Integratingextrachromasomalarrays

intotheC.eleganschromosomes

Whyandhowtodoit,

byMichaelKoelle

4/6/94

Whatisthebenefitofintegratinganextrachromosomalarray?Extrachromosomalarrayssufferfromthreeproblemsthatmaybesolvedtodifferentextentsbyintegrationintothechromosomes:

1)Extrachromosomalarraysarelostfromthegermlineatsomefrequency,somaintainingalinecarryingthearrayinvolvesconstantlyselectinganimalsatleasteveryfewgenerations.Forsomephenotypesthisisdifficult.Integratingandhomozygosingthearraycompletelyeliminatesthisproblem.Havingsuchastablelineisconvenientforavarietyofexperiments.Forexample,whendevelopingantibodiestoaprotein,itisusefultohaveoverexpressingwormstostain,andtomakeextractsforwesterns.Youcanusetheoverexpressortoworkoutconditionsforusingtheantibody,andtoconfirmwhichbandontheWesternisreallyyourprotein.

2)Extrachromosomalarraysarelostmitoticallyatsomefrequency,sothatanimalsthatcarrythemaremosaicsinwhichonecannotnecessarilydeterminewhichcellshavelostthearray.Thispatternofmosaicismvariesindifferentindividualworms.Thustransgenesonextrachromosomalarraysmaynotbeexpressedconsistentlyinthesetofcellsthatonewouldlike.Thisproblemisonlypartiallysolvedbyintegratingthearray.Resultsfromintegratinganumberofbeta-galactosidasereporterconstructsshowthatgenesinintegratedarraysmaybeartifactuallyactivatedorshutoffincertaincells.Presumablyduetopositioneffect,theintegratedpromotersmaynolongergiveexactlythesameexpressionpatternastheendogenouspromoterinitsnormalchromosomalposition.Inaddition,thepatternofexpressionfromtheseintegratedconstructsisnot100%reproducIBLefromanimaltoanimal;somecellsshowvariableexpression.Overall,however,theexpressionfromintegratedarrayscanbemorereliablethanthatfromextrachromosomalarrays.Basically,you"rereplacingtherandomerrorinexpressionthatcomesfrommitoticlossoftheextrachromosomalarraywithamoresystematicerrorthatcomesfrompositionaleffectsontheintegratedarray.

3)Extrachromosomalarrayscanchangetheirpropertiesovertime.Thedataforthisisthatifanextrachromosomalarrayhasacertainmeasuredtransmissionfrequency,andanumberofindividualwormscarryingthisarrayareusedtoestablishnewlines,thesenewlinesmayhavetransmissionfrequenciesfortheirarraysthatstronglydifferfromeachotherandfromtheoriginallymeasuredfrequency.Theexpressionofgenesfromtheseextrachromosomalarraysthereforepotentiallyalsosuffersfromthiskindofvariation.Thiscreatesthefollowingkindofproblem:ifatransgeneistobecrossedintoanumberofgeneticbackgroundsandtheresultingphenotypesaretobecompared,howcanyouknowthattheextrachromosomalarrayhasn"tsufferedsomesortofchangeduringthestrainconstructions?Itisassumed(thoughwithnodataI"mawareof)thatintegratedtransgenesarestable.Therefore,whenyouneedyourtransgenestohaveconsistentpropertiesovermanygenerations,itispreferabletointegratethem.

Howtodoit:

Summary:AstrainbearinganextrachromosomalarrayisirrADIatedwithgammaraysorx-rays,severalhundredF1Progenycarryingthearrayarepickedtoindividualplates,andforeachF1severalF2carryingthearrayarepickedtoindividualplates.Theseplatesarescreenedfor100%transmissionofthearraytotheF3;suchastrainishomozygousforanintegratedarray.

Stategy:ManypeoplepickafewhundredF1s,andthenpick4-5F2sfromeachF1.ThelargenumberofF2platesinvolvedusuallynecessitatesdoingthescreeninafewbatchestoavoidkillingyourselfsettingupalltheF2satonce.However,boththeoreticallyandpractically,thisisnotthebeststrategy.Statistically,youwillminimizethetotalnumberofplatessetup(F1andF2summed)perintegrantrecoveredifyoupickmoreF1s,andpickonly2F2sperF1.Thisalsohasthepracticalbenefitofspreadingtheworkoutmoreevenlybetweenthetwogenerations.Mypracticeistopick250F1soneachoftwoconsecutivedays.Aftertwodaysoff,Ithenpick2F2sfromeachF1.ForthoserareF1platesthatseemtohavethrownthearrayatamuchhigherfrequencythantherest,Ipick4F2s.After1-2daysoff,IscoretheF2platesoverthreedays.ThehardestworkispickingtheF2"s;Ifindthatpicking500oneachoftwoconsectutivedaysisaboutasmuchasishumanlypossibletodo.

1.StartbyusingmicroinjectiontogenerateextrachromosomalarrayscontainingthegeneofinterestandanappropriateMarkergene.Indesigningtheexperiment,considerusingacoinjectionmarkerlikeunc-76,dpy-20,orlin-15,ratherthanthepopulardominantrol-6.Withrol-6,theanimalscarryingthearrayarelesshealthythantheanimalsnotcarryingthearray;withtheothermarkersthereverseistrue.I"vefoundlin-15tobeagreatmarker;theMuvanimalsstickoutlikeasorethumb,makingthescreenforintegrantsasnap.

2.Selectinganextrachromosomalarraytointegrate:

Whenidentifyingtransmittinglines,besuretokeeplinesthattransmitthearrayatalowfrequency.About20%transmissionisveryconvenient;thisisenoughtransmissiontoallowyoutoeasilygetenoughanimalstoirradiate,butitismuchlowerthanthe75%transmissionfrequencyyouwilllookforcomingfromanimalsheterozygousforanintegrant.However,don"tgetbentoutofshapelookingforsuchalowtransmitter,sincethisactuallyonlymakesamarginal(30%?)differencetothescreen.Unfortunately,individualsfromsuchlinestendtoshowhugevariationsintherateatwhichtheytransmittheirarrays,sothatananimalthattransmitsthemarkeratahighfrequencyusuallydoesn"tcarryanintegratedarray.Ifallyouhavearehighfrequencytransmittinglines,don"tworryaboutitandforgeahead;youcanstillgettheintegrantbyblindlypickingF2sfromeachF1andlookingforanF2thattransmitsthearrayto100%ofitsprogeny.

Anotherconsiderationisthestrengththedesiredphenotypecomingfromyourgeneofinterestinthearray.Youpresumablyshouldstartwithanextrachromosomalarraythatgivesyouaverystrongphenotype.However,thisisnoguaranteethatalltheintegratedlinesgeneratedfromitwillgiveastrongphenotype;positioneffectsseemtointroducelotsofvariation,sothatbothstrongandweakintegratedlineswillbeproduced.

3.YouwillmutagenizeL4hermaphrodites.InordertogetenoughL4P0scarryingthearray,youmayneedtosetuplotsofplates,especiallyifthearrayistransmittedatlowfrequency.Formutagenesis,use~30L4animals(carryingthearray)perlargeplate.Youwillneedabout500F1s(carryingthearray)togetanintegrant,sodependingonthetransmissionfrequencyofyourarrayandthenumberofintegrantsdesired,youshouldscaleupaccordingly.(Data:usingthisscreeningstrategyIgot3integrantsfrom1500F1sforonearraytested).

Whenusinglin-15asacoinjectionmarker,youarefacedwiththeannoyingproblemthatlin-15cannotbescoreduntiltheadultstage.Therefore,youshouldproducestagedP0animalstomutagenizebytransferringlotsofwormstoanewlayplatetwiceaday.ThenyoucaneasilypicklotsofL4progenytonewplatestomutagenize.Thedayaftermutagenesis,pickthenon-Muvmutagenizedanimalstonewplates.Iput30onasmallplate,andtransferthemtwiceadayfor2daystoproducewellstagedF1animals,sothatnon-MuvF1scanbeeasilypickedasyoungadults.

4.Mutagenesis.Eitherx-raysorgammaraysseemtowork;use3600-4800Radswitheithertype.TheHorvitzlabhasbothanx-raymachineandagammasource.ShaiShahamclaimsthateithertypeofradiationworksaboutthesame.YishiJinsaysshegetsconsistentresultswithgammarays,andthatshegotpoorresultstheonetimesheusedx-rays.Itriedbothx-raysandgammaraysonthesamearrayandIgot2integrantsfrom300F1swithgammarays,and0integrantsfrom200F1swithx-rays.

5.WhenF1scanbescoredforthepresenceofthearray,picktheF1stonewplates(oneF1perplate).Asmentionedabove,youshouldpickabout500F1stogetoneintegrant..MostpeopleavoidpickingtheveryfirstF1progenythattheP0sgenerate:thesemaybederivedfromgermcellsthathadalreadycompletedmeiosisatthetimeofmutagenesis.Theoretically,(andwithsomeempiricalevidence),youcanrecoverahigherfrequencyofmutantsbywaitingabit.Typicallythemutagenizedanimalsareaged24hrsat20deg.,thentransferredtonewplates,andtheF1sfromthissecondsetofplatesareused.

7.WhentheF2canbescoredforthepresenceofthearray,pickF2animalstoindividualplates.Asdiscussedabove,Ipick2F2sperF1.

Ifyouarestartingwithanextrachromosomalarraythatistransmittedatalowfrequency,youcanalsolookthroughtheplatesforpotentialintegrants.Ifyoustartedwithanarraywitha~20%transmissionfrequency,youexpectmostofthebackgroundF1stohavethrownfewanimalswiththearray.F1animalscarryinga(heterozygous)integratedarray,incontrast,willthrow75%F2scarryingthearray.Ifyouseeaplatelikethis,pickalot(~8)ofF2sfromit.Notethatyouwillgetsomefalsepositivesandfalsenegativeshere.Falsepositives:thearraymaystillbeextrachromosomal,butmayhavechangedtoahighertransmissionfrequency.Falsenegatives:duetoapositioneffectthemarkeronthearraymaybeexpressedonlyweakly,andmaynowberecessiveinsteadofdominant.Anotherproblemisthatsomearrayscarrygenesthatmakethewormssick.Forexample,theRolphenotypemakeswormsgrowslowly.Thusiftheplatesarescoredtooearly,thepercentageofanimalscarryingthearraymaybeunderestimated.Onethingtodointhissituationistowaitafewextradaystoscoretheplates,andthenlookfortheplatesthatarethelasttostarveout;theseprobablycontainedahighproportionofsickF2worms(carryingthearray).

8.WhentheF3sareoldenough,lookforaplateonwhich100%oftheF3scarrythearray.Thisplateshouldcarryahomozygousintegratedarray.Iactuallydon"tthrowtheplateoutunlessIsee2wormsthathavelostthearray;asinglewormlackingthearraymightjustbeduetoanunrelatedL1thatwasinadvertantlytransferredtotheplatealongthewithadultF2.

9.Sinceyouputyourwormsthroughmutagenesistogeneratethisstrain,youshouldbackcrossitseveraltimestocleanupthebackground.