RT-F-6

RESTRICTION FRAGMENT LENGTH POLYMORPHISMS AND GENETIC IMPROVEMENT

P o lim o r f is m o de l a r e s t r i c c i 6 n p o r l o n g i t u d d e fr a g m e n to s y m e jo r a g e n S t i c a

M. SOLLER*J . S . BECKMAN**

ISRAEL

Methods have recently been developed, based on the use of restriction enzymes and cloned DNA, that allow a new class of genetic polymorphisms: restriction fragment length polymorphisms (henceforth, RFLPs) to be detected (Botstein et al■, 1980; Solomon and Bodmer, 1979). Large number of RFLPs have already been demonstrated in humans (Bell et al. , 1981; Jeffreys, 1979; Wyman and White; 1980) and in other organisms (Petes and Botstein,1977).

The purpose of this paper is to draw attention to the potential use­fulness of this class of polymorphisms in genetic improvement. We will describe (1) the nature of RFLPs and the methodologies involved in their detection, (2) the implications of a virtually unlimited source of genetic polymorphisms for breeding practice, in particular for improvement of strain crosses, and for within populaiton selection, and (3) present costs of RFLP methodologies and anticipated costs of their applications.

I. RFLP METHODOLOGIESRestriction fragments

Restriction endonucleases are enzymes that recognize specific nucleic acid sequences in DNA and cleave the DNA at these sites or at adjacent sites (Zabeau and Roberts, 1979). Depending on their specificity, most restriction endonucleases will have 4 or 6-base recognition sequences. Such sequences can be expected to occur once in every 4^ or 4*> nucleotides, respectively.

The more prevalent the recognition sequence is in the DNA, the more frequently it will be cleaved by the enzyme which recognizes that sequence.The DNA fragments so formed, can be separated by gel electrophoreses since smaller fragments migrate more rapidly through the gel than larger fragments. When the DNA from a higher organism is digested by restriction enzymes, so many different sized fragments are produced that a continous smear is formed on the gel. Specific fragments arp t-hgrpfnrp__datected by the use of an appro- priate probe} This is carried out as follows: The DNA pattern is transferred

'Trom troTgei to a solid support, such as a nitrocellulose filter (Southern,1975). The DNA, now bound to the filter is then exposed to a radioactively labelled probe under conditions that promote DNA-DNA hybridization. The probe used would be a cloned DNA sequence homologous to a particular DNA fragment or some portioa of it. The unhybridized radioactivity is then washed away and the filter is dried and placed against photographic film for auto­radiographic exposure. After film development the specific fragment that hybridized with the probe will be visualized as a band on the film.

*Department of Genetics, The Hebrew University, 91904, Jerusalem, Israel.**Department of Field and Garden Crops, Agricultural Research Organization,

50200 Bet Dagan, Israel.

396

By using specific probes Csay reverse transcribed unique DNA) a DNA fragment that occurs as rarely as once in a million or less can be detected. The probes used in RFLJ? detection do not have to be homologous with known genes. Any unique sequence will suffice as long as it hybridizes with some part of one of the DNA fragments formed by the restriction process.Restriction fragment length polymorphisms

Base changes can alter the sequences that are recognized by restriction enzymes, abolishing sites or creating new sites for particular enzymes. De­letions or transpositions of large elements into a site will make simultaneous changes in the restriction patterns of a number of enzymes. As a result, a given restriction enzyme will not always cleave a given DNA molecule at the same point in two individuals. In this case fragments of different length will be formed when the DNA of the two individuals is digested. The unequal sized fragments will travel at different rates through the gel, and following hybridization and autoradiography, the bands formed will be located at differ­ent points on the film. In this manner a restriction fragment length poly­morphism will have been demonstrated. A hypothetical example is illustrated in Figure 1.

Figure 1. Hypothetical example illustrating a restriction fragment length polymorphism. A region of homologous chromosomal DNA sequence from two individuals is diagrammed. Both segments contain a sequence (indicated by the box with hatched lines) that has been cloned. The site for Enzyme R on the right side of the box in individual 2 has been abolished by alteration of the DNA sequence:A larger fragment will therefore be produced when DNA from individual 2 is digested with enzyme R. Auto­radiographic visualization of the fragments containing the cloned sequences are shown in the lower half of the figure.

