We have investigated the use of the transposable element Mu1 as a molecular tag for cloning. The major advantage of the Mutator system is its high rate of mutation. This could make Mutator the system of choice for tagging previously unmarked genes. However, the high mutation rate is probably due in part to the high copy number of Mu1 elements (about 30) in Mutator stocks and their ability to maintain their high copy number upon crossing to non-Mutator lines (Allemand and Freeling, pers. comm.). This high copy number of Mu1 elements militates against the use of Mu1 as a transposon tag.
There has been considerable interest expressed in the maize genetics community about cloning genes with a Mu1 probe. However, to our knowledge, no one has succeeded in using Mu1 homology alone to identify a target gene. We present below details of an unsuccessful attempt to clone the bronze gene using Mu1 as a probe that was initiated before bronze was cloned using an Ac tag, and suggestions arising from this work on how to overcome the high Mu1 copy number problem in the future.
I. A failed attempt to clone the bronze locus using Mu1 as a probe: The bronze locus was considered an ideal model system to test cloning with Mu1, since it had been well documented genetically, it was assumed to be present in single copy, and several mutable bronze alleles had been isolated by Don Robertson. We will designate mutable bronze alleles arisen in a Mutator stock as bz-Mum. We attempted to clone the bz-Mum4 allele isolated by Robertson in a Mutator line that had about 30 copies of Mu1, using Mu1 as a probe.
As a first step, we attempted to determine which Mu1-hybridizing band in a genomic Southern segregated with the bz-Mum4 allele. The high number of bands makes the analysis difficult if individual plants are assayed, so we developed a pooled seedling assay, described below, that averages out the newly arisen Mu1 bands that occur at each generation.
In a Mutator line with 30 copies of Mu1, approximately 15 of them will be passed to any one offspring in a mendelian fashion. Each individual in the progeny will receive a different subset of these parental copies. In addition, since Mu1 tends to maintain its copy number at approximately 30, each offspring will also have about 15 new copies of Mu1. Analysis of individual offspring would show about 30 copies of Mu1, making it very difficult to follow segregation of a particular band because of the high background. However, if one pools the DNA from many plants which are selected for the mutable allele of interest and compares the pooled DNA with that of pooled sib plants not carrying the mutation in question, then the Southern analysis becomes slightly easier. The newly arisen copies of Mu1 will not be the same in each plant. Therefore, in the pooled DNA these new bands will be diluted by the number of plants used, i.e., if 30 plants are used each new band will be 1/30th as dark as a regular band. Each parental band segregating independently of the selected mutable allele will be present in half of the progeny, and therefore will be half as intense as the band belonging to the selected gene. In addition, sib seeds not carrying the mutable allele will have all of the parental bands except those segregating with the targeted locus. Thus, the Southern of pooled DNA from plants selected for mutable vs. stable expression should show an intensity band difference.
Using this approach it was still not possible to positively identify one band in genomic Southerns of bz-Mum4 plants as corresponding to the bronze locus. However, a tentative assignment was made. A lambda library of 800,000 phage was constructed from a partial Sau3AI digest of genomic DNA and screened with a Mu1 probe kindly provided by Mike Freeling. Twenty-seven independent phage were isolated and characterized by Southern blots. Several different phage had restriction fragments similar in size to the putative bz-Mum4 Southern band. At about this time, we isolated a bz-specific probe from bz-m2, an Ac-mutable allele, so we used this probe in Southern blot hybridizations and to screen the Mu1-hybridizing phage. However, though the Southern analysis of bz-Mum4 digests confirmed that a Mu1-hybridizing band also hybridized to the bz-specific probe, none of the isolated phage were found to contain bz homologous sequences.
We concluded that even if one could identify by Southern analysis which copy of Mu1 was cosegregating with the target locus, cloning that gene would be very difficult because of the high copy number of Mu1. Similarly, in the cloning of the a1 locus, O'Reilly et al. (EMBO J. 4:877, 1985) found 1 correct clone out of 35 isolated and characterized. It is clear that if the Mu1 copy number can be lowered, the job of cloning will be greatly reduced. The following report outlines how this might be achieved.
II. Reduction of copy number of Mu1: Observations made during the course of our genetic analysis of spotted bz-Mum alleles suggest that it may be possible to reduce the Mu1 copy number. In our standard backcrossing scheme of Sh bz-Mum Wx to our W22 tester sh bz-R wx, we noticed in most ears a variable number of stable bronze seeds with the outside markers of the bz-Mum allele at too high a frequency (1 to 5%) to be double crossovers (Figure 1). Analysis of these stable bronze seeds showed that upon further backcrossing to our W22 tester or in selfs the phenotype remained bronze-stable; but upon crossing to sh bz-R wx sibs (i.e., extracted from a "Mutator background"), spotted seeds were again produced, although at a low frequency. This suggests that the change from a bronze-mutable to a bronze-stable phenotype was caused by the inactivation of the Mu1 element's ability to transpose, and that subsequent exposure to a Mutator background reactivated this ability.
The inability of the Mu1 element to transpose somatically apparently also reflects an inability to transpose germinally and thus an inability to maintain a high copy number. Test-cross progeny of the stabilized bronze plants described in the previous paragraph have a reduced copy number of Mu1 elements, as low as 8, and Southern analysis of the individual offspring clearly shows that only one Mu1-hybridizing band is common to all (Figure 2). Hybridization with a bronze-specific probe confirms that this band corresponds to the bronze gene.
Modification of the Mu1 element has been reported to be correlated with the stabilization of mutable bz2 phenotypes (Chandler and Walbot, MNL 57:96, 1985) and with the loss of Mutator activity (Bennetzen, unpub.). We have verified that the Mu1 element becomes modified when bz-Mum alleles change from a mutable to a stable null expression. At the bronze locus, the modification does not seem to extend beyond the Mu1 element itself, because the HinfI sites in bronze are still cleaved in bronze-stable plants while the HinfI sites inside Mu1 are not. In addition, we have found that, upon restoration of the mutable phenotype, most if not all copies of Mu1 are cleaved again by HinfI.
The change from a mutable bronze to a stable bronze phenotype is very easy to detect even though the transposition events in bz-Mum4 occur very late in seed development and, thus, the spotting pattern is very fine. The change from mutable to stable in a less easily scorable phenotype, (e.g., knotted, dwarf, nitrate reductase minus, etc.) would be hard, if not impossible, to detect. Thus, we propose to use the bronze-mutable to bronze-stable change as an indication of Mu1 element activity, and envision the following scheme for identifying and isolating genes harboring a Mu1 insertion:
1. Isolation of a mutant allele of the desired locus in an active Mutator line.
2. Crossing of the new mutant to a bronze-stable (ex bz-Mum) tester line where the Mu1 element has become modified and does not transpose.
3. Backcrossing to the tester and selection of seeds in which Mu1 has become modified as indicated by the stable bronze phenotype. One half of the plants derived from bronze seeds should be heterozygous for the mutation of interest and can be identified by selfing.
4. Repetition of the backcrossing scheme, determining at each generation the Mu1 copy number of the individual offspring by Southern analysis. The Mu1 copy number should be halved every generation, thus allowing the cloning of the desired locus with a Mu1 probe 2 to 3 generations following the isolation of the mutant allele.
Figure 2. Southern blot of Pst-cut "miniprep" genomic DNA of progeny of a stabilized bz-Mum plant probed with Mu1. Lanes 1, 3, and 4: testcross progeny of a stabilized bz plant. Lane 2: plant with restored bronze-mutable phenotype obtained from a cross between plants A and B in Figure 1. The arrow marks the position of the common bronze hybridizing band.
Ed Ralston and Hugo Dooner
Return to the MNL 60 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page