QTL Conditioning Yield and Fruit Quality Traits in Cucumber (Cucumis sativus L.)

admin

Cucurbit Genetics Cooperative Report 22:8-10 (article 4) 1999

Arian Dijkuizen
Firma RAGT, Hauptstrasse 2, 94533 Ottmaring-Buchhofen, Germany

Jack E. Staub
Vegetable Crops Research Unit, USDA/ARS, Departments of Horticulture, University of Wisconsin-Madison, WI 53706 U.S.A.

Introduction

Recent molecular appraisals of quantitative trait loci (QTL) have documented the importance of genotype (G) x environment (E) interactions in several crop species (Paterson et al., 1991 ; Stuber et al, 1992). Some of the observed differences in QTL expression have been attributed either to the statistical method used in evaluation of GxE (Dudley, 1993) or to experimental sampling biases (Beavis et al., 1994). Such GxE interactions for quantitatively inherited traits can occur in cucumber (Cucumis sativus var. sativus L.) (Wehner et al., 1989).

A genotype that has been used by cucumber breeders for the purpose of increasing yield in cucumber is C. sativus var. hardwickii (R) Alef. (hereafter referred to as C.s. var. hardwickii), a feral relative of C.s. var. sativus (Horst and Lower, 1978). This genotype possesses a multiple lateral-branching and sequential-fruiting habit not present in C.s. var. sativus lines.

Both RFLPs (Kennard et al., 1995) and RAPDs (Serquen et al., 1997) have been used to identify QTL for yield and quality in cucumber. However, the effects of growing environment on the action of QTL conditioning cucumber fruit yield and quality has not been assessed using molecular marker technologies. Therefore, we designed an experiment to identify QTL affecting yield and fruit quality traits in progeny derived from a cross between C.s. var. sativus and C.s. var., hardwickii.

Methods

An F2 mapping population was constructed by crossing the gynoecious C.s. vr. hardwickii accession PI 183967 (P2). Parental matings produced seed from which a single F1 plant was selfed to obtain an F2 bulk population, and F2 plants were subsequently self-pollinated to produce 200 F2S1 families. In addition, 60 F3 plants (different from those used to derive the F2S1 families) were backcrossed to both parents to generate BC1(P1) and BC1(P2) families for a North Carolina Design III evaluation.

The F2S1 families were evaluated for fruit yield and quality components in 1991 and 1992. Experimental units (plots) consisted of 2.1 m rows positioned on 1.5 m row centers. Individual plots were over-seeded and thinned to9 or 18 plants to obtain the desired planting density of 29.000 or 58,000 plants ha-1, respectively. Backcross families were evaluated in 1992, using a randomized complete block design with four replications. Plots consisted of .5 m single-row plots set on 1.5 m row centers. Plots were over-seeded and thinned 13 plants per plot (~44,000 plants ha-1 ).

Data were collected on days to anthesis, fruit number and weight, and fruit length and diameter (L/D). Best linear unbiased predictions (BLUPs) for F2S1 and BC family means per environments were calculated using PROC MIXED (SAS Institute, 1992).

Variance components and standard errors (S.E.) associated with F2S1 families ( ơ 2g), F2S1 family x year interaction ( ơ 2gy) or family x density ( ơ 2gd), and residual variances ( ơ 2), and their standard errors (S.E.) were calculated using PROC MIXED (SAS Institute, 1992). QTL were identified using an interval approach described by Lander and Botstein (1989) using the computer program, MAPMAKER-QTL (Lincoln and Lander, 1990).

Bulked leaf tissue from each F2S1 (utilized for genotyping F2 individuals) and BC family was collected, extracted, and Southern blot hybridizations were performed according to Kennard et al. (1994). DNA was digested with Dra1, EcoRI, EcoRV, or HindIII (BRL, Gaithersburg, MD or Promega, Madison, WI). Digested DNA was electrophoresed, gels were stained in ethidium bromide, and DNA was transferred to Zetaprobe membranes (Biorad, Waverly, MA) according to (Sambrook et al., 1989). Previously identified cloned DNA fragments showing polymorphisms between the two parents were radio-labeled by random hexamer priming.

Results

Earliness, and fruit yield and quality components of cucumber were investigated by examining cross-progeny (BC and F2S1) derived from a wide mating [gynoecious cucumis sativus L. var. sativus line GY 14 x monoecious C. sativus var .hardwickii (R) Alef. PI 183967] (Table 1). A molecular marker map constructed from F2 individuals was used to identify quantitative trait loci (QTL) for each trait examined, and to assess the consistency of QTL over years (1991 and 1992) and planting density (29,000 and 58,000 plants ha -1), QTL affecting earliness (days to anthesis and number of barren nodes), fruit yield (fruit number and weight at two harvest times) and shape [length (L), diameter (D), and L:D ratio] were identified. The traits examined were less affected by planting density than by year. While earliness and yield traits were largely under non-additive control, components of fruit shape exhibited additive genetic variance resulting in high values for narrow sense heritability estimates. While the number and map location of some QTL was relatively consistent over environments (years and planting density), differences in their number and location were found in F2S1 and BC families. Some of these differences could be attributed to disparities in population size, dominance and the amount of genotypic information available ( F2S1 BC). Fruit L and D, and to a lesser extent L:D ratio, are developmentally dependent, and thus map placement of QTL was affected by the physiological stage of fruit development. QTL evaluation of the F2S1 generation revealed that earliness is determined by relatively few genes, and that the genetic control of early yield resides in the same chromosome regions as does days to anthesis. Positive genetic correlations were identified when plants of similar physiological age were compared at different harvest times in each of the environments (years) and genetic backgrounds ( F2S1 , BC1P1 and BC1P2 families) examined. Thus, these factors which should be considered when assessing C. sativus var. hardwickii-derived germplasm and QTL profiles in cucumber.

Table 1. Multi-QTL models for cucumber traits evaluated at each of three environments (29,000 and 58,000 plants ha-2 in 1991 and 58,000 plants, ha-1 in 1992), final LOD ratio (LOD), and proportion of variance explained (R2) by the model.

Linkage

Group

 1991 F2S1 1991 F2S1 1992 F2S1
29,000 ha-1 58,000 ha-1 58,000 ha-1
Marker x l a d Marker l a d Marker l a d
Anthesis (days)
B F 24 1.9 -0.6 F 24 1.9 -0.1 No data
E CsC029 65 3.9 -3.6 CsC029 65 3.6 -3.4
F Per 11 1.4 0.4
LOD = 10.0, R2 = 0.41 LOD = 10.0 R2 = 0.46
Fruit number at harvest (10,000 ha-1)
A CsP357 32 3.1 -1.9
B CsP287 67 2.8 9.4 CsP193 0 -2.4 9.6 F 24 24 -7.7 1.9
E CsC613 45 -5.9 11.2 CsC613 50 -5.4 11.0
E CsC029 66 -8.3 2.7
LOD = 7.9, R2 = 0.33 LOD = 5, R2 = 24 LOD = 15.1 R2 = 0.56
Fruit weight at harvest (Mg Ha-1)
B F 21 -2.3 4.4 F 20 -3.4 4.3 F 21 -4.8 0.6
B CsP024 60 -3.2 -2.6
E CsP215 44 -7.3 6.1
E Pep_pap62 -6.2 2.9
E CsC029 68 -5.6 -0.8
F CsC443 7 -2.9 0.2
F CsP130 3 -3.3 0.2
LOD = 9.7 R2 = 0.38 LOD= 10.4 R2 = 0.42 LOD=22.5 R2 = 0.69
Fruit L:D ratio (cm cm-1 ) at late harvest
A dm 70 0.06 0.05
B CsP024 52 -0.09 0.01 CsP024 59 -0.10 0.03
C CsP056 42 -0.16 0.00 CsP056 42 -0.16 -0.03 CsP056 38 -0.14 -0.01
D CsE060 8 -07 -0.01
E CsP211 37 -0.11 0.01
E CsP475s30 -0.9 -0.01
G CsP280 8 -0.07 0.00 CsP280 11 -0.9 -0.03
H CsC166 23 -0.06 -0.08

x Nearest (marker) to the putative QTL, map location (l) in cM, estimates of average effect of substitution of a GY14 allele by a Pl 183967 allele (a), and dominance deviation (d).

Literature Cited

  1. Beavis, W.D., O.S. Smith, D. Grant and R. Fincher. 1994. Identification of quantitative trait loci using a small sample of topcrosses and F4 progeny from maize. Crop Sci. 34:882-896.
  2. Dudley, J.W. 1993. Molecular markers in plant improvement. Manipulation of genes affecting quantitative traits. Crop Sci. 33:660-668.
  3. Horst, E.K. and R.L. Lower. 1978. Cucumis hardwickii: A source of germplasm for the cucumber breeder. Cucurbit Genet. Coop. Rpt. 1:5.
  4. Kennard, W.C.., K. Poetter, A. Dijkhuizen, V. Meglic, J.E. Staub and M.J. Havey 1994. Linkages among RFLP, RAPD, isozyme, disease-resistance, and morphological markers in narrow and wide crosses of cucumber. Theor. Appl. Genet. 89:42-48.
  5. Kennard, W.C. and M.J. Havey. 1995. Quantitative trait analysis of fruit quality in cucumber: QTL detection, confirmation, and comparison with mating design. Theor. Appl. Genet. 91:53-61.
  6. Lander, E.S. and D. Botstein. 1986. Mapping complex genetic traits in humans: New method using a complex RFLP linkage map. Cold Spring Harbor Symposia on Quantitative Biology 51:49-62.
  7. Lincoln, S.E. and E.S. Lander. 1990. Mapping genes controlling quantitative traits using MAP-MAKER/QTL. Whitehead Institute for Biomedical Research, Cambridge, MA>
  8. Patterson, A.H., S. Damon,J.D. Hewitt, D. Zamir, H.D. Rabinowitch, S.E. Lincoln, E.S. Lander and S.D. Tanksley. 1991. Mendelian factors underlying quantitative traits in tomato: comparison across species, generations, and environments. Genetics 127:181-197.
  9. Sambrook, J., T. Maniatis and E. Fritsch. 1989. Molecular cloning manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.
  10. SAS Institute. 1992. Technical report P-229 SAS/Stat software changes and enhancement. SAS Institute, Cary, NC
  11. Serquen, F.C., J. Bacher and J.E. Staub. 1997. Mapping and QTL analysis of horticultural traits in a narrow cross in cucumber (Cucumis sativus L.) using random amplified polymorphic DNA markers. Mol. Breeding. 3:257-268.
  12. Stuber, C.W., Lincoln S.E., D.W. Wolff, T. Helentjaris and E.S. Lander. 1992. Identification of genetic factors contributing to heterosis in a hybrid from two elite maize inbred lines using molecular markers. Genetics 139:823-839.
  13. Wehner, T.C., R.L. Lower, J.E. Staub and G.E. Tolla. 1989. Convergent-divergent selection for cucumber fruit yield. HortScience 24:667-669.