--Donald S. Robertson and Philip Stinard
In the 1985 (MNL 59:14-15), 1987 (MNL 61:10-11) and 1988 (MNL 62: 21-22) Newsletters, evidence was presented suggesting that Mutator could induce mutants in either the male gametophyte or the zygote. The principal support for such a conclusion came from the observation of a number of white endosperm discordant kernels generated in the cross of y1 y1 wx wx gl1 gl1 x Y1 Y1 Wx Wx Gl1 Gl1 Mu2. The frequency of such discordant kernels was 16.69 x 10-5.
In the 1988 report, we presented the results of tests for discordant kernels among the yellow kernels from this same cross. Plants from this discordant class would be expected to be y1 y1. In order to score a large plant population, yellow kernels were sown in an isolation plot and the resulting plants allowed to open pollinate. Plants from discordant kernels would segregate yellow to white kernels in a 1:1 ratio, while those from nondiscordant kernels would segregate in a 3:1 ratio. Seven putative 1:1 ears were found in a population of 37,975 plants screened (frequency = 18.43 x 10-5). However, of these seven ears, all but one were semi-sterile. It was suggested that the induction of an ovule lethal mutation closely linked to Y1 would result in such semi-sterile ears with 1:1 ratios. Also, a discordant kernel that was heterozygous for a translocation would produce a plant with such an ear.
In our studies of over 200 independent Mutator-induced y1 mutants, 71.94% have been shown to have the pastel (pale green - zebra) phenotype when grown at 35 C. Thus if the 1:1 ears on plants from putative discordant yellow kernels are the result of a Mutator mutational event, there is a good likelihood that the white kernels should give pastel mutable plants. If the 1:1 ears indeed carry a Mutator-induced y1 mutant, one half of the yellow kernels should carry this mutant. Yellow kernels from the normal 1:1 and two of the semi-sterile 1:1 ears were sown and the resulting plants self-pollinated. The yellow and white kernels were separated and seedlings tested at 35 C for the expression of the pastel phenotype. Three of the nine selfed ears of plants from yellow kernels of the 1:1 ear with normal seed set segregated for mutable pastel seedlings. Similar tests of nine plants each from yellow kernels of the two semi-sterile 1:1 ears failed to segregate for pastel seedlings. In addition, white kernels from these same 1:1 ears were sown and the resulting plants outcrossed to Y1 Y1 standard plants. For each, 4 to 5 different outcross progeny were grown and the plants self-pollinated and seedling tested as above. Only the selfs of outcrosses from the 1:1 ear with normal seed set segregated for white-endosperm, pastel-seedling mutants.
The occurrence of a 1:1 ear with normal seed set demonstrates that the reciprocal discordant class (yellow endosperm - y1 y1 embryo) does occur as would be predicted if Mutator induced y1 mutations in the male gametophyte or embryo. However, the question is still open as to the origin of the semi-sterile 1:1 ears. The fact that pastel mutants have not been found in this class of mutants does not prove that they are not Mutator-induced y1 mutants, because about 28 percent of Mutator-induced y1 mutants do not have the pastel phenotype. The two semi-sterile ears tested could be carrying such a non-pastel mutant along with a reciprocal translocation, or some other genetic phenomenon (e.g., a large unlinked inversion, unlinked ovule lethal, etc.) could be responsible for the semi-sterility. Alternatively, these ears may not involve a y1 mutation, but instead result from the induction of an ovule lethal tightly linked to the Y1 allele. Evidence supporting the idea that some of the semi-sterile 1:1 ear ratios are due to an ovule lethal closely linked to Y1 is the occurrence of several semi-sterile ears in both sets of reciprocal tests with a very low number of white kernels per ear (2-7). These may represent the same ovule lethal mutant linked to the y1 allele.
If there is only one true yellow endosperm discordant kernel out of 37,975 tested, the frequency of this class of discordant kernels (2.63 x 10-5) is significantly different (at the 5% level of significance but not at the 1% level (c2 = 6.4193)) from the frequency of white-seeded discordant kernels (16.69 x 10-5). Because the difference is significant at the 5% level and just escapes significance at the 1% level, this difference might be real. If so, it may have a biological basis such as preferential fertilization, if gametophytic mutations indeed are occurring. However, because rare events are involved and because there is a seven fold difference in population size, sampling error could have a profound effect. If one more 1:1 normal ears had been found, the difference between these two populations would not have been significant at the 5% or 1% levels.
