Table. 8. Segregation of plant populations resulting from the two test-crosses presented in Table 7

1+ stands for heterozygote or dominant phenotypes, - for homozygote or recessive phenotypes.
Ss stands for semisterile pollen, Ft for fertile pollen'

Because the number of cross-overs in the region studied was greatly suppressed (Table 8, cross 1), that sample was insufficient to determine linkage relationships. Therefore, we performed testcrosses analogous to those in cross 1 by choosing from cross 1 plants heterozygotes for all the loci considered and pollinating them with homozygotes for all the markers derived from the line DRC. The progeny from these crosses are referred to as cross 2 in Tables 7 and 8. In cross 2 the distance r-Sca obtained appears to be about twice as large as that for the initial cross WL1479 x DRC x DRC (Table 7). Comparison of the values with the aid of Fisher j-transformation and Student criterion suggests a weak significance of differences at the 5% level, but the values are too small for even j-transformation to make an approach to normality enough for a correct use of the Student criterion.

We obtained 150 plants from the seeds of cross 2. The data on segregation of markers  and the breakpoint for both crosses are given in Tables 7 and 8. A shortage of the markers derived from the line DRC can be noted. The sample of plants generates a somewhat smaller r-Sca distance than was observed in the full sample of seeds, but for the sake of mutual correspondence, we use the former value in the diagram below of linkage relationships obtained from the combined sample of plants of the two crosses. No cross-over event was observed between the breakpoint and the gene His1.

In this cross we faced the following phenomenon: Nine of 428 seeds obtained from the cross 1, and 4 of 247 seeds from the cross 2 (Table 7), had a strange electrophoretic pattern of the SCA protein: both allelic variants, SCAs and SCAf, were present but the former looked twice as abundant as the latter while in normal heterozygotes the variants are equally abundant. None of these seeds germinated in a greenhouse. This phenotype cannot be ascribed to tertiary trisomics, because one dose of Sca^f always comes from the testerline while an unequal resolution of a chromosomal cross in a structural heterozygote, being a female parent, cannot provide a gamete with a double dose of the same Sca^s allele. Such a phenotype could be exhibited by a trisomic resulting from self-pollination. But self-pollination was almost completely prevented in this testcross. Only one clear example of a seed derived from selfing (phenotype R Sca^s) was found among the 429 seeds considered, and it was excluded from our analyses. So, we putatively ascribe these problematic phenotypes to some rare alteration of a relative expression of the Sca alleles and consider these individuals as heterozygotes.

Both series of crosses demonstrate the same trend of a progressing shrinkage of recombination distances between markers in a following order: normal karyotype; translocation breakpoint on one side of the markers; and breakpoint between the markers. Thus, one can see a reduction of the distance cri-gp from 11.9% to 4.4%, to almost zero in a sequence of three crosses between the same testerline Yellow Crispa and WL110 (no translocation), WL1718 (Nilsson translocation, with a breakpoint on one side of these markers), and HT-2 (Hammarlund translocation, with a breakpoint between these markers), respectively. Similarly, in the series of crosses involving the line DRC the recombination length of the r-Sca segment decreases from 27.5% (cross with WL110), to 8.4% (cross with WL1393—breakpoint outside the segment), to 3.9% (cross with WL1479—Winge translocation with a breakpoint within the segment). This regularity suggests that cross-over rates are suppressed near translocation breakpoints. It would be of interest to learn to what an extent this effect varies in different translocations, but this is a matter of special studies involving a chain of closely set markers in relevant chromosomal regions, which are still missing in pea.

An interesting point is that we have two translocations with breakpoints that cosegregate with loci coding for histone H1. Here we find a perfect association between His1 and the Winge translocation, and Gorel et al. (5) report cosegregation of His(2-6) and the Hammarlund translocation breakpoint was observed in a sample of 800 families. One of such loci, His(2-6), is a cluster of five tightly linked genes (10). The large sample allowed us to identify a case of cross-over within the His(2-6) cluster, known to be about 1cM long (10) . This recombination event provided an exchange H1 haplotype 2123 on a normal karyotype, suggesting that the breakpoint is very tightly linked to the particular histone gene His6 within the cluster. The cluster His(2-6) may contain repetitive DNA sequences which could somehow facilitate chromosome breakage. In addition, this cluster is located close to the centromere. However, the locus His1 seems to contain a single histone H1 gene and is located far from the centromere, although there may be some structural similarity of its chromosomal surroundings to those of His(2-6) which could have facilitated the origin of Winge translocation.

In two crosses, WL1718 x Yellow Crispa and WL1393 x DRC, we applied a maximum likelihood method for evaluation of recombination distances between markers and a translocation breakpoint in F2 populations. The breakpoints were followed by pollen fertility (reflecting structural homozygosity) versus semisterility (structural heterozygote). Equations developed for the two cases of co-dominant and dominant markers are presented in Kosterin et al. (11). The former is simple, the latter can be easily solved numerically by a computer using the program T-BREAK.

In the cross WL1718 x Yellow Crispa we have obtained a map segment where the three distances (cri-gp, cri-Bp, gp-Bp) correspond well to each other. In the cross WL1393 x DRC the distances between the breakpoint, on the one hand, and the markers, on the other hand, (above the bold line) suggest the same gene order, but appeared quite different: the long distances were much longer and the short ones were shorter. This discrepancy may be ascribed to the fact that the maximum likelihood method is sensitive to deviation from the implied models of single gene segregation (8). In the F2 seeds resulting from our cross, there is a substantial shortage of the alleles Sca^f and, especially, r, which came from the line DRC (Table 4). This shortage could be explained by the presence in the vicinity of the allele r of some factor lowering the competitive ability of gametophytes. This imbalance is even stronger in the population of plants grown from them (Tables 5 and 6) and may affect differently the results obtained with the maximum likelihood method for different models of inheritance, marker-marker and marker-breakpoint joint segregations. Nevertheless, in spite of the discrepancies between breakpoint-marker and marker-marker distances, they suggest the same order of the genes and breakpoint.

Finally, in a paper by Hall et al. (6) the position of the breakpoint of a translocation in the line JI61 (which is considered to be equivalent to L83) is as follows:

This result, in combination with the data from a second communication of the same authors (7), suggests the order:

—Fs——5S/1—Gp—Vc2——Tl—R—

or:

1. Berdnikov, V.A., Gorel, F.L., Bogdanova, V.S., Kosterin, O.E., Trusov, Y.A. and Rozov, S.M. 1999. Genetical Research 73: 93-109.
2. Berdnikov, V.A., Gorel, F.L. and Kosterin, O.E. 1999. Pisum Genetics 31:1-4.
3. Blixt, S. 1972. Agri. Horrtique Genetica 30: 1-293.
4. Folkeson, D. 1990. Hereditas 112: 249-255.
5. Gorel, F.L., Kosterin, O.E. and Berdnikov, V.A. 1999. Pisum Genetics 31:5-8.
6. Hall, K.J., Parker, J.S., Ellis and T.H.N. 1997. Genome 40: 744-754.
7. Hall, K.J., Parker, J.S., Ellis, T.H.N., Turner, L., Knox, M.R., Hofer, J.M.I, Lu, J., Ferrandiz, C., Hunter, P.J., Taylor, J.D. and Beird, K. 1997. Genome 40:755-769.
8. Immer, F.R. 1930. Genetics 15: 81-98.
9. Kosterin, O. 1993. Pisum Genetics 25: 71.
10. Kosterin, O.E., Bogdanova, V.S., Gorel, F.L., Rozov, S.M., Trusov, Y.A. and Berdnikov, V.A. 1994. Plant Science 101: 189-202.
11. Kosterin, O.E., Pukhnachev, N.V., Gorel, F.L. and Berdnikov, V.A. 1999. Pisum Genetics 31:33-34.
12. Lamm, R. 1949. Hereditas 35: 203-214.
13. Lamm, R. 1974. PNL 6: 29.
14. Lamm, R. and Miravalle, R.J. 1959. Hereditas 45: 417-440.
15. Snoad, B. 1966. Genetica 37: 247-253.
16. Weeden, N.F., Ellis, T.H.N., Timmerman-Vaughan, G.M., Swiecicki, W.K., Rozov, S.M. and Berdnikov, V.A. 1998. Pisum Genetics 30:1-4.

---