Increased outcrossing in hothead mutants

Abstract

Arising from: S. J. Lolle, J. L. Victor, J. M. Young & R. E. Pruitt Nature 434, 505–509 (2005); Lolle et al. reply

Lolle et al.¹ report that loss-of-function alleles of the HOTHEAD (HTH) gene in Arabidopsis thaliana are genetically unstable, giving rise to wild-type revertants. On the basis of the reversion of many other genetic markers in hth plants, they suggested a model in which a cache of extragenomic information could cause genes to revert to the genotype of previous generations. In our attempts to reproduce this phenomenon, we discovered that hth mutants show a marked tendency to outcross (unlike wild-type A. thaliana, which is almost exclusively self-fertilizing²). Moreover, when hth plants are grown in isolation, their genetic inheritance is completely stable. These results may provide an alternative explanation for the genome wide non-mendelian inheritance reported by Lolle et al.

Main

Initially, we constructed hth-12 gl1-4 double-mutant plants in the Columbia ecotype, reasoning that HTH and GL1 should revert independently because they are on different chromosomes. hth-12 DNA carries a transfer-DNA (T-DNA) insertion (SALK_024611) and gl1-4 is a guanine-to-adenine (G-to-A) transition mutation (like that shown previously to revert¹) that changes the start codon of the trichome gene GL1 (ref. 3) from ATG to ATA. Among 1,597 progeny of hth-12 gl1-4 plants, 10 were phenotypically GL1 (normal trichomes). Genotyping based on polymerase chain reaction showed that nine were heterozygous for gl1-4, and one was GL1/GL1. Surprisingly, the nine GL1/gl1-4 plants were also heterozygous for hth-12, and the GL1/GL1 homozygote was homozygous for HTH. These observations are most easily explained by pollen contamination (nine heterozygous plants) and seed contamination (one homozygous plant). We also found a single hth-12 heterozygote that was still homozygous for gl1-4, which could be explained by pollen contamination from nearby gl1-4 plants.

To test whether pollen contamination could be a source of apparent hth genetic reversion, we grew homozygous hth-12 plants either in a mixed population (near to, but not touching, plants with varied genotypes) or in an isolated room containing only hth-12 plants. In one experiment, the progeny of plants grown in the mixed-growth room showed 19/245 revertants (Table 1). Eighteen of nineteen revertants segregated the erecta phenotype in the next generation, suggesting that they arose from pollen contamination by nearby erecta-containing plants.

Table 1 Outcrossing in hth mutants

In a second mixed-population experiment, 18/415 plants were phenotypically HTH. All 18 contained a BIN2-1::GFP transgene⁴, which was present in other plants grown in the room (Table 1). In contrast, not a single revertant was found among 932 progeny of hth-12 plants grown in isolation.

We repeated these experiments with the originally reported hth-8 and hth-5 alleles in the Landsberg erecta (Ler) ecotype^1,5. We found that hth-8 plants grown in mixed populations yielded 156/994 progeny with a HTH phenotype. Most were either ERECTA or contained BIN2-1::GFP (Table 1). However, hth-8 plants grown in isolation gave exclusively hth progeny, none of which was ERECTA (Table 1). Similar results were obtained with hth-5.

Our results indicate that hth mutants are particularly susceptible to pollen contamination, possibly because the hth floral organ fusion defects lead to inefficient self-pollination and exerted stigmas¹, or because of changes in cuticle composition⁵. This tendency to outcross may provide an alternative explanation for the apparent genetic instability of hothead mutants.