A mutator system (Mu) has been previously described (MCGNL 45:81-87, 1971; 49:73-79, 1975; 50:68-70, 1976; 51:32-36, 1977). One question of interest to be answered about Mu has to do with its time of action. Are mutations induced during sporophytic cell divisions? If so, do they occur early in development or late? Or are new mutations induced only in meiosis? If mutations are meiotic (occurring any time after premeiotic S), one would expect only one occurrence of any particular mutant in the selfed outcross progeny of any Mu plant. However, if mutations occurred premeiotically (occurring any time prior to premeiotic S) there could be from two to half of the offspring that would carry any given mutation. Two mutants would be expected if the mutation occurred in the last premeiotic cell cycle. A mutation that occurred very early, in a cell destined to give rise to the whole tassel, would be transmitted to half of the offspring. Mutations occurring between these extremes would result in tassel sectors carrying the mutant and thus could give rise to a situation where varying numbers of outcross progeny may carry the mutation. If mitotic mutations occur, phenotypically similar mutants would be expected in the selfed outcross progeny of Mu plants. These should all prove to be allelic. However, two mutants are not necessarily allelic just because they have the same phenotype. Many instances of phenotypically similar mutants that are not allelic are known.
In our mutator tests we frequently found more than one mutant of a given phenotype segregating on ears of sibling outcrossed plants that had been selfed. Similar mutants from given outcross progeny were allele tested. The results are given in Table 1. Seventy-eight percent of the putative allelic situations turned out not to be.
Among families that had only two mutants of similar phenotype tested, 23-81% (5/21) were allelic while in families with more than two similar mutants tested 50% of the families had allelic mutants.
These results indicate that both meiotic and premeiotic mutants occur. The bulk of the mutants seem to be meiotic (or perhaps very late mitotic). The probability of a given testcross family having premeiotic mutants increases if more than two similar mutants occur. Even in instances where premeiotic mutants occur they must be relatively late events giving rise to small tassel sectors, since none of the outcrosses to date with mutants of similar phenotypes approach the 50% mutant frequency expected if a mutation occurred in the progenitor cell of the tassel. Other tests for early mitotic mutations have also proven negative. Mu lines heterozygous for selected recessive loci have failed to show more sectoring for the recessive phenotypes than the non-Mu controls. The following recessive loci have been tested: yg2, y, c sh bz wx, a sh2 and a2 bt.
If one goes back to the raw data analyzed in Table 1, a total of 69 mutants were included in the tests. Of these, 22 were involved in positive allele tests. Thus, 68.12% of the mutants involved in potentially allelic situations proved to be nonallelic. In all estimates of Mu activity to date, the occurrence of 2 or more mutants with similar phenotypes was counted as the result of a single mutation event (counted once) in our estimates of the total number of different mutations. However, in all of our tests the total numbers of mutants (irrespective of phenotype) were also recorded. The difference between these two totals represents the mutants of similar phenotype that were not counted as resulting from an independent mutation event. Data presented above would indicate that 68.12% of these would be the result of independent mutations. Table 2 summarizes the result of all Mu tests to date.
In column 6 the mutation rate is estimated as 6.4%. However, this does not reflect the true mutation rate since two or more mutants with similar phenotypes in a given outcross were counted as being the result of only one mutational event. If we take 68.12% of the difference between column 3 and column 5 (all mutants including those with similar phenotype) we will get an estimate of the putative allelic mutants that are actually non-allelic (are the result of an independent mutation), 49.73. If these are now added to the total of column 5 we have a better estimate of the number of different mutations. This new total now gives an estimated mutation rate of 9.6%, which is 3.2% more than our previous estimate and is some 48-fold higher than our best estimate of the spontaneous rate.
Donald S. Robertson and Peter N. Mascia
Return to the MNL 52 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page