EVALUATION OF GENELINKAGE AIID ITS USE

IN TÏIE DETER}TINIATION OF TIIE IIEREDMYOF TRATTS IN ST]NFT,OWER

KOVÂCIK, A., SKALOUD V.Research Institut for Crop ProductionRuzyeé-Prague, Czechoslovakia

INTRODUCTION

The r,ange of problem,s studied by the Ge-netics Section of the Prague Research Insti-tute of Crop Production, includes trhe problemsof sunflower genetics and cytology. Theresearch in this crop, used as ,a model repre-senting allogamous plants pollinated by insectsand suitable for heterosis breeding, is concen-trated in the following three fields :

a) study of the hereditarily conditionedquantitative traits of production importance ;

b) the use of cytological methods in thestudy of pollen sterility;

c) the determination of the heredity ofmarker traits, i.e. morphological and ,anatomictraits easy to define by sight and potentiallyappiicable to the disclosure of other, lessmarked tr,aits.

The determination of the relationships tothe traits which are to be disclosed is a prere-quisite for the use of the marker traits. At thelevel of the genes determining the respectivetraits, such relationships are realized throughgene linkage.

Some methodical and interpretation possi-bilities ,of ttre evaluation of gene linkage aresubjected to the study described in this paper.

METHODS

The deter,rnination of the dependence of twotraits and the e:qpression of the linkagestrength between the genes conditioning thesetraits can be done by several v/ays. Two basicprinciples available for this purpose are ,seenin the procedure based on the segregationratios transformed into the contingency tableand in the procedure in which phenotype fre-quencies are used for the calculation withoutfurther transformation. Both procedures offersome a'dvantages ; the iatter one is currentlyindicated in the basic geneti,c handbooksHruby, 1961; Neèâsek and Cetl, t9B0).

The first procedure can be divided into tùrefollowing ,steps :

1. determination of the dependence of twotraits from the 2X2 or 2xxn contigency table ;

2. ,determination of the extent of depen-dence by means of the association coefficientor contingency coefficient ;

3. determination of the presence of an inter-action ;

4. determination of the strength of the lin-kage by ,conversion of the association coeffi-cient to the Morgan nurnber.

Th-e first procedure offers the advantage ofenabling some continuity in the system of theevaluation of quantitatively and qualitativelyconditioned traits, i.e. the traits determinedby a small number of genes are evaluated bycalculation,s based, in the essence, on the samêprinciples as ttre calculation procedures usedin quantitative genetics. In addition to this, inthe calculation of the strength of the linkagei" & generation it is not necessary to usehardly ,accessible tables ,or a cômplicatedcalculation. If two traits demonstrated onlyby two phenotypic categories are involved, thêtests of dependence are based on the 2Y.2contingency table and on the cal,cuLation of thetesting criterion X2. If rbhe ,oarlsujl,aûed. Xlry valueexceeds the table value, we believe that thegenes underlyinrg both traits ,cannot be freelycom,bined and that they can be expected tohave a gene linkage. îhe extent of this linkagecan be deterrnined ,by means of the so-callédassociation,coefficient whose calculation ensuesfrom the relation :

V= (ad-bc)

li@It i9 also advantageou,s to use a simplifiedmethod of the conversion of the assoèiationcoefficient to the Morgan number, determin-ing the strength of ithe linkage. The overallcalculation of the strength of the linkage isquite simple. For instance, ,when linkage isdetermined from the F2 generation in ttre eisstage (i.e. at increased phenotype frequencies

35

a, d) and in the absence of interaction,sequence of calculations looks as follows :

xâr =n(àd-bc)2

(a*b) (c*d) (a*c) (b*d)

tt -:;v-vxirr p:t-l/3v+l

Vn Y 2

The dependence between two tr'aits 'characte-rized by several phenotypic categories can betested by ,rneans of the contingency coeffi-cient. In the genetic ,analysis of the traitsconditioned by a small number of genes, theuse of the contingency coefficient is of im-portance in two cases : first, in the oase oftesting the linkage with a trait conditionedby a single gene but realized rby three pheno-types (incomplete ,dominance) and second, in atrait in which a multi-phenotypic manifestationis conditioned by interaction.

The calculation of the 'contingency coeffi-cient is based on the value of Xlly, determinedby the following procedure :

xt. - {-4-ç,.. " )- 'b''^(r) [ b,b, a r:t ^,.

J b2

":wI xit,*nThe mentioned metho'ds of the determinationof the strength of linkage with 'contin'gencytables require previous testing of the segre-gation ratio aimed at finding whether a changein the distribution of phenotypes is not causedby interaction, since the conversion of theassoci,ation coefficient to the Morgan num'berchanges with the type of interaction ; this, ofcourse, fully corresponding with differentways of the calculation of the Morgan numberfor different interactions with the use of thes€corrd procedure without the transformationof data into ,contingency table.

The second procedure of testing of thepresence, of linkage also includes the causeof the change in the segregation : whether itis a linkage or an interaction. When testingthe linkage according to the segreg,ation ratioof phenotypes in the F2 generation, the se-quence of c.alculations is as follows :

. o (ar *az-3 a3-3 aa)2xirl :-- 3r, t

the testing criterion serves for the verificationof the non-interacti,on segregation ratio for thefirst trait.

--, - (at-3 a2*a3-3 aa)2Àirl - g"

verifies the non-interaction. segregation ratiofor the second trait.

__, (a1-3 a2-3 a:*9 ad2xu) 9

'''covers the presence of a gene linkage.

36

The sum of all three values of X121 shouldbe in congruence with the value of X"6y deter-mined at testing the free combining abilitywith dominance, i.e. the dihybrid segregationlatio.

If the value of Xâr for the first and secondgene is insignificant, the cause of the changein segregation ratio lies in linkage. A changein segr.egation is caused by linkage as wel.l asby interaction in those ,cases in which theXf1; is significant for linkage and at the sametime one or both Xfiy for the verification ofthe non-interaction segrogation ratios.

The strengths of the linkage determined bydirect calculation of the Morgan number differfor each linkage stage cis-trans and for casesof separate interactions ; in the stage of ci,s inthe F2 generation the Morgan nurnbe,r valueis determined from the relationship :

the

àz'àt -àt'àt

2 (1-p)2 + (1-p) I

1-2(1-p)2+ (1-p)4

in which the ensuing from the right side ofthe equation is usually tabulated for the non-interaètion segregation ratio (H r u b y, 1961)as well as Jor the inter,action segregationratios (Immer, 1930). The left side of theequation is called the product ratio and whencalculating the strength of the linkage on'l'ythis part of the relationship is calculated. Thep coiresponding to it can be seen from therespective table. In thg cis ,stage, with a non-intèraction segregation ratio, a simplified cal-culation can be used for the determination ofthe strength of the linkage: p: t_2]/ aan.

In the trans stage ,the determination of thestrength of linkage from F2 is ,rather difficuitand when the linkage is strong, i.e. the fre-quency of phenotypic category expressed byfrequency a4 is very low, such a determina-tion is factually impossible. The product ratio

for the tran.s stage }} is again convertedâz'à3

to the Morgan number, using special tables.In our opinion, when iinkage strength is

deterrnined from the segregation ratio in theF2 generation it is better to use, instead ofthe product rations, the salculation of the pvalue from the association coefficient or asimplified calculation based on the phenotypefrequency of the twice recessive gene set a4

as proposed in this study.

R,ESULTS AND DISCUSSION

Four cases of linkage between the genesunderlying the so-called marker traits havebeen found in sunflower until now. the mostimportant of them is the linkage between thegene for ,anthocy,anic colour and the gene forpollen sterility. Further, a gene linkage wasfound between the anthocvanic colour of the

plants and the colour o{ ligulate flowers. Thelinkage with pollen sterility gained somepractical importance in the heterosis breedingof sunflower, based on the production of hy-brids from gene-conditioned pollen,sterility.

The linkage between the T gene for antho-cyanic colour and the pollen sterility gene Mswas discovered by Le,clercq (1966). Thisauthor presumed the ,genotype set of a fertileplant with anthocyanic oolour to be TTMsMsand that of a sterile green plant ttm,sms. Thehybridization o both genotypes produced an!'t generation of fertile red coloured speci-mens defined as TtMsms. In the F2 genera-tion the segregation ratio of 157 red fertileto 37 green sterile to 1 green fertile plant wasobtained. The segregation ratio in the B1 gene-ration was 51 : 36 : 1, which also showed arare occurrence of ttMs-recombinants and sug-gested that the T-msms genoty,pes did notoccur at all. L e cl e r c q believed the Ms gene,characterized by monofactorial heredity, andthe T gene are in a linkage which is veryclose. The author enumerated the recombi-nants and got a value close to 10/6.

On the basis of the linkage relationship, ascheme of the marintenance of sterile genotypeswas determined for hybrid production. In thisscheme the heterozygotic fertile red pheno-type with the TtMsms genotype set, crossedwith .the twice rrecessive genotype of a greensterile plant ttmsms, gives two phenotypiccatqgories in the progeny. About one half ofthe plants should be green and sterile, i.e.suitable for the production of heterozygotichybrids, and the other half of useless plantswould refnain red and fertile. Hence from thegenetic point of view, the principle of themaintenance of sterile genotype ,correspondswith repeated back crossing. Some compli-cation in the use o anthocyanic disclosure isseen in the recombinants of fertile plantswithout anthocyanic colour. The originallydetermined 10/e proportion of such phenoty,lreswould not do much harm but the actualnumber of the plants with un'desired pheno-typic manifestation was often ,as high .as fiveper cent.or higher.

The discrepancy between the assumptionand the fact was attributed to a redu,ced pene-tration of the ms rgene unrder field ,conditions.Change in the strength of linkage as the eauseof this phenomenon was not taken into consi-' deration ibeoause such a situation would neces-sarily include an insreased occurrence of thefourth phenotypic category of red sterileplants whose existence was not demorr-strated.

Our opinion concerning the explanation ofthe increasing proportion of green fertileplants in .the progeny of crosses used for themaintenan,ce of sterile genotytrles includes theassumption of a strong linkage between the Tand Ms genes, requiring the involvement of ahigher num,ber of genes in the realization of

the phenotypic manifestations of a trait.-Thestarting points are the determination ,of theexistence of a higher number of Ms genesand an apparent difference in the ,gene basi,sof the phenotypes with the anthocyanic colourof whole pl'ants and phenotypes with redhypoc,otyl.

The explanation of the incrêased proportionof green fertile phenotypes can be demonstra-ted as follou's : the crossing of the genotypesAsTtMqmslMs2ms2 ;1 Asttmslmslms2trrs2, i. e. f er-ti1e, red X sterile, green with red hytrlocotyl,gives, in the B1 gener,ation, rise to the pheno-type categories red, fertile : green, fertile :

green, sterile,in a 3 : 3 : 2 ratio. Al1 green fer-tile planrts have a red hypocotyl ; of the green

Isterile plants ; have ,a non-coloured hypo-cotyl. The presence of a ,single dominant geneMs suffices for the m,anifestation of fertilityand only one sterility €ene is in linkage with_gene T. The general exilstence of these prere-quisites lvas dem'onstrated by some authors.The proposed explanartion reckons with afairly high proportion of green ertile pheno-types. However, in fact, when back crossingwas frequently repeated, the proportion offertile green plants was observed in somecases to increase rapidly to the detriment ofsterile plants. The consideration of the invol-ve,menÉ of a higher nurnber of genes enablesthe explanation of the increase in the pro-portion of one recombinant category withoutthe presence of the other recombinant cate-gory even in cases when the linlcage ren:tainsvery close. This f'act is hard to accept by thehypothesis in which only two genes are consi-dered. Among other things, concurrent pre-sence of red plants with green and red hypo-cotyl was observed during exact study of phe-notype rnanifestations after back crossing ; therecombinants characterized by fertility mostlybelonged to the coloured hypocotyl phenotypecategory. This finding is also in agreementwith our opinion 'described above.

The linkage between the genes underlyingthe colouring of ligulate flowers with the geneunderlying the anthocyanic colour of tubularflowers, or whole plants, was determined inour trials.

We studied the results oJ ,the crossing ofgreen plants with yel1ow ligulate and redplants with whitish yellow ligulate flowers.Another cross consisted of the combination ofgreen 'piants wil,h orange yellow flowers an'dred plants with whitish yellow flowens. Forthe first case, only red plants with yellowligulate flowers occurred in the Ft gener,ation.The progeny of the secorld :combination eon-sisted of red plants with or,ange yellow flo-'ffers. Segregation in the F2 generation isshown in Table 1.

37

The table shows that in both eross typesthe belsic segrogation ratio 'of the dihybrid wasdistorted by linkage. The 'strength of the lin-kage could not be determined, since in thetrans stage the number of plants was too lowfor oalculatiqg the percentual proportion ofrecombinants in the F2 generafiion. Besides theundeniable fact of the existence of linkagebetween genes L (or Lo, la) and T (or t), theoccurrence of phenotypes in the seg'regatinggenenation âlso clearly indicates that the genesresponsible for orange yellow and yellowcolôur h,ave a,bout the sam'e 'distance fromgene T ; this supports the hypothæis of theinvolvement of the multiple allelism rela-tionship among alleles L, lo, Ia.

The deterrn:ination of the strength of thelinkage was enabled by experiments with hy-bridization ,analysis in cis stage and with anincreased number of individuals in the F2 ge-ner,ation, in the tr,ans'stage. Crossing in the cisstage was only performed in the com'binationof red, yellow X ,green, whitish yellow, ,5inceonly one plant with a twi,ce recessive tUalagenotypes was available. Fon the combinationof green, onange yellow X rd, whitish yellowin the trans stage, the amount of plants usedfor initial cr,ossing rvas increased. Both theseanalyses demonstrrated the existence of linkageand gave ,a prerequisite for approxim'ate deter-mination of the strength of the linkage bet-ween rthe genes for the colouring of ligulateflowers and anthocyanic colouring of the pl,ant.The crossing of a red plant with yeltrow ligu-late flowers and a green plant with whitishyellow flowers wàs followed, in the F2 gene-ration, by segreg,ation rn the r,atio of144:lB:13 :53 which corresponds to linkageexpressed by the 130/6 value bf reoombinants.Segregation of progenies in the F2 generationproduced by the crossing of a green plantwith orange yellow flowers and ,a red plantwith whitish yellow flowens is shown inTable 2.

The summarized segregation ratio of403 :177 :230 :1 corresponds to linkage whiohcan be expressed as Tofs of recombinants.

38

Table 2

160x220 243

244

245

246

247

220x160 248

245

250

251

252

160 X 220 suin220X160 sum

45

54

43q

29

46

42

44

50

43

1?8

229

16

26

20

t6

16

20

LI

22

20

82

95

2s

25

20

B

10

33

t,

27

29

31

BB

t42

0

0

0

0

0

0

0

0

I0

0

1

Although the results suggest some diffe-rences between the position of the L and logenes from the ;point of view of linkagestrength, it can be said that the compari'sonof the Morgan nu,mber, determined in twocontrasting stages, may ah,vays indicate cer-tain difference in ,the obt'ained value. In addi-tion to this, the possible occurrence of another(at leest one) phènotype of green, whitish yel-low would increase, in the trans stage, theproportion of r'ecom,binants to as much as100/6. The range of the linkage determinedforgenes T and L is fr,om 7 'to 190/s ; in the caseof genes T and lo tLre exactness of the deter-mination of linkage strength is within thenange fr,om 4 to 1007t. Hence these 'data coin-cide, and it can be generally cmrcluded thatthe vatrue expressing the strength of the lin-kage between the genes for the anthocyaniccolou,ring of the plant ,an'd genes for the co-louring ôf tgulate flowers ranges about 100/e.

The experirnents ,also included the studyo{ the linkage relationship of the trait of ligu-late flowers shape in two alternatives of ma-nifestation, i.e. normal' flowers and tubularliguiate flowers. No relationship was foundbetween the traits of ligulate fl,ower sh,apeand colour. The phenotype r,atio iin the F2 ge-neration oorresponds to a linkage strengthexpr.essed as 44.20fs of reoombinaints, and thisis, in the essence, a value characterizing freecombining ability. Simil,arly, between the geno-types un'derlying flower shape and ,ant'hocyaniccolouring, the alleles 'are freely 'combinable,because 45.2010 of the reoombinants ensuingfrom the segregation natio of 92 :37 : 23 : 13fully proves this conclusion. This finding wel!cor,responds with the data asserted by S t o e -nescu and Vrânceanu (1976) who founda linkage between the Ms2 genes for pollensterility and Fl for ligulate flower sh,ape, andthis linkage had a strength of 20olo. It can begeneraily stated that the Ms1, T and L genes

Table 1

Parent phenotypesRed

yellow(orange)

Green, orangeyellowXred,whitish yellow

.SumGreen, yellow Xred, whitishyellow

Sum

28

46

74

19

15

27

61

t424

38

13

5

t735

It,

31

1t

tu31

0

0

0

1

0

0

1

5.30

10.34*

t4.78*

6.10

1.?B

10.04r

14.89+

are looated on one chromosome and genesMsr and Fl occur jointly on a,nother chro-mosome.

In relation with the linkage relationshipsan example can be given, in which the deter-min.ation of linkage becomes the deci,sive ,cri-terion for the determination of the correct-ness of one of 'two or ,r'nore different opinionsconcerning the heredity of a tr,ait. In sun-flower a discrepancy was f,ound between ouropinion which is in agreement with the con-clusions published,by Lecler'cq (1968) andStoenescu, (1 974), and the corn'ception oftrait ,as propose'd by Fick (19?6). The pointof contr,adiction is the eolour of the ligulateflowers, maniiesting itself in the variant ofyellow, orange yellow and whitish yellow. Anallelic rel'ationship of genes was 'derived f,romour results, whereas Fick introduced a two-gene hypothesis. The results of both worksare the same in the case of the crossing ofyellow X or,ange yellow anrd orange yeilow ;whitish yellow, in which the monogenic basisof the phenotypic rnanif,estation is involved.In our trials, like in those performed byStoenescu, ,a monogenic mranner of hereditywas determined ùn the ,cases of the cr.ossingof yellowywhitish yellow. F i c k believed thatthere was an interaction of two genes of reces-sive epistasis type. F i,c k's opinion is based 'onthe analy'sis of the F2 gener,ation in which heallegedly obtained a segregation natio of216:79:86 and the orange yellow phenotypeoccurred after the autogamic poliinatio,n of theyellow F1 generation produced from the ,sros-sing of yellow y whitish yellow. The segre-gation ratio shows good corresp,on'dence withthe theoretical r:atio of 9:3:a ()i'T:1.?1).Wi'th the existence of such a relationship, eachcross between the alternative genotypes couldbe expressed by the following genotype sets :

yetrlow X orange yell,ow AABB 1 aaBB, orangeyellowywhitish yellow aaBB;aabb, yellowtwhitish yeilow AABBXaabb. Then genotypesA-B are realized in yellow ligulate flowers,aaB by orange yellow, and A-bb are whitishyellow.

This consideration is apparently acceptableand can resist, to some degree, to the argu-ments available to those who ,believe that atrait is ,conditioned by multiple allelism. Thecrossing of orange yellow ;1 whitish yellow,evaluated as allelism test, would ,produce aprogeny with yeliow flowers in the case ofthe aaBBXAAbb combination, but when theparental genotypes are aaBB y aabb it is im-possible to judge whether the ailleles belongor do not belong to the same gene. However,our results concerrnirr-g the linkage between theanthocyani,c colouring of the lplant and thecolour of ligulate flowers clearly indicate thatif the studied genes are not the alleles oJ thesame locus, then they must be located close toeach other on the chromosome, This firm fact

inevitably suggests that even if the ,two-genehypothesis is assumed to hold good, the cros-sing of the yellow and whitish yellow phe-notype cannot yield orange yellow ,colour. Thisis due to the fact that the heterozygotic geno-type in the F1 generation would produce, inthe AB gene linkage, only two types of ga-metes, AB, ab, whose combination gives ayellow to whiti,sh yellow ratio of 3 : 1 in theF2 generation. It is entirely impossible in thecase of allelism 'or linkage of two genes forthe colouring of ligulate flowers, to imaginethe occurrence of the orange yellow pheno-type after the hybridization of plants withthe yellow and whitish yellow m'anifestationof the traits. Hence, the only thing to believeis that F i c k confused the orange yellow colourwith a shade of yellow which only slightlyrsminds .of rich orange colour. Some variabi-lity in the shades of the yellow ,colour of ligu-late flowers is generally known. However, ifthe proclaimod onange yellow phenotype was,in fact, yellow, then the results of F i c k'sexperiment corresponds to the monogenic basisof the trait even in the case of the yeliow ywhitish yellow 'combination. The two-genetheory appears not to be realistic. Thereforethe relationship of multiple altrelism of genesL, lo, la is more probable than their ,cI,ose

linkage. Hence the^ opinion ,asserteil by uS,concerning the hereditary basis of the colour-ing of ligulate flower,s, is valid.

The considenations on the heredity of 'theyellow, whitish y,ellow and orange yellowcolour ,can be further evolved. It has been said,so far, that the genes for the three mentionedvariants of the tr,ait are alleles of the samelocus : hence F i c k',s opinion in which a geno-type denoted as LLLaLa underlines the ye1lowcolour, LI.lal,a or lllala is related with whitishyellow colour, and llI,al,a with orange yellowcolour, is not justified. However, it'was reve-lead during the study of the colour of ligulateflowers that the yellow, whitish yellow andorange yellow colours of the petals can bedetermined by two genes. Such a findingcould not be obtained in direct crossing of thestudied phenotypes, but onl.y indirectly 'incro,ssing with the red coloured flowers. Howe-ver, the two considered g'enes underlying theheredity of the non red ,colour do not bearahy relation to the ,two genes presented inF i c k's hypothesis. This can be explained asfollows : in the crossing of whitish yellow xyellow, genotypes aab2b2CC x aaBBcc give anF1 gener.ation with a uniform yellow colour,aaBb2Cc, and in the F2 generation the yellowto whitish yellow natio is 72 :4.

This situation suggests two f acts. First,the results from which F i c k derived the opi-nion con,cerning two genes remain explainedin the one-gene way in the form of the Bb2set, and second, it is evident that the interac-tion of genes B and C cannot be demonstrated

39

by the crossing. of whitish yellow and yellow,sinrce the !2 ;4 ra1,;'o cannot be forrnally dif-ferentiated from the monohy,brid ratio of 3 : 1.Only crossing with a genotype includingall the thnee genes, ABC (red-coloured pheno-type), it is possitble to determine ,the preserr-ceof gene C and, consequently, the two-gene in-teraction of B and C. Analogically, the cros-sing of orange yellow x yeLl,ow aablblCC xaaBBc'c produces in the F1 generation theyellow phenotype aaBblC,c and in the F2 gene-ration the 12 :4 segregation ratio apparentlyindicates rnonogenic heredity. The combinationof whitish yellow x orange yellow aa'b2b2CC xaablblCC gives orange yellow aablb2CC in theF1 generation and in the F2 generation the se-gregation is really monohybrid, 3 : 1. The pre-vious finding concerning the non red colour ofligulate flowers justifies the'conclusion that thealleles L, lo,la are congruent with B, b1, b2 andare located in the same place on the chromo-some. Gene C which may also be involved inthe manifestation of the genotype, is in dupli-cate relation with gene L in the determinationof the yellow phenotype and cannot be foundin the direct analysis of the variants of the nonred colour. .In addition to this, gene C is in nolinkage with gene t, so that there is âIso noIinkage with gene T. Both cases - the linkageof genés L and.T, as well as the indirect deter-mination of gene C,.can be used as exemples ofcomplicated heied.ity of .an apparently verysiriiply hereditarily conditioned trait whosebacliground cannot be identified by currentanâlytic method based on'ttre crossing of twoaltêrnative genotypes

. REFERENCES

fFi'ôk ç. .N.,. 7916, Genetics ot floral colour anil' morphologg in sunf.louers, Journ. of Heredity,

87, 4.

H r u b f K., 1961, .Genetika,. Acadernia, Praha ,196,1.

Imrnet Fi'R., t93O, Fôrnxulae'and, tables for calcu''lating llnko,geintensities, Genetics, 15. .

Leclercq P., -11966, 'Une stérilité rndle utili.sablepour la production èI'hubricles simples d,etournesol, Ann. Amélior. Plantes, 16, 2.

Leclercq P., 1968, Hëréd.ité ile quelques cara'c-tères qualitatifs chez le tournesol, A,nn. Amé-Iior. Plantes, 18, 3.

Neëâsek, J., Cetl I., L980, Obecna genetika, SPN.Praha. 1980.

Stoenescu F.. L974, Genetica in: Floarea-soarelui.Edit. Acad. R.S.R., Bucureçti.

Stoenescu F., Vrânceanu A. V., 1976, Li'nkagestudi,es betloeen lns genes and lour markergenes in sunflower, Proc. VII Intern. SunflowerConfer., Krasnodar.

EVALUATION DU LINKAGE GENIQUEET SON UTILISATION

POUR DETERMINER L'HERÉDITEDES CARACTÈRES CHEZ LE TOURNESOL

Résumé

Le présent ouvrâge analyse les différentes méthodesd'évaluer le linkage entre les gènes et relève l'i'm-portance de la méthode basée sur le calcul de I'in-tensité du linkage en F2 [a phase,,cis") à I'aide ducoefficient d'association. Les auteurs pas,sent enrevue les groupes de linkage connus chez 1e tour-nesol entre Msl T et L, et entre Ms2 et Fl, avec desconsidérations sur I'association des gènes I et L. Lephénomène de linkage est utilisé pour analyser cer-t.aines opinions concernant I'hérédité de la couleurdes fleurs ligulées, conditionée rpar Ie gène L, ainsique pour expliquer les écarts ,des rapports prévuspour 'certains phénotypes, en tenant corn-pte de larelation entre les gènes Ms1 et T. Cette étude faitpartie des thèmes de recher,che en collaboration avecle sous-réseau FAO rpour la génétique du tour-nesol.

EVALUACION DEL LINKAGEDE LOS GENES Y SU EMPLEO

EN DETERMINAR LA HEREDIDADDE LOS CARACTERES AL GIRASOL

Res{trnen

In este papel hay consideraciones sobre lâs dife-rentes métodos de determinaci6n 'del linkage (enla-zarniento) de 'los genes. Esté evidenciado el métodopor el .cual se ca,lcula la intensidad del linkage en 1ageneraciôn F2 (fas ,,cis") con Ia ayuda del coeficientede asociaciôn. A continuaciôn. se pasan de revistaIos grupos de linkage conoci,dos del $irasol, entre losgenes : Ms1 T y L, Ms2 y Fl, y se hacen apreciacionessobre la asociacidn entre la gene TyL. El fenômenofle hnkage estâ empleado para confrôntar algunasopiniones concerni.entes a la heredidad del color delas flor'es liguladas condicionada por la gene L y ala vez para explicar las desviaciones de las rela-ciones esrperadas de unos fenotipos, tomando enconsideraciôn la relaciôn entre la gene Ms1 y T. Esteestudio es parte de la temâtica de investigaci6n encooperaciôn de Ia subred FAO 'para la genética delgi rasol.

are looated on one chromosome and genesMsz and Fl occur jointly on another chro-mosome.

In relati.on with the linkage relationshipsan example can be given, in which the deter-mination of linkage becomes the decisive cri-terion ,for the determination of the correct-ness of one of two or rnore different opinionsconcerning the heredity of a trait. In sun-flower a discrepancy vlas found between ouropinion which is in agreement with the con-clusion,s published by Lecler'cq (1968) andStoenescu, (1974), and the conception oftrait ,as proposod by Fick (1976). The pointof contr,adiction is the oolour of the ligulateflowers, manifesting itself in the variant ofyellow, orange yellow and whitish yellow. Anallelic rel,ationship of genes was derived ,fromour results, whereas Fick introduced a two-gene hypothesis. The results of both worksa.re the same in the case of the crossing ofyellow X orange yellow arrd orange yellow ywhitish yelIow, in which the monogenic basisof the phenotypic ,rnanif,estation is involved.In our trials, like in those performed byStoenescu, ,â monogenic manner of hereditywas determined rin the 'cases of the cr'ossingof ye11owlwhitish yellow. Fick believed thatthere was an interaction of two genes of reces-sive epistasis type. F i,c k's opinion is bas,ed ,on

the analysis of the F2 gener,ation in which heallegedtly obtained a segregation natio of216 :79 : 86 and the or,ange yellow phenotypeoocurred after the autogamic pollination of theyellow F1 generation produced fr:om the ,sros-sing of yellow X whitish yellow. The segre-gation ratio shows good correspondence withthe theoretical ratio of 9:3:4 (Xz'? :1.?1).Wiûh the existence of such a reletionship, eachcross between the alternative genotypes couldbe expressed by the following genotype sets :

yellow X orange yell,ow AABB y aaBB, orangeyellow;whitish yellow aaBB;aabb, yellow;4whitish yellow AABBXaabb. Then genotypesA-B are realized in yellow ligulate flowers,aaB by orange yellow, and A-bb are whitishyellow.

This consideration is apparently acceptableand can resist, to some 'degree, to the argu-ments availabie to those who 'believe that atrait is ,conditioned by multiple allelism. Thecrossing of orange yellow y whitish yellow,evaluated as allelism test, would produce aprogeny with yellow flowers in the case ofthe aaBBXAAbb combination, but when theparental genotypes are aaBB ;4 aabb it is im-possible to judge whether the alleles belongor do not belong to the same gene. However,our results ,concerrnirr-g the linkage between theanthocyanic colouring of the plant and thecolour of ligulate flowers clearly indicate thatif the studied genes are not the alleles of thesame locus, then they must be located cl'ose toeach other on the chromosome, This firm fact

inevitably suggests that even if the 'two-genehypothesis is assumed to hold good, the cros-sing of the yellow and whitish yellow phe-notype 'cannot yield orange yellow ,colour. Thisis due to the fact that the heterozygotic geno-type in the F1 generation would produce, inthe AB gene linkage, only two types of ga-metes, AB, ab, whose combination gives ayellow to whitish yellow ratio of 3 : 1 in theF2 generation. It is entirely impossible in thecase of allelism ,or iinkage of two genes forthe colouring of ligulate flowers, to imaginethe occurrence of the orange yeilow pheno-type after the hybridization of plants withthe yellow and whitish yellow manifestationof the traits. Hence, the only thing to believeis that F i c k confused the orange yellow colourwith a shade of yellow which only slightlyreminds of rich orange colour. Some variabi-lity in the shades of the yellow 'colour of ligu-late fLowers is rgenerally known. However, ifthe proclaimed onange yellow phenotype was,in fact, yellow, then the results of F i c k'sexperiment corresponds to the monogenic basisof the trait even in the case of the yellow ywhitish yellow combination. The two-genetheory appears not to be r'ealistic. Thereforethe relationship of multiple altrelism of genesL, lo, la is more prohable than their closelinkage. Hence the opinion ,asserted by us,concerning the hereditary basis of the colour-ing of ligulate flower,s, is valid.

The corrsiderations on the heredity of ,theyellow, whitish y,ellow an'd oran'ge yellowcol.our ,can be further evolved. It has been said,so far, that the genes for the three mentionedvariants of the tr,ait are alleles of the samelocus : hence F i c k's opinion in which a geno-type denoted as LLLaLa underlines the yellowcolour, Ll,lai,a or 111a1a is related with whitishyellow colour, and lllaI,a with orange yellowcolour, is not justified. Hovrever, it'was reve-lead during the study of the colour of ligulateflowers that the yellow, whitish yellow andorange yellow colours of the petals can bedetermined by two genes. Su,ch a findingcould not be obtained in direct ,crossing of thestudied phenotypes, but only indirectly 'incrossing with the red coloured flowers. Howe-ver, the two ,consi'dered genes ,underlying theheredity of the non red ,colour do not bearany relation to the ,two genes presented inF i c k's hypothesis. This can be explained asfollows : in the crossing of whitish yellow xyellow, genotypes aab2b2CC x aaBBcc give anF1 gener,ation with a uniform yellow colour,aaBb2Cc, and in the F2 generation the yellowto whitish yeltrow natio is 72 :4.

This situation suggests two f acts. First,the results from which F i c k derived the opi-nion con,cerning two genes remain explainedin the one-gene way in the form of the Bb2set, and secon'd, it is evident that the interAc-tion of genes B and C cannot be demonstrated

39