Interaction of a new leaf mutation ins2 with af, uni^tac and tl^w

Berdnikov, V.A., Gorel�, F.L.,                           Institute of Cytology and Genetics, Russian Academy of Sciences

Bogdanova, V.S. and Kosterin, O.E.                  Acad. Lavrentiev ave. 10, Novosibirsk 630090, Russia

      In 1959 Lamprecht described a recessive mutation insecatus (ins), with apices of the first pair of leaflets incised. The midvein in the zone of incision was not only retained but also overdeveloped sometimes being transformed into a tendril. Unfortunately, this character had incomplete penetrance and highly variable expression, as well as showing reduced fertility (5).   Thus genetic mapping of the relevant gene has been difficult.

        In 1992 in the progeny of a Twt trisomic (see [1]) there was found a plant with a normal karyotype and a phenotype closely resembling that of insecatus, although the incisions of the leaflet apices were deeper and midveins frequently transformed into ramified tendrils. The first incisions appeared at the first pair of leaflets of the nodes 5-6, counting the first scale leaf as node 1. At higher nodes, incisions became deeper and appeared at the second pair of leaflets (Fig. 1a).  Sometimes the first pinna pair position possessed three or four leaflets forming a butterfly-like structure (Fig. 1b). At the nodes 8-12, roughly at the first flowers, the expression of the mutant phenotype reached its maximum: the leaflets being replaced by compound tendrils and the phenotype may become indistinguishable from afila (Fig. 1c, d). On higher nodes the mutant phenotype gradually decreases. The mutant plant was selfed for 15 generations, producing an isogenic line AFD. 

      Pea lines WL021, WL025, WL1759, and WL1898 from Weibullsholm collection (Landskrona, Sweden) are designated as carrying the gene ins.  However, we found poor visible expression of the mutant character in these lines and were unable to perform an analysis of allelism with the newly found mutation. We called the new mutation ins2. In the crosses of ins2 plants with wild type plants, the F₁ progenies had a normal phenotype, thus indicating the recessive mode of inheritance of ins2. It should be noted that a thorough analysis of the heterozygotes ins2/Ins2 reveals a aberrant phenotype weakly expressed in some leaves, most frequently manifested in the pointing of the apical part of the leaflets.

      The similarity of ins2 and afila phenotypes may be evidence of allelism of the relevant genes. To test this possibility we crossed the line AFD (Af, I, ins2) with WL1746 (af, i, Ins2). F₁ plants displayed the insecatus phenotype, suggesting that ins2 and af might be allelic.  However, segregation data for these traits in the F₂ are in contradiction with this hypothesis (Table 1).  One can see that all seven plants grown from i/i seeds possessed the afila phenotype, confirming a close linkage of i and af. We therefore used the color of the cotyledons in seeds produced on the plant to postulate the genotype (homozygous dominant or heterozygous) of the plant at the Afila locus. Of nine I/I plants (and, most probably, Af/Af) three displayed the insecatus phenotype, indicating non-allelism of the genes af and ins2. In addition, the data of Table 1 suggest the hypothesis that heterozygosity at Af causes the ins2 allele to become dominant, manifesting its effect in the heterozygous state.

Table 1. Segregation for the afila and insecatus-2 phenotypes among the F₂ progeny of the cross 
AFD (Af, I, ins2) x WL1746(af, i, Ins2). 

¹+ = dominant �wild type� alleles.

Table 2. Segregation for the afila, uni and insecatus-2 phenotypes among the F₂ progeny 
(further designated as �VAR�) of the cross AFD (Af, I, Tl, Uni, ins2) x GRT (af, i, tl^w, uni^tac , Ins2). 

¹+ = dominant �wild type� alleles

     To study interaction of ins2 with genes affecting development of a compound leaf we crossed the line AFD with an earlier synthesized stock GRT carrying the markers i, af, tl^w and uni^tac derived as F4 from a cross cv. Frontier x line A45 (kindly supplied by Prof. E. T. Gritton). As expected, F₁ plants had the phenotype ins.  Phenotypes of the F₂ plants (144 in total) are given in Table 2. For convenience, this F₂ population will be referred to as VAR. All 35 plants of this population grown from i seeds possessed the afila phenotype and were excluded from the further analysis. Of 109 plants of VAR grown from I seeds 51 (47%) had leaflets incised to various extents.  The greater than expected proportion of plants with the ins phenotype can be accounted for by an effect of heterozygosity for the gene af.

Of 58 VAR plants without incisions 32 displayed the phenotype uni^tac, and 26 were of the normal phenotype Uni. At the same time the phenotypic class (ins uni^tac) was completely absent (no plant both possessed a leaflet in place of the terminal tendril and had the first pair of leaflets incised).

      The impression is that the mutant genes uni^tac and ins2 either are tightly linked or are antagonists in their effect on the leaf development. Since all the plants with the uni^tac phenotype lacked incisions in the leaflets despite variation of the dose of ins and af alleles, we could hardly expect appearance of the ins trait in their off-spring. For this reason the descendants of plants homozygous for the genes uni^tac or af were excluded from the further analysis. From VAR there were chosen 41 plants that produced no less than 12 seeds. Fifteen plants lacked incisions (Ins), eight had weak incisions (ins-w), seven had strong incisions (ins-s) and the remainder (eleven) had medium expression of the trait (ins-m). An analysis of the progeny of these 41 plants permitted the determination of the genotype of these plants with respect to the loci Af, Uni and Ins2 (Table 3).

      Although classification of plants according to leaflet incision was made visually and should not be treated as completely accurate, several general conclusions can be made.  All seven plants with deep incisions were found to be homozygous for ins2 and heterozygous for af, six of these being homozygous Uni/Uni. Heterozygosity for both ins and af was typical for the phenotype ins-m. Plants with weak incisions were heterozygotes ins2/Ins2 in the homozygous background Uni/Uni or Af/Af. The majority of plants without incisions were homozygous for the wild-type allele Ins, but in five cases they were heterozygotes ins2/Ins2 and homozygous Af/Af. We can reject the hypothesis of the tight linkage of the genes uni and ins2 because four of the plants analyzed from VAR were double homozygotes (Uni/Uni Ins/Ins). At the same time we failed to observe double homozygotes (uni^tac/uni^tac ins2/ins2). Probably, the missing class should be sought for among the plants of the uni^tac phenotype.

      Further interesting information was gained from analysis of a particular plant (#99) (phenotype ins-m, Uni, Af) from VAR. None of its 17 descendants had the afila phenotype, 11 of them with the phenotype Uni had incisions, while the remaining six lacked incisions and were uni^tac.  Thus, the genotype of the plant #99 most probably was (Af/Af uni^tac/Uni ins2/ins2).  If such were the case, the progeny with the uni^tac phenotype should be double homozygotes (uni^tac/uni^tac ins2/ins2).  However, the small size of progeny of the plant #99 may have prevented the observation of the Ins2 phenotype and hence been a source of an error in genotype determination.  For this reason we extended our analysis by examining the progeny of the eight most productive descendants of plant #99 with the insecatus phenotype. All the progeny of four of them had the phenotype (Uni ins), while segregation for uni^tac  was observed in the progeny of each of the other four families.  All 12 plants with uni^tac phenotype lacked incisions, while the remaining 35 plants had incised leaflets. This result confirms that the genotype of plant #99 was (uni^tac/Uni ins2/ins2). A thorough examination of the plants of the uni^tac phenotype revealed a number of peculiarities: the mutant trait was clearly expressed only at the first three nodes, at a higher level leaves did not differ from normal ones, and only at some flowering nodes was the terminal tendril replaced by a leaflet.  Thus, in the double homozygote (ins2/ins2 uni^tac/uni^tac ) the uni^tac trait is not expressed in all leaves.

        In the present work we have shown that the gene Ins2 acts as a synergist of the gene Af and as an antagonist of the gene Uni. Antagonistic interaction of the genes Af and Uni has been demonstrated by several authors (2, 6, 7) and recently confirmed by Hofer and Ellis (4). The striking phenotype pleiofila of the double homozygote (af/af tl/tl) is well known. Taking into account synergism of Af and Ins2, we might expect a similar phenotype to be displayed by the double homozygote (ins2/ins2 tl/tl).

      In VAR we found several double homozygotes (ins2/ins2 tl/tl). These plants had in the leaflet incisions instead of an overdeveloped midvein a tiny leaflet at a tiny petiolule. As the size of incision increased, the midvein was replaced by a small compound leaf, its size correlating with the size of an incised leaflet that produced it (Fig. 2a, b). At higher nodes the incisions became deeper and, finally, the first pair of leaflets disappeared and in place of the second and third pairs there appeared compound leaves of the second order (Fig. 2c, d) closely resembling the leaves of plants with (af/af tl/tl uni^tac/uni^tac ) genotype (3). The impression is that the effect of the homozygote ins2/ins2 is equivalent to that of the double homozygote (af/af uni^tac/uni^tac ).

      In the Fabaceae, bipinnate leaves are not found in the subfamily Papilionoideae, but they are very common in two related subfamilies - Mimosoideae and Caesalpinoideae. A function of the gene Ins2, as well as that of the gene Af seems to consist in decreasing the rank of branching of the leaf axis. Weakening of this function together with inactivation of the Tl gene makes the structure of the compound pea leaf similar to what is perhaps a more archaic bipinnate form.

      A thorough examination of leaves of different pea samples often reveals incisions of the insecatus type. It appears to be an evidence of existence of a number of genes that can produce the insecatus phenotype. Recently we have found another mutation with expression similar to ins2. At present it is being investigated.

 Acknowledgements: We express our gratitude to Prof. E.T. Gritton for providing us with a rich and useful germplasm, including cv. Frontier and line A45. For technical assistance we thank Ms. O.F. Fialko. The work was partly supported by the Russian State Program �Russian Fund for Fundamental Research�.

1.   Berdnikov, V. A., Gorel�, F. L. and Temnykh, S. V.  1994.  Pisum Genetics 26: 9-10.

2.   Goldman, I. L. and Gritton, E. T.  1991.  J. Hered. 82: 479-483.

3.   Hofer, J.M.I. and Ellis, T.H.N.  1996.  Pisum Genetics 28: 21-23.

4.   Hofer, J.M.I. and Ellis, T.H.N.  1998.  Trends in Plant Science 3: 439-444.

5.   Lamprecht., H.  1959.  Agri Hort. Genet. 17: 26-36.

6.   Marx, G.A.  1987.  Plant Molec. Biol. Reporter 5: 311-335.

7.   Sharma, B.  1981.  Pulse Crops Newslett. 1: 56-57