Restriction Sites in Chromosomal DNA

— ; e r a j FR R R

2) i — T---------m -------------------------------------r-R R

Restriction fragments visualized after hybridization with a cloned probe

(-)

Origin 1 2

a(+)

397

Expected frequency and mode Of Inheritance o f RFLPsA large v a r ie ty of r e s t r ic t io n enzymes having d if fe r e n t s p e c i f i c i t i e s

are a v a ila b le and RFLPs in both coding and non-coding DNA sequences can be detected w ith appropriate probes. Thus the RFLP method can p o te n t ia lly monitor a s ig n if ic a n t fra c tio n of the to ta l genome and should be v ir tu a lly unlim ited in i t s a b i l i t y to d etect large numbers o f polymorphisms.

Careful consideration of Figure 1 shows th at in a heterozygote for both types o f DNA m olecu le , both bands would be present in the autoradiograph.Thus RFLPs can be expected to show a codominant mode o f in h er ita n ce , and th is has indeed been found in a number of in stan ces (P etes and B o ts te in , 1977;Wyman and W hite, 1980). Furthermore, s in c e RFLPs w i l l o fte n be located in flanking sequences or in in tr o n s , one would expect them to la ck secondary p le io tro p ic e f f e c t s on q u an tita tive ch aracters. This cla s s o f polymorphism _ should therefore be w e ll su ited for purposes o f genet i c a n a ly s is and improvement of q u a n tita tiv e t r a i t s .

The number of RFLPs required for complete genome coverageA minimum estim ate o f the number of properly placed markers for complete

genome coverage would be T /c , where T i s the to t a l map length in cM, and c i s the spacing between markers in cM. For purposes o f g en etic a n a ly s is o f quanti­ta t iv e t r a i t s , a spacing of 40 cM between markers would be s u f f ic ie n t (S o ller e t a l . . 1976) . For purposes of marker a ss is te d in tro g ress io n or s e le c t io n , a spacing of 20 cM between markers would be d es ira b le (S o lle r and P lo tk in ,1978).

The to ta l number of polymorphic markers th at need to be detected in order to obtain a w e ll spaced subset has been considered in d e ta i l for the" human genome by Bishop e t a l . (.to be pub lished). From th e ir ta b le s i t appears that roughly tw ice the minimum number of polymorphisms w i l l g iv e a subset o f markers that w i l l cover 90-95% of the genome at the desired spacing . On th is b a sis for dairy c a t t le say (T=3000), 75 w e ll placed markers would be required for a 40 cM spacing and 150 for a 20 cM spacing, w h ile the to ta l number o f polymorphisms that would have to be detected to y ie ld an appropriately spaced subset would be about 150 to 30U ,resp ectively .

I I . RFLPs AND GENETIC IMPROVEMENTIn th is se c t io n we w i l l describe two p o ss ib le a p p lica tio n s o f RFLPs

to gen etic improvement in animal breeding. The f i r s t , w i l l d ea l w ith improve­ment of e x is t in g s tr a in crosses or o f crosses between inbred l i n e s , as in p ou ltry . The second w i l l deal w ith increased e f f e c t iv e n e s s of young b u ll se le c t io n in dairy c a t t le breeding.

Improvement of hybridsThoday (1961) and h is co-workers, using sp e c ia l t e s t e r s tr a in s in

D rosophila have shown how gen etic markers can be used to lo c a te and evaluate linked lo c i a ffe c tin g q u an tita tive t r a i t s . Subsequently, S o lle r et a l . (1976) concluded that r e la t iv e ly modest experim ents, in vo lv in g 1000-2000 F-2 or backcross in d iv id u a ls out of a cross between two inbred l i n e s , d if fe r in g in one or more marker l o c i could enable d ifferen ces in the q u a n tita tiv e value of chromosomal segments adjoining the d if fe r e n t ia t in g markers to be determined.In f a c t , i f the two s tra in s d if fe r in many markers, w e ll spaced along the ge­nome,a s in g le F-2 generation w i l l allow a to ta l a n a ly s is o f th e en tire genome, including main e f f e c t s and dominance e f f e c t s ,fo r a l l q u a n tita tiv e tr a i t s scored.

