Additional Isozyme Loci in Tuber-Bearing Solanums: Inheritance and Linkage Relationships D. S. Douches and C. F. Quiros

From the Department of Vegetable Crops. University 01 California, DaVIs Dr. Douches IS currf'nlly at the Department 01 Crop .nd Soil Science, Mldngan State University, East lansIng..~cldress rep',nt requests 10 Dr. Douches. Department of Crop a,a) So,1 SCIence, M,chigan Slale Univers,ty, [a~t LansLng ~11 45824.

Journal 01 Heredity 1988;79:377-384; OI)~2-1503!88!$2.00

Horizontal starch gel electrophoresis was used to pertorm genetic analysis to confirm the inheritance of various isozyme loci and report on additional loci in diploid Solanum species. Aps-I, Dis-I, and Idh-1 were identified as new loci, and tetrasomic seg­regati on was reported at the tetraploid level for Mdh-2. Of 15 loci stu died, d Istarted segregatIons were observed in five diploid test crosses involving the Got-I, Idh-1, Pgi- 1, an d A dh-1 Ioc i. In these ca ses, fou r of the five pa rents were 01 inte rs pec ific originj the family segregating for Adh-1 was the exception. Utilizing various clones and interspecific combinations between S. phureja, S. tuberosum, and S. chacoense, further test crosses were made to determine linkage relationships between these isozyme loci and the yellow tuber flesh gene (Y). Two linkages were detected among these markers. Estimates of the Idh-1JSdh-1 linkage ranged from 10.4 map unIts (m.u.) to 36.8 m.u., whereas a tight linkage was confirmed lor Pnc-2!Prx.-3 (0.6 m.u.). Comparing putative homologous loci for the Idh-I!Sdh-I linkage implies some con­servation sInce the divergence of Lycoperison, So/anum, and, to a degree, Capsicum. As a result of codominant ex.pression, these isozyme markers provide new oppor­tunities for further genetic studies of tuber-bearing Solanums.

Several genetic markers have been re­ parison 01 the chromosomal changes that ported in the potato, Solanum lubeTosum have taken place in the evolution ot the L.: tuber shape,2; russetling,2.JO morpho­ two genomes.3' In the same sense, il link­logical and chlorophyll markers,9.,2.J3.37 age relationships and chromosome posi­embryo seed spot,1O anthocynanin pig­ tions could be determined on the pOlato. ments,J·s deformed anthers,7 yellow tuber these electrophoretic markers could be flesh. 11 and tuber proteins.' Aside from the exploited to develop more critical studies utilization of the embryo seed spot markerS of genetics, breeding, and evolution. More and the yellow tuber flesh marker,2s the specifically, the identification ot an iso­value of these markers has not been dem­ zyme marker close to the centromere is onstrated beyond their initial develop­ needed to apply haH-tetrad analysis (HTA) ment. to determine the mode of 2n egg lorma­

More effective chromosome markers are tion. J7 Furthermore. if linkage could be de­needed il further progress is to be made tected between two loci involved in this in the genetics a/luber-bearing Solanums. analysis, the power 01 HTA would in­As revealed by starch gd electrophoresis, crease. '6 These Iinkage re lat ionships can isozymes have been a source of chromo· also be applied to the characterization of some markers innumerous species.6 These meiotic mutants (e.g., sy3/sy.1, ps/ps) markers afford several advantages over through genetic analysis. 1/ segregating morphological ones in genetic and breed­ diploid populations could be derived lrom ing studies because of their codominant plants saturated with identified isozyme nature. In the past few years, genetic anal· loci, associations might be revealed be­ysis of isozymes has been initiated in the tween these markers and important ag­potato.J4·J5.IS,U,27 ronomic traits. The tagging of traits with

The application 01 isozyme linkage re­ economic importance by specific markers lationships to breeding and genetic stud­ would facilitate the introgression 01 genes ies has been demonstrated in the tomato, lrom the wild species.2,.JO For the purpose Lycoperslcon 23,33-35 In Capsicum, another of utilizing 2n gametes in modern polato member of the Solanaceae, a linkage map breeding schemes, it would be worthwhile has begun to be revealed, allowing a com- to determine genes linked close to their

377

Table l. Diploid and tetraploid Solanum selections used 10 the study

Parental clones Taxonomic species

84510 S4S11 84512 H4 M5 845022 84SD5-7 84509·2 84509-4 845D9-6 84509-9 845013-7 845DI3-) 7 845DI3-29 A66133·2 ND0277-2 NDD47-1 Centennial Russet

S pl'/Jreja S phurejo S. pllUreja (phu x hap) x (phll x hap) (phll x hap) x (phll x hap) phu x chc' hap x che S phurejo S phurejo S. pllUrejO S phurejo (phu x ehe) "- (rhu x che) (phu"-chc) x (phuxe!>c) (phu xehe) x (phu x chc) S lu!Jero,ulll S lu!>erosum S tu!>erosum S. tubero$um

"chc - S, churnerlse; phll = S phurej<r. hap = dihaploid S. tuberowm subsp lubero$um.

centromere, in that these traits would breed true in the 4x-2x first division res­titution (FOR) progeny.22

We report the Inheritance of additional isozyme loci as determined through dip­loid, tetraploid-diploid, and tetraploid se­grations and also confirm at the diploid level the inheritance of loci thai have been analyzed at the tetraploid level. Linkage relationships between these loci and the tuber Oesh color locus (Y) are also ex­amined.

Materials and Methods

Plant material. Various clones of S. phureja Juz. et Buk. along with hybrid combina­tions between S. phureja and S. chacoens€ Bitt, and S. phureja and dihaplolds of S. /uberOSlimwere used (Table I), Tetraploid clones of S. /uberosum were supplied by Dr. J. J, Pavek (U.S.D.A., Idaho) and Dr. R. E. Voss (University of California, Davis), and the diploid clones MS, H4, and T704 were supplied by Dr. S. J. Peloquin (Uni­versity of Wisconsin, Madison).

We made all c:rosses in a greenhouse. The resulting true seeds were treated with 1,000 ppm GA for 24 hOIlTS, Germinated seedlings were transplanted in trays (50 plants per tray). We sampled roots or leaves for electrophoretic analysis from 4­to 6-week-old seedlings that were healthy and were growing Vigorously. For tuber­specific enzyme:; (APS and ADH), the seedlings were allowed to tuberi7.e in the trays. Small mature tubers were harvested in 3 to 4 months.

Electrophoresis We obtained crude pro­tein extracts by crushing an approximately 120-mg sample 01 pC):ato tissue in trays over ice. One hU[!L:red twenty /11 ollreshly

50urce Plo,dy

MUlIl 2999.79-207 MLOO 2999 79·209 Minn 2999.79·294 Sturgeon Bay. WI Sturgeon Bay. WI Madisoll. W1 Selec:tlon from 82565-3 selted SelectIon lrom 82554-26 selted Selection (rom 82S54·26 selted Selection (,om 82554.26 selted SelectIon from 82554-26 selted 5ele(lion lrom 82S59-21 x 61-23 5e1ecllun fro", 82559-21 x 61-23 SelectIon f ('(HI( 82559-21 x 61-23 J J Pavek, USOA. 10 R.£ voss. U c., DaVIs R. E_ Vnss. U.C., Davis R. E voss, U,C., DaVIS

prepared 0.1 M Tris-HCl buffer, pi' 7.8, containing 2% glutathione was added to the tissue sample before crushing, The ex­tracts were absorbed onto two 3 x 8-mm Whatman 3MM wicks and stored over­night at - 20Q C.

We performed electrophoretic assays in two different gel systems, Tris-Citrate, pH 7.8,20 and Histidine-Citrate, pH 5.7.' Gel slabs consisted 01 10% potato starch. An­odal peroxidase (PRX), phosPhogluco­mutase (PGM), alcohol dehydrogenase (ADH), glutamate oxaloacetate transami­nase (GOT), triose phosphate Isomerase (TPI), and diaphorase (D/A) were as­sayed using the first system. Phosphoglu­coisomerase (PCI), 6-phosphogluconate dehydrogenase (6-PCDH), shikimic add dehydrogenase (SDH), malate dehydro­genase (MDH), isocitric acid dehydroge­nase (fDH), and add phosphatase (APS) were assayed using the second system. We prepared enzyme activity stains according to the method of Vallejos..1; Techniques concerning the procedures (loading, elec­trophoresis, slicing, and staining) were

Table 2.

Locus

Idh-l 6-Pgdh-3 6-Pgdh-J Pgm-2 Pg,n-2 GOI-I SJil·1 Pgl-1 Adh-I Prx-."! Mdh-l Mdh-l

2n = 24 2n = 24 2/l = 24 2n = 24 2/l = 24 2n·24 2n = 24 2n = 24 2n = 24 2f1 = 24 2n·24 2n = 24 2n = 24 2n - 24 2n = 48 20 = 48 2n a 48 20 = 48

described by Quiros.20 The nomenclature used to describe the isozyme system. en· codhg lad, and alleles was in accordance with Quiros and McHale."

Linkage analysis_ Pairwise comparisons 01 segregati ng lad were analyzed using the LI NKAGE-I program developed for the mi­crocomputer,29

Results

Euzyme Variation [n this study, 49 alleles for IS enzyme­coding loci were identified in 11 enzyme systems. In addition, segregation data for these loci revealed three linkages,

Isocitric acid dehydrogenase (IDH). Ac­tivity of the enzyme 1D1-I in potato tissue was previously reported by Sanford et al.;~"

however, genetic analysis was not per­formed, We observed one zone of electro­phoretic activity in the :DH zymogram, named Idh-I. In the cultivated Solanum species studied, we identified only two al­lozymes at this locus, Idh-I ' with the far­thest anodal migration and Idh-F A third al\ozyme, Idh-F, with the slowesl migra­tion, was identified in an accession of the wild species S. pinniatisecium (p.r. 275232) (Figure IA). The expression of this en­zyme was greatest in leal and pollen tissue, whereas poor activity was associated with the use 01 tuber tissue.

We observed six diploid families that followed test cross segregations (Figure IB and Table 2). Heterogeneity among families was observed Cx2 ~ 15.417, P < .01). Deviation from the I: I ratio was ob­served in family 855038. H4, the segre­gating parent in thiS cross, was derived Irom tuberosum x phureja hybrids (Table I). Distorted ratios (or this locus were not found in other families in which the seg­regating par~nts were of S. chocoense x S. tuberosum or strictly S phun?)Q origin,

6-Phosphogluconafe dehydrogenase (6­

Pooled segregation dllla for severt'l Isozyme loci

Number of Parental familLes genotypes

liP x Jill

8 3'3' x 3'3' 2 3'3' x 3'3' 6 2'2' x 2'2' I 212:! x 222;} 3

6

I:J] ~ x 1::J}:1

6 I')" x 1'1' 3 1'1' x 1'1' I 1'1' x I')' 3 3"3' x 3'3" 4 1'1' x 1'1' 2 1'1' x 1'1'

upc·,·ted Observed segregation ratio x' P

138(1 '1').199(1'1') 1.1 026 .63 216(3'3');213(3'3' ) \1 om 75

20(3'3') '45(3' 3'):26(3'3') I ! I 080 72 13' (2'2'): 152(2'2') I J I 27 29 25(2'2'):23(2 '2'-) 10(2'2').15(2'2~) 1:1:1;1 684 08 55(1'] '):55( 1'1') 11 000 1.00 l81(1'1'),21~(1 'I ') I:] 2.76 .10 75( I'l '):60( I' 1') II 145 2S 5( I'1')_7(1'1 '):5(1'1 ') 1:2 I 0.99 .63

69(323')37(3'3') I I 9-07 0025 93(1 ' 1') 75(1'1') 1'1 122 .20 29(1'1 '):58(1' 1');27( 1'1 ') 12'1 035 99

378 The Journal QI Here0ity 1ge'.' '9( '"

edabc

A

3~

3L

B

C

a b c d

... I

6 Pdgh 2< 0 ,, ­

6 Pdgh 3 __

1L 1L...· 21_ E

22 -3

2 -­

F

a c d e f 9

Flgu,"" l. Progen les segregating lor various enzyme-( oding loci Anodal direction is above (A) fDH profile i1lustrat,ng allozymes 01 \Ill' IJh-llocu,; wnes a Ihrough C show in(llv,duals homozygous for Idh-I', and lanes d and e refer to typ'cal three-banded phenotypes of the Idh-I' I' heterozygote_ In lanes f through i, the lIld'Vlduals hom"'v~o"s for Ie/h-I' are observed (B) Test noss (lJplo,d progeny 855D50 segregating 1: 1 at the Idh-I locus lor I:)~ "lides I' and p (e) Profile of 6--PgJh LllustratlOg allozymes of the 6-Pgdh-J locus, The lower live bands

PGDHj. The 6-PGDH zymogram revealed three zones of electrophoretic activity, Of the three, we identified one locus, 6-Pgdh-J, in the segregating families stud­ied. Previously, Martinez·Zapaler and Olivier" studied the inheritance o( PGD-C in two tetraploid (amilies. Based upon their nomenclature for the segregating parents. 6-Pgdh-3 corresponds to PGD·C. which is closest to the origin_ For this Jocus, we observed two allozymes as single bands in the segregating families. The laster al­lozyme was assigned 6-Pgdh-3I

, and the slower band was assigned 6·Pgdh·,12, These correspond to the alleles PGD-0 and PGD­Cb, respectively, of Martinel-Zapater and Olivier's nomenclature, 14 A thi rd allozyme, 6-Pgdh-3', with an even slower migration than that of 6·Pgdh"3", was observed in accessions of S. sparsipi/um (P,l. 246536) and S chacoense(P.1. 320283) (figure Ie),

We scored a total of 520 progeny in the 10 diploid (amilies studied, Test cross and F2 (1:2:1) segregations were observed for this enzyme (Figure ID and Table 2), We fou nd homageneity among the fam iIies for both the test cross and F1 dala (x:' "9.4 1, x 2 = 1,325). The pooled data fit their re­spective segregations.

Phosphoglucomutase (PCM). PGM iso­zymes were determined to be coded by two loci, Pgm-T and Pgm-2, which corre­sponded to Marlinez-Zapater and Olivier's" PGM-A and PGM-B, respective­ly, We studied Pgm-T in (amily 86SD3L The progeny followed a typical test cross seg­regation pattern (or monomeric enzyme (40 Pgm-l ' J2:36 Pgm·P F; x2 = 0,118, P ,72) (Figure IE). Two singlc"banded allo­zymes were observed, Pgm-l ' and Pgm-J2, with Pgm-J2 assigned to the less anodal one, Enzyme aclivi ty for this locus was high in all tissues, but the bands were clearly

ill lanes a and b correspo"d to t,iallellc triploid plant~

of the phenotype 6·Pgdh-J'YJ3 The lower three bands in lane c reveal the dIploid heterozygote 6-Pgdh-J1 J'. and lane d corresponds 10 the d'l/tOld heterorygote 6-Pgdh-JIY (0) The d.plol(J tesl cross lam II} 855D33 segregahng lor Ihe 6-Pgdh-J.' and Y allozymes (E) PGMrymogram reveals two monomeric enzyme-cou­tng lod se~regat illg 10 the d,ploLd lamLly 855D31 The more anodal locus, Pgm-I, is above. segregalLng In ~

test cross manner, The upper two ba'I(b I" lanes a through d correspo"d lO the Pgm-I'I' hetnozyg(JI<" and the upper band Ln lane e correspond~ to the In­dividuaJ homorygous lor the Pgm·I' alJo"yme The Pgm-2loeus is segregating for thn'" alleles The lower two alleles In lane 8 reler \0 the Pgm·2' P two-banded helerozygote, and lane g reve~1 s the Pgm-2·· :'.. two­bande<l heterozygote. (f) Adh zymogram as revealed from tuber t,ssue_ The diploid progenIes of 86501 ) segregating in a lest CrOSS manne' lor Ad"-I' and I' allozymes

Douches and Quiros • Isozyme LOCI In Tubc··Bearing Solariums 379

--

Ii­

12_ A 13

a b c

13 --. 41 ­ B

15_

a c d e f

c

abcde '9

D

E

Ft~re 2. Progenies segregatJng for various en.lYme-coding loci. Anodal direcUon is above. (A) Aps-J locus as revealed from tuber t,ssue, Lanes a and h correspond to the di ploid Aps- I' I' J'1', and lane c reveals tile Aps-!'l' Ihree-banded indiv,dtl~1 (8) Variuus lJil'loid progenies for the Gol-! locus, The upper three bands of lane a correspond to the COl.! ./' heterMygote, A nonseqregatmg. nonallelic band overlaps with the COl' I' allele, giving the four-banded phenorype In 1~)1es b, c. d, e. ""d I The next three slower-migrating bands In these lanes correspond with the Cot·!' /., heterozygote. (C) Sdh- / lo~"s segregating Ill. the d,ploid progenIes of 855D48 .n a ,.] manner for the Sd"-I' and 1" alleles (D) 855031 fan,;ly segregating for the monomeric isozyme locus Dlo-I In a test cross manner for the alleles D,Q'!' and j2 (£) The more anouallocus. Mclh·2, IS segregating I: I for the five-banded phenotype ll,fdh·2'2' (lanes b, c. d. and f) versus the three-banded phenOlype Mdh·2' (lanes a. e, and g) in the tetraploId pro~"nJes of 86SDI The Mdh-!Iocus (below) is segregaling in a 4x-2x (FDR) test cross manner

380 The 'c·ur",' 01 Heredi\y 1988;79(5)

resolved only when we obtained enzyme extracts from tuber tissue.

The locus Pgm-2, also coding for mo­nomeric allozymes, segregated indepen­dently of Pgm-/ (Figure lE). We observed three a!lozymes-Pgm-21

, Pgm-:P, and Pgm· 2!-in order of decreasing electrophoretic migration. Pgm-23 was unique to the S. cha­coense accessions studied, whereas Pgm­21 and Pgm-:?were common to all the Sola­num species utilized in this study. In the family 865031, these three alJozymes were observed to segregate. We observed test cross segregations in six families (Table 2). Pooled data for 284 progeny (P = .62) fit a I: I segregation (131 Pgm-211': 152 Pgm­2223, Xl = \.271, P = .29).

Glutamate oxaloacetale transaminase (GOT). We detected two zones of activity in the GOT zymogram. The zone closer to the anode, Cot-/, correspo nds to the GO ]"·A locus described by Martinez-Zapater and Olivier. J. A lower zone 01 activity, desig­nated Got-2, was not amenable to analysis because 01 inconsistent expression in the tissue sampled and the difficulty of clearly defining the genotypes 01 the parents as a result of alleles expressing as multiple­banded phenotypes. We identified three alleles at Got-! in lour segregating families: Got-]), the most common allele, corre· sponds to the most anodal migrating ai­lozyme, whereas Got-/ 4 and Got- l' refer to less anodal migral ing allozymes. Gol-/5 was unique to the species S chacoense acces­sions (Figure 28). Alleles Got-] , and Got­12 were identified in other Solanum acces­sions that we re not com rna n to the paren ts in this stUdy. Families 855033 and 855018 fit expected test cross segregations Cor a dimeric enzyme at the locus; however, two other Camilies-855025 and 865031-had distorted ratios (Table 3). Highly signifi­cant X:' values for these families (x2 4.8 and x2 = 22.61, respectively) precluded pooling individual family data. All the seg­regating parents in these four families were of interspecific hybrid origin, either chu­coense x phureja or c1lGcoense x tubao­sum.

In the 4x-2x family 865030, segregation for the Got-J locus was observed in the tetraploid progeny. Based on relat ive banding intensities, NOD227-2 (2n = 4x 48) was assigned the genotype Got- fJ P {41", and 855039-36 (2n = 2x = 24) was as­signed the genotype Go/-/' F', With these parental genotypes, a 1:4: 1 segregation ra­tio is expected if tetrasomic chromosome type segregation is assumed. The data on lamily 865030 easily fit these expectations (41 Cot-jJjJfJFl61 Got-JJ jJfJj4:48 Cot­

Tab.~ 3. Distorted segregalioo ratios

Locus Family Parental croSs Observed segregation

Ex-peCle<J rallQ x' P

Idh-I 855038 H4(1'l') x 84510(1'1') Col-I 865031 845 (O( J'I') x 845D22( l' J') Gal-I 855025 845DI3-7(1 '1') x 13-17(1 'I') Pgi-J 855018 845 II (J' 1') x 84SD l3-29( l' IJ) Mh-I 86SDI J 84512(('(') x 84S11(l'l')

F ]3/4]4, X2 0.9777, p;o .65), suggesting ;0

a proximal chromosome arm position lor the Got-jlocus.

Shikimic ucid dehydrogenase (SDHj. We observed zones of enzyme activity lor this enzyme, but Sdh-j, the locus of more an­odal expression, had consistently strong actiVity in leaf tissue. The lower-migrating zone, 5dh-2, had inconsistent expresSion ,'X(l'pt in an accession of S stolonirerum (P,l. 195166) and a derived triploid hybrid progeny involving this allotetraploid and ,';_ phureja (unpublished data). Enzyme ac­tIvity for this enzyme was highly ex­pre.~sed in leaf tissue, w1-!ereas root tissue was inconsistent. Tuber tissue expression / of this enzyme was generally weak.

5ix families segregated in a test cross lashion for various paired combinations 01 three alleles. For each family, the segre­gations fit the expected Mendelian ratios for monomeric enzymes (Table 2). Our al­lozyme nomenclature is in accordance with that 01 Quiros and McHale," who recently reported limited segregation data at the tetraploid level for thi~, locus. Allele Sdh­j', which has a two-banded phenotype, migrates most a-Jodally, whereas alleles Sdh- P and Sdh- P migrate in a slightly less anodal lashlon (figure 2C).

Phosphoglucoisomerase (pel). The in­heritance 01 the Pgi-j locus had been elu­cidated a: the tetraploid level by Staub et i'L21 and at the diploid level by Quiros and JVIcHak,"Z In our study, diploid test cross segregations for the Pgi-; locus were ob­served in a number of families that were analyzed for linkage relationships (Table 2). Three of the lour families showed good fits to the 1:1 ratio. The one deviant seg­regation involved a phl/reja x chocoense hybrid, 84S013-29 (Table 3).

Alcohol dehydrogenase (AOHj. One lo­cus, Adh·l, was detected lor which activity was specific to tuber tissue, We observed two single-banded allozymes in the seg­regating fami lies, wh ich we designated A dh­I' and Adh-J3, the former referring to the 'llOre anodal migralillg allele. These al­leles respectively correspond to the AOH­A k;-us alleles N and A' described by Mar­

3(1'1')'18(1' I') 1-1 685 >_02 GO( l'P)-18( I'1') 1.1 22.61 :> 0001

9(1 "1'):21 (1'1 'J 1'1 4_8 032 10( 1'1 '):22( I" I') II 4.661 ,035 47(1' 1'):26(1'1') 1'1 598 025

tinez-Zapater and Olivier. H In the clone USW-S295.7 (genotype Adh-/ I F), a third allozyme, AdIJ-j', of more anodal migra­tion than Adh-J2 was identified. Adh-P is equidistant from Adh-]' and Adh-P.

Two families segregated for this enzyme (Tables 2 and 3). 855D33, an F2 progeny, segregated 1:2:1, as expected (Xl 0.999, p= .63), whereas 865011, a test cross seg­regation (Figure I F), deviated from ex­pectations (x2

- 5,98, P = .025). This was the only locus we observed with a dis­torted ratio in which the segregating par­

·ent was not derived from interspecific hy­brid ization s.

Diaphorase (OIA) The diaphorase zy­mogram in the potato revealed three zones of activity studied in the Tris-Citrate pH 7.8 gel buffer system. Of the three zones, only the most cathodal one, governed by the locus Dia-I, revealed polymorphism. Expression of the DIO-/locus was greatest in leaf tissue, whereas the extraction 01 DfA from root or tuber tissue resulted in weak or variable expression.

For the Diu-I locus, we observed two single-banded allozyrnes-designated Dia­II and Dia-P-with the more anodal band given the lower number (Figure 2D). In the lamily 865031, this locus segregated in a :: 1 manner. assuming a monomeric en­zyme (43 Dio-]! J-'):35 DIG-/ I /1, x2 = 0.821, P . ,32).

And phosphatase (APS). We identified one APS-coding locus in potato tuber tis­sue through the Histidine-Citrate pH 5.7 gel buffer system (Figure 2A). Crude pro­tein extracts from root or leaf tissue were not resolvable. Segregation lor this locus was studied in one family, 865011, This dimeric enzyme segregated as expected in a test cross (or the alleles Aps-j' and Aps­[2 (38IJlF):351' 'II), x2 ;o 0.055, p;o ,89).

Malate dehydrogenase (MDH). MDH no­menclature and allozyme descriptions can be found in Qu iros and McHale.22 The most cathodal locus, Mdh-], was observed to segregate in six diploid families studied. Families 855045 and 865031 showed bor­derline x 2 values for a 1:\ ratio; however, the pooled data of four families showed a

good fit to a test cross segregation (93 Mdh­{I F:75 Mdh-F/2, x 2 = 1.22, p, .20). Two F, families, 865011 and 855D39, segregat­ed in a Mendelian fashion (Table 2). The pooled data fit a 1:2:1 ratio (29 Mdh-J' I': 58 Mdh-I ' F:27 Mdh-J2]2, x' - 0,035, P = .99).

We were able to define a second locus, Mdh-2, 01 more anodal expressioll than Mdh- I. Quiros and McHale" found no vari­ation lor this locus among their segregat­ing families except among some acces­sions 01 S. sparsipilium that were fixed Cor a laster-migrating phenotype designated Mdh-22. In the tetraploid family 86501 (de­rived from a 4x-2x cross) the segregating 4x parent, A66133-2, was scored as the variant Mdh-22 phenotype, whereas the diploid parent, 84512, was homozygous for Mdh-2J (based on other crosses involving this parent). The 4x progeny segregated for the Mdh-2! and Mdh-2k phenotype in a 1: 1(61 Mdh-2'2 'J

I 22:65 Mdh-2'2' 212' ) man­

ner, implying a simplex heterozygote ge­notype, Mdh-2'212122

, lor the 4xparent. We observed a third possible phenotypic class, Mdh-2'212222, hased on dosage eRects, lor two plants 01 the 128 progeny (Figure 2E). If it is assumed that the genotype can be inferred from relative band intensities, these two plants of possible duplex ge­notype might have arisen through double reduction (alpha 0.06250), The fre­quency of these putative duplex genotypes would be based on the assumption that this locus is not tightly linked to the cen­trOinere,

Yellow tuber flesh gene (Y). In a cross between a yellow-Oeshed and a white­fleshed diploid potato, the lamily 8650 II segregated in a 1:1 fashion Cor yellow and white tuber flesh (40 Yy:33)Y, x2 • 0.493, P -, .50). The resulls are in accordance with the hypothesis that a single dominant allele determines the yellow-nesh trait. ,o ,2A

With this inlormation, the parent 84512 was designated Yy for the yellow luber nesh locus.

Linkage Relationships of Isozyme Loti Table 4 summarizes linkage data for the marker loci reported. We determined link­age relationships from the segregating families 86S031, 865D11, 855033, and 855048. Missing entries in the table indi­cate that segregation data were not avail­able for those paired combinations.

Based upon test cross and F2 segrega­tion data, most 01 the 15 loci showed in­dependent assortment in pairwise com­binations, For the sake of brevity, we discuss only combinations of lod for which

Douches and QUirOS' Isozyme LOCI In TutJel·Beallng Solanums 381

Table 4. Two-way contlngency lest for linkage ~tween segregallng Isozyme loci'

Aps'! Dia·} Go/·I 10"-1 Mdfl'! 6-Pgo·] Pgi·! Pgrn-l Pgm·} Prx·;] Fh-J Sdh·1 y

Ad"·! Aps·/ Dia·} COl·! ldh·l

" (73)

. (78)

• (67)

• (78) , (78)

, (73) • (73) , (78) , (78) '(78)

• (70) , (70) 32 eM'" (71) • (71) • (71 )

· (73) · (73) , (73)

· (7S) 38.2 ~M' (76) , (76)

, (72) · (72) • (73) · (73) • (73)

'(77) 416eM'(n) '(77)

• (67) 40.3 eM' (67) • (67)

• (73) • (~1) • (II) , (77) 10.4.36 eM" (77. 166)

• (73) • (73) , (7S) • (75) , (7S)

Mdh·l {j·Pgd·.?

• (71) • (73) · (71)

• (76) - (71)

• (71) · (71)

- (77) • (7ll)

· (67) · (60)

- (77) , (71)

• (73) , (''')

Pgi·1 · (72) • (72) • (72) , (62) · (73) • (70) Pgm-! Pgm·}

• (7:3) - (75) , (72)

· (65) • (62)

· (76) '(,1)

• (73) • (72)

Pr.y·J I 5 cM" (66) · (76) • (74) Prx·'] · (61;) • (1;4) SDH·I '(::1) y

" umbers in parentheses - n.

h' ~ nOl significant.

, eM ~ eentimorgan.

J Significant at 1% level.

• Significant at 5% level.

significant deviations from independent assortment were found.

The Idh-ljSdh-1 linkage was observed in two segregating diploid families, 865031 and 855048 (Table 4). The two estimates of map distance between those loci were significantly different from each other (10.4 map units [m.u.] vs. 36.8 m.u.). The ~eg·

regating parents in these two families were 845022 and H4, respectively. Both parents are of interspecific hybrid origin (Table I)

The linkage between Prx-2 and Prx-3was reported by Quiros and McHale22 after no recombinants out of 110 progeny were found between these two Joci. OUf data, obtained from the family 865D31, are in accordance with their data, with only one recombinant found among 66 progeny (combined estimate = 0.6 m.u.).

In the family 86SD31, we detected a pos­sible linkage between 6·Pgdh-3 and Dia-1 (X2 = 9.37, P= .002) with a recombination frequency of 32.4 ± 5.6%. The segregating parent in this cross was 845022. Because this parent was previou~ly involved in the conflicting Idh-1jSdh-1 linkage estimates mentioned earlier, further crosses neerl to be made using segregating parents o( uif­ferent genomic constitutions, In this way. any biases of the map distance caused by interspecific genomic combi nations would be revealed.

Discussion

We have confirmed through genetic anal· ysis the inheritance of isozyme loci and have reported additional loci at the dip­loid level. Under proper electrophoretic conditions. codominant expression of al­lele products is found. Possibly as an in­dicator of the immense diversity available

382 The Journal of Herp,dity 1988.79(5)

in the potato species, (our or mOfe alleles have been identified for the Sdh·l, Mdh-I, Col-l, Pgi-I. and Pgm-2 Joci. In a number o( instances, alleles were expressed as two­banded or doublet phenotypes_ Alleles Mdh-l', Mdh-I'. and Mdh-1 4 are attributed to posttranscriptional modification, are expressed as doublets, and breed true in selled progeny from homozygous ptants.'" For the Go/-I, Sdh·l, and Pgm·210ci, doub­let phenotypes were expressed by alleles GOI-F. Sdh-I I

• Sdh·]4, and Pgm-24. Unlike the Mdh-I alJozymes, each band 01 the doublet was exprb.;ed wilt· equal inten­sity, Inferences o( the genotype based on these phenotypes ffitly be in error for these loci. As a rule. multiple-banded phc-JO­types should be progeny-tested to confirm the genotype or reveal what other single­banded alleles may be hidden.

One zone of electrophoretic enzyme ac­tivity was detected for PCI Previous stud­ies have examined this enzyme in a num­ber 01 bufler systems: pH 6.1 ,2~ pH 7.8,26 and pH 8,3. 1' The resolution of the Pgi-1 locus was most clear using the Histidine­Citrate pH 5.7 system. When the PCI en· zyme was separated in the pH 'i',8 and 8.3 systems. a more anodal zone of activity was observed, which leads us to believe that other PCI-coding loci may exist. How­ever, in a pH 5.7 system, the more anodal zone of activity was resolved as the 6-PGDH zymogram.

Quiros and McHale22 synthesized their data for the PRi- J locus in accordance with Stallh et al.": and Olivier's" group, As a refereoce, the most common allozyme, Pgi­I", is equivalent to CI of Staub et al. and PGI-!3< of Olivier's group, which is the phe­notype of the cuJt\vars "Red Pontiac" and "Nooksack." In our study, thediplol(j 84510

had the genotype Pgi·jl J2 and 8450 13-29 had the genotype Pgi· J2 J1.

We observed that the stains used for alkaline phosphatase (AKP) and APS re­solved identical phenotypes lor all the clones tested. We (ell that the APS and AKP stains resolved the same enzyme lo­cus because o( the common stain sub­strates. [n addition, both loci were ex­pressed only when tuber tissue was used, Staub et aLl? studied the inheritance of AKP. Our results and observations are consistent with their analysis. Staub's group identified three alleles (Figure' 2A), As a comparison 01 the two enzyme sys­tems, one of the parents studied-Nor· chip-was characterized as homozygous (N' ,), whereas in our system it was also homozygous, Aps-/' 111/ ]1, by our nomen­clature. Meallwhile, another cultivar, "Kennebec," was characterized as A" ,A 'J

(simplex genotype) by Staub's group. whereas we identified it as Aps·11 If 11 / 2.

We preler the APS system because of the greater slain activity.

Loose linkages were detected between Gol-IjPrx-2. Cot- 1j Prx-3, and Got·ljPgm·1 in the family 865031. The validity of these linkages is questionable because of the very high recombination frequencies of 41,6%, 40.3%, and 38.2%, respectively. As a result of sample size, these values may not be significantly different from independence. Second, the segregation (or the COl-/locus in this family was highly distorted and in effect could have biased the observed re­combination frequency as a consequence of the Jow frequency of certain progeny classes. Third, gene-centromere map dis­tances from COI-/ (0.9 m.u.) and Prx-3(18.0 m.u.) (unpublished data) do not correlate with our data. A much tighter linkage woul d

•

"

be expected if there were an actual phys­ical linkage. In addition, no linkage was found between Prx-2 and Pgm- J or be­tween Prx-3 and Pgm-I. Further crosses must be made to study the associations between Got- J, Pgm- J, and Prx-2J Prx-3 be­fore linkages can be established conclu­sively.

