NORMAL, ILLINOIS

Illinois State University

TUCSON, ARIZONA

University of Arizona

Recombinant inbred pairs (RIPs) as a more efficient RFLP mapping population

--David Weber and Tim Helentjaris

Several public and private labs currently have active RFLP mapping programs in maize. While RFLP mapping, one needs to subdivide the genome into as many compartments or “bins” as possible utilizing a fixed number of individuals in the mapping population. Two types of mapping populations are used almost exclusively for this purpose at this time: 1) F2s and 2) recombinant inbreds (RIs). Both of these populations have important virtues for mapping. In our discussions to identify the most efficient population to use for RFLP mapping, we concluded that a third type of mapping population, recombinant inbred pairs (RIPs), might be a preferable population to use for certain types of RFLP mapping because it would provide resolution that is essentially twice as great as with either F2s or RIs.

We realize that there are differences of opinion about how one should calculate the resolution of mapping using different mapping populations, and the discussion below is designed to compare the relative efficiencies of different mapping populations for developing RFLP genetic maps. Hopefully, these considerations are correct. More importantly, we hope this will stimulate discussion, and we look forward to your comments about this proposed approach.

Most RFLP mapping is carried out with DNAs from F2 plants. Numerous F2 populations are available for immediate use, and each F2 plant contains two different recombinant gametes. Furthermore, each F2 plant can also be selfed to produce a population of F3 kernels which contains all of the alleles that were present in the F2. By planting a relatively small number of these F3s and pooling the DNAs from these plants, one is able to obtain a large amount of DNA which should be enough for numerous studies. These have been described as “immortal F2s or IF2s”.

One can calculate the number of "bins" that are present in a gamete produced by a F1 plant in the following way: 1) 10 bins are present without recombination because 10 chromosomes are present, 2) assuming that the maize genome is 2300 map units long, an average of 2.3 recombinational events will involve each of the 10 chromosomes; therefore, an average of 33 bins (10 + 10 x 2.3) will be found in each gamete produced by an F1 plant. Each F2 plant is formed by the fusion of two F1 gametes; therefore, an F2 plant will contain an average of 2 x 33 or 66 bins.

Mapping with recombinant inbreds in maize has been pioneered by Burr et al. (Genetics 118:519-526, 1988). Recombinant inbred lines are produced by inbreeding the progeny of an F2 for a sufficient number of generations to achieve homozygosity or near homozygosity. Recombinant inbred families, therefore, are a permanent population in which segregation is complete or nearly complete that can be used indefinitely for mapping.

One can calculate the number of "bins" that are present in a recombinant inbred plant using the assumptions given above. Both of the haploid genomes in a F2 plant would contain 33 bins and they are 50% homozygous; therefore, a F3 plant produced by a F2 would contain 33 bins + 16.5 (33/2) informative recombinants per chromosome or 49.5 bins. Each successive selfing reduces the amount of heterozygosity by half; therefore, a F4 plant would contain 33 + 16.5 + 8.25 bins, a F5 would contain 33 + 16.5 + 8.25 + 4.125 bins, F6 would contain 33 + 16.5 + 8.25 + 4.125 + 2.0625 bins etc. so that the number of bins in a recombinant inbred approaches 66.

Thus, the genome is divided into 66 bins in each F2 plant and into essentially 66 bins in each RI plant. In our discussions, we realized that by using the DNA from a pair of recombinant inbreds (RIPs) in each lane on a gel, one would have 132 bins in each sample analyzed. One could either mix DNAs from pairs of recombinant inbreds and place these DNAs in individual lanes or extract DNA from F1s between two different recombinant inbreds; the results would be the same. In this simple way, one would nearly double the number of bins. The analysis of this material could be carried out with the Mapmaker program using an appropriate correction factor.

We believe that if this simple approach were used, the resolution would be essentially doubled using a fixed number of DNA samples for mapping.

Return to the MNL 66 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage