Backcrossing

Introduction

Backcrossing is used extensively to incorporate alleles for novel traits into elite germplasm for a cultivar development program. The novel alleles may be natural mutations or may be the result of mutagenesis or genetic engineering.

The strategy for conducting a backcrossing program is dependent on the type of cultivar being developed.

Pure-line and hybrid cultivars

Chapter 28 in Principles of Cultivar Development provides a thorough discussion of the principles involved in conducting a backcrossing program for plant species in which novel genes are incorporated into elite germplasm that will be used as pure-line cultivars or as inbred lines for hybrids. The goal of a backcrossing program is to recover a pure line or inbred that will contain the novel allele and be as good as the recurrent parent for all other important traits. Table 1 provides an illustration of a backcrossing program at Iowa State University to incorporate a gene for aphid resistance into the soybean cultivar IA3027.

It is common in commercial pure-line and hybrid cultivars to pyramid or stack multiple novel alleles. For pure-line cultivars, the novel alleles must be stacked together. For example, assume that the breeder wants to stack a novel dominant allele for herbicide resistance (H) with a novel dominant allele for insect resistance (I). The backcrossing program must lead to a cultivar with the genotype HHII.

For hybrid cultivars, it is possible to put one novel dominant allele, such as herbicide resistance (H), in one inbred line and a different dominant allele in another inbred line, such as insect resistance (I). When the two inbreds are mated to produce a commercial single-cross hybrid, the F1 plants grown by the farmer would be heterozygous HhIi and have both herbicide resistance and insect resistant.

Clonal cultivars

Heterozygosity is important for the performance of clonal cultivars. Self-pollination or the mating of related individuals results in homozygosity, inbreeding depression, and unacceptable performance. Because backcrossing to a single recurrent parent results in homozygosity, breeders of clonal cultivars generally change the elite parent every backcross generation. The procedure is referred to as modified backcrossing because there is not a single recurrent parent. The backcrossing program results in a population from which individuals with the novel allele and other desirable traits can be identified for subsequent clonal evaluation as potential cultivars.

Synthetic cultivars

Heterozygosity and heterogeneity are major considerations in a backcrossing program involving synthetic cultivars of forage and turf species. As with clonal cultivars, the breeder tries to avoid the inbreeding depression associated with homozygosity. This is accomplished by changing the elite parent each backcross generation. To adequately sample the genetic heterogeneity of synthetic cultivars, as many different plants as possible of the elite parent are used each backcross generation.

For autopolyploid species, such as alfalfa, it is not practical to incorporate a novel allele into every individual plant of a synthetic cultivar. Instead, the goal of the breeder is to have an adequate percentage of plants in a cultivar with the novel allele. For example, a transgene for glyphosate resistance, referred to as the Genuity® Roundup Ready® Gene, has been incorporated into synthetic cultivars of alfalfa. The majority of the plants in the synthetic cultivars will contain the transgene. The seeding rate for the cultivar can be increased by the farmer to account for the plants that will die when sprayed with the herbicide.

Managing a successful and efficient backcrossing program

Step 1. Selection of appropriate donor and recurrent parents.

The donor parent parent selected for the program should have as many traits that are similar to the recurrent parent as possible to minimize the number of backcross generations needed to recover a desirable cultivar. In the illustration provided in Table 1, LD05-15601 was considered by the breeder at the Univ. of Illinois to be the highest yielding line available in his breeding program that had the Rag1 allele. Other traits that could have been considered in choosing the donor parent were seed size and protein content as similar as possible to that of IA3027.

The recurrent parent should be one that will continue to have competitive performance compared with other cultivars when the backcross is completed. The backcrossing program to develop IA3027RA1 began in 2006 and commercial quantities of seed were not available until 2010. IA3027 was used as the recurrent parent because the breeder believed in 2006 that the backcross-derived version with the Rag1 gene would be a competitive cultivar in 2010. It is common for breeders to use as recurrent parents elite experimental lines that are likely to become cultivars in the future. This reduces the number of years between the commercial use of the original recurrent parents and that of the backcross-derived version. The disadvantage of this approach is that the backcross program may have to be discarded if the experimental line used as the recurrent parent fails to be accepted for commercial use.

Step 2. Minimizing the number of backcross and self-pollinated seeds needed.

The use of artificial hybridization to obtain backcross seed can be a limiting factor for some self-pollinated species. Chapter 28 has a table that is very useful for calculating the minimum number of plants or seeds needed to obtain the required number with the allele(s) of interest at the 95% or 99% probability level. Understanding how to make this calculation is critical for managing a successful program with the least amount of resources possible. The calculation requires that the breeder understand “q”, the expected genotypic frequency of individuals with the desired allele(s); “r”, the number of individuals with the desired allele(s) that the breeder needs; “p” the probability of successfully obtaining the “r” number of individuals; and the germination of the seeds when they are planted.

The number of backcross seed required is directly related to the number of alleles that are to be incorporated into the recurrent parent. When two or more alleles are involved, the breeder must decide whether to backcross both alleles individually and combine them together when backcrossing for each is completed or backcross them together in the same program. Less backcross seed is required each generation if individual programs are conducted for each allele; however, an additional season may be required to combine the alleles, if all of them are not in the original donor parent.

A breeder may choose to obtain more than the minimum number of seeds or plants indicated by the calculation in order to select among seeds or plants for genes of the recurrent parent. For example, if only 10 BC1F1 plants are needed for crossing, the breeder may choose to obtain 40 plants, genotype them with molecular markers, and choose the 10 for crossing that have the greatest frequency of markers that match the recurrent parent. This selection may reduce the number of backcross generations needed to recover the desired percentage of the recurrent parent genotype.

Step 3. Minimizing the number of years to complete the backcross

The likelihood that a backcrossed-derived version of the recurrent parent will be competitive is enhanced by utilizing as many seasons per year as possible for crossing. In some cases, this may require that backcrosses are made to plants before it is known if they have the desired allele(s), which means that backcross seed obtained from plants found to be without the desired allele would have to be discarded. Most breeders would prefer to do the extra work in order to reduce the number of years.

The use of multiple seasons a year for backcrossing can result in problems if environmental conditions in some seasons are not highly favorable for artificial hybridization. The program described in Table 1 encountered problems with obtaining hybrid seed in Puerto Rico and with adverse conditions during germination at Ames. It is not uncommon for a breeder to revise a backcross plan to deal with such unforeseen circumstances.

The number of years can be minimized by utilizing the fewest number of backcross generations as possible. This is strongly influenced by the similarity of donor and recurrent parent, the degree to which the backcross-derived version must perform like the recurrent parent, and the testing of the backcross progeny.

With regard to testing of backcross progeny, there are two alternatives.

• Alternative 1: The breeder may choose to grow multiple backcross-derived lines; select those that are similar for phenotypic traits with a high heritability; and bulk seed of the selected lines for release of the new cultivar. The advantage of this strategy is that it requires less time and resources than the second alternative. The disadvantage is that the new cultivar may not perform as well as the recurrent parent for important quantitative traits. In the example of Table 1, a bulk of phenotypically similar lines was made without yield testing. It was found that the bulk did not perform as well as the recurrent parent.
• Alternative 2: For this alternative, replicated tests are conducted of the individual backcross lines, as was described in Table 1, and seed of only those lines with acceptable performance is bulked to form breeder seed of the new cultivar or inbred line.

The choice between the two alternatives is influenced by the experience of the breeder in working with the allele from the donor parent. If it is the first time the breeder has backcrossed with the allele, the decision may be to conduct replicated tests of individual backcross lines to determine their similarity to the recurrent parent for important quantitative traits. Also, replicated testing may be preferred when the donor and recurrent parent differ for multiple quantitative traits, as was the case for the example in Table 1. The donor parent LD05-156021was significantly lower than the recurrent parent IA3027 in seed size and protein content, both of which are quantitative traits. By testing the 30 individual lines, the breeder was able to discard 12 that had lower seed size and protein content than the recurrent parent.

Table 1. Development of the soybean variety IA3027RA1.

Year	Activity
2006	The cross of IA3027 x LD05-15621 was made in Puerto Rico during March to obtain F1 seeds. The objective of the cross was to backcross the Rag1 for aphid resistance from the line LD05-15621 into the cultivar IA3027. IA3027 was chosen as the recurrent parent because it had the highest yield of cultivars with large seed and high protein that are used by the soyfood industry. The donor parent LD05-15621 was developed by the Univ. of Illinois and IA3027 was developed by Iowa State University.
2006	The F1 seeds were planted in the field at the Iowa State University Agricultural Engineering and Agronomy Research Center near Ames, IA, in May. The F1 plants were crossed to IA3027 to obtain BC1F1 seeds.
2006	The BC1F1 seeds were planted during October in Puerto Rico for the next backcross; however, no BC2F1 seeds were obtained because of unfavorable environmental conditions. Leaves were harvested from each BC1F1 plant to identify plants heterozygous for the Rag1 gene based on molecular analysis for a SSR linked to the gene. The heterozygous BC1F1 plants were harvested individually.
2007	The BC1F2 seeds from the heterozygous plants were planted during January in Puerto Rico. Each BC1F2 plant was evaluated with the SSR to identify plants homozygous for the Rag1 gene. Five homozygous resistant plants were backcrossed to IA3027 to obtain BC2F1 seeds. The five homozygous resistant plants were harvested individually.
2007	The BC2F1 seeds and the five BC1F2:3 lines were planted in the field during May in Ames. Due to adverse weather conditions, none of the BC2F1 plants survived. The five BC1F2:3 lines were screened for aphid resistance in the greenhouse. Two lines with scores similar to the donor parent LD05-15621 were crossed to IA3027 to obtain BC2F1 seed.
2007	The BC2F1 seeds were planted during October in Puerto Rico. Each BC2F1 plant was evaluated with the SSR to confirm that they were heterozygous for the Rag1 gene. The heterozygous plants were crossed to IA3027 to obtain BC3F1 seeds.
2008	The BC3F1 seeds were planted during January in Puerto Rico. Each plant was evaluated with the SSR to identify those that were heterozygous for the Rag1 gene. The heterozygous plants were harvested individually.
2008	The BC3F2 seeds were planted during May in Ames. Each plant was evaluated with the SSR to identify homozygous resistant individuals. In August, plants were scored for aphid resistance in the field when a natural infestation of the insect occurred. Plants with aphid resistance and maturity similar to IA3027 were harvested individually.
2008	The BC3F2:3 lines were planted individually in Puerto Rico for seed increase.
2009	The BC3F2:4 seed of each line was planted in the greenhouse during March to test for aphid resistance. A sample of seed from each of 30 BC3F2:4 lines was bulked for yield testing.
2009	a. The bulk of lines was evaluated for seed yield and other characteristics in the Iowa Specialty Test in three replications of four-row plots at five Iowa locations. The 30 BC3F2:4 lines also were tested individually in two replications at each of four locations. b. The BC3F2:4 lines were grown at Ames in a seed increase and each of them was harvested individually. Seeds of 18 lines with similar agronomic and seed characteristics as IA3027 in the yield trials were bulked as breeder seed of IA3027RA1.
2009/10	The variety was licensed to interested companies by the Iowa State University Research Foundation. The companies produced foundation seed in Argentina for planting in the Midwest during 2010.

 

You are going to backcross the B allele for increased beta-glucan content in oat seed from a donor parent into an elite pure-line cultivar with the genotype bb. You can differentiate BB and Bb seeds from those that are bb by analyzing the beta-glucan content of the part of the seed without the embryonic axis and saving the part with the embryonic axis for planting. You cannot differentiate BB and Bb seeds from each other with the test. For any seed you select, the part with the embryonic axis can be used to obtain a plant for crossing or self-pollination.

1. What characteristics would you consider important in selecting your donor and recurrent parent?
2. Which parent would you use as female to produce F₁ seed? Why?
3. You cross the F₁ plants to the recurrent parent. Would you prefer to use the F₁ plant as the male or female in the backcross cross to assure that a plant grown the following season that has the B allele is a BC₁F₁ and not a F₂ resulting from an accidental self-pollination? Would your answer be the same for every subsequent backcross generation when deciding whether to use the recurrent parent as the male or female for the backcross?
4. If you want to be 95% sure of obtaining 10 BC₁F₁ plants with the genotype needed to continue the backcrossing program and your germination percentage is 70%, how many BC₁F₁ seeds would you have to obtain?
5. You test the B₁F₁ seeds for beta-glucan content. What genotypes would you expect to find in the BC1F1 seeds and in what frequency? What would be the genotype of the seeds you select for planting?
6. You cross the selected BC₁F₁ plants to the recurrent parent. If you want to be 99% sure of obtaining 6 BC₂F₁ plants with the genotype you need to continue the backcrossing program and your germination percentage is 60%, how many BC₂F₁ seeds would you have to obtain?
7. You test the BC₂F₁ seeds for beta-glucan content. What genotypes would you expect to find in the BC₂F₁ seeds and in what frequency? What would be the genotype of the seeds you select for planting?
8. You self-pollinate the selected BC₂F₁ plants. What genotypes would you expect to find in the BC₂F₂ seeds and what would be their frequencies? If you want to be 95% sure of having at least 15 BC₂F₂ seeds with the genotype BB, how many BC₂F₂ seeds would you have to test?
9. How many BC₂F₂ plants would you have to grow to be 99% sure of finding at least 10 that have the genotype BB? What would be the average percentage of the recurrent parent in the BC₂F₂ plants?
10. How many progeny would you have to test from each BC₂F₂ plant to be 95% sure of identifying those that have the BB genotype?

 

You are responsible for developing pure-lines cultivars of barley for Minnesota. You have a high-yielding cultivar Ada to which you would like to add a major recessive allele for insect resistance (r) and a major dominant allele for fungal resistance (F). The two alleles are at independent loci. The r and F alleles are in different donor lines. There is a direct SNP marker that can be used to differentiate the r and R alleles. There are no molecular markers yet developed for the f and F alleles.

Outline your breeding program season-by-season for developing a backcross version of Ada with the genotype rrFF in the shortest possible time and with the least amount of work. You have two seasons a year. Season 1 is in the field in Minnesota and season 2 is in the field in Arizona. At both locations, you can do hybridization and selfing in the field. All tests for insect or fungal resistance are done in the greenhouse in Minnesota. You cannot produce enough hybrid seed on a plant to use that hybrid seed for progeny testing. However, you can have progeny test results before flowering occurs. Your laboratory results from molecular marker analysis are available before flowering.

In outlining your program, indicate for each season the number of plants grown, genotype of plants used for crossing, and number of hybrid seeds obtained on each plant. If progeny testing is involved, indicate the number and genotype of plants to be progeny tested and the genotype and number of their progeny that need to be evaluated. Use the following assumptions.

1. Begin the backcrossing program by making the single cross in Arizona
2. Probability of recovering the desired genotype is 0.95
3. Germination is 90%
4. Maximum number of hybrid seed that can be obtained on a plant is 12
5. Maximum number of selfed seed on a plant is 100
6. The new cultivar should have at least 90% of its genetic background from Ada.
7. You will not yield test before producing breeder seed
8. Breeders seed will be produced from a composite of 40 progeny rows that are homogeneous rrFF.