Spontaneous mutations (reversions) from male-sterile to male-fertile condition occur in plants with S male-sterile cytoplasm (cms-S) . This event occurs more frequently in some inbred lines than in others, and may involve a change in the cytoplasm, corresponding to a change from cms-S to normal condition, or may involve a nuclear event, following which fertility is inherited in Mendelian fashion (J. R. Laughnan and S. J. Gabay-Laughnan, 1978, Maize Breeding and Genetics, D. B. Walden, ed.). We have also shown that nuclear integration of the fertility element may occur at different chromosomal sites, a property which led us to describe the element as an episome.
It is now known that some genetic backgrounds are highly favorable to this mutation and others are very stable. Moreover, among genetic backgrounds that exhibit rather high frequencies of the mutational event, there are striking differences in the proportions of mutations that are cytoplasmic versus nuclear in origin. In an attempt to determine the relative influence of cytoplasm and nucleus on both the frequency of the event and its origin, we undertook a backcrossing program involving a number of inbred lines carrying different subgroups of S cytoplasm as the nonrecurrent female parent, and the inbred line M825 as the recurrent male (maintainer) parent. Seven inbred lines that do not restore cms-S, that is WF9, 38-11, N6, K55, M14, 1153 and IllA, and five subgroups of S cytoplasm, S, Vg, I, ML and Rd, were represented among the nonrecurrent female parents. In all, 14 different line-cytoplasm combinations were employed. Each of these, with the possible exception of WF9-Rd, can be characterized as highly stable, and for most, although they have been grown for a number of generations, no mutations from male-sterile to male-fertile have been observed. WF9-Rd plants have occasionally produced fertile tassel sectors but the frequency of this event is less than 1 percent. On the other hand, M825-Vg male-sterile plants, when crossed with the M825 maintainer that was used as the recurrent parent in this experiment, produce an abundance of mutations to male-fertile condition. Most of these occur as plants with fertile tassel sectors, many fewer as plants with entirely fertile tassels (entires), and the vast majority of these mutations are cytoplasmic in origin. Among 7,558 plants of this type that were analyzed for male-fertile mutations, 748 (9.9%) had fertile tassel sectors, and 74 (0.97%) had entirely fertile tassels.
The backcrossing program referred to above was undertaken without selection for mutational ability and is now in the ninth backcross step. In the summer of 1978 we replanted the second (3X) and sixth (7X) backcross generations of this experiment and looked for mutations from male-sterile to male-fertile condition. These exceptions were scored with respect to whether they were entires or sectors, and were pollen checked in the field to determine whether the pollen was essentially normal or half-aborted; the former is a preliminary indication of a cytoplasmic origin for the fertile revertant, and the latter signals a nuclear basis for the change.
The results indicate a strong influence of the genome of the M825 recurrent parent on both frequency and origin of the fertile event. The data in Table 1 indicate an increase in mutations after three generations of crosses by the M825 recurrent parent. Since the data from the 14 different line-cytoplasm nonrecurrent sources have been combined in Table 1, it is useful to know that the mean number of plants per source in the 3X group was 152, and that the numbers ranged from 95 to 312 per source. At least one fertile mutation occurred in 13 of the 15 sources; only the WF9-Vg and the 38-11-Vg sources, with 106 and 102 plants to score, respectively, had none. After an additional four generations of backcrossing by the M825 male parent (7X series) there is an overall fourfold increase in mutation rate, from 2.2 to 8.3 percent. All but one of the 14 sources showed an increase in mutation rate from the 3X to the 7X generation. The single exception was the IllA-S entry whose mutation rate in the 3X generation, 15/309 (4.9%) was highest among the fourteen sources. The rate for the IllA-S source in the 7X generation was 9/269 (3.3%), a reduction that is not significant. On the other hand, 5 of the sources exhibited striking increases in mutation rates in the 7X generation: WF9-Vg from 0/106 in the 3X generation, to 19/135 (14.1%) in 7X; WF9-Rd from 7/100 (7.0%) to 23/112 (20.5%); N6-Vg from 3/107 (2.8%) to 25/159 (15.7%); K55-Vg from 4/306 (1.3%) to 26/254 (10.2%); and M14-ML from 1/141 (0.7%) to 19/196 (9.7%). The overall mutation rate of 8.3% for the 7X generation approaches, but is still significantly lower than, the mutation rate of 10.9% for the M825-Vg strain whose maintainer is the recurrent parent in this experiment. Nevertheless, two of the 14 line-cytoplasm sources, WF9-Rd and N6-Vg (see above) had significantly higher rates of mutation than the M825 strain, and three of the sources had rates that were not significantly different from the M825 rate. To summarize, the near replacement of the inbred line nuclear genomes of 15 line-cytoplasm sources with the M825 nuclear genome results in an increase in spontaneous mutation rate from virtually zero to an average rate of 8.3%, which is close to the rate characteristic of the recurrent M825 strain itself.
As indicated above, an attempt was made to analyze the pollen character of each male-fertile mutation. We were able to score over 90% (201/219) of mutations for pollen type. Mutations that are cytoplasmic in origin produce mainly normal pollen which, when involved in test-crosses with cms-S male-sterile testers, produces all male-sterile progeny (except for newly occurring mutations). On the other hand, 50% of the pollen produced by a nuclear revertant is shriveled and aborted; since the normal grains in such a sample carry the equivalent of an Rf restorer allele, this pollen, when used on male-sterile testers produces all male-fertile progeny whose pollen characteristics are the same as those of the male parent. Among several strains of maize whose cms-S versions produce frequent mutations to male fertility, one designated wx bm4 has a high frequency of nuclear and a low frequency of cytoplasmic events. Another, M825/Oh07, has about equal numbers of both. The M825-Vg strain whose maintainer counterpart was used as the recurrent parent in this experiment produces overwhelmingly cytoplasmic reversions to male fertility, the nuclear events being estimated at less than 5%. The summarized data on pollen character in Table 1 indicate that the nuclear genome exerts a strong influence over the type of mutational event. Though testcross data are not available yet, it can be inferred from pollen analysis that male-fertile mutations in the 3X generation are cytoplasmic and nuclear in origin, in equal proportions. In the 7X generation, that is, after four additional backcrosses using M825 pollen, the mutations are mainly cytoplasmic, less than 5% being nuclear in origin. This trend is apparent in each of the 14 pedigrees. Moreover, this shift in proportion of cytoplasmic:nuclear cases is due to both an absolute increase in frequency of mutations with a cytoplasmic origin as well as an absolute decrease in frequency of revertants with a nuclear origin. Appropriate adjustments for the relatively few mutations for which pollen analyses are not available indicate that there is more than a 6-fold increase in the frequency of cytoplasmic mutations from the 3X to the 7X generation, from 1.1% to 7.3%. The corresponding values for the nuclear mutations are 1.0% and 0.36%.
Members of the S group of male-sterile cytoplasms carry two unique species of satellite mitochondrial DNA (D. R. Pring, C. S. Levings, W. W. L. Hu, and D. H. Timothy, PNAS 74:2904-2908, 1977). Recent studies (see North Carolina State contribution this News Letter) of five cytoplasmic male-fertile mutations by the agarose gel electrophoresis method indicate that the mutation in each case was associated with disappearance of these unique mitochondrial DNAs. It is apparent that the genetic basis for the S male-sterile trait is localized in the mitochondrion. Molecular hybridization studies carried out with labeled satellite mtDNA (B. D. Kim, C. S. Levings, D. R. Pring, M. F. Conde, J. R. Laughnan, S. J. Gabay-Laughnan and R. J. Mans, FASEB abstract, 1979) suggest that mutation of cms-S to male fertility is correlated with integration of at least some of the satellite DNA sequences into the mitochondrial circular genome. These studies are therefore consistent with the view that the fertility element in cms-S systems is an episome whose integration into the mitochondrial genome leads to a spontaneous cytoplasmic reversion to male fertility. Although there has as yet been no similar analysis of the spontaneous nuclear revertants to male fertility in cms-S strains, the above evidence supports the notion that these represent corresponding integrations of the F episome(s) into one or another chromosomal site. Of special interest in this connection is the finding (C. S. Levings III, W. W. L. Hu, D. H. Timothy and D. R. Pring, MNL 52:96-98) that both satellite DNAs from mitochondria of S male-sterile maize form lariat-like (lollipop) structures when they are denatured and allowed to renature under appropriate conditions. They are therefore similar to the numbers of insertional (IS) elements that have been described in numbers of prokaryotes, and, we like to think, admirably suited for the integration role they presumably play in the mutational events described here.
Whatever the details of the integration phenomena may be, the evidence from the study reported here indicates that the recurrent male parent, in this case M825, has the predominant influence over the frequency and localization of the initial event. In other words, the nucleus, not the cytoplasm, has primary control over both the frequency and the site of integration.
J. R. Laughnan and S. J. Gabay-Laughnan
Return to the MNL 53 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage
