Inheritance of somatic instability in Mutator lines

We have previously reported that somatic instability in the mutable bz2-mu1 allele recovered in a Mutator background can be lost (MNL 58:188; MNL 59:98). This loss can occur in both self and outcross progeny. The loss of somatic instability is correlated with de novo modification of Mu elements in these lineages detected by a lack of restriction by HinfI (V. Walbot, et al., 1985, Plant Genetics, ARCO-UCLA Symposium, M. Freeling, ed.; V. L. Chandler & V. Walbot, 1986, PNAS, in press). Robertson (Mol. Gen. Genetics 191:86, 1983) previously reported loss of the high forward mutation rate characteristic of Mutator stocks, and molecular analysis of such stocks indicates an increased level of Mu element modification as well (J. L. Bennetzen, 1985, Plant Genetics, ARCO-UCLA Symposium, M. Freeling, ed.). Thus, two indicators of transposable element activity-high mutation rate and somatic instability of alleles--are both affected by modification of Mu elements. For the purpose of this discussion Mutator lines which are losing or have lost somatic instability at bz2-mu1 will be termed OFF lines.

Initial experiments in 1983-1984 demonstrated that OFF lines remain off through subsequent outcrossing to non-Mutator lines and upon selfing. Attempts to "reactivate" the mutable phenotype by crossing an OFF line as female by an active Mutator line gave a low fraction of somatically unstable kernels in the progeny; more prominent was inactivation of a heretofore active Mutator line on crossing with an OFF stock. A few reciprocal crosses in which the OFF line was the male parent showed substantial reactivation of the bz2 mutable phenotype suggesting that there could be a maternal effect.

During summer 1985 larger tests for the stability of the OFF and active states of Mutator lines were conducted as well as tests for maternal effects on inactivation of an active Mutator line by an OFF line and on reactivation of an OFF line by an active one.

The original bz2-mu1 kernel was derived in 1982 from the cross of bz2 tester (W23/K55) by an active Mutator line (B139-7). The resulting kernel was planted in the greenhouse, and it transmitted the spotted kernel phenotype normally in a self (3:1 segregation sp:cl K) and when used as the pollen parent to the standard bz2 downtester or to the an bz2 deletion stock (1:1 segregation sp:cl K). In 1983 progeny from the self and outcross ears were planted and it was in this generation that loss of the spotted kernel phenotype was noted and correlated with the presence of modified Mu elements. The first question addressed here is what fraction of progeny from each type of cross showed aberrant segregation.

As shown in Tables 1-3, loss of the spotted kernel phenotype occurred in the second generation self and outcross progeny of the original mutable plant. The most severe loss occurred in the self lineage (Table 1), in which about three-fourths of the second generation progeny failed to transmit the expected ratio of sp:cl kernels. Five of 10 OFF ears were completely colorless, and the average level of spotted kernels was less than 10%. In the outcross progeny of the original mutable plant (Tables 2 & 3) fewer switches to OFF occcurred (15/43 ears) and the average percent of spotted kernels on affected ears was considerably higher than in the self lineage. Considering all of the data together it is clear that an active Mutator state is labile in these lineages and that it can be lost in a substantial percentage of the progeny from an ear with a normal segregation pattern in the previous generation.

The state of Mu elements has been reported for eight plants and their progeny in the C230 lineage; there is a highly significant correlation between the loss of the spotted kernel phenotype and the presence of modified Mu elements (V. Chandler & V. Walbot, in press, PNAS). Most lines with normal segregation had only Mu that could be completely restricted by HinfI; lines with abnormal segregation often had all modified elements, although spotted K from such ears always had at least a few unmodified Mu elements. Although molecular analysis is still in progress on the subsequent generations described below, I would like to report some of the genetic data at this time.

First, how stable is OFF? To test this colorless kernels from an ear of bz2 tester x C230-1 in which there were few spotted kernels (63:2842) were planted (E302) and progeny tested (Table 4). Of the colorless K, one-half should be bz2/bz2-mu1, but loss of somatic instability masks the mutable phenotype. At the molecular level 10/10 colorless K tested had modified Mu elements. In tests of this and other lineages (C230-3(X) and C231 plants) in three subsequent generations there have been no instances of spontaneous reactivation. Consequently, the OFF condition appears to be very stable.

Second, how stable is an active Mutator line? To test this progeny of bz2 tester x C230-7 and tester x C230-8 have been examined. These two plants were each crossed to >20 tester plants and each gave the expected 1:1 ratios. At the molecular level 5/5 C230-7 sp K tested showed complete Mu element restriction with HinfI; for C230-8 18/18 sp K and 16/16 cl K seedling DNA preps showed complete restriction. These two lines were our best-characterized active Mutator lines with no evidence for loss of the somatic mutability character or of Mu element modification. The data for part of the C230-8 lineage are reported in Table 5 (from bz2 x C230-8 = CF49 E301 E306).

Based on the 1:1 segregation of sp:cl kernels on the bz2 tester x C230-8 plant, colorless kernels were picked as likely to be simply bz2; on subsequent outcrossing no evidence of "spontaneous reactivation" was found, however, the test is very small compared to a similar test for reactivation in an OFF lineage (Table 4). Spotted K from the C230-8 active Mutator line transmitted the spotted kernel phenotype in the expected fashion in most self (95%) and outcross progeny (79% as male, 95% as female). Similarly, a second active line, C230-7, showed 66% normal segregation upon selfing (2/3), 78% on crossing to tester (29/37), and 100% normal ears when crossed by tester (10/10). The switch to OFF upon selfing was only 5% for C230-8 selfed compared to an average of 40% for the C230 family as a whole in the previous generation (Table 2). Thus, lines picked because of their normal segregation pattern and absence of modified Mu elements are more likely to remain active. However, the characteristics of the previous generation do not guarantee that a line will remain active, because both C230-7 and -8 produced some OFF progeny in the next generation.

Further evidence that active lines are labile comes from tests with subsequent generations. For example, the selfed progeny in Table 5 (CF49) showed 95% expected segregation. When 10 sp K from each of these ears (and four similar ears from another family) were tested by selfing again, 17/25 or 68% showed the expected 1:0 and 3:1 ratios. The ears with a deficiency in spotted K progeny showed close to the expected percent of spotted K.

Combining all data in tests of the stability of the active state of Mutator, it is clear that lines can turn OFF at any time and once OFF tend to remain so. The turning OFF is progressive, noticed first as a deviation in the percentage of ears showing expected sp:cl K ratios; the abnormal ratio ears often have a substantial percentage of spotted K. In the next and subsequent generations both the percentage of normal ears and the percentage of spotted K/ear fall.

I also noticed a maternal effect. Active lines are more likely to switch to OFF when used as male onto tester plants than when crossed by tester pollen. For C230-7 & -8 used as female 34/35 ears (97%) showed normal segregation and the one abnormal ear retained 26% sp K. In reciprocal crosses where these active lines were used as male, 57/71 ears (80%) showed normal segregation; on the ears showing a deficiency of spotted K there were still about one-fourth sp K. One might speculate that transposase or other factors required to maintain the activity of the Mu elements is found in a higher concentration in the maternal nucleo- or cytoplasm than in the pollen. Alternatively, we could be seeing the effects of Mu dosage in these crosses. The bz2 tester (W23/K55 hybrid) contains only a few highly modified copies of Mu elements and is unlikely to contribute any activities or dosage of Mu. Thus, the crosses of an active line as male may dilute the copy number of Mu below a threshold required for Mutator activities.

A maternal effect was also noted in the maintenance of the spotted K phenotype in the rare sp K (63:2482) from bz2 tester x C230-1. All of the cl K from this cross had modified Mu elements; of the 5 sp K examined at the molecular level, one contained modified elements but four did not. In the next generation (E303) these sp K differentially transmitted the sp K trait depending on the direction of the cross (Table 6). Although nearly all ears showed abnormal transmission, the percentage of spotted K was much higher when the Mutator plant was used as male. This effect is the opposite of that seen in the C230-7 & 8 lineages. In the C230-1 lineage most progeny contained modified Mu elements in the previous generation; thus, it was expected that in the subsequent generation most progeny would be OFF. The two hypotheses given above about the nature of the maternal effect could apply here also. However, a third alternative is also suggested, namely the maternal effect in this case might be explained by dilution of the "modification system" by transmission through pollen.

Further evidence that the C230-1 lineage contains an active modification system, responsible for switching active lines to OFF at high frequency, comes from examination of the spotted K on the C230-1 selfed ear which showed 124:59 sp:cI K ratio (p <0.04); at the molecular level 2/5 spotted K and 5/5 colorless kernels contained modified Mu elements. In subsequent generations almost no spotted K were recovered from crosses involving spotted K from C230-1 selfed (CC2 & 3, D90 & 91) or from the progeny of these ears (E305) (Table 7). Thus, in the C230-1 selfed lineage the rate of turning OFF is exceptionally high, even though a near normal sp:cI K ratio was found on the parent ear. The OFF condition is again very stable.

The last question is, what is the impact of crossing an active and an OFF line together? The first experiment that addresses this point involved crossing the C230-8 and C230-1 lineages. cl K of bz2 tester x C230-8 (E301), which should be simply bz2 in an active Mutator stock, were crossed as male and female by cl K taken from bz2 tester x C230-1 (E302). 10/10 seedlings from this C230-1 stock contain some modified elements. In exact reciprocal crosses the ability of individual E301 plants to reactivate the suppressed sp K phenotype of E302 individuals was examined. As shown in Table 4 the "spontaneous reactivation" of this family is very low. In crosses with E301, however, 20/36 plants showed at least a few spotted K, close to the expected fraction (18/36) of these plants that should be bz2/bz2-mu1. Averaging all ears (E302 as male or female) in which spots were found, the average percentage of spotted K was 27%. of 50% expected to be spotted. Twelve pairs of exact reciprocal crosses were completed between E301 and E302 plants in which spotted K were found in the progeny. Of these two instances were found in which only the E301 as female had any spotted K (22:278. 7:282 vs. 0:>300 for E302 x E301). Additional support for a positive maternal effect bv active E301 plants on reactivation of the cryptic bz2-mu1 somatic mutability comes from a comparison of the average percent of spotted K: with E301 as female. the average was 31%, with E302 as female the average was 25%.

The final evidence for a maternal effect comes from an "inactivation" experiment in which spotted K progeny of bz2 tester x C230-8 (E306) were crossed by cl K from the C230-3 x bz2 tester ear; this latter ear gave a 0:168 phenotypic ratio. 12/12 seedlings tested contained modified Mu elements, and in subsequent crosses to bz2 tester 22/22 ears were completely OFF. Thus, the C230-3 lineage is by all criteria an OFF line. As in the reactivation experiment, each individual was crossed as male to tester, and then exact reciprocal crosses were completed between active and OFF individuals (Table 8).

In the "inactivation" experiment there is again a strong maternal effect. All ears derived from an active line crossed by an OFF line showed the expected 1:1 segregation for sp:cl kernels. In the reciprocal crosses, however, only 3/13 ears showed normal segregation. Those showing abnormal segregation had an average of 28% spotted kernels suggesting that the switch to OFF is just beginning. It will be interesting to determine the transmission of the spotted K phenotype in subsequent generations and the correlation to modified Mu elements in this stock.

In summary, these data suggest that an active Mutator line can lose activity measured as somatic instability at a reporter mutable allele at a reasonably high frequency. The somatic instability of the previous generation is, in fact, a poor indicator of the behavior of the stock in subsequent generations: for example, essentially all spotted K of the C230-1 selfed ear failed to transmit sp K. Once OFF a Mutator line seems to remain off, and there are only rare instances of "spontaneous reactivation." Several kinds of maternal effects were noted in these materials. In lines showing just a minor loss of Mutator function (C230-7 & 8) such losses occurred preferentially when these plants were crossed as male onto tester. This maternal effect suggests that a positive factor(s) required to maintain Mutator activity is being diluted. In contrast, lines such as C230-1 in which the switch to OFF is occurring at high frequency are more likely to retain activity when crossed as male onto tester; it is tempting to speculate that they are escaping an active modification system. These interpretations are supported by the reactivation and inactivation experiments. When active and OFF lines were reciprocally crossed, an active line (C230-8) shows completely normal sp:cl K segregation when used as the female parent but not as male (Table 8); I hypothesize that positive factors in the active female line promote Mutator activity, while negative factors in an OFF line suppress it. Similarly, an OFF line is more likely to be reactivated when crossed as male onto an active line, again suggesting that the state of the female parent is critical in determining the extent to which the active and OFF states are maintained.

Until the nature of the activities encoded by autonomous and non-autonomous Mu elements have been delineated, it is impossible to know precisely what the relationship between element modification and the OFF state really is and whether this relationship depends on the number of modified and unmodified copies in a stock. However, it does appear that there are two aspects to the regulation of Mutator activity: active lines contain factors which positively regulate activity and OFF lines contain factors which negatively regulate activity. These activities condition the state of the embryo sac producing maternal effects on Mutator activity. A simple explanation for this would be that active lines contain transposase and OFF lines contain the DNA modification system that renders Mu elements inactive and that both the transposase and modification system transmit poorly through the pollen.

Table 1. Progeny (C231) of the Selfed Ear of the Original Mutable Plant

Table 2. Progeny (C230, MF11) of the Outcross Of the Original Mutable Plant

Table 3. Progeny (C233, C234) of the Cross of the Original Mutable Plant to an bz2

Table 4. Test for spontaneous reactivation of somatic instability in the C230-1 lineage.

Table 5. Transmission of the spotted kernel phenotype in progeny of bz2 tester x C230-8.

Table 6. Transmission of the spotted K phenotype by the C230-1 lineage.

Table 7. Transmission of the spotted K phenotype from progeny of C230-1(X).

Table 8. Effect of crossing active and OFF Mutator lines.

Virginia Walbot

Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors.

Return to the MNL 60 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page