MicroRNA Expression Profiling by Bead Array Technology in Human Tumor Cell Lines Treated with InterferonAlpha2a相关交流-蚂蚁淘在线

MicroRNAsarepositiveandnegativeregulatorsofeukaryoticgeneexpressionthatmodulatetranscriptabundancebyspecificbindingtosequencemotifslocatedprevalentlyinthe3′untranslatedregionsoftargetmessengerRNAs(mRNA).Interferon-alpha-2a(IFNα)inducesalargesetofproteincodinggenesmediatingantiproliferativeandantiviralresponses.Hereweuseaglobalmicroarray-basedmicroRNAdetectionplatformtoidentifygenesthatareinducedbyIFNαinhepatoma-ormelanoma-derivedhumantumorcelllines.Despitetheenormousdifferencesinexpressionlevelsbetweenthesemodels,wewereabletoidentifymicroRNAsthatareupregulatedbyIFNαinbothlinessuggestingthepossibilitythatinterferon-regulatedmicroRNAsareinvolvedinthetranscriptionalrepressionofmRNArelevanttocytokineresponses.

KeyWords:MicroRNAs-OligonucleotideArraySequenceAnalysis-Interferons-Melanoma-Hepatoma-ReverseTranscriptasePolymeraseChainReaction-SuppressorofCytokineSignalingProteins

Introduction

Thegeneexpressionpatternsoftumor-derivedcelllinesdiffergreatly,asdotheirresponsestoantiproliferativeeffectsofinterferons(IFNs).Thecauseofthisvariationhasbeenunderinvestigationformorethan40years,butonlybasicregulatorymechanismsofinterferonsignalingareunderstoodtoday.SmallregulatorygenomeencodedRNAs,suchasmicroRNAs,haverecentlyattractedattentioningenomicresearch.Newmethodstoanalyzethelevelsoftheseregulatoryelementsarenowcommerciallyavailable,butthepowerofthesetechniquesisstilldiscussedextensively.OurstudywasdesignedtocomparetwomethodsformicroRNAdetectionwithrespecttousefulnessindefinedcellcultureassays.TheexperimentaldesignassessesvariationbetweenthetwocelllinesandthetreatmenteffectsofIFNα.

AhallmarkofthetherapeuticactivityoftypeIinterferonsistheinductionofantiproliferativeactivitymediatedbytheupregulationofseveralhundredresponsegeneswithpleiotropicfunctions(1).Thesegenescanbedividedintotwomajorclassesbasedonthekineticpropertiesofinduction(2).Primaryresponsegenes(PRGs)areupregulatedwithin24hafterthecytokinesignalandthesecondaryresponsegenes(SRGs)areinducedfollowingday 1whentheactivityofthePRGsdecays.IncontrasttoSRGs,allPRGsstudiedtodatecontainbonafideinterferonresponseelementsinthepromoterregion,whicharerequiredforbindingoftheinterferon-stimulatedgenefactor3(ISGF3)complexandforjanuskinase/signaltransduceroftranscription(JAK/STAT)-pathway-mediatedsignaling.

ExpressionofPRGsisturnedoffbyproteinstermedsuppressorsofcytokinesignaling(SOCS)(3).Asthenomenclatureindicates,thisclassofpolypeptideshasthecapacitytointerfereandsilenceothercytokine-inducedsignalingcascades(forreviewsee(4)).SOCS1forinstanceispartoftheearlyinducIBLePRGclusteranddownmodulationoccurstogetherwiththeothergenesbeforeonsetofSRGexpression.ItisbelievedthatfeedbackinhibitionofJAK/STATsignalingbySOCS1repressestranscriptomemodulationofIFNαsignaling(5).RegulationofSOCSproteintranslationbyinterferon-regulatedmicroRNAs(IRmiRs)wouldenhancethepotentialofcytokinefineregulation.IthasbeenreportedthatmiR-19antagoNISTsleadtohigherSOCS1levelsandmiR-19mimicscanrepressSOCS1reporterconstructs,thusobviouslysupportingthebioinformaticpredictionsthatSOCS1isadirecttargetofmiR-19(6).InhibitionofSOCSactivitycouldforinstanceprolongthedurationofcytokineactivity,whichhasobviousclinicalimplications.

FollowingthediscoveryofmicroRNAsinvirtuallyallhighereukaryoticorganismssignificantresearcheffortswereinitiatedtoaddressthefunctionofthesecatalyticoligonucleotideswhicharethenaturalcounterpartsofsyntheticsmallinhibitoryRNAs(siRNAs)usedforexperimentalgenesilencing(forreviewsee(7)).MicroRNAsarepositiveandnegativemodulatorsoftheexpressionofentiregeneclustersthatcontaincomplementarymicroRNArecognitionsequencemotifsinthe3′-UTR.Today,predictionofmicroRNAtargetgenesbyhomology-basedalgorithmsisstillambiguous(8).TheactivityofoneorseveralmicroRNAscouldexplainsuppressionoftheentirePRGclusterprovidedthatmicroRNAabundanceisregulatedbyIFNα.Alternatively,microRNA-mediateddegradationoftranscriptsencodingnegativeregulatoryproteinswouldalsoabolishPRGexpressionandrestoreIFNαresponsiveness.

Somerecentreportsshowedthatinterferonbeta(IFNβ)stimulationcanboostmicroRNAlevelsincellculturetogetherwithinhibitionofviralreplication(9).AtthispointitisanopenquestionwhetherthisinductionisIFNβspecificorasharedfeatureofalltypeIinterferons.ToinvestigatewhethermicroRNAarealsoinvolvedinregulationofIFNαresponse,weusedtwohuman-tumor-derivedcelllines:themelanomalineME-15(10)andthehepatomalineHuH7(11).Wehavechosenthesecelllinesasmodels,becausewehaveagoodunderstandingoftheIFNαresponsesatthemRNAandtheproteinlevelsinthesecelllines.FurtherwechosetouseamelanomacelllinebecauseIFNisalsousedfortreatmentofthiscancertype.HuH7iscommonlyusedasamodelfortestingantiviraleffectsofIFNintheHCVrepliconsystem.InbothmodelsefficientresponsestoIFNαhavebeenshownatthefunctionalandtranscriptionallevel.IFNαresponsegenescarryresponseelementsintheirpromoterregionandthesemotifsareresponsibleforgeneexpressionwithsimilarefficiencyinmanycelltypes.ThereforeweexpectedtofindasimilarregulatedsetofgenesinbothlinesgiventhatIRmiRgenesareregulatedbythesamemechanism,whereassomeconstitutivelyexpressedmicroRNAgeneswereexpectedtobecelltypespecificforfunctionalreasons.WehavechosenaDNA-microarray-basedtechnology(Illumina)forthemulti-parallelexpressionanalysisofallknownhumanmicroRNAs(http://microrna.sanger.ac.uk/;Release10.0:August2007).ThismethodallowedustoprocesstotalRNAastemplate,allowingthepossibilityofmRNAgeneexpressionprofilinginfurtherexperiments.Briefly,annealingofmicroRNAspecificprimerscombinedwithenzymaticpolyadenylationallowsmulti-parallelpolymerasechainreaction(PCR)-mediatedamplificationofindividualmicroRNAs.TheoutputofthisstepisaDNAampliconlibrarythatreflectstoalargeextenttheoriginalstoichiometryofmaturemicroRNAsinacellortissue(12).PCRamplificationisperformedwithfluorescentlylabeledprimers,whichallowsquantitativesignaldetectionbyconventionalconfocallaserscanning.

MaterialsandMethods

CellCulture,InterferonTreatment,andRNAPrecipitation

Melanomacells(ME-15)wereculturedinRPMI1640withL-Glutaminesupplementedwithnon-essentialaminoacidsandsodiumpyruvate(1mM)andhepatoma(HuH7)cellswereculturedinDMEM+GlutaMAX.Bothmediacontained10%FBS.AllcellculturereagentswerepurchasedfromInvitrogen(Gibco®).Roferon(Interferonalpha2a,Roche)wasdilutedinfreshmediumtoafinalconcentrationof1,000U/mLandcontrolculturesweregrownwithoutcytokine.Cellswereculturedat37°Cinahumidifiedatmospherecontaining5%CO₂.TotalRNApreparationwascarriedoutusingTRIZOL(Invitrogen)totalRNAextractionusing1/2volumeof1-bromo-3-chloro-propane(molecularBIOLOGygrade,SIGMA)aschloroformsubstitute.ForefficientrecoveryofsmallRNAs,DNALoBindtubes(Eppendorf)wereusedandallcentrifugationstepswereperformedatmaximumspeedand4°CinanEppendorf5417Rcentrifuge.TotalRNAwasprecipitatedwith2volof2-propanol(Fluka)at−20°Cforatleast16h.TheRNApelletwaswashedwith75%ethanol(Merck),dried,anddissolvedinDEPC-treatedwater(Ambion).TheRNAwasquantifiedwithQuant-iT™RiboGreen®RNAAssay(Invitrogen)assuggestedbyIllumina.

IlluminaBeadArrayMicroRNADetection

Startingwith500ng/sampleoftotalRNA,maturemicroRNAswereamplifiedwiththeIlluminahumanv1MicroRNAexpressionprofilingkitcontainingprimersfor743humanmicroRNAs.Theresultingampliconswerehybridizedtoa96sampleuniversalprobecapturearrayandfluorescentsignalsweredetectedbyconfocallaserscanning.AllstepswereperformedaccordingtoIllumina’sinstructionsmanual.

DataProcessingandStatisticalAnalysis

ThedatawasprocessedwithBeadstudiosoftware(version3.1.3,geneexpressionmodule3.3.8)includingthecalculationofdetectionpvaluesbasedonnegativecontrolbeadsignals.Log-transformation,loessnormalization(13)andstatisticalanalysiswereperformedwithR(2.8.1)(14)usingthepackagelumi(1.8.3)(15)andsoftwarecontainedtherein,inparticularlimma(2.16.4)(16).Statisticalmodelswerechosenasfollows:alinearmodel(limmatstatistics)withtwoseparatecoefficientsforHuH7andME-15cellswasusedfortheselectionofdifferentlyexpressedgenesshowninFig.1andTable1.Statisticsrepresentedinthetableswerecalculatedbytestingthetwoindicatedconditionsasindependentfactors.InTable1,pvalueswereadjustedbythefalsediscoveryratemethod(17).TreatmenteffectsshowninTable2bandFig.3aweremodeledwithtwocoefficients(cellline,treatment)fortimepoint4h,pvaluesarisefromtstatistics.Normalizedrelativefluorescencelevelswerecalculatedby2^mean(oflog2transformed,loessnormalizedvalues).Changefactors(CHF)werecalculatedasfoldchangeonthelinearscaleminus1aspreviouslydescribed(2).Rawdata,non-normalized,andnormalizedmicroRNAexpressiondatahavebeensubmittedtotheGeneExpressionOmnibuswithaccessionnumberGSE16421.

MediaObjects/12575_2009_9012_Fig1_HTML.gif

Fig.1DifferentialmicroRNAexpressioninhumanmelanoma(ME-15)andhepatoma(HuH7)cells.microRNAexpressionlevelswerecomparedintwocelllinesattwodifferenttimepointsandcorrectedforthetreatmenteffect.The50mostsignificant(pvaluebelow10⁻¹²)microRNAexpressionvaluesfromuntreatedsamplesareshowninaheatdiagramincludinghybridizationcontrolsasreferencefortechnicalvariance.Whiteindicatesnoiselevels,yellowindicatesthefirstquartile,orangethemedian,redthethirdquartile,andblackmaximumexpressionlevels.Theintensitydata,significancevaluesandtheIFNα-dependentexpressionlevelsaresummarizedinTable1.

Table1CelllinedifferencesinmicroRNAexpression

364

Listofcalculatedrelativefluorescentlevels(withoutbackgroundcorrection)andcalculateddifferences(CHF)betweenhepatoma(HuH7)andmelanoma(ME-15)cells.AlinearmodelwithME-15andHuH7asseparatefactorswasusedtoestimatecelllinedifferencesinmicroRNAexpression.Thegeneswiththehighestsignificance(andlogfactorchange>0.5)arelistedintheuppersectionandthebottomshowsdataforthehybridizationcontrolsasreference.Changefactorsshowthecelllinedifferencesoftheretransformedmeantothelinearscale.pvalues(limmattest)wereadjustedbyfalsediscoveryratecorrection.Significancecodesaredefinedbytheintervals:‘***’<0.001≤‘**’<0.01≤‘*’<0.05

Table2ModulationofmicroRNAexpressionbyIFNα—4and24hafterstimulation

	ME-15	HuH7	CHF	ME-15	HuH7	CHF	ME-15	HuH7	CHF	ME-15	HuH7	CHF
	Control						Interferon-alphatreated
	4h			24h			4h					24h
microRNAsratedhigherinHuH7
hsa-miR-122a	685	30,912	44.14***	704	32,058	44.51***	717	30,459	41.46***	876	29,044	32.17***
hsa-miR-224	359	10,085	27.07***	345	7,358	20.31***	358	8,183	21.87***	333	7,483	21.44***
hsa-miR-483	915	23,671	24.87***	811	18,073	21.29***	849	17,784	19.95***	802	17,554	20.88***
hsa-miR-200a	517	8,392	15.22***	667	13,543	19.29***	722	11,317	14.68***	500	12,649	24.30***
hsa-miR-218	405	5,654	12.96***	417	6,527	14.64***	419	5,764	12.76***	431	6,155	13.29***
hsa-miR-618	700	7,549	9.78***	649	8,128	11.52***	651	8,042	11.35***	616	7,416	11.03***
hsa-miR-215	638	6,826	9.71***	829	6,013	6.25***	663	6,641	9.02***	723	6,501	7.99***
hsa-miR-192	3,956	34,724	7.78***	3,401	34,181	9.05***	4,021	29,382	6.31***	3,343	33,293	8.96***
hsa-miR-194	4,410	32,950	6.47***	4,142	32,991	6.96***	4,700	33,182	6.06***	4,410	32,422	6.35***
hsa-miR-182	1,575	11,480	6.29***	2,541	13,088	4.15***	2,227	12,512	4.62***	3,067	13,907	3.53***
hsa-miR-452	434	2,284	4.27***	707	3,326	3.70***	530	3,378	5.37***	636	3,028	3.76***
hsa-miR-183	1,763	8,748	3.96***	2,123	8,945	3.21***	2,000	8,984	3.49***	2,129	8,258	2.88***
hsa-miR-200b	1,934	7,551	2.90***	1,692	9,633	4.69***	1,705	9,393	4.51***	1,911	10,855	4.68***
hsa-miR-143	1,355	5,087	2.75***	1,775	7,167	3.04***	1,698	5,688	2.35***	1,765	7,519	3.26***
hsa-miR-624	4,852	13,418	1.77**	4,188	11,593	1.77***	4,479	11,718	1.62**	3,872	10,539	1.72**
hsa-miR-99a	8,653	20,232	1.34**	8,887	20,645	1.32***	9,064	18,019	0.99**	8,870	19,506	1.20**
hsa-miR-27b	10,067	22,544	1.24**	11,924	23,914	1.01***	11,564	23,044	0.99***	12,182	24,515	1.01**
microRNAsratedhigherinME-15
hsa-miR-146a	36,976	15,721	−1.35**	33,410	16,103	−1.07***	35,166	15,137	−1.32***	35,688	19,962	−0.79***
hsa-miR-422b	19,264	7,759	−1.48**	19,962	7,832	−1.55***	20,641	7,462	−1.77***	17,453	7,152	−1.44***
hsa-miR-149	5,868	2,246	−1.61**	6,195	2,030	−2.05***	6,061	1,743	−2.48***	5,477	2,418	−1.26**
hsa-miR-510	1,051	368	−1.86***	1,489	396	−2.76***	1,435	366	−2.92***	1,053	384	−1.74***
hsa-miR-30a-5p	10,664	3,717	−1.87**	10,223	3,700	−1.76***	10,686	3,477	−2.07***	10,466	4,538	−1.31***
HS_98	3,274	1,065	−2.08**	3,225	859	−2.76***	3,445	965	−2.57***	2,850	750	−2.80***
hsa-miR-340	4,583	1,443	−2.18**	4,218	1,220	−2.46***	4,664	1,061	−3.40***	3,078	933	−2.30***
HS_182.1	1,965	615	−2.2**	1,687	606	−1.79**	1,903	562	−2.39***	1,844	534	−2.45***
hsa-miR-378	4,619	1,371	−2.37**	4,975	1,364	−2.65***	4,994	1,213	−3.12***	4,188	1,237	−2.38***
HS_305_b	4,105	1,207	−2.40*	3,853	833	−3.63***	4,278	977	−3.38***	3,688	1,076	−2.43***
hsa-miR-505	1,312	343	−2.83***	2,070	342	−5.05***	2,050	360	−4.70***	2,030	362	−4.60***
hsa-miR-10a	1,570	350	−3.48**	1,592	350	−3.55***	1,828	360	−4.07***	1,880	352	−4.34***
hsa-miR-10b	1,695	378	−3.49**	2,929	371	−6.89***	3,564	411	−7.67***	4,297	382	−10.26***
hsa-let-7g	10,720	2,302	−3.66**	16,046	3,077	−4.21***	13,792	2,854	−3.83***	15,435	2,568	−5.01***
hsa-miR-584	11,643	2,261	−4.15***	12,717	1,873	−5.79***	12,394	1,912	−5.48***	12,616	1,674	−6.54***
hsa-let-7i	11,701	2,255	−4.19**	16,410	2,707	−5.06***	15,319	2,251	−5.80***	15,718	2,563	−5.13***
hsa-miR-330	9,029	1,653	−4.46***	9,684	1,670	−4.80***	9,432	1,397	−5.75***	7,773	1,380	−4.63***
HS_307_b	2,453	444	−4.52***	4,188	443	−8.45***	3,669	407	−8.01***	3,318	448	−6.41***
hsa-miR-34c	2,292	412	−4.56**	3,303	426	−6.75***	3,524	381	−8.25***	2,403	404	−4.95**
hsa-miR-133a	9,270	1,594	−4.82***	7,942	1,662	−3.78***	8,488	1,435	−4.92***	6,495	1,783	−2.64***
hsa-miR-361	5,234	828	−5.32***	7,641	755	−9.13***	7,356	875	−7.41***	7,904	706	−10.19***
hsa-miR-508	2,552	325	−6.86**	4,280	376	−10.39***	3,965	357	−10.09***	4,013	336	−10.96***
hsa-miR-9*	8,372	1,039	−7.06***	10,114	940	−9.76***	10,027	1,263	−6.94***	9,948	843	−10.81***
hsa-miR-31	5,967	661	−8.03***	5,742	720	−6.97***	6,619	657	−9.07***	5,978	728	−7.21***
hsa-miR-506	5,855	625	−8.37***	6,125	713	−7.59***	5,916	665	−7.90***	5,602	675	−7.30***
hsa-miR-211	5,147	476	−9.81***	3,218	490	−5.57***	3,833	471	−7.14**	5,575	464	−11.01***
hsa-miR-296	6,194	570	−9.86***	5,913	504	−10.72***	6,441	557	−10.57***	4,369	532	−7.21***
hsa-miR-598	4,445	401	−10.09***	4,475	386	−10.59***	4,934	424	−10.63***	4,127	367	−10.24***
hsa-miR-199b	9,081	775	−10.71***	9,478	916	−9.35***	9,629	865	−10.13***	8,059	911	−7.85***
hsa-let-7b	8,236	539	−14.28***	12,563	475	−25.44***	12,150	500	−23.30***	12,309	452	−26.24***
hsa-miR-509	10,413	536	−18.44***	10,346	512	−19.22***	9,966	548	−17.20***	9,311	529	−16.62***
hsa-miR-9	12,587	422	−28.85***	15,870	447	−34.50***	15,502	418	−36.09***	16,402	429	−37.26***
hsa-miR-514	12,065	−32.16***	13,112	358	−35.64***	12,201	315	−37.74***	12,355	355	−33.83***
HyBDridizationcontrols
array_hyb_con4	6,404	7,110	0.11	7,724	6,684	−0.16	7,031	6,739	−0.04	6,964	6,746	−0.03
array_hyb_con3	6,923	7,681	0.11	7,833	7,368	−0.06	7,422	7,054	−0.05	6,906	7,365	0.07
array_hyb_con1	10,779	11,868	0.10	11,812	11,728	−0.0.01	11,436	11,382	0.00	11,059	11,522	0.04
array_hyb_con2	10,258	11,145	0.09	11,382	10,829	−0.05	11,075	10,650	−0.04	9,997	10,283	0.03

MicroarraydataforRT-PCRvalidatedmicroRNAs,seedescriptionofTable1.Averagerelativefluorescentlevelsandtheirchangefactors(CHFs)areshownforhepatoma(HuH7)andmelanoma(ME-15)cellsattwotimepoints.Geneswereselectedbysignificancecalculatedinalinearmodelwithinteractionassumingthatthechangemaynotoccurinbothcelllines(e.g.miR-10bhasnodetectablelevelsinHuH7).pvaluesignificancecodes:0‘***’0.001‘**’0.01‘*’0.05‘.’0.1‘‘1.SectioncshowsthemicroarrayexpressionlevelsofmicroRNAsassayedbyRT-PCR(treatmentanalysis)

QuantitativePCRandDataProcessing

microRNAlevelsweremeasuredusingTaqMan®microRNAassays(AppliedBiosystems)usingtheTaqMan®MicroRNAreversetranscription(RT)kitwithTaqMan®2×universalPCRmastermix(NoAmpErase®UNG)asrecommendedbythesupplier.TennanogramsoftotalRNAwasusedasinputforamplificationusingthesamplesusedformicroarrayanalysis.Reversedtranscriptaseproductswerediluted1:15andmeasuredonanABI7900HTfastreal-timePCRsystem.Technicalreplicateswererunonthreedifferentplates(onewith40cyclesandtwowith50cycles)andthresholdforcyclingtime(CT)calculationwassetforallprobesto0.2.ForestimationofendogenoussmallRNAcontent,thenucleolarRNARNU48wasusedascontrolandreference.

Standarderror(Δx)wascalculatedbytheaveragestandarderroroftreatedanduntreatedMNEforbiologicalreplicates.

ResultsandDiscussion

Thefollowingtechnicalaspectshavetobeconsideredforresultinterpretation.ThedatasetofthemicroRNAbeadarrayassayisnotdirectlycomparabletogeneexpressionarrayswhereinvitrotranslatedtranscriptsaredirectlyhybridizedtotheprobes.MoreoverIllumina’sbeadarraytechnologytendstohavehigherbackgroundfluorescencelevelsandlowerchangefactorvaluesthanGenechipsfromAffymetrix.Background(averageofnegativecontrolsignals)andnoise(standarddeviationofnegativeprobesfromeachsample)were528 ± 60and229 ± 67,respectively.Thedensityofallsamplesshowsabimodaldistributionpeakingaroundthebackgroundfluorescentlevelsandtherobustlevels(approximately12,000).Thecurveisskewedtotherightandpeakdensityheightisfoundintheratio4:1consideringallprobes(datanotshown).Thedistributionofprobesdetectedinallsamples(detectionpvaluethresholdat0.01)hasaplateaurangingfromabout2,000closetothedetectorsmaximumcapacityof2^16relativefluorescentunits(12).Asexpected,thecorrelationofdatacomingfrombiologicalreplicatesr² = 0.952 ± 0.028(notnormalized)andr² = 0.956 ± 0.022(afterloessnormalizationandlog-transformation)waslowerthanfortechnicalreplicatesr² > 0.97(12).Wepreferredloessnormalizationtoquantilenormalizationbecausethelaterwastooaggressiveforthegivensmallprobenumbers.

AsafirststepwewishedtoaddresstherobustnessofthemicroRNAarrayinprobedetectionbyselectingmicroRNAgenesthataredetectedunderallexperimentalconditionswithhighstatisticalsignificanceinallbiologicaltriplicates(detectionpvalue < 0.01).Ineachsetoftriplicatesamples(control,4hor24h,IFNαstimulation)wedetectapproximately270genesthatfulfilltheabovecriteria.ThiscorrespondstoroughlyathirdofmicroRNAsavailablefordetectionintheassaysystem.FurThermore,thisresultsuggestsindirectlythatIFNαtreatmentdoesnotinduceglobalchangesinmicroRNAgeneexpression,butitmodulatesrathertheexpressionofindividualgenes.

TodayitiswellestablishedthatmicroRNAexpressionpatternsarecellandtissuetypespecific,whichisconsistentwitharoleincelldifferentiationandbiologicalfunction(8).Thusweexpectedtodetectgeneswithpreferentialexpressionineitherhepatomaormelanomacellsasthesecelllinesarederivedfromdifferenttumors.Indeed,whenallexperimentalconditionsanddatapointsareincludedinthedataanalysisabout150microRNAsgenesshowpreferentialexpressionineitherHuH7orME-15cells(Fig.1).Table1showstheexpressiondataforthemostsignificantgenesincludingchangefactorsandsignificancescoreasreference.Amongthesedifferentiallyexpressedgenestherearethreemembersofthelet-7family,whichhaspropertiesoftumorsuppressorgenes(forreviewsee(18)).Thereforeitisnotsurprisingthatthemembersofthiswell-knownmicroRNAgenefamilyarederegulatedintheanalyzedcancercellstoo.Furthermore,thedifferentdevelopmentalstageofourcancercelllinesisexpectedtohaveleftagenomicfingerprintwheresomemicroRNAgenesareexpressedinonebutnottheothercellline(19).Consistentwiththis,expressionofsomemicroRNAsisstrictlycelltypespecificandbarelydetectableintheothercelltype(Fig.2a),forexampletheliver-specificmiR-122aandmiR-192(20).

MediaObjects/12575_2009_9012_Fig2_HTML.gif

Fig.2CelllinespecificmicroRNAexpressionVolcanoplot(a)displaydemonstratesthemulti-variantbiologicaldiversityofmicroRNAexpressioninME-15orHuH7cells.Theestimatedfold-changevalue(changefactor)isplottedontheX-axisagainstthepvalue(limmatstatistics)inlogarithmicscaleontheY-axis.Alinearmodel,usingtheexpressionvaluesofuntreatedME-15cellsat4hasbasetogetherwiththreeparameterstoestimatedifferencesintime,treatment,orcellline.ThetoprankedandqPCRmeasuredmicroRNAsareannotated.bQuantitativePCRvalidationofmicroarraydatausingeightselectedmicroRNAs.InputtotalRNAcamefromindependentcellcultures.Dataareshownasrelativecyclingtimes(ΔCT)calculatedwithendogenouscontrolRNU18forME-15(color-filledbars)andHuH7(gray).ErrorbarsrepresentΔCT ± Δx(standarddeviationofbiologicalreplicates).ΔΔCTarenotedabovethebarstogetherwiththesignificancecodesfortstatistics(0‘***’,0.001‘**’,0.01‘*’,0.05‘.’,and0.1‘‘1).

Bead-array-basedmicroRNAdetectiontechnology,includingthebio-statisticanalysis,iscurrentlynotwellestablishedorwidelyusedandwehaveappliedacommercialPCR-basedassaytoconfirmthearraydataforsomemicroRNAsthatcoverdifferentexpressionlevelsandchangefactors.IncontrasttomRNAprofiling,whereRT-PCR-basedassaysareconsideredasgoldstandardfordatavalidation,newgenerationdeepsequencingisconsideredasthemethodofchoiceformicroRNAquantificationbutisnotavailableinourresearchinstitute.ForthemicroRNAslet-7a/b,miR-19a/b,andmiR-203,thePCR-basedquantificationmethod(Fig.2b)confirmedthedirectionofchangefoundwithmicroarraytechnology(Table2a).ExpressionofmiR-130bandmiR-455wasatsimilarlevelsinbothassays.ThecorrelationcalculatedfortheeighttestedmicroRNAswasacceptable:multipler²fromftestofmeanrelativecyclingtimes(ΔCT)tomeanlog2microarrayexpressionvalueswas0.9279.DifferencesofabsolutelevelsbetweenthemicroRNAtargetsprobablyresultsfromdifferenthybridizationpropertiesofthemicroarrayprobesandvariationintheperformanceofTaqmanprimersforthespecificmicroRNAontheotherside.

AssumingthatanyIFNαrelevantmicroRNAwillhavethesamekineticsasthemRNAforPRGs,welookedattheregulationofmicroRNAgenesinourexperiment.TheseIRmiRsshouldrespondtoIFNαstimulationpreferentiallyinbothcelllines,becausethiswouldbeagoodindicationofageneralmechanismintheIFNαresponse.Withinthe25mostsignificantlyregulatedgenes(Table2b),onlyonegene(HS_250)isdownregulated.AgeneralupregulationoftranscriptsisconsistentwithclassicIFNαsignalingseenformRNAs.However,themaximalobservedchangefactorwithhighsignificancewas1.84(miR-33binTable2b)whichisclearlylowerthanthevaluesseenforproteincodingmRNAs(2).Wealsoincludedanexpressionanalysis24hafterIFNαstimulationinordertodetectmicroRNAgenesthatshoweitherdelayedinductionorremainactivatedatcomparablelevelstothe4hstimulus.Basedonourdataset,themajorityofthemicroRNAresponsegenesshownofurtherinduction,butrathermoderatedownregulation24hafterinduction.ThisfindingisnotsurprisingasweexpectedimmediateearlyimpactofIFNα-mediatedprimarysignaling.

WealsomeasuredtheIFNαresponseinthesameexperimentandforthesamemicroRNAs(Table2c).WhenweanalyzetheIFNαeffectattheearlytimepointinbothcelllineswefindallthevalidatedmicroRNAstobeupregulated(Fig.3a).ThemagnitudeofupregulationandthebasalexpressionlevelsofthemicroRNA-19aand19baresimilarinbothcelllines(Fig.3b,top).ThisandthefindingthatmiR-19regulatesSOCS1(4)mayberelevantfortheregulationofcytokinesignaling.let-7aandlet-7bhadhigherlevelsinthemelanoma-cell-line-derivedsamplescomparedHuH7,buttheinductionbyIFNαinME-15couldnotbereproducedbyRT-PCR(Fig.3b,bottom).Inbothassaysaccuratefoldchangesaredifficulttocalculate,ifthebaselineexpressionlevelisclosetobackgroundnoiseorthedetectionlimit.AnexampleofageneatthedetectionlimitismiR-203,whichisnotdetectablewithoutIFNαtreatmentinHuH7cells(Table2c).UponIFNαstimulation(24hinHuH7)themicroRNAisdetectableabovebackgroundsuggestingminimalinduction.Consequentlyasolidchangefactorcannotbecalculated,whichisconsistentwiththehighvarianceobtainedbyqPCR(ME-15).Thisresultisinfactnotsurprising,becausebothtechnologiesrelyonlogarithmicPCRamplificationofmicroRNAtemplates.Atlowexpressionlevels,bothtechnologiesshowrelativelyhighvariationinbiologicalreplicates,whichshouldbeconsideredfordatainterpretation.Interestingly,miR-203hasaputativebindingsiteforISGF3inthepromoterregion,whichwouldenableIFNα-dependentupregulation.miR-30hasbeenreportedtobeIFNβinducible,althoughthesubclassmeasuredwasnotspecifiedbytheauthors(9).Wedecidedtoanalyzethemostpromisingcandidate(miR-30e-5p)presentinourmicroarraydataset(Fig.3aingray).DetectionofmiR-30efailedinME-15cellsduetotechnicalproblems,butinductioninHuH7wassimilartomiR-19a/b.

