ABSTRACT

Pedigree breeding has been widely used to improve agronomic quality in oilseed rape varieties. However, it has often resulted in poor performance due to the undetected introgression of segments containing undesirable inferior genes adjacent to the alleles of potential benefit. Substitution lines that contain isolated chromosome segments of donor genotype in the genetic background of an elite variety allow the detection of beneficial donor alleles separated from undesirable portions of donor genotype. A marker accelerated backcrossing has been used to produce a set of chromosome segment substitution lines in Brassica napus. The development of substitution lines is an extremely powerful technique for evaluating the potential benefit of wild germplasm to crop breeding programs and for accelerating the incorporation of novel alleles into crop varieties. Substitution lines are valuable for accurate assessment and mapping of QTLs, improving the study of quantitative traits (such as yield, glucosinolate and erucic acid content) in replicated environments. It will be possible to detect significant phenotypic differences of each substituted segment. Mathematical models can be developed to improve the design of marker-accelerated backcrossing programs. In addition, high resolution mapping of genes controlling the quantitative traits will then be applied.

Introduction

Oilseed rape (Brassica napus L.) is a crop of major economic importance, grown primarily for its oil in wide areas through the world. In terms of breeding programmes, breeders have put considerable effort in the improvement of its qualitative and quantitative traits such as yield, oil and protein content and reduced glucosinolates (Thompson and Hughes, 1986). Most of the important agronomic characters are controlled by several genes referred to as quantitative trait loci (QTLs). They exhibit continuous variation and are influenced by the environment. Although their biological role is poorly understood (Mohan et al., 1997), they can be effectively manipulated using marker-assisted breeding. Paterson et al. (1991) established the utility of RFLP markers in the improvement of quantitative traits.

Through the use of RFLP maps (Lander and Botstein, 1989) have become available to identify and localize the genetic factors contributing to the traits showing quantitative variation and to estimate their effects. In general, it is difficult to determine the precise location and gene action of individual QTLs. In segregation populations QTLs segregate simultaneously and the effect of the genetic background cannot be controlled. The problem can be overcome by construction of substitution lines, each line containing a single segment from an exotic donor genotype isolated in an otherwise constant commercially elite genetic background.

1.1 Implications in modern plant breeding

Crop domestication and the development of modern plant breeding methodologies over the past century has usually produced highly productive varieties. As a result from crosses amongst genetically related modern species, genetic variation of the crops has been reduced (Tanksley and McCouch, 1997). Wild and unadapted germplasm is a likely source of new and valuable genes to increase the genetic variation and improve traits important to humans. With the advent of molecular linkage maps in plant breeding, the introgression of desirable traits has been greatly enhanced. It allows selection for beneficial genes and against all the deleterious ones.

Pedigree breeding has been used to develop improved oilseed rape varieties. However, agronomically inferior genes adjacent to the genes of interest often reduced agronomic performance of elite cultivars derived from introgression of exotic donor germplasm into established varieties. This was also case of "Tapidor", double low quality variety, which was selected from a cross between “Bienvenu” and a low glucosinolate donor parent (Sharpe and Lydiate, 1997).

Marker-assisted backcrossing together with selection used to produce a set of true breeding substitution lines can effectively evaluate exotic germplasm and eliminate portions of all unwanted donor genotype from beneficial alleles. We have produced such lines, where each line will contain an individual segment of "Bienvenu" genotype in a pure "Tapidor" background and these lines will allow the accurate calculation of the mean effects of each substituted tract.

Plant Material

Parents used in generating F₁ population were two oilseed rape winter varieties "Tapidor" and "Bienvenu". The recurrent parent was known as TapDH1, which was a doubled-haploid derivative of a "Tapidor" microspore-culture (Howell et al., 1996). TapDH1 were employed to pollinate individuals from all subsequent backcross generations

RESULTS

Genetic analysis in the first backcross population

The methods and RFLP probes used were as described by Sharpe et al. (1995). Initially, 162 pre-selected Brassica clones were used to assay polymorphism in B₁ individuals from "Tapidor" x "Bienvenu" cross, of which 64 were informative. The probed DNA was digested by five different restriction enzymes. An RFLP map, comprising 77 polymorphic loci detected by 64 informative probes was constructed, based on 60 B₁ individuals (Sharpe and Lydiate, 1997). The map has identified 15 residual segments where "Tapidor" differs from "Bienvenu", covering approximately 30% of the genome. These segments have been introgressed into "Tapidor" during pedigree breeding from inferior variety "Bronowski", the primary low glucosinolate donor. Some of these segments were probably selected, because they include alleles for low seed glucosinolate or other desirable characters, but many of them reduced productivity.

Three loci influencing the amount of seed glucosinolates have been mapped by QTL analysis (Howell et al., unpublished data) to linkage groups N9, N12 and N19. Together they control approximately 90 % of the variation of glucusinolate levels.

