Skripkina, T.A., Kosterin, O.E., and Berdnikov, V.A.

Institute of Cytology and Genetics, Novosibirsk, 630090, RUSSIA

It's well known that the presence and degree of development of bracts is highly variable in the garden pea (1). Four genetic factors have been reported so far to determine bract presence: Br, Bra, brac, and sifl (1). Br and Bra are thought to be polymeric, i.e. each of them is a dominant factor producing some bracts while their simultaneous presence brings about much greater development of the bract (2). Br and Bra were associated with current linkage groups III and IV, but this association is now considered as doubtfull (5). brac and sifl are recessive mutations, the latter inhibiting flower formation. brac is attributed to the d-segment of the linkage map in this volume (4). Besides, the bract development is affected by genes determining the first flowering node (mostly lf), as the bracts tend to deplete and disappear on higher. Trisomy for a short exchange chromosome of the well-known Hammarlund's translocation (K-line) also causes a substantial bract enlargement (3, and our observations). It may be concluded that the bract development is influenced by combination of environmental conditions, a physiological status of an inflorescence, and a number of genetic factors.

The aim of this work consisted in clarification of a complicate mode of action of the presumed genes Br and Bra. For this reason we took three lines, WL1081, WL1238, and WL1255, constructed by Lamprecht and assumed to have the genotypes Br/Br bra/bra, br/br Bra/Bra, and Br/Br Bra/Bra, respectively, and crossed them with lines without bracts.

Table 1 shows the results of segregation for the presence versus absence of bracts of F2 plants in four crosses with the lines WL1018 and WL1238, having either Br or Bra involved, with the lines TS2, TS52, and TS34, derived from the same cultivar Ramonsky 77, which lack bracts.

Table 1. Segregation for the presence versus absence of bracts in F2

  populations of four crosses.

 Cross         Factor thought   Plants                Models
              to be involved  with bratcs      3:1            9:7
                             present absent  Chi-sq.  P    Chi-sq. P

TS34 x WL1081      Br          94     68     24.9  <0.0001  0.2 <0.0001
Wl1081 x TS2       Br          39     65      8.7  <0.005   1.7  >0.1
TS52 x WL1238      Bra         40     54     15.4  <0.0001  0.05 >0.8
WL1238 x TS2       Bra        174    432      4.5  >0.025   55.7 <0.0001

One can see that this segregation did not fit the 3:1 model. In three cases there were more plants without bracts than plants with bracts. The first and third cases are closer to the 9:7 model (complementarity), but with opposite dominance relationships (see the share of plants with bracts). These data evidence for a polygenic inheritance of bract development.

As the bracts are variable we need a convenient measure of their development. The number of main radial veins of a bract can serve as such a measure. In the lines WL1081 and WL1238 and in the progenies of the above mentionned crosses the bracts were rather little developed as represented by small lobes mostly with one main vein rarely with two or three veins. In the line WL1255 the bracts are well developed, especially on lower nodes, and have mostly 5-10 main radial veins.

This line was crossed with the lines WL851, WL2224, WL 2715, wich lack bracts, and with the line WL1081 having small bracts and assumed to bear the gene Br. The resulted F1 plants are analyzed in Table 2.

Table 2. The presence of bracts in F1 plants in four crosses.

 Cross             Total number     Bracts         Number of main
                                present  absent   radial veins
WL1255 x WL2715       11           0       11          -
WL1255 x WL2224       16          15        1         1-3
WL1255 x WL851        17          17        0         2-3
WL1255 x WL1081       32          31        1         1-4

In three crosses majority of them had bracts, that indicates at a dominant nature of this trait. Their bracts were rather little variable, they were rather small and had 1-3 main veins. It implies an intermediate inheritance of the degree of bract development. The bracts in the F1 plants of the cross WL1255 x WL1081 (which, according to Lamprecht, have a genotype Br/Br Bra/bra) did not differ from those of the F1 plants in the crosses of the line 1255 with the lines WL2224 and WL851 (assumed to be Br/br Bra/bra). The absence of bracts in the cross WL1255 x WL2715 makes us to suppose the presence of a supressor of bract development in the genome of the line WL2715.

The F2 populations were studied in two crosses of the line WL1255 with lines without bracts (Table 3).

Table 3  Segregation for the presence versus absence of bracts in F2 

 populations of two crosses involving WL1255.

 Cross             Total number     Bracts          Chi-sq.   P
                                present  absent      3:1  
WL1255 x WL2715      215          160      55       0.04    >0.8
WL1255 x TS2         282          204      78       1.06    >0.2

The ratio of plants with and without bracts corresponds excelently to 3:1 model. The bract size in these F2 populations varied widely from hardly noticeable scales to giant bracts which resembled stipulae and had 8-10 main radial veins (Fig. 1).

Fig.1. Variation of bracts of the F2 plants of the cross WL1255 x WL2715.

All the data lead us to the following conclusions:

1) The line WL1255 has a major monogenic dominant factor determining large bracts in homozygous state and less developed bracts in heterozygote.

2) The bract size is greatly influenced by some other factors of polygenic nature, which can produce small bracts without the above postulated factors (as in the lines WL1238 and WL1081).

3) It would be more convinient to abandon the concept of two polymeric loci, Br and Bra, and rather say on one major locus and a number of undetermined factors with weak and inpersistent effetcs.

A question remains what name, Br, Bra, or a new one, should be applied to the major factor. Further experiments aimed to find linkage of it with the markers to which Lampreht (1) had linked his Br or Bra could help in this respect. So far, in the cross WL1255 x WL 2715 this factor was proved not to be linked to the loci i, af, r, tl, wlo, Pl, n, p.

This work was partly supported by the Russian State Program for Fundamental Research.

References:

1. Blixt., S. 1972. Agri Hortique Genetica. 30: 1-293.

2. Lamprecht, H. 1963. Agri Hortique Genetica. 21: 159-165.

3. Pellew, C. 1940. J. Genet. 39: 363-390.

4. Rozov, S.M., Gorel', F.L., and Berdnikov, V.A. 1996. Pisum Genetics 28:

5. Weeden, N.F., Swiecicki, W.K., Ambrose, M., and Timmerman, G.M. 1993. Pisum Genetics 25: 4 and cover