MediaObjects/12575_2009_9012_Fig3_HTML.gif

Fig.3IFNα-dependentmodulationofmicroRNAexpression.aVolcanoplotdisplayofIFNαinducedmicroRNAupregulation4hafterinduction.Thechangefactorvalues(2^logfactorchange−1)areplottedontheX-axisagainstthepvalueinlogarithmicscaleontheY-axis.Top-ratedmicroRNAsareannotatedtogetherwiththelet-7familymembers.bConfirmationofIFNαeffectforselectedmicroarraydatabyqPCR.TheCT-valuesaretheaverageofthreetechnicalandthreebiologicalreplicatesandchangeswerecalculatedwith2^ΔMNE(meannormalizedexpressionvalues).Errorbarsshow2^ΔMNE ± Δx(averagestandarderroroftreatedanduntreatedMNE)frombiologicaltriplicates.miR-30efailedtoamplifyinME-15andmiR-203wasbelowthedetectionlimitinHuH7.ExpressionvalueswerenormalizedagainstendogenoussnoRNARNU48levels.

Sometechnology-relatedquestionsremainopen.ThemicroRNAassaymeasuresessentiallythenumberofampliconsgeneratedbyRT-PCRforeachtranscript.ThusthesignalisanindirectmeasurementoftranscriptabundanceascomparedtoclassicalmRNAmicroarrayplatforms,wherethetargetmRNAisdirectlylabeledduringlinearamplificationbyinvitrotranscription.Asaconsequence,changefactorcalculationsforamplicon-basedassaysareambiguous.

Insummary,Illumina’sbeadarraytechnologyiswellsuitedformulti-parallelprofilingofmicroRNAsexpressedindifferentcelltypesortissues.WewerealsoabletodetectIFNα-induciblemicroRNAgenesalthoughthechangesobservedweremoderateandbiologicalsignificanceremainstobeproven.Likemostmicroarray-baseddetectiontechnologiesthetechnicalvariabilityamongidenticalsamplesislowcomparedtobiologicalvariationsofindividualcellcultures.Atthispointitisimportanttonotethatvariationamongbiologicalsamplesoccursandisindependentoftheparametersthataremeasured.ConsistentwithIFNα-dependentinductionofmRNAswefindthatvirtuallyallmodulatedmicroRNAgenesareupregulated.However,theIFNα-inducedchangesdetectedinourstudyarerelativelysmallcomparedtothechangesinducedbyIFNβinHuH7cells(9).Finally,itisnoteworthythatIRmiRshavesimilarkineticpropertiestotheirmRNAcounterparts.miR-10bforinstanceisinducedearlyinME-15andremainsupregulated,whilemiR-19abundanceceasesafter24h.Ingeneral,themajorityofIRmiRgeneswereresettobasallevelsafter24handfurtherstudiesareneededforkineticclassification.Thus,ourstudyaddsanotherlevelofcomplexitytothedynamicregulationIFNαsignalingandothermechanismslikeepigeneticpromotermethylationarecurrentlyunderintenseinvestigationinourlaboratories.

AcknowledgementsWethankDr.GuidoSteinerandAndreasBuness(F.Hoffmann-LaRocheLtd.)fortheirsupportinbioinformaticsandstatistics,Dr.MartinEbeling(F.Hoffmann-LaRocheLtd.)forthecomparativegenomicevaluationofmicroRNA-targetedtranscriptsandProf.Dr.GiulioSpagnoli(UniversityHospitalBasel)forthegiftoftheME-15melanomacellline.Finally,wearegratefultoHeatherHinton(F.Hoffmann-LaRocheLtd.)forcriticalreADIngofthemanuscriptandtoDr.LauraSuter-Dick(F.Hoffmann-LaRocheLtd.)forsharinglabspaceandintroductionintoGCPsampling.

================ 蚂蚁淘在线 ================

免责声明：本文仅代表作者个人观点，与本网无关。其创作性以及文中陈述文字和内容未经本站证实，对本文以及其中全部或者部分内容、文字的真实性、完整性、及时性本站不做任何保证或承诺，请读者仅作参考，并请自行核实相关内容