The b asic approach i s shown in Table 1 . I t can be seen th at a q u an tita tive d iffe r e n c e , 2 ( l - 2r)d w il l be found between the mean v a lu e o f the two homozygous marker genotypes in the F—2 . S im ila r ly , the q u a n tita tiv e v a lu e o f the homo­zygote marker genotype, expressed as a d ev ia tion from the mean o f the two homo­zygous marker genotypes w i l l be equal to ( l - 2r)^h , and th is parameter, dividedby h a lf o f the main e f fe c t w i l l be an estim ate o f ( l - 2 r )2l , th e r e la t iv ed

398

Table 1 . E valuation o f main e f f e c t and dominance r e la t io n s ofq u a n tita tiv e l o c i in the F-2 o f a cross between inbred l in e s d if fe r e n t ia te d at marker and q u a n tita tiv e l o c i .

Mean v a lu e o f genotypes a t the q u a n tita tiv e locus

Yalue ^d________________ 1 ____ h________ +dGenotype aa 0 Aa AA

Cross between' inbred l in e s d if fe r in g a t marker locu s (M»m) and q u a n tita tiv e locus (A,a)

P arenta l genotypes MAMA x ma

maValue +d -d

FI genotype MAma

"Value h

F2 genotype MM Mm TTTTT1

Value ( l - 2r)d+ 2r (l-r )h .

{ ( l - r ) 2+r2 }h - ( l - 2r)d+2 r ( l - 2r )h

E stim ates o f main e f f e c t s and dominance at the q u a n tita tiv e locus

Main e f f e c t s MM - mm = 2 (l-2 r )d

Dominance Mm - (MM + mm) = ( l-2 r )2 h

R ela tiv edominance

Dominance *5 Main e f f e c t s ( l - 2r) £

dominance. The s ig n o f the dominance estim ate w i l l g iv e a notion as to the d ir e c t io n of dominance, but the magnitude o f the estim ate w i l l be le s s than the actu a l dominance by an amount 2r. In favorable c a se s , some notion as to the va lu e o f r , and the ab solu te va lues o f d and h can be obtained by maximum lik e lih o o d methods (W eller et_ a l . , to be published) .

I f the q u a n tita tiv e a l l e l e o f in te r e s t i s included between two c lo s e ly linked marker a l l e l e s (a marker bracket say , MM-SS), then i t can rea d ily be shown that the expected mean value o f the three homozygous marker genotypes in the F-2,MMSS, MmSs, and mmss, w i l l c lo s e ly approximate d , h and -d , respec­t iv e ly . That i s , in t h is case very good estim ates o f main e f f e c t s and domin­ance e f f e c t s a t the q u a n tita tiv e locus can be obtained d ir e c t ly from the data .

I n i t ia l r e s u lt s in our laboratory (Weller et a l . , to be published) and elsew here (Tanksley e t a l , . 1981; Zhuschenko e t a l . , 1978) show that a wide v a r ie ty o f q u a n tita tiv e e f f e c t s can in fa c t be found a ssoc ia ted w ith s p e c if ic g en etic markers in cro sse s of th is s o r t . Maximum lik e lih o o d methods show that

3 9 9

in some c a se s , a t l e a s t , the r e su lts are due to lin k a g e rather than p le io trop y . The r e s u lts in our laboratory suggest that in tomato a t l e a s t , th e m ajority of m arker-linked q u a n tita tiv e e f f e c ts are of a magnitude th at could read ily be d etected in experiments o f about 1000 F-2 in d iv id u a ls .

Thus, the methods for gen etic an a lysis described above can be u t i l iz e d to id e n t ify s p e c if ic marker-linked segments or brackets showing dominance or overdominance in the F-2 of a cross between two inbred l i n e s . This would enable an e x is t in g commercial cross to be analyzed in terms o f th e number o f chromosomal segments involved in producing th e observed hybrid v ig o r , the type o f g en etic e f f e c t involved ( i . e . , complementary dominance, or over­dominance), and the magnitude o f e f f e c t of such segm ents. In a d d it io n , other inbred l in e s could be tested against the standard l i n e s , and screened for the presence o f ad d ition a l dominant or overdominant segm ents. These could then be incorporated in to the a ltern a te standard l in e by marker-based in trogression techn iques, as described in S o lle r and P lo tk in (1977) and Tanksley and Rick (1980). In th is way chromosomal segments con tributing to h e te r o t ic e f f e c t s could be c o lle c te d from a wide v a r ie ty of resource l in e s and introduced in to the standard l i n e s , allow ing d irec t and s p e c i f ic con stru ction o f l in e s showing maximum degrees o f hybrid v ig o r .Marker a s s is te d s e le c t io n in segregating populations

In s p e c ie s , such as dairy c a t t l e , where very la rg e progeny groups can be obtained from in d iv id u a l m ales, gen etic markers or h ea v ily marked chromosomal segments can a lso a s s i s t in w ith in population s e le c t io n . The b a s ic procedure i s aimed at evaluating a ltern a tiv e h eav ily marked homologous segments (denoted Mi ,M2 ) of a chromosome pair in a given male. S e le c t io n could then be carried out among fu ture o ffsp r in g o f such an evaluated male on the b a s is o f the estim ated breeding value o f the p articu lar marked segments th a t the offsp rin g in h er ited from h is s i r e . Although the contribu tio n o f any s in g le marked segment to gen etic progress in th is manner would be sm a ll, th e jo in t con trib u tion o f a l a r g e number o f s n o b segments could b e appreciab le.

The procedure would be as fo llo w s. We denote the male parent as MjM2 and the fem ale parents as MpMp, where M i, M2 and Mp are homologous chromosome segm ents, 20-30 cM in length and marked by 3 or more polymorphic markers. I t i s assumed that in the male parent, M]_ and M2 d i f f e r a t a t le a s t one o f the marker lo c i included in the segment. The Mp are a l l o f the corresponding homologous segments in the female paren ts , and may be sim ilar or d if fe r e n t than the M i, M2 segments. The cross o f that male to a s e r ie s o f fem ales w i l l be o f the nature M1M2 x MpMp, and the con trast HjTIp - HJFIp in th e o ffsp rin g w i l l g iv e inform ation as to the q u a n tita tiv e v lau e o f the Mq and M2 segments in the te s ted m ale. The biom etric equations d e fin in g the in crea se in breeding va lu e o f young s ir e s chosen on the b a s is o f the p a rticu la r marker segments they received from th e ir s i r e (fo llow in g S o lle r , 1978) are g iven in Table 2.

As a p a rticu la r marked segment i s follow ed from a s i r e to h is sons and evaluated again in the o ffsp r in g o f th ese son s , a d d itio n a l inform ation w i l l accumulate on the breeding va lu e o f th is segment as compared to other homolo­gous marked segments in the population . As inform ation accum ulates, future evaluations o f the breeding value of a lte r n a tiv e chromosomal segments in a s ir e w i l l be based on past inform ation as w e ll as on new inform ation obtained in the daughters o f that s i r e . This w i l l lead to more exact marker-based eva lu ation s o f young s ir e s .

E ventually as th is process continues the frequency o f favorably evaluated segments w i l l in crease in th e population to the p oin t where e s s e n t ia l ly a l l segments carrying the favorab le marker combination w i l l have o r ig in a ted from the o r ig in a l marked segment in the o r ig in a lly evaluated s i r e . In ad d ition , accumulated inform ation w i l l g iv e a fa ir ly p r e c is e r e la t iv e eva luation of th e various homologous marked segments segregating in th e popu lation . At th is

400

T able 2 . B iom etric equations d e fin in g th e in c r e a s e in breed ing v a lu e o f young s i r e s chosen on th e b a s is o f marker a n a ly s is o f th e ir s i r e s .

D = observed phenotyp ic d if fe r e n c e between daughter groups r e c e iv in g a lte r n a te marker a l l e l e s or b rackets from th e s i r e .

2----- = error v a r ia n c e o f D.n

o^p * 2pqd^ = g e n e t ic va r ia n ce o f D, fo r a s i r e heterozygous a t both endso f a marker b r a ck e t, p and q are freq u en cies o f q u a n tita t iv e a l l e l e s , o i s d e fin ed as in T able 2 .

= 2p q (l-2r)2d ^ «• g e n e tic va r ia n ce o f D, fo r a s i r e heterozygous a t a s in g le m arker, w ith recom bination p r o b a b ility r , between marker and q u a n t ita t iv e lo c u s .

<y o 2o 2°D = °AD + ~n~ = Phenotypic va r ia n ce o f D .

bo “ aAD aD = re 8ress:*-on breed ing v a lu e for D on phenotypic v a lu e fo r D.

J5D = mean d e v ia t io n from th e o v e r a ll daughter mean (fo r th a t s ir e ) o f th ose daughters r e c e iv in g a p a r tic u la r marker b ra ck e t.

M ■ bj|OsD) = expected breed ing v a lu e a sso c ia ted w ith a p a r tic u la r marker or marker b ra ck e t.

B = breed ing v a lu e o f dam of young s i r e .

I = If M- + B = index v a lu e o f young s i r e in c lu d in g breed ing v a lu e o f h is dam and o f k evaluated markers.

2 . k A A A0£ = h b” + o | - varian ce o f index v a lu e s .

i = in t e n s it y o f s e le c t io n (F a lcon er , 1970).

S=i(dj;-ag) = in c r e a se in breed ing va lu e o f young s i r e as a r e s u lt o f marker a s s is t e d s e le c t io n .

p o in t , th e s i r e s can b e s e le c te d on th e b a s is o f th e ir t o t a l genotype w ith re sp ec t to t h i s segm ent. I f a number o f such segments are seg reg a tin g sim u l­tan eou sly in th e p o p u la tio n , young s ir e s would be s e le c te d on th e b a s is o f an index composed o f th e sum o f th e ir "known segment" v a lu e s , and th e ir dam breed ing v a lu e . At t h i s p o in t marker a s s is t e d s e lc t io n would be th e major rou te to g e n e t ic improvement in th e p o p u la tion . T able 3 g iv e s b iom etric equations d e fin in g th e in c r e a se in breeding va lue o f young s i r e s chosen on th e b a s is o f "known segm ents", and T able 4 g iv e s a num erical example i l l u s t r a ­t in g the in c r e a s e in breed ing v a lu e o f young s ir e s th a t m ight be achieved in th is manner.

401

Table 3. Biometric equations defin ing the increase in breeding value of young s ir e s chosen on the b a sis o f "known segment" se le c tio n .

a = d + (q-p)h = average e ffe c t o f a gene su b stitu tio n .~ breeding value of i t h "known segment".= 2qa, (q -p )a , and - 2pa for the M]M , and H2M2 "known

segments", resp ectively . k

I = £ K. + B = index value o f young s ir e including breeding value ofdam and of a l l "known segments".

k k jo2 (.E . K ) = 2 E. P iqidi = to ta l variance of "known segment" breeding

v a lu es.2 2 k 2= 0 ( . £ K ) + Og = variance of index values o f young s ir e s ,

including dams breeding va lu e .S = iCoi-Og) = increase in breeding value of index se lec ted young as

a re su lt of "known segment" se le c t io n

Table 4. Numerical example i l lu s tr a t in g the ap p lica tion of marker-a ss is ted se le c tio n and of "known segment" se le c t io n of young s ir e s in dairy c a t t le .

Assumptionsa = phenotypic standard deviation of milk production = 1000 kg.k = number o f polymorphic marked segments = 30.

= number of "known segments" = 20

P “ q “ 0-5.d = 10 0 , h = 0

n = 500 daughters evaluated per marker segment (1200 daughters per s ir e in a l l ) .

i = 2.0Then for:marker a ss is ted s e le c tio n known segment se le c tio n

20

cr Ci= l Ki ) = 316

Oj = 331

S = 460 kg

marker bracket, s in g le marker (r=0 . 1 )

a marker bracket, a s in g le markera marker bracket s in g le marker

2oz 2—----- 4000 kgn

0? = 5000 kg2ADa2 = 9000 kg2

bp = 0.56 for a 0.44 for a

a3 - 10 ,0 0 0 kg2DOj- = 176 kg for

143 kg forS = 150 kg. for

86 ke for a

402

I I I . COSTS

Costs per polymorphism determ inationThree independent s e t s o f a c t iv i t i e s are involved in demonstrating RFLPs.

These are: (1) DNA ex tra c tio n , (2) Preparation of probes, and (3) DNA analy­s i s , including r e s t r ic t io n , e lectrophoresis and b lo ttin g o f DNA, and hybridi­zation and autoradiography o f the b lo t s . DNA extraction i s a major c o s t , coming to $15.00 per sample at current p r ic e s . Although a given DNA sample can be used for up to 240 RFLP determ inations (20 s lo t s per sample x 4 hybrid­iza tio n s per b lo t x 3 probes per h yb rid iza tion ), most ap p lications would require far fewer determ inations. Preparation of probes i s a rather minor c o s t , coming to about $0.05 per RFLP determ ination. DNA a n a ly s is , again , i s a major c o s t , coming to $0.30 per RFLP determ ination. The o v era ll co sts o f an RFLP ap p lication would vary from $0.41 to $1.10 per polymorphism, depending upon the number o f polymorphisms determined per DNA sample.

Costs per ap p lica tionTable 5 g ives estim ated costs for the various RFLP ap p lication s described

in the t e x t , as a function o f the number of DNA extractions required (equal to the number o f in d iv id u a ls tested ) and the number of RFLP determ inations carried

Table 5 . Costs for various RFLP a p p lica tion s, as detefmined by thenumber of in d iv id u a ls tested and the number o f polymorphisms analyzed per ind ividual te s ted .

Number ofind iv idualstested

Number of polymorph­isms per individual

CostsA pplication DNA

extraction 3($)

DNAan alysis '5

($)

Total($)

I n i t ia l screening of population for RFLPs 20 1000 1200c 700 1900

I n i t ia l mapping of RFLPs 100 200 1500 7000 8500

Genetic a n a ly s is of a cross between inbred l in e s

1000 50 15,000 17,500 32,000

Introgression of a q u an tita tive a lle le ^ 500 2 7,500 350 7850

Marker a ss is te d se le c tio n per evaluated s i r e e

1000 50 15,000 17,500 32,500

Identify ing a known locus^

per locus 10 ,0 0 0 3 150,000 10,500 160,500per 20 lo c iS 40,000 60 600,000 840,000 1,640,000

S election on the b asis of 20 "known- lo c i" . Der ind iv idual

1 60 15 2 1 36

aBased on $15 per DNA ex tra c tio n , see Table 1. bgased on $0.35 per RFLP d eter- mination, including $0.05 probe cost and $0.30 cost o f DNA a n a ly s is , see Table 1. cBased on 4 DNA ex traction s per ind iv idual to supply su f f ic ie n t DNA for 1000 RFLP determ inations. ^Based on 5 generations of backcrossing, 100 in d iv i­duals typed per generation . eBased on 1000 typed daughters per evaluated s ir e . fSee te x t . SAssuming each ind iv idual tested contributed to evaluation o f 5 marker-brackets.Cost o f 1 the en tire program would be expended over 5-10 years.