The successive backcross populations

In successive backcross generations, the marker assisted selection consisted of a number of criteria, including the inheritance of a large proportion of the recurrent parent genome and the inheritance of relatively large, single and compact donor segments. The selected individuals with the inheritance of a small number of crossovers were preferred.

The number of introgressed segments has been steadily reduced from 5-10 present in B₂ families, through 1-6 present in B₃ families to a single segment in subsequent generations. Lines heterozygous for a single "Bienvenu" segment have been recovered after three or four rounds of backcrossing with marker-assisted selection followed by one round of self-pollination to "fix" the desired genotypes as homozygotes, with RFLP analysis of the resulting progeny.

Evaluation of lines

The lines containing homozygous substituted segments of donor genotype have been multiplied and are being assessed in field trials alongside "Tapidor", their recurrent parent. Trialing is conducted at four independent field sites within Europe, in two consecutive growing seasons. It will be possible to detect significant phenotypic differences of each substituted segment. A full set of 27 substitution lines will be ready in time for multiplication in autumn 1999 (Fig.1). Together they are spanning the whole donor genotype.

We have obtained the first results from the 1998/99 trial for establishment (Fig. 2). Other characters from this set of lines will be scored in a near future. Statistical analysis of trial data will follow to help to identify the number, position, location and range of magnitude of effects of QTLs for a number of important agronomic characters.

Conclusion

The potential for developing new cultivars in plant breeding is increased by the application of the newly emerging technologies of molecular genetics such as marker-assisted breeding, selection and introgression of genes controlling useful traits. These will enhance possibilities in practical plant breeding in the future as they become more cost effective and accessible to breeders. They will reduce the cost of breeding and selection and greatly increase efficiency.

A set of true breeding chromosome segment substitution lines were produced in Brassica napus, using the old variety "Bienvenu" as a donor and the relatively modern variety "Tapidor" as recipient. The produced substitution lines will provide more uses as a material for commercial trialing and a material for more detailed genetic analysis.

Marker-assisted backcrossing has been used to produce such substitution lines in B. napus (Howell et al., 1996) from a cross between "Tapidor" x "Victor" varieties. In both cases, heterozygous substitution lines were identified in the B₃ and B₄ generations. The assessment of both lines in conjunction will significantly contribute to an understanding of genetic control of quantitative production traits such as yield and stress resistance in oilseed rape. Substitution lines will help to identify and characterise genetic factors controlling quantitative variation. For this reason, intervarietal substitution lines were successfully used in wheat (Law, 1967; Snape et al., 1985). In B. oleracea, where the genome is approximately half size, lines were obtained in the B₂ and B₃ generations (Ramsay et al., 1996).

Cost of the process of producing substitution lines is considerable. To optimize the design of marker-assisted backcrossing programmes, mathematical models will be developed and tested with our experimental data. The result will be both improved design and an improved framework for cost benefit analysis.

Acknowledgements

This work was supported BBSR Agricultural System Directorate and the four breeding companies: CPB/Twyfords Ltd., Nickerson Seeds Ltd., PBI Cambridge Ltd. and Zeneca Seeds Ltd. We also wish to thank Dr. Mike Kearsey and Dr. Malcolm Burns from University of Birmingham for their valuable input into whole project and organization of field trials.

References

1. Howell, P.M. et al. 1996. Towards developing intervarietal substitution lines in Brassica napus using marker-assisted selection. Genome 39: 348-358

2. Lander, E.S. and Botstein, D. 1989. Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121: 185-199

3. Law, C.N. 1967. The location of genetic factors controlling a number of quantitative characters in wheat. Genetics 56: 445-461

4. Mohan, M. et al. 1997. Genome mapping, molecular markers and marker-assisted selection in crop plants. Molecular Breeding 3: 87-103.

5. Paterson, A.H. et al. 1991. DNA markers in plant improvement. Advances in Agronomy 46: 39-90.

6. Ramsay, L.D. et al. 1996. The construction of a substitution library of recombinant backcross lines in Brassica oleracea for the precision mapping of quantitative trait loci (QTL). Genome 39: 558-567

7. Sharpe, A.G. et al. 1995. Frequent nonreciprocal translocations in the amphidiploid genome of oilseed rape (Brassica napus). Genome 38: 1112-1121

8. Sharpe, A.G. and Lydiate, D.J. 1997. The genetic consequences of pedigree breeding in oilseed rape (Brassica napus). Submitted to Genome

9. Snape, J.W. et al. 1985. Genetical analysis of chromosome 5A of wheat and its influence on important agronomic characters. Theoretical and Applied Genetics. 71: 518-526

10. Tanksley, S.D. and McCouch, S.R. 1997. Seed banks and molecular maps: Unlocking genetic potential from the wild. Science 27: 1063-1066

11. Thompson, K.F. and Hughes, W.G. 1986. Breeding and varieties In: Oilseed Rape. Edited by D.H. Scarisbrick, and R.W. Daniels. William Collins Sons and Co. Ltd., London. pp. 32-82.