University of Florida

New shrunken2 mutants derived from an Ac/Ds stock with totally unexpected properties
--J.R. Shaw, D. R. McCarty and L. C. Hannah

In an attempt to use Ds as a highly efficient mutagen of the shrunken2 locus, we have uncovered an interesting and unexpected set of mutants. In brief, an analysis of 13 newly?derived mutants reveals that all are homozygous lethal, all are transmitted through the male parent with 50% expected frequency, and those monitored are associated with loss of function of the closely linked A1 gene. However, none is a deletion for the Sh2 locus. A brief summary of the results is given below.

Initially 44 revertants from the Ds?containing allele sh2?m1 were collected. This mutant is described in Giroux et al., Proc. Natl. Acad. Sci. 91:12150, 1994. Revertants were made homozygous and approximately 1/2 of these were placed in a detasselling block and pollinated by a sh2 mutant, sh2?R in the sweet corn 'Florida Stay Sweet'. This mutant contains a very large insertion in the 5' end of the gene (Shaw and Hannah, unpublished). Approximately 150,000 kernels from each revertant were monitored for the occurrence of new sh2 mutants. One and only one revertant, Rev25, was found to produce new sh2 mutants at very high frequencies (several per thousand seed). While the vast majority of the resulting mutant seed did not germinate, we did find about 25 new, heritable mutants. Most of these were analyzed as described below.

Plants derived from mutant seed of the detasselling block were self?pollinated and crossed onto wild-type plants. Seeds of the cross were planted, numbered, selfed and crossed onto 'Florida Stay Sweet'. Leaf samples were also taken and Southern analysis was used to identify plants containing the mutant sh2?R allele. The large 5' insertion in this mutant leads to diagnostic restriction fragments. A more complicated crossing scheme was used in those cases in which the initial outcross to wild-type was not performed.

Unexpectedly, selfed seed from the heterozygotes above did not include typical sh2 seed. Rather, these ears contained paper?thin seeds composed almost entirely of pericarp. In cases of excellent seed set, these aborted seeds were seen only after shelling. That these plants were sh2 heterozygotes however, was seen in the test cross. Typical shrunken2 seed were found in the cross to 'Stay Sweet', albeit at frequencies of approximately 25%. Of the approximately 25 new mutants initially isolated, we failed to identify any that behaved as typical sh2 mutants (the standard mutant phenotype and normal transmission through the male). In those families not exhibiting the lethal seed phenotype, Southern analysis showed that all plants contained the sh2?R allele. Additional plants from these families will be analyzed in the future; however, we cannot exclude the possibility that the new mutant in each of these families totally lacks male transmission.

Homozygous lethality and reduced male transmission are hallmarks of mutants of this region of chromosome three that are thought to be deletions. Three X?ray induced mutants, a-x1, a-x2 and a-x3 are known. These lack A1 and Sh2 function and are associated with either reduced or no male transmission. These mutants are also homozygous lethal and presumably lack a gene necessary for normal seed development. These mutants have been interpreted as deletions for that portion of chromosome 3 that contains at least these three genes.

In an attempt to extend the analysis of the new mutants and their apparent similarity with the a-x series of mutants, we monitored a number of the newly created, homozygous lethal sh2 mutants for A1 function. The revertant, Rev25, from which most but not all of the newly?created sh2 mutants were derived, contains a functional A1 allele. However, crossing of a number of the new sh2 mutants derived from Rev25 to an a1 tester showed that these mutants lacked A1 function. We conclude that the mutation that gave rise to the loss of Sh2 function also abolished A1 function as well as the gene needed for seed development. In this regard, these mutants are identical to the a-x series of mutants.

Surprisingly, while these mutants appear from more classical types of analyses to be deletions, Sh2 sequences are still present in these mutants. Because the new mutants are homozygous lethal, paired samples (wild-type/new mutant and sh2?R/new mutant) were analyzed on Southern blots. The large 5' insertion in sh2?R gave rise to unique 5' fragments, and RFLPs 3' to the gene were used to monitor variation in that region. Of the 13 new mutants analyzed, all contained 5' Sh2 fragments. Clearly, these are not deletions for this gene.

