Genetic study of the loss and restoration of Mutator transposon activity in maize: evidence against dominant-negative regulator associated with loss of activity.

The Mutator system of transposable elements is characterized by a family of transposons called Mu transposons that share common termini and are actively transposing in Robertson's Mutator (Mu) lines of maize. Mu lines lose transposition activity during propagation by either outcrossing or inbreeding. This loss of transposition activity, which can occur at non-Mendelian frequencies, is in the form of loss of forward transposition activity resulting in a decrease in the generation of new mutations, as well as the loss of mutability of Mu transposon induced mutations, and it has been correlated with hypermethylation of the Mu elements. Previous studies have concluded that restoration of Mutator transposon activity by crossing inactive lines back to active lines is incomplete or transient, and depends upon the sex of the inactive parent. Further, it has been proposed that the inactive system is dominant to the active system, with the dominance possibly mediated through a negative regulatory factor that is preferentially transmitted through the female. In this study, we have examined the frequencies of loss and restoration of Mu transposon activity using a Mu line carrying an insertion in the bronze 1 locus. We find that transmission of Mu transposon activity to non-Mu plants can occur at high rates through males and females, but individual cases of decreased transmission through the male were observed. We also find that in crosses between inactive-Mu and active-Mu plants, reactivation was efficient as well as heritable, regardless of the sex of the inactive parent. Similar results were obtained whether the inactivation occurred in an outcross or a self. In all cases examined, loss of Mu transposon activity was correlated with hypermethylation of Mu elements, and reactivation was correlated with their demethylation. Our results indicate that an inactive Mu system does not exhibit dominance over an active Mu system. We conclude that contrary to current models, inactivation and its maintenance is not obligatorily associated with a dominant negative regulatory factor whether nuclear or cytoplasmic, and we propose a revised model to account for these and other observations.