Because the two cells the female parent contributes to the endosperm come from two different cell lineages in the developing embryo sac, discordant white endosperm kernels are expected only very rarely or not at all in the reciprocal cross (i.e., Y1 Y1 Wx Wx Gl1 Gl1 Mu2 x y1 y1 wx wx gl1 gl1). However, because only one cell, the egg cell, is contributed by the female to the embryo, discordant yellow kernels can be expected in this cross. In the female gametophyte, there are three cell divisions involved in the production of the egg. In each division there is the potential for a mutational event, and thus yellow discordant kernels are expected 2.7 times as frequently in this cross as in the reciprocal cross in which the Mu2 parent is a male. In the past summer, yellow kernels from the crosses with Mu2 plants as females were planted, the resulting plants allowed to open-pollinate and the ears were scored for the presence of a 1:1 ratio (Table 1). The same three classes of ears (1:1 ratio of yellow to white, normal seed set; 1:1 ratio, semi-sterile seed set; and 2:1 ratio, semi-sterile seed set) seen in the reciprocal cross reported upon in the 1988 MNL were observed in this cross. Each class occurs in a higher frequency when the Mu2 parent is used as a female. Class 1 (1:1 ratio, normal seed set) is six times as frequent. Class 2 (1:1 ratio, semi-sterile seed set) is 6.4 times as frequent. These two classes are those that are most likely to include mutations at the y1 locus (see MNL 62: 21-22 for an explanation of the putative origin of these classes). Class 3 ears (2:1 ratio, semi-sterile seed set) occur with a frequency 1.5 times greater in the female crosses compared to the male crosses. Until all 1:1 ears from both crosses are tested further for the presence of pastel mutants, a clear estimate of the frequency of y1 mutations in these crosses will be impossible. However, whether Mutator is inducing mutants at the y1 locus or linked loci (e.g, ovule or gametophytic lethal loci) on chromosome six, these mutations are occurring in a higher frequency when the Mu2 parent is a female than when it is a male. The presence of two classes of discordant kernels when Mu2 plants are crossed as males, and the presence of only the yellow endosperm, y1 y1 class of discordant kernels when Mu2 is used as a female supports the hypothesis that Mutator can induce mutants in the development of the gametophyte or in the zygote and triple fusion nucleus. However, the results from the crosses when Mu2 plants are used as females suggest that few if any of the Mu-induced mutants studied are being induced in the zygote. If the mutations found in these studies were occurring at this time (i.e., the zygote), the frequencies of mutations of classes 1,2 and 3 would be expected to be equal in the reciprocal crosses. This is not the case. They are more frequent when Mu2 plants are used as females. Again, this is expected because there are three times as many divisions in the development of the female gametophyte, each with the potential for producing discordant kernels, as there are in the development of the male gametophyte. In the latter, only in the division of the generative nucleus would a mutation occur resulting in a situation in which discordant kernels would be produced.
Table 1. Results of test for putative yellow discordant kernels from
the cross of Y1 Y1 Wx Wx Gl1 Gl1
Mu2 X y1 y1 wx wx gl1 gl1
and its reciprocal cross (data from MNL 62:21-22, 1988)
| Class 1 | Class 2 | Class 3 |
| Normal or near normal | Semi-sterile or near semi- sterile seed set | Semi-sterile seed set or near semi- sterile seed set |
Y1 Y1 Wx Wx Gl1 Gl1 Mu2 X y1 y1 wx wx gl1 gl1
| 1:1* ratio | 1:1* ratio | 2:1* ratio |
| 3 (8.9 x 10-5) | 34 (101.3 x 10-5) | 9 (26.8 x 10-5) |
% of total white kernels for each class
| 47.77 | 49.00 | 36.24 |
Total ears scored - 33,561
y1 y1 wx wx gl1 gl1 X Y1 Y1 Wx Wx Gl1 Gl1 Mu2
1 (2.6 x 10-5) 6 (15.8 x 10-5) 7 (18.4 x10-5)
Total ears scored - 37,975
* The ears in these categories do not differ from the indicated ratio
at the 1% level, and the vast majority do not differ at the 5% level (only
three ears differed from the expected ratio at the 5% level).
