The a1-eap allele of a1 represents a specific marker for the selection of embryogenetic mutants

--Gavazzi, G, Stucchi, C, Pilu, R

In some of our stocks, we occasionally observed seeds with a colorless aleurone and a ring of color surrounding the embryonic axis. A closer scrutiny under low magnification shows that this ring is due to pigmentation of a thin layer of cells of the scutellum surrounding the embryonic axis along its entire length or its apical portion (Fig. 1).

Sometimes an additional layer of cells is formed on the side of the scutellum facing the endosperm. After several rounds of selection we obtained a line, referred to as eap (embryonic axis profile), with this trait expressed in all seeds. When this line was crossed to an ACR line, homozygous for all genes required for aleurone pigmentation, the F1 was uniformly colored, whereas the F2 segregated colored and colorless seeds, the latter again showing the embryonic trait, in a 3 to 1 ratio. This observation suggested that this trait was due to a monogenic recessive color mutant. We then applied pollen of a single eap male parent to silks of five color tester lines, each one homozygous recessive for one of the genes required for seed pigmentation and homozygous dominant for the other color genes. These lines are here referred to as c1, c2, a1, a2 or r testers. The resulting F1 seeds were homogeneously colored, and the F2 progenies segregated colored and colorless seeds in a 9 to 7 ratio, as expected if two genes responsible for seed pigmentation are segregating in each of the families scored (Table 1). A χ² test for 9:7 segregation fits the data except for the F2 where a2 is segregating. Since in this case there is a significant excess of colorless seeds over the expected 7/16, this discrepancy could indicate preferential transmission of the a2 marked chromosome. If we assume that the eap phenotype requires at least one dose of each of the color genes, the colorless seeds recovered in the F2 should be eap in 3 out of 7 cases, whereas the remaining 4/7 should be totally devoid of pigment. However the proportion of eap among the colorless seeds was consistently less than the expected 3/7. This could be the result of poor expression of this marker in a different genetic background.

An unplanned cross of eap females with a heterozygous TB-3La stock revealed unexpectedly that the eap phenotype is uncovered by this translocation. Since a1 maps on 3L, we tried to establish the chromosomal location of eap with those microsatellites that define bin 3.09 where a1 is positioned. An SSR analysis was performed on the F2 progeny of the cross ACR-r × eap. Two of the eight microsatellites tested on this segregating population showed polymorphism. Of the 62 individuals tested with the p-umc2008 marker only one recombinant was recovered, whereas five out of 36 recombinants were obtained with the p-umc1641 marker. These results place eap in a region where a1 resides, with a recombination frequency between eap and umc2008 and umc1641 of 1.2 and 6.8 cM, respectively. The original a1 tester employed in the complementation test was derived by conversion of a1 into an ACR-r W22 inbred line.

We then crossed the eap line with other sources of a1, i.e. homozygous a1R-sc, a1Mum2, a1Mum3, a1sh2 stocks, and we obtained progeny ears consisting of homogeneously colorless seeds. These results, together with the assignment of eap to the same bin where a1 is located, support the conclusion that the eap phenotype is due to a recessive allele of a1 that we accordingly named a1-eap. The apparent complementation of a1-eap with the a1 tester in the W22 background remains to be elucidated.

Because of its phenotype the a1-eap allele can be used as a visible embryonic marker for the detection of mutants impaired in the development of the apical shoot. To find out the feasibility of this strategy we made use of the recently isolated sml gene (Pilu et al., Plant Physiology 128:502-511, 2003). The shootless phenotype caused by disruption of SAM formation, is due to the synergistic interaction of mutations of two genetic loci. Seedlings homozygous for both sml (shoot meristemless) and the unlinked dgr (distorted growth) loci have a SAMless phenotype, seedlings Dgr/- sml/sml are impaired in their morphogenesis to different extents, whereas the dgr mutation alone does not lead to a recognizable phenotype. The F2 obtained by selfing the progeny of homozygous a1-eap Sml Dgr females to A1/A1 dgr/dgr Sml/sml male parents consists of two kinds of progeny ears, one with all seeds homogeneously eap and another one segregating eap/Eap in a 3 to 1 ratio. Upon germination, seeds of the first class gave rise to normal seedlings whereas those of the second class yielded normal or a majority (80%) of abnormal (sml or dgr) seedlings following their separation into eap and Eap, respectively.

We take this result as evidence that the homozygous double mutant a1-eap sml exhibits a negative interaction, causing suppression of the embryonic axis pigmentation. Presence of a functional SAM thus seems necessary for cells around the embryonic axis to become competent to express pigment or, alternatively, that a functional SAM or a signal it elaborates, is required for pigment formation. We think that this observation represents useful information for the selection of mutants impaired in the shoot meristem formation or establishment. Mutants of the SAM could in fact be searched for in the selfed progeny of homozygous a1-eap seeds following chemical or transpositional mutagenesis by selecting exceptional M2 ears exhibiting segregation of the eap trait.