403

out per in d iv id u a l te s te d . In m ost ca se s , c o sts are commensurate w ith the economic va lue o f the breeding o b jec tiv e . For s i r e s e le c t io n , c o sts are h igh , but in the case o f outstanding proven s ir e s whose sons w i l l en ter serv ice as young s i r e s , eva luation of the proven s ir e and m arker-assisted se le c t io n among h is sons might be econom ically f e a s ib le . The investm ents required to bring marker brackets to the stage o f "known lo c i" seem high although they would bq expended over a period of 5-10 y ea rs , but th is a p p lica tio n might become more a ttr a c t iv e as experience accumulated on the v a lu e o f m arker-assisted s e le c t io n among sons o f outstanding s i r e s , and on the magnitude of marker-linked quanti­ta t iv e e f f e c t s in dairy c a t t le populations. Twenty "known lo c i" might w e ll account between them for the bulk o f gen etic v a r ia t io n in m ilk production in dairy c a t t l e , and th e ir f u l l ex p lo ita tio n over a r e la t iv e ly short period o f time might y ie ld eventual production in creases on the order o f 2000 to 3000 kg. per la c ta t io n . A return o f th is magnitude would be more than adequate repay o f the investment involved .

REFERENCESB e ll , G .I . , J.H. Karam and W.J. R utter. (1981). Polymorphic DNA region adjacent

to the 5' end o f the human in su lin gene. P roc. N a tl. Acad. S c i . USA, 78: 5759-5763.

B o ts te in , D ., R.L. W hite, M. Skolnik and R.W. D avis. (1980)-. C onstruction of a g en etic linkage map in man using r e s t r ic t io n fragment len g th polymorphisms. Am. J . Human G ene., 32: 314-331.

Falconer, D .S. (1960). Introduction to Q u an tita tive G en etics. The Ronald P ress C o., New York.

J e f fr e y s , A.F. (1979). DNA sequence varian ts in the Gamma-, Agamma-, D e lta - and B eta-glob in genes of man. C e ll, 18: 1 -10 .

P e te s , T.K. and D. B o tste in . (1977). Simple mendelian in h er ita n ce o f the re itera ted ribosomal DNA of y e a s t . Proc. N a tl. Acad. S c i . USA, 74: 5091-5095.

S o lle r , M. (1978). The use of lo c i a ssocia ted w ith q u a n tita tiv e tr a i t s in dairy c a t t le improvement. Anim. P rod ., 27: 133-139.

S o lle r , M ., A. G enizi and T .) rody. (1976). On the power o f experimentaldesigns for the d etection of linkage between marker l o c i and q u an tita tive lo c i in crosses between inbred l in e s . Theor. Appl. G en et., 47: 35-39.

S o lle r , M. and J . Plotkin-H azan. (1977). The use o f marker a l l e l e s for the in tro g ress io n o f linked q u a n tita tiv e a l l e l e s . Theor. Appl. G enet., 51: 133-137.

Solomon, E. and W.F. Bodmer. (1979). Evolution o f s ic k le v a r ia n t gene. L ancet,1 (#8122): 923.

Southern, E.M. (1975) . D etection o f s p e c if ic sequences among DNA fragments separated by g e l e lec tro p h o resis . J . Mol. B io l . , 98: 503-517.

Tanksley, S.D. and C.M. R ick. (1980). Isozymic gene lin k age map of the tomato: a p p lica tio n s in g en etic s and breeding. Theor. Appl. G en et., 57: 161-170.

T anksley, S .D ., it. Medina-Filho and C.M. R ick. (1981). The e f f e c t o f isozyme s e le c t io n on m etric characters in an in t e r s p e c if ic backcross of tomato: b a s is of an ea r ly screening procedure. Theor. Appl. G en et., 60: 291-296.

Thoday, J.M. (1961). Location o f polygenes. N ature, 191: 368-370.Wyman, A.R. and R. White. (1980). A h ighly polymorphic locu s in human DNA.

Proc. N a tl. Acad. S c i. USA, 77: 6754-6758.Zabeau, M. and R. R oberts. (1979). The r o le o f r e s t r ic t io n endonucleases in

m olecular g e n e t ic s . M olecular G enetics 111:1—63. (J.H . Taylor E d .).Academic P ress , New York.

Zhuschenko, A .A ., A .P. Samovol, A.B. Korol and A.B. Andryuschenko. (1979 ).Link­age between lo c i o f q u a n tita tiv e characters and marker l o c i . I I . In flu en ce o f three tomato chromosomes on v a r ia b i l i ty of f iv e q u a n tita tiv e characters in backcross progen ies. Translated from Genetika 15: 672-683. In: S ov ie t G en etics .

404