[PDF] rams of two breeds J. Pelletier D. H. Garnier

á ° x a U ¬ * * a x - - * à à

byelectricalclocks.Theseclockswereadjustedevery2-4daysaccordingtothemagnitudeofthedailychangeofdaylength.AtthelatitudeofNouzilly,47°N,daylightvariesfromabout8hinDecembertoabout16hinJune.Ineachroom,four40W,white,fluorescenttubesprovidedabout300luxattheleveloftherams"eyes.Animalswereallowed1monthtoadjusttothenewenvironmentbeforethefirstperiodofbloodsamplingandthenstayedcontinuouslyintheroomsduringtherestoftheexperiment.Foodwasprovidedeachmorningat08:30-09:00handconsistedof400gdehydratedlucerne,250gdehydratedmaize,300goatsand25gmineralsupplementalongwithstrawandwateradlibitum.ThemalesexualseasonextendedfromtheendofJunetotheendofDecemberinIle-de-FranceramsandfromthebeginningofJunetotheendofDecemberinPréalpesduSudrams(Thibaultetal,1966;G.Colas,personalcommunication).Theramswerebledinsuccessiveyearswithalternationandreplicationforeachbreed.Theperiodsofbloodsamplingandthetotalnumbersofanimalswereasfollows:endofDecember-41Ile-de-Franceand75PréalpesduSudmid-February-11Ile-de-Franceand9PréalpesduSudbeginningofApril-11Ile-de-Franceand9PréalpesduSudbeginningofJune-11Ile-de-Franceand9PréalpesduSudInaddition,bloodsampleswerecollectedattheendofSeptemberfrom5Préalpesramsononeoccasion.Ineachsamplingperiod,bloodwascollectedhourlyfromthejugularveinfor24h(25samplesfrom09:00hto09:00hthenextday).Duringthenight,bloodwascollectedindarkness.Plasmasampleswerekeptfrozenat-14°CuntilradioimmunoassaysforLHandtestosteronewerecarriedout.LHassayLHwasmeasuredinduplicateaccordingtothemethodofPelletier,Kann,Dolías&Rosselin(1968)whichwasslightlymodifiedinthat:(1)125I(RadiochemicalCompany,Amersham,U.K.)wasusedforlabellinginsteadof131I;(2)afterlabelling,iodinatedLHwasseparatedfromfreeiodineonaP-10G25Sephadexcolumn,andthemostradioactivefractionwasrepurifiedfurtheronaK30G100Sephadexcolumn,equilibratedwith0-025M-veronaland0-25%humanserumalbuminbuffer;(3)thesecondantibodywasanti-guinea-pig globulinserumobtainedbyimmunizationofamare;(4)allthesamplescollectedduringagivenyearwereassayedsimultaneouslyinanautomaticLKB2071sampleprocessor;(5)resultsexpressedintermsofLH-M3(=l-8xNIH-LH-S1)wereobtainedon-linefromanLKB1270Rack-gammacounter;(6)specificitywasverified,especiallyforcross-reactionwithTSH(Freychet,Pelletier&Rosselin,1969);(7)within-assaycoefficientsofvariationofthebaselineconcentrationvariedaccordingtoexperimentsbetween4and10%;and(8)theinterassaycoefficientsofvariationwere10and8-3%forbaselineandpeaklevelsrespectively.Sensitivityundertheseconditionswas0-1ng/ml.TestosteroneassayTestosteronewasmeasureddirectlyin0-05mlplasmabytheradioimmunoassaydescribedbyGamieretal(1978).Thewithin-assaycoefficientsofvariationwere8-4,7-8and7-9%for1-0,4-8and8-2ng/mlrespectively,andtheinter-assaycoefficientsofvariationwere14,10-5and13%.Sensitivitywas0-2ng/ml.

á ° x a U ¬ * * a x - - * à à

DeterminationofLHandtestosteronepeaksContinuousbleedingexperimentscarriedoutonramsduringthebreedingseason(Davies,Main&Setchell,1977)orduringboththenon-breedingseasonandthebreedingseason(Terqui,Gamier,Pelletier,Ravault&Ortavant,1978;Terqui,Gamier,deReviers,Huet&Pelletier,1980)haveshownthatmostLHandtestosteronepeakslastmorethan60min.Indeed,only19%ofLHpeaksand4%oftestosteronepeakshaveadurationof40-60min.Theprobabilityofdetectingtheseshorterpeakswithasamplingintervalof1histhereforeatleast4/6andthehourlysamplingintervalshouldgivedetectionofatleast93-5%ofallLHpeaksand98-6%ofalltestosteronepeaks.Apeakwasdefinedasanincreaseinhormoneconcentrationtoavaluegreaterthan4timesthecoefficientofvariationofthebaselineintheassay,followedbyadecrease.Becausethepercentageincreaseofthehormonallevelabovethebaselineatthetimeofapeakwashigh(100-1000%),thedifferencesinthewithin-assaycoefficientsofvariationswerenotconsideredtohavebiasedthecomparisonofthenumberofpeaksbetweenbreeds.ThisisreflectedbytheverycomparablepercentageofcoincidentLHandtestosteronepeaksfoundinthetwobreeds(see"Results").LHandtestosteronepeakswereconsideredascoincidentpeaksifthetestosteronepeakwassimultaneouswithoroccurredjustafteranLHpeak.ResultsWhenplasmaLHandtestosteronelevelsweresummarizedasfrequencydistributions,thesedistributionswereunimodalandstronglyasymmetric(Text-fig.1).Inaddition,plasmaLHandtestosteroneforagivenramandagivenhourofsamplingcannotbeconsideredasindependentvariables.Therefore(i)theunimodaldistributionoflevelscouldnotbeusedforthedeterminationoftheLHandtestosteronepeaksassuggestedbyChristian,Everson&Davis(1978)whofoundabimodaldistributionofGHlevelsincattleand(ii)classicalparametricanalysis,andtheexpressionmean±s.e.m.,wasinadequateandonlyseasonalvariationinthenumbersofLHandtestosteronepeaksper24h(expressedasmedianvaluesratherthanmeanvalues)aregiven.Effectsoftimeofyearandbreedonnumbersofpeaksper24hweretestedstatisticallybythenon-parametricMann-WhitneyUtest(1947).Testosterone( =4504)>10ng/mlText-fig.1.DistributionsofplasmaLHandtestosteronelevelsexpressedasrelativefrequencies(%)-

á ° x a U ¬ * * a x - - * à à

CorrelationbetweenLHandtestosteronepeaksAsageneralrule,apeakofLHwasfoundsimultaneouslywith,orfollowedby.atestosteronepeak(Text-fig.2).Outofatotalof670LHpeaksand647testosteronepeaks,therewere96-4%coincidentLHandtestosteronepeaks(404LHand394testosterone,97-5%,inPréalpesduSud,and94%inIle-de-Francerams).Thisfigureincludedafewresults,mainlyoccurringinJune,inwhichtwowell-definedLHpeakswerefollowedbyasinglesteady,highleveloftestosterone.Therewere48isolatedpeaks,i.e.3-6%ofthetotal;23werepeaksofLHwithoutanytestosteronepeakand25weretestosteronepeakswithoutanyLHpeak.PréalpesduSudDecembergLJune- - - - - - - - - - - - - - - - - - - - - - - - -09:0013:0017:0021:0001:0005:0009:0011:0015:0019:0023:0003:0007:00HoursText-fig.2.PlasmaLHandtestosteronevariationthroughoutthedayinPréalpesduSudandIle-de-FranceramsinDecemberandinJuneSeasonalvariationinLHandtestosteronepeaks:influenceofgenotypeText-figure3showsthatthefrequencyofLHandtestosteronepeaksincreasedbetweenDecemberandJuneforramsofbothbreeds.InPréalpesduSudrams,thenumberofLHandtestosteronepeaksincreasedsignificantlyinFebruaryandApril(P<0-05)comparedwiththeDecembergroup.ThedifferencebetweenDecemberandJunewasstillgreater-a200%increaseforbothhormones(P<0-01).However,thenumberofpeaksinJunewassimilartothatobservedinSeptember.Bycontrast,theincreaseinthenumberofpeaksofbothLHandtestosteronewasmoderateinIle-de-Franceramsandthedifferencewassignificant(P<0-01)onlyforthe66%increasethatoccurredbetweenDecemberandJune.Finally,whilethenumbers

á ° x a U ¬ * * a x - - * à à

ofpeaksweresimilarinthetwobreedsinDecember,theyweresignificantlydifferentinJune(P<0-025).PréalpesduSudIle-de-France(9)(75)(9)FT)(9)(5)UDecFebAprJuneSept LHVITestosterone(41)(11)(11)IDecFebAprJuneText-fig.3.SeasonalvariationinthefrequenciesofpeaksofLHandtestosteroneinPréalpesduSudandIle-de-Francerams(nosinparentheses).Valuesaremedians(seetext).DiscussionFromastatisticalpointofview,plasmaLHandtestosteronevaluesdidnotappeartobesampledfromnormallydistributedpopulations,asshownbyasymmetricaldistributions,varianceheterogeneity,andlackofindependenceofplasmahormoneconcentrationswithinanygivenram.Fromaphysiologicalpointofview,theoccurrenceoflargeandrapidchangesinplasmaLHandtestosteronevaluesfrom0-5-1-0upto10-20ng/ml(Terquietal,1980)alsoprecludestheuseofmean+s.e.m.values.Ontheotherhand,thenumberofLHpulsesappearstobecloselyrelatedtothedevelopmentofthetestes(Lincoln,1979).Forthisreason,wechosetoexpressthedataasthenumberofLHortestosteronepeaksobservedinbloodsamplescollectedover24handtouseanon-parametrictestfortheanalysis.Fromthepresentresults,itappearslikelythateachLHpeakisaccompaniedbyatestosteronepeak.Continuouscollectionofbloodfor2dayshasconfirmedthecloseconnectionbetweentheLHandthetestosteronepeaks(Terquietal,1980).FindingsofanapparentabsenceofrelationshipbetweenLHandtestosteronelevelsinramsmayonlybetheresultoftooinfrequentsampling(Falvoetal,1975).DuetothedurationofLHpeaksandtheuseofhourlysampling,somepeakscouldhavebeenmissed,butthehighpercentageofcoincidentLH-testosteronepeaksobservedexcludedthisasamajorsourceoferror.ThepresentresultsconfirmthatthenumbersofLHandtestosteronepeaksincreasefromthenon-breedingtothebreedingseason(Sanfordetal,1974,1977;Schanbacher&Ford,1976;Wilson&Lapwood,1978).butindicatefurtherthatthepeaknumbersincreasebeforethestartofthebreedingseasoninbothbreeds,andrisecontinuouslyinPréalpesduSudramsduringtheseasonofincreasingdaylength.AnincreaseofplasmaconcentrationsofFSHandtestosteronebeforethesummersolsticehasbeenpreviouslymentionedforSoayramsbyLincoln&Short(1980).UnlikeSanfordetal(1977),wedidnotobserveanyincreaseinthenumbersofLHpeaksortestosteronepeaksattheendofthebreedingseason(December).Therewasacleargenotypicinfluenceonthepatternofrelease,withthepercentageincreasebetweenDecemberandJunebeingthreetimesasgreatinthePréalpesduSudramsasinIle-de-Francerams.Thisresultindicatesthat,instudiesofphotoperiodiceffects,genotypemustbetakenintoaccount.

á ° x a U ¬ * * a x - - * à à

á ° x a U ¬ * * a x - - * à à