Pisum Genetics   

Volume 25

1993

Research Reports

pages 23-26

Genes a and d may not be in the same linkage group

Kosterin, O.E.

Institute of Cytology and Genetics
Novosibirsk 630090, Russia

Recently, the gene His7, coding for the histone H1 subtype with the highest electrophoretic mobility, was mapped in linkage group 1 of pea (4). The following arrangement of genes was proposed:

Since that paper was written, the scheme was corroborated in several additional crosses. According to the hitherto constructed pea linkage maps (1, 6, 7), the gene His7 should be located fairly close to gene d (maculum). In order to reveal the arrangement of genes His7 and d, I carried out several crosses. Surprisingly, the two genes segregated independently in all populations examined.

First, a testcross (WL1393 x WL1688) x WL1393 was made. Line WL1393 has genotype d, allele 2 of gene His7, and haplotype 1323 of the cluster His(2-6) of the linked histone H1 genes (4). Line WL1688 has allele Dco, allele 3 of His7, and haplotype 1121 of His(2-6). The progeny of the testcross segregated as follows: 47 Dco His73/2 : 41 Dco His72 : 39 d His73/2 : 49 d His72; these data give a recombination fraction of 45.4 ± 3.8% but the deviation from 50% is not significant (P > 0.2). However, line WL1393 turned out to carry a chromosome rearrangement and, because this work was done on seedlings, I had no opportunity to check this progeny for semisterility. To exclude the possibility that the rearrangement of WL1393 resulted in the decoupling of genes His7 and d, three other testcrosses were made between parents known to have a normal karyotype. The descriptions of these crosses are given in Tables 1, 2, and 3. In all three crosses no linkage was found between d and either His(2-6) or His7.

Taking into account that His7 is located about 26-27 cM from gene a in the direction of the putative position of d, (4) these results imply that genes a and d would also not be expected to exhibit linkage. However, the widely accepted assignment of a and d to the same linkage group appears to be based primarily on direct segregation data. Lamprecht (5) tested the joint segregations of a and d in numerous crosses. In some of the crosses the genes segregated independently, whereas in others a weak linkage was detected. Gene d is hypostatic to a, so, the phenotypic class a gave no information on gene d. In order to calculate the recombination fractions and their standard errors by the product ratio method, Lamprecht deliberately divided the a class into the classes a D and a d, letting the former to be equal to the reciprocal class A d. For example, the segregation data 258 A D : 93 A d : 103 a (repulsion) became 258 A D : 93 A d : 93 a D : 10 a d; and 215 A D : 125 A d : 101 a (repulsion) became 215 A D : 125 A d : 101 a D : 0 a d ! In the latter case the entire class a was less than A d. This procedure introduces artifactual information not resulting from the experiment, and hence reduces the standard errors. For this reason a "significant" linkage was detected in several cases where the correct calculation does not reveal it. Thus, for the former of the above examples the calculation by the method of maximum likelihood for the case of cryptomery provides a recombination value of 45.23 ± 7.92% instead of the reported 32.4 ± 3.98%. The maximum likelihood estimation of the recombination value for the second example is zero with an infinite standard error, implying that the segregation provides no information on linkage. The processed data, with one zero class, in this case could also not be treated by the product ratio method, and the reported value of 9.1 ± 4.71% was obtained by substituting 1 in place of 0.

When the data reported in (5) were recalculated by the method of maximum likelihood, all but one of the crosses gave recombination values which deviate from 50% by less than twice their standard errors. The exceptional cross 684, which gave a recombination fraction of 23.5 ± 3.7%, might involve some chromosome rearrangement. Nevertheless, the combined data for all crosses, including those published prior to (5), seem to provide evidence for a very weak but significant linkage. The total segregation data for all coupling crosses gives a recombination value of 40.55 ± 2.06%; that for repulsion phase crosses gives 37.8 ± 2.46%. However, the combined data should be considered with caution, since any aberrant cross can bias the whole sample.

In the same paper (5), Lamprecht reported data on gene au (aurea) suggesting its location between a and d. Two crosses were reported with the following segregations for classes Au A, Au a, au A, and au a: 194:85:86:0 (repulsion) and 179:15:15:36 (coupling). Since au plants probably could not be classified for a as they die before flowering, it appears likely that the au class was artificially divided into au A and au a classes, in the same way as in the case of cryptomeric genes a and d. The proper estimations of the recombination values are 28.03 ± 17.5% and 12.36 ± 3.37%, respectively. The latter case suggests a significant linkage between a and au. Moreover, for the latter cross (Number 880) the segregation 165 Au Dco, 69 Au d, 57 au Dco, 0 au d (repulsion) was also presented. It should be noted that the au class could not be classified for d because alleles au and a were in the coupling phase, therefore, these data should be treated by the cryptomery model. (Besides, the Au class is expected to have contained about 8% of a plants, which were probably added to the class Au d). The maximum likelihood estimation of the recombination fraction for the above segregation is 34.0 ± 12.2% (instead of the reported 14.1 ± 5.72%). This result does not provide evidence for the linkage of genes au and d and thus does not prove the presence of genes a and d in the same linkage group.

Table 1. Segregation for genes d, His(2-6) and His7 in the progeny of testcross A1.

His(2-6)

1123/1221

 

1221/1221

His7

2/2       2/3

 

2/2       2/3

Dco

24        11

 

10        37

d

23        10

 

14        29

 

Recomb. fract.

Joint c2

Probability

His7 – d

53.8 ± 4.0%

0.83

P > 0.3

His(2-6) – d

49.4 ± 4.0%

0.01

P > 0.9

His(2-6) His7

28.5 ± 3.6%

28.21

P << 0.0001

1 Line RT-1 (Dco, His(2-6)1123 – His72) was pollinated by WL1238 (d, His(2-6)1221 His73) yielding four fully fertile F1 plants. An extra rapid line 6-14 (d, His(2-6)1221, His72) was fertilised by the pollen of the F1 plants. Line RT-1 was obtained by S.M. Rozov as F6 of cross 6-14 x 5-11 (which is F11 of VIR7036 x WL1018). Line 6-14 was obtained by S.M. Rozov as F14 of cross VIR320 (Palestine) x Sprint-1 (which is F25 of cross VIR7036 (Nepal) x cv. Avanti). The Dco allele in RT-1 causes not only the single anthocyanin ring in the axils but also red spots at the leaflet bases.

Table 2. Segregation for genes d, His(2-6) and His7 in the progeny of testcross B1.

His(2-6)

1323/1221

 

1221/1221

His7

2/2       2/3

 

2/2       2/3

Dco

32        15

 

13        49

d

20        13

 

14        43

 

Recomb. fract.

Joint c2

Probability

His7 – d

49.2 ± 3.5%

0.25

P > 0.5

His(2-6) – d

52.3 ± 3.5%

0.85

P > 0.3

His(2-6) – His7

27.6 ± 3.2%

35.78

P << 0.0001

1 Analogously to testcross A, line 6-14 plants were pollinated by fully fertile pollen of F1 hybrids between WL1238 and an individual plant (Dco, His(2-6)1323, His72) which resulted from the complex testcross [(a single plant in the F2 of WL1292 x WL1688) x WL102] x WL102, i.e. Cross 1 in (4).

Table 3. Segregation for genes d, His(2-6) and His7 in the progeny of testcross C1.

His(2-6)

1221/1121

 

1121/1323

His7

2/2        2/3

 

2/2       2/3

Dw

12         3

 

0          9

Dco

10         3

 

5          11

 

Recomb. fract.

Joint c2

Probability

His7 – d

50.9 ± 6.9%

0.05

P = 0.9

His(2-6) – d

40.5 ± 6.8%

1.61

P = 0.2

His(2-6) – His7

28.8 ± 3.3%

18.13

P < 0.0001

1 Two fully fertile F1 hybrids of cross SGE (Dw, His(2-6)1221, His72) x individual plant (Dco, His(2-6)1323, His73) were pollinated by WL102 (Dco, His(2-6)1121, His72). Line SGE was derived by V.A. Berdnikov from VIR6135 (Greece), VIR320 (Palestine), and Sprint-l (see Table 1). The individual plant came from Cross 1 of (4) (see Table 2).

Surprisingly, I failed to find information on the existence of any other markers linked both to genes a and d. The accumulated results provide evidence for two groups of subsequently linked genes traditionally attributed to linkage group I, one containing gene a and the other genes d and i. The assignment of these two groups to the same chromosome seems to be problematic. Thus, it cannot be excluded that the pea linkage map may require more major reconstruction even after the recent alterations (2, 3, 7).

1.           Blixt, S. 1972. Agri Hort. Genet. 30:1-293.

2.           Folkeson, D. 1990. Hereditas 112:249-256.

3.           Folkeson, D. 1990. Hereditas 112:257-264.

4.           Kosterin, O.E. 1992. Pisum Genet. 24:56-59.

5.           Lamprecht, H. 1956. Agri Hort. Genet. 14:66-106.

6.           Weeden, N.F., Ambrose, M. and Swiecicki, W.K. 1991. Pisum Genet. 23:
cover.

7.           Weeden, N.F. and Wolko, B. 1990. In Genetic Maps, Ed. S.J. O'Brien,
Cold Spring Harbour Laboratory Press, Cold Spring Harbour, New York, pp.
6106-6112.