	Control						Interferon-alphatreated
	4h			24h			4h					24h
a	ME-15	HuH7	CHF	ME-15	HuH7	CHF	ME-15	HuH7	CHF	ME-15	HuH7	CHF
ValidatedmicroRNAs
hsa-miR-19a	13,617	6,726	−1.02	18,318	17,178	−0.07	19,226	17,409	−0.10	14,425	14,079	−0.02
hsa-miR-19b	13,365	9,406	−0.42	25,463	22,438	−0.13	21,532	20,625	−0.04	18,039	17,440	−0.03
hsa-miR-30e-5p	9,497	6,838	−0.39	13,045	11,321	−0.15	12,643	10,887	−0.16	13,230	11,809	−0.12
hsa-let-7a	15,244	4,335	−2.52.	28,304	6,181	−3.58***	22,226	6,386	−2.48***	30,449	5,612	−4.43***
hsa-let-7b	8,236	539	−14.28***	12,563	475	−25.44***	12,150	500	−23.30***	12,309	452	−26.24***
hsa-miR-203	907	325	−1.79**	1,155	348	2.32***	1,302	359	−2.63***	1,186	5,757	−1.07**
hsa-miR-130b	5,700	8,316	−0.46	13,023	13,880	0.07	9,994	14,055	0.41	12,053	12,432	0.03
hsa-miR-455	2,677	2,604	−0.03	3,944	5,285	0.34*	3,437	5,927	0.72**	3,262	5,342	0.64*
b	ME-15						HuH7
	4h			24h			4h					24h
	−IFNa	+IFNa	CHF	−IFNa	+IFNa	CHF	−IFNa	+IFNa	CHF	−IFNa	+IFNa	CHF
Interferon-regulatedmicroRNAs
hsa-miR-33b	1,226	3,484	1.84**	2,888	2,318	−0.25	476	1,113	1.34**	1,306	1,851	0.42***
hsa-miR-33	2,631	7,352	1.79	9,774	5,764	−0.70*	1,887	6,645	2.52.	8,893	2,726	−2.26
hsa-miR-126*	2,221	5,409	1.44*	5,578	6,111	0.10	4,749	6,410	0.35.	5,687	5,803	0.02***
hsa-miR-10b	1,695	3,564	1.10*	2,929	4,297	0.47**	378	411	0.09	371	382	0.03
hsa-miR-551b	2,085	4,169	1.00*	4,423	3,419	−0.29	1,804	4,979	1.76*	5,221	4,865	−0.07
hsa-miR-137	1,037	1,966	0.90	2,523	2,872	0.14	1,613	3,155	0.96**	3,429	3,598	0.05
hsa-miR-138	2,158	4,074	0.89*	4,709	3,701	−0.27	725	828	0.14	903	859	−0.05
hsa-miR-130b	5,700	9,994	0.75	13,023	12,053	−0.08	8,316	14,055	0.69**	13,880	12,432	−0.12
hsa-miR-101	6,701	11,387	0.70	13,088	11,688	−0.12	3,569	10,871	2.05**	10,445	10,796	0.03
hsa-miR-140	7,339	12,236	0.67	15,512	15,931	0.03	3,058	5,976	0.95*	6,135	7,221	0.18
HS_92	829	1,356	0.64.	1,246	1,179	−0.06	407	587	0.44	492	478	−0.03
hsa-miR-362	1,382	2,234	0.62*	2,144	2,276	0.06	1,907	2,737	0.43*	2,456	2,519	0.03
hsa-miR-19b	13,365	21,532	0.61**	25,463	18,039	−0.41	9,406	20,625	1.19*	22,438	17,440	−0.29
hsa-miR-130a	13,031	20,878	0.60.	27,571	23,048	−0.20	17,884	31,012	0.73*	33,801	28,875	−0.17
hsa-miR-579	552	816	0.48.	992	1,053	0.06	729	1,362	0.87*	1,250	1,308	0.05
hsa-miR-29b	23,138	34,148	0.48.	32,691	29,710	−0.10	8,737	17,989	1.06*	20,599	17,405	−0.18.
hsa-miR-19a	13,617	19,226	0.41*	18,318	14,425	−0.27	6,726	17,409	1.59.	17,178	14,079	−0.22.
hsa-miR-338	1,298	1,813	0.40.	1,870	1,603	−0.17	2,363	4,603	0.95.	5,109	4,088	-0.25
hsa-miR-590	1,403	1,949	0.39.	1,545	1,367	−0.13	669	1,068	0.60.	884	847	−0.04
hsa-miR-545	1,455	1,973	0.36.	2,149	1,720	−0.25*	829	1,371	0.65**	1,573	1,393	−0.13
hsa-miR-30e-5p	9,497	12,643	0.33	13,045	132,030	0.01	6,838	10,887	0.59*	11,321	11,809	0.04
hsa-miR-570	1,989	2,576	0.30.	2,397	2,610	0.09	2,213	3,739	0.69*	4,312	3,708	−0.16
hsa-miR-301	13,621	17,531	0.29*	16,321	15,142	−0.08	5,646	10,460	0.85*	10,925	10,295	−0.06.
hsa-miR-561	517	621	0.20	568	464	−0.22**	683	1,157	0.69**	1,149	941	−0.22
HS_250	4,284	1,813	−1.36.	740	983	0.33	3,688	2,337	−0.58	1,195	1,303	0.09
c	ME-15						HuH7
	4h			24h			4h					24h
	−IFNa	+IFNa	CHF	−IFNa	+IFNa	CHF	−IFNa	+IFNa	CHF	−IFNa	+IFNa	CHF
ValidatedmicroRNAs
hsa-miR-19a	13,617	19,226	0.41*	18,318	14,425	−0.27	6,726	17,409	1.59.	17,178	14,079	−0.22
hsa-miR-19b	13,365	21,532	0.61**	25,463	18,039	−0.41	9,406	20,625	1.19*	22,438	17,440	−0.29.
hsa-miR-30e-5p	9,497	12,643	0.33	13,045	13,230	0.01	6,838	10,887	0.59*	11,321	11,809	0.04
hsa-let-7a	15,244	22,226	0.46	28,304	30,449	0.08	4,335	6,386	0.47.	6,181	5,612	−0.10
hsa-let-7b	8,236	12,150	0.48*	12,563	2,309	−0.02	539	500	−0.08	475	452	−0.05
hsa-miR-203	907	1,302	0.44*	1,155	1,186	0.03	325	359	0.10	348	575	0.665**

在线客服咨询
	您好请问有什么可以帮您的？
	取消	点击对话