Mapping is a process in which each new locus provides another tool for future studies. As in all species, the initial development of a partial map for pea was tedious and drawn out. Once sufficient markers had been identified to develop a primitive linkage map, the major problem became identifying a technique that would expose simple, genetic polymorphism. Initially, phenotypic polymorphism was the method of choice, first selecting natural mutants and later using mutagens to induce changes in gene expression. As biochemical techniques developed, protein and allozyme polymorphism enabled a number of laboratories to map over 50 new loci coding these proteins. The advent of DNA technology permitted the resolution of all changes in DNA sequence and, thus, any genetic variation that was present between two genotypes.

One of the dilemmas encountered in selecting between DNA-based techniques used in mapping is whether to employ a technique that is based on conserved sequences with relatively low polymorphism or highly polymorphic sequences that cannot be extended to other crosses or genera. With the amplification of specific DNA fragments by PCR becoming the most commonly used approach for the development of markers, a number of laboratories have tried to develop methods that combine polymorphic and conserved sequences. Probably the best example of this combination is microsatellite or simple sequence repeats (SSRs), in which the repeat itself is expected to display a high level of polymorphism, while the primers are designed to be homologous to much more highly conserved flanking regions (5). A second approach has been to design primers to cDNA sequences, and cut those sequences using restriction enzymes with 4-base recognition sites in order to detect internal polymorphism. This latter approach has been called ‘cleaved amplified polymorphic sequences’ or CAPS (4). The SSR approach has proven highly effective within a genus (8), and CAPS has also been very useful, particularly in wide crosses (1). The ideal marker would be a be an SSR intron within a gene. In such a situation the primers could be designed to the conserved regions of the coding sequence; yet the highly polymorphic SSR would still be amplified and any polymorphism resolved by electrophoresis or other technology. Following this logic, we felt that comparative mapping would best be served by a set of primers designed to very highly conserved regions of sequences coding polypeptides with a second condition that the amplified fragment would contain at least one intron. We initiated our effort with the sequence encoding the large subunit of ADP-glucose pryophosphorylase, and the result of this work has been published in a brief communication (9). Here we present data on six additional coding sequences, most of which have not been previously mapped.

Each of the coding sequences mapped to a unique and specific region on the JI1794 x Slow linkage map. The distances are given between the “new” locus and the closest marker that was identified by name on the consensus map (10). However, in all cases, there were markers (usually RAPDs) closer to the coding sequence, and in several cases the RAPDs cosegregated with the amplified sequence.

After cutting with RsaI, the Rpl22 product gave three fragments of which only the larger, 550 to 600 bp fragment was polymorphic. The polymorphism confirmed the position of this gene (as previously determined by RFLP analysis) as being about 6 cM distal to Acp4 (Table 2) on the upper portion of linkage group VI on the consensus map (10).

The plastid-specific fructose bisphosphatase produced a fragment 700 to 800 bp larger than the coding sequence presented in Hahn et al. (2) indicating the presence of at least one intron. The gene containing the coding sequence, Fbpp (for fructose bisphosphatase–plastid specific) mapped approximately 6 cM from Acp1 towards Pgdc. This location would place Fbpp close to Cri.

The primers for phosphoenolpyruvate carboxylase were taken from the first third of the coding sequence in order to take advantage of two introns that were identified in this region of an alfalfa phosphoenolpyruvate carboxylase sequence (6). The primers were designed using a nodule-enhanced cDNA (7). Joint segregation analysis placed Pepc approximately 4 cM from the ISSR marker B812a on linkage group III (Table 2). It is interesting to note that several other nodule-related genes (ENOD 3, ENOD 12A, 12B, and Sym7) are also in the region, although not tightly linked.

The coding sequence for the mitochondrial-specific manganese superoxide dismutase also mapped to linkage group III, albeit on the other arm. The primers closely flank the first large intron (~1260 bp) near the 5’ end of the gene (3). Thus, the amplified product is nearly all intron sequence. Nearly all restriction enzymes tested gave polymorphism between the J11794 and Slow parents, suggesting considerable sequence divergence in the intron.

The two Paal genes are known to be tightly linked to each other (separated by about 6000 bp). We designed primers that would amplify both genes simultaneously and obtained the expected 2720 and 2260 bp fragments. Restriction digests of these fragments gave numerous subfragments, with a number of these fragments polymorphic between the two parents. In the RILs, the parental patterns appeared to segregate as intact units, confirming tight linkage between the two genes and indicating that no other intact Paal genes exist in the pea genome, at least with sequences homologous to the primers used. These genes are located between R and Bt, a region in which markers are relatively scarce. We feel this gene cluster will be valuable for marking this section of the genome not only in various pea crosses but also in comparative mapping among cool season legumes.