No abstract

Most angiosperms are thought to share strict maternal inheritance of both plastids and mitochondria. Exceptions have been described and analysed, especially for plastids. However, the lack of phenotypic markers and the use of RFLPs on small samples may have biased the prevailing view of organelle inheritance by underestimating the occurrence of low-frequency paternal transmission of organelles. According to Muller's Ratchet, some recombination among organelle genomes is required, which would necessitate at least occasional biparental transmission. Uniparental inheritance can reduce the spread of selfish genetic elements and maintain good combinations of alleles. However, this does not explain why organelles transmitted by both parents have not invaded populations with uniparental inheritance. A link between outcrossing reproductive systems and the occurrence of biparental transmission