Probing of the 3' portion of the gene proved less straightforward. A unique fragment was found in 12 of the 13 mutants. This fragment was of a different mobility than that found in sh2?R, Rev25 or in any of the wild-type alleles analyzed. While differing from these other alleles, the 3' fragment was indistinguishable in mobility from all 12 mutants. Eleven of the 12 mutants arose from Rev25 while the 12th mutant was derived from another revertant, Rev31. Two other features are interesting about the 3' portion of Sh2 in the new mutants. First, the fragment is always of less intensity than that of the corresponding fragment found in the heterozygote. This is not the case with the 5' fragment. Secondly, the 3' fragment was present in the heterozygote of the new mutant sh2?M15 with sh2?R but was absent in the paired wild-type/sh2?M15 sample. Thus it appears that this fragment is genetically unstable. The results from the paired samples involving sh2?M15 may suggest that the fragment is meiotically separable from the sh2 locus. However, since the DNA used in those blots revealing lesser intensity of this band was always derived from single plants, it would appear that this DNA maybe mitotically unstable.

The data to date show that while the 5' portion of Sh2 is always present in these mutants and is of the intensity expected, in comparison to the paired allele, this is not the case with the 3' portion of the gene. This fragment is always of less?than?expected intensity and is missing in some plants. The point delimiting these regions has not been identified although it must lie within or close to the 3' portion of intron 13 of the gene. Intron 13 is the largest within Sh2 (1.8 kbp). Surprisingly, probing of Southerns with this intron and selected subclones derived from it reveals that these sequences are very highly repetitive in the maize genome. Whether this fact is related to the mutant formation is currently unknown.

The transposable element Ds may not be involved in the creation of these new mutants. We also analyzed a new mutant derived from a wild-type allele of uncertain origin. The properties of this new mutant -- homozygous lethality, reduced male transmission, reduced intensity of the 3' fragment -- are indistinguishable from those of the mutants described above. We (Clancy, unpublished) sequenced selected portions of the wild-type allele and can detect no evidence that this allele is related to the progenitor of sh2?m1. Therefore, we have no evidence that Ds is involved in the creation of these new sh2 mutants.

These data are relevant for a number of reasons. First they call into question the reliability of using reduced male transmission and the loss of function of more than one gene as definitive criteria for deletions. In our case, we repeatedly isolated sh2 mutants that behaved like deletions by these two criteria, yet the shrunken2 gene sequences are still present. We are currently examining a-x1 and a-x2 at the molecular level to determine if any Sh2 sequences are present in these stocks. The data also point out a problem of using a closely linked transposable element to efficiently isolate new mutants of a particular gene. We screened approximately 25 revertants of a Ds?induced mutation and obtained no evidence that any could produce new mutants via Ds reinsertion.

The mechanism giving rise to these new mutants is clearly not presently understood. One possibility that explains all of the unexpected observations is the following: sequences within intron 13 of Sh2 pair with homologous sequences that lie distal to the A1 locus and the gene required for normal seed development. Recombination then occurs, giving rise to a linear chromosome and a ring containing all sequences lying between these two homologous sequences. Following recombination, the 5' portion of Sh2 would remain on chromosome 3 whereas the 3' portion as well as A1 and the seed development gene would be found on the ring. This model is in agreement with the orientation of A1 and the transcriptional unit of Sh2, as recently demonstrated by Schnable and colleagues, explains the inferred somatic instability (instability of ring chromosomes) of the 3' but not the 5' portion of Sh2, exploits the repetitiveness of the sequences found within intron 13 of Sh2, and explains the fact that this mutation is recurrent and independent of the transposable element Ds. It is currently under test.

The phenotype exhibited by these new sh2 mutants, paper-thin seed composed almost entirely of pericarp, is common in maize genetics. With this letter, we ask for wild-type seed from selfed ears segregating for this phenotype. In return, we will determine whether these lack Sh2 function and, in doing so, gain some insight into the frequency of this event in maize. 

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

Return to the MNL 69 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page