With the availability 012n pollen in some of the diploid selections in this study, gene­centromere relationships can be readily determined through 4k2x crosses. '1 This independent source 01 linkage data can easily be compared to the diploid segre­gations concerning these loci. Estimates of gene-centromere map distances will be presented in another paper. Data gener­ated from 4x-2xcrosses can be pooled with the diploid segregation data to confirm linkages. In addition, information can be drawn about the distribution of this set of isozyme markers in the potato genome, With linkages known Irom diploid segre­gation data, 4x-2x data should also yield gene order for these loci and provide in­sights into recombination interference. Lo

Alnong the 15 loci studied, distorted segregation ratios were observed in only four loci (Table 3), In all instances the segregating parent was identified, and in four of the five families the segregating parent was of interspecific hybrid origin (phureja x e,hm:oense, luberosum x cha­coense, and phurejo x tubero.';um). Dis­torted segregations have been reported for Prx-2and Mdh-J in the potato.'2Invariably, iLe distortions involved hybrid combina­tions between S. phlJrr>ia, S. chocoense, and S. tUberosum dihaploids. Distorted ratios also have been reported in tomatoes" and alfalfa./) for PRX loci, Quiros and McHale.12 attributed this to modification by genes located at other loci, which I;lo.y occur more frequently when genomes of interspecific origin are involved,

Segregations of the GOI-! locus were se­verely distorted from th e expected I: 1 test cross ratio in the family 86SD31. At times, distorter segregations may be a clue that gametophytic or zygotic selection has oc­curred in a cross, The segregating parent, R4SD22. is an FL selection Irom all S tuber­usum x S. chocoense cross. As it result 01 this species combination. heterozygosity may be expected lor a 1ll.. lllber of loci linked to Got-1. The heterozygous nature of the potato may conceal recessive lethal and sublethal loci in the genome. Linkage be­tween Got-I and a locus 01 this nature can IldVe a selective disadvantage at the ga­metophytic stage or the zygotic stage. Lam and Erikson lJ lound a gene causing albi­

nism in S. chacoense and suggested that zygotes homozygous lor this locus may be lethal at an early embryonic develop­menlal stage. WagenvoortJ7 and Hermsen el al.~ described four lethal genes in seg­regating potato populations that may af­lectthe segregation ralio 01 the other linked loci th rough actions in various stages of embryonic developmenI. It is poss ible that a linkage between the Got-J locus and another heterozygous locus of this nature, selected against gametically, can lead to distorted ratios lor the Col-1 locus. Tanks­ley and Loaiza·Figueroa3' recently detect­ed a linkage between the sell-incompati­bility locus and both Idh-l and Prx-J in the tomato based on distorted segregation ra­tios for these loci. Further crosses and analyses should be undertaken to support this type of gametophytic selection in the potato.

Since the potalo, the pepper, and the tomato all belong to the same family, the Solanaceae, linkage relationships may re­veal regions 01 the genome that have re­mained intact since the divergence of these genera. The linkage between Idh-J and Sdh-J in the potato reveals a linkage con­servatio n among these three genera, Tanksley and Loaiza-Figueroa32 placed this linkage on chromosome I in the tomato, Tanksley3L placed Idh-J and Sdh-J near the breakpoints 01 two chromosomes in­volved in a reciprocal translocation in a Capsicum annum x C. chinense hybrid. Re­cently. Quiros and McHale>" also noted the conservation of the Prx·2 and Prx-Jlinkage block between the potato and the tomato since their divergenee. In the potato, a tight linkage exists for these loci (0.6 m.ll.), sim­ilar to the 0.14 m,ll. that separates them on chromosome 2 in the tomato.

The family 865D31 was a cross involving the d ipi oid clones 845 I 0 and 84SD22. which were selections of S. phureja and an 5. lu­berosum x S. chacoense F, hybrid, respec­tively. As a result of allelic differences be­tween these species, 12 isozyme loci, along with morphological traits such as flower color and tuber flesh pigmentation. seg­regated in this cross. Considering the lim­ited number of linkages observed between these loci, a large number of the chro­mosomes probably were tagged. Wi th seg­regating families of this nature, it now may be possible to utilize these markers to gain insight into the genetic bases of important agronomic traits in the potato.

Ref<:reu~es

1 Cardy, B J, C W. Stuber, and M, M, Goodman Techniques for slarch gei eiectrophoresis of enzyme!

lrom maize (lea mays L) InSI Slat. Mineo Series 1317, North Carol "a State Universlly, RaleLgh, J981

2 Dejong, H Tnll':"lan~e 01 rnsselmg m cuil,vated dlplo'd potatoes Potato Re~ 24 :309-313, 1981,

3. Dejong, H" and P, R. Rowe. Genelic markers 'n inbred clones of cullivaled diploLd potalOes Potato Res, 15:200-208, 1972­

4, Oesborougb, S L Potato (Solonum tu/x'rnsl1m L) In Isozyme, in plant genelics and breedmg, Pari B (S 0, Tanksley and T J, Orion. eds) Els€\~er, New York, 1983. PI'. 167-188

5, Dodds, K S., an(l D H Lon'! Th,' mhefltance oj color in diplntd polalocs. 1 Types of anthocy,\l' LdJns and their gendLc ioCl J Genet. 54: 136-149. 1955

6. Gottheb. L D Cons",rvatJon "Il<l duplLcatwLl of iSO­zymes '" plants, SClence 216.0,3-380, 1982

7, Grun, P, Cyloplasrr",· steriiltles that separale the cultivated potato from ,ts putat,ve dlplold ancestors Evolution 24 7S0~ 753, 1970

I; Hermsen, J G T. New approJches 10 hree<lmg for Ihe pOlalo for the year 2000 In Proceedings 01 Ihe InlemallOIlal Congress "Research lor the Year 2000" \'1<' ), Hooker, ed) ClP. Lima, Peru, i98:1. PI' 29-32

9, Hermsen, J G T, M 5 Ramanna. an,] .I Vogel. The localiun of a re<:essrve gelle for chlorophyll delecll'" cy In dlpioid Sot~num tuberosurIJ by LlJeM'S oj tflsomLC analysis Can J Genel CytQi 1~807-813. 1973

10. Hermsen, J G T , and J Verdenius SelectIOn Irom Solan urn luberosum grou p Phuwju of genor)' I"'S ,·om· bin,ng high frequen~y haploLd mductiOll Jor embryo spot Euphyl,ca 22:244-259, 1973,

11, Iwanaga, M , and S J Peloquin Synapl ic mutant "fle~ting only megasporogenesLs m polatoes, J H~red.

70:3-85-3-89, 1979,

12, Kessel, R" and P, R, Rowe Inhenlance ollwO qual­,Iallve lralts and a proposed genetlC map Jor their lmk­age group in d ,ploid potaloes, Potato Res 17:283-295, 1974,

13. Lam, S.. and H T ErJ~kson LocalLon 01 a mutant geLl" ~ausmB alb'msm Ln a d Lploid potato, J, Hered, 62:207-208, 1971.

14, Marll nez-Zapaler, J M, and J 1. 01 ivie r. Genetic analysis of isozyme 10cl in telraploLd potatoes (So/u­num luberosum L,) GenellCs 108:669-679, 1984

15. Martinez-Zarater, J M" and J Olivier Idenlifi­calion of potato varieties, An isozyme approach, In Solanaceae: Biology and systematics (W G D'Arcy, ed ). Columbia UniverSity Press. N~ York, 1986, pp 457-467

16 Masson, M, Mappmg. combLnlflg a b,1 LtLes, herita­bilities. and helemsi! wilh 4x-2-< (ros~es 11\ pOlato, Ph. D Dissertation, lJniv 01 Wisconsin, Mad 'son, 1985

17 Mend Lburn, A 0, and S. J, Peloquin Gene-centro· m<:,r~ mapping by 4.>:-2-< matings in potatoes Theor Appl Genel '15:21-25, 1979.

18 Olrvier. J L.. and J. M Martlflez-Zapater A genelic classl!icalion 01 pOlalo cuitl\'MS base<J Oil allozyrne p.alterns Theor Appl Genet 6930.0-3 ii, 1985

19, Pavek, J, J, Sluthes on the rUSSel-skin character in polalo Progress Report to I he North Central RegiOn &l Polalo Genetics Technicai (Ommlttce Me"tmg Chi­cago, December 2-3, 1980

20, Qu trOS, C. F Starch gel "'. 'Clrophol'!:'SlS tee hniques used with alfalfa and other Mec1i<:ogo speCies Can J. Plant Sc' 61 74S- 749, 1981,

21, Quiro,. C F. Al fal fa, I" Icern~ (MedicQgo satwo L ) In iso")' Ines ,n piant gen~tKS and breeding. Part B. (5 D, Tanksley and T J Orton, eds.) Elsevier. New York, 1983, pp, 253-294

22. QUirOS, C I', and N /o,kHale Genet,c anaiysls of Isozyme vananls in dip 'OLd and tet raploid potaloes, Genellcs 111 :\31-145. 1985.

23 Rick. C M" an<l J F Foks Association 01 an al­lozyme w,th nematode reSlS!ance Rep. Tomato Genel Coop, 24:25. 1974

Douches and QUiroS· Isozyme LOCI I~ Tuber-Beaflng Solanum, 383

24. Rick. C. M.. S. D. Tanksley. and J. F. Fob .5. A p.<eu­dodupJication in Lycopersicon p/mpmellifolium Proe. Nat!. Mad Sci. USA 76:3435-3439. 1979.

25. Salaman. R. N Potato varieties. Cambridge Uni­versity Press. Cambridge. 1926_

26 Sanford. J c.. N. F. \ eeden. and Y. S. Chyi. Re­garding the novelty and breeding value of protoplast­derived variants of Russet Burbank. Euphytica 33:709­71 .1984.

27. Staub. J E. L. J Kuhns, P. Grun. and 8. May. Genetlc basis for isozyme variation lor alkaline phosphate and lucosephosphate isomerase in Solanum. Theor ....ppl

Genet. 67:505-513. 1984.

28. Stelly. D. M.. and S. J. Peloquin. Diploid female gametophyte formation in 24-ehromosome potatoes: Genetic evidence for the prevalence of the second mlliotic division restitution mode. Can. J. Genet eytol 28:101-108. \986

29 SUIter. K A.. J F. Wendel. and J S Cran~ LlNKAGE­I A PASCAL computer progr<lm lor the Jetection and analysis 01 gener,c linkage. J Hercd 74.203-204,1983_

30. Tanksley. S D Moll'I'ular markers m plant breed­mg Plant Mol Bioi Rep \ 3-8. 1983,

31 Tanksley. S D. Lin1<age relallonshlps anti chro­mosomallocalions of enzyme-codlllg genes '" pepper, CapSICum annum_ Chrornosorna 89:352-360, 1984

32 Tanksley. S. D.. anu F. Loaiu-FIRueroa. Gamelo­phyt,c sell'lIleompatlhihty IS conlroll:'d by a sLngle major lo(.us on chromosome I In Lycopersrcon peru­UIOllum Proe Nall Acad. Sci. USA 82:5093-5095. 1985.

33 Tanksley. S D.. H Medllla-Filho, and C. M. RIck. The efleet o( Isozyme seleclJon on metTle characters In an Interspecific backcross of tomalo-basls of an earlyscreenmgprocedure Theor App!. Gelid. 60:291­296, 1981

34 Tanksky, S D. H Medina-filho. and C. M. Rick. Use o( nalura,lIy-o<:c\lrTlng {',,",yme <'anation to deteel and map genes controlling 'luanlilahve ITa lIS III an mlerspecific backcross of tu",alo Heredity 49:) 1-25, 1982

~S Tanksley. S. D.. and C. M. RIck. Isoryme gene hnk· age mal' 01 tl,,' tomalO Apphc<llions in genehcs and brel;<Jing Th.eor Appl Genet. 5; I!) 1-170, 1978

36 Vallejos, C E. Enzyme actIVity Slalnmg In Isozymes in plant genellcs and bre('lling. Vol A (5 D. Tanksley and T J. Orton. eds.). El'evier. New YOrk, 1983, pp ~69-516

:>7 Wagenvoort. M. LocatIon of till' recessl\" gene ym (yellow margin) on chromosome 12 of dIploid Solo­tlum tu/)erosum by means 01 trtsomic analysis. Theor_ Appl Genet 6) :239-243. 1982

384 The Journal of Heredity 1988:79(5)