When Ac2/Ac2; bz2-m/bz2-m plants are used as male parents in crosses to bz2-m ac testers, the endosperms of ac ac Ac2 constitution show no dots. Occasional kernels are found with numerous dots and these have proved either to possess a more potent Ac2 allele (which we have designated AcAc) or to have two unlinked Ac's. These alternative types were equally frequent and we were unable to predict from the number of Bz dots whether a kernel had AcAc or possessed two unlinked Ac's. We have attempted to test the hypothesis that the potent Ac2 allele is actually a duplication. These tests involve the closely linked marker genes, virescent-16 and male-sterile-8, flanking the Ac2 locus on chromosome 8 (recombination between v16 and ms8 in several compounds varied from 3.3% in female heterozygotes to 7.4% in male heterozygotes with the Ac locus showing about 0.6% recombination with ms8.) Crosses of V16 AcAc Ms8/v16 ac ms8 females with v16 ac ms8/v16 ac Ms8 males give a 1:1 ratio for kernels with many dots to kernels with no dots, as well as occasional kernels (about 1-2%) with a low dot number. The low-dot cases were at first believed to result from crossovers separating the two Ac elements. If so, they should possess either v Ms or V ms markers. Of course, half of the plants with the V ms crossover would be phenotypically V Ms due to the unavoidable heterozygosity for ms in the male parent, but these could be identified by testcrossing. The observed marker distribution in a group of 42 low-dot cases was: 8 V Ms: 19 v Ms: 0 v ms: 15 v ms. Further tests were made to confirm that the low-dot endosperm cases possessed a weak form of Ac2 in the embryo as well as in the endosperm, to establish the linkage relations of the Ac allele, and to check the classification of the v16 and ms8 phenotypes. At this time, 16 of the low-dot cases have been analyzed. Some bias is introduced in the composition of the tested group of plants due to greater viability of green over virescent plants and to the time needed in genotyping plants of v ms phenotype where dominant marker alleles must be introduced before linkage of Ac on chromosome 8 can be tested. The 16 fully analyzed plants consisted of nine plants with Ac linked to chromosome 8, 8 V Ms and 1 u Ms, and seven plants with transposed Ac, 6 v Ms and 1 v ms.
There was no evidence that crossing over between v and ms was involved in the origin of the low-dot cases. The near equality of v Ms and v ms phenotypes in the group of 42 (due to the Ms/ms constitution of the male parent) suggests that most of the v plants are noncrossovers of tr-Ac/ac constitution arising when one component of the presumed duplication was transposed to a heterologous chromosome. Tests for linkage indicated that seven of the eight virescent plants possessed a transposed Ac (tr-Ac). The absence of the V ms phenotype indicates that the V Ms plants are either V Ac Ms/v ac Ms or V Ac Ms/v ac ms and not V Ac ms/v ac ms. The linkage phase of the flanking markers in the V Ms plants is the same as in the parent chromosome but one Ac component has been lost from the AcAc allele. Linkage tests with eight such plants showed the Ac was linked to chromosome 8 in every case.
Although attempts to demonstrate the duplicated nature of strong Ac alleles by association of crossing over with the origin of weak derivative alleles gave negative results, we still favor the interpretation of strong Ac alleles as duplications or triplications for the following reasons:
1. Similarity of the dot number in the endosperm of kernels with Ac/ac/ac; tr-Ac/ac/ac constitution and that found in AcAc/ac/ac kernels. When unexpected dotted kernels occur in testcrosses of Ac/Ac male parents, the phenotype of those kernels shown to possess two independent activators is similar to that of individuals having the presumed duplication at the original locus. In female testcrosses, the Ac/ac; tr-Ac/ac plants produce 1 high-dot: 2 low-dot: 1 no-dot kernel ratios while AcAc/ac plants give 1 high-dot: 1 no-dot ratios. Kernels in the high-dot classes in two testcrosses, although differing in genotype, cannot be distinguished phenotypically from each other.
2. Few, if any, one-step changes from a potent allele to a null form. In testcrosses of V AcAc Ms/v ac ms female parents, a change from AcAc to ac by excision of the AcAc as a unit would give a chromosome 8 with V ac Ms markers. From 1031 no-dot kernels, there were eight green normal and thirteen green male-sterile plants in addition to 894 virescent individuals. If the AcAc was completely excised, all the green plants should have normal tassels. Since more than half of the green plants were male-sterile, the 21 green plants can be accounted for as crossovers possessing the V ac ms chromosome with either the v ac ms (the thirteen green male-sterile) or the v ac Ms (eight green normal) chromosome contributed by the male parent. The absence of plants with a V ac Ms chromosome argues against the simultaneous loss of both Ac components and is in agreement with the map of v16 - Ac - ms8, a region so short that no double crossovers are expected.
Further evidence of the same sort comes from testcrosses of putative V AcAc ms/v AcAcAc Ms female parents where one chromosome possessed a noticeably stronger Ac than the homologue. Twenty-five kernels with low or no dots were selected from a testcross population of 897. Twentyone of these produced plants possessing a V Ac ms chromosome (11 green normal and 10 green male-sterile) indicating a stepwise loss of a single component from the weaker of the two parental alleles. Tests of linkage with v16 showed the remnant Ac is still on chromosome 8. The remaining four individuals consisted of one v Ms, two v ms, and one V Ms plants. The two v ms plants are crossover individuals derived either by loss of one component of the AcAc allele with a concomitant crossover between Ac and v16 or by a crossover within the Ac complex following oblique synapsis of the components. The v Ms individual can also be accounted for by the same events but it received the Ms marker from the male parent. The linkage of Ac has not yet been determined in these three plants. The only exception to the proposed hypothesis of stepwise loss of Ac components is the single V Ms plant which seems to have lost both Ac components from the V AcAc ms chromosome but possesses the Ms allele from the male parent. In the great majority of cases where changes occurred in the Ac allele, a stepwise loss of a single component appears most likely. Similarly, newly transposed Ac alleles seem to consist of a single component. Cryptic transpositions of Ac from AcAc/ac parents were identified by repeated testcrossing of all the dotted progeny and searching for ears with 3 mutable: 1 stable ratios. These represent AcAc/ac; tr-Ac/ac individuals and the tr-Ac's have turned out to be weak Ac alleles with only one exception. A strong Ac allele consisting of a single element should be able to give rise to equally strong transposed alleles. For these reasons, we favor the duplication hypothesis as the explanation of strong Ac2 alleles, but confirmation or rejection must await molecular analyses.
M.M. Rhoades and Ellen Dempsey
Return to the MNL 61 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page