof

neuroleptospirosisbynext-generationsequencing.N.Engl.J.Med.370, 2408–2417.