If, as the results of these studies suggest, Mutator does induce mutants in the developing gametophyte, progeny of Mutator plants used as female parents would be expected to exhibit more mutants overall than those in which Mutator plants are used as male parents. This has not been observed to be the case even when the male data are corrected for the presence of white discordant kernels. In fact, just the opposite is found to be true (D. S. Robertson, MNL 61:10-11, 1987). This may only be an apparent contradiction, however, because the new mutants found in these outcross progenies include those from other possible stages in development during which mutations can occur (i.e., premeiotic and meiotic).
In the crosses of the Mu2 parents as a male, the corrected mutation frequency ( i.e., without discordant kernels included) is 3.22 x 10-4 (MNL 61: 10-11, 1987). When the Mu2 plants are the female parents, the frequency of white mutant kernels is 1.99 X 10-4. There is a 1.6 fold frequency difference in favor of the male. The female data only counted mutations giving rise to a mutant ear sector as one event. In the male crosses, however, similar sectors occurring in the tassel could not be recognized and thus all kernels resulting from all mutant gametes produced by a sector were counted as separate events. Thus the male mutation frequency would be expected to be higher. The exact increase in frequency expected as a result of not being able to distinguish kernels derived from tassel sectors can not be determined. It is known, however, that ear sectors are usually small. Fifty-seven percent are two-kernel sectors, while 86% consist of 5 or fewer kernels. Tests of male transmitted mutants suggest that sectors in the tassel are equally small (D. S. Robertson, Genetics 94:969-978, 1980). Because the difference between female and male derived mutants favors the male (i.e., male 1.6 times that of female), it could well be that the higher male frequency results from not being able to distinguish mutants coming from tassel sectors.
The data from the tests for gametophytic mutants, however, create another
problem. Data from Table 1 suggest that events that seem to be the result
of mutations happen 3 to 6 times as frequently in the female as male. However,
at this time we do not fully understand the nature of the genetic change
responsible for some phenotypes observed. If classes 2 and 3 are the result
of some event independent of the y1 locus (e.g., a translocation,
or an abortive ovule mutant, etc.), there is no a priori reason to consider
these as gametophytic mutants. All or part of them could be meiotic or
premeiotic in origin. The Class 1 (Table 1) mutants seem to be the most
likely candidates for gametophytic mutants, and these occur in the female
with a frequency 3.4 times that of the male. Thus the overall female mutation
rate would be expected to be higher. It may be that the sample of 1:1 normal
ears (Class 1) is too small to support any definite conclusion in this
regard. It should be noted, however, that female gametophytic mutants may
not account for the bulk of transmitted mutants. If this is so, gametophytic
mutants may not markedly affect the balance between the mutants transmitted
by the two sexes in reciprocal crosses. If it is assumed that the 1:1 ears
with normal seed set in Table 1 (Class 1) are the result of gametophytic
mutants (male data suggest this is a reasonable assumption), then it is
possible to estimate the frequency of gametophytic events relative to pre-gametophytic
events. In MNL 61:10-11, 1987, it was determined that the pre-gametophytic
events occurred in a frequency 1.99 X 10-4 (Mu2 as female). The
tests reported in Table 1 involved a subset of the same population of kernels
scored to obtain the frequency of 1.99 X 10-4. Thus, if this frequency
is applied to the population 33,561 (Table 1), it is possible to estimate
how many pre-gametophytic mutants would be expected in that population.
This number is 6.7. Thus only 30.9 percent (3/(3 + 6.7)) of the total mutants
that are found in the eggs of Mu2 plants are gametophytic. Thus
over two thirds of all mutants are pre-gametophytic. However, our database
is too small, and much more data are needed before definite conclusions
can be made on the full effect of gametophytic mutants on the mutation
frequency of Mutator plants.
Return to the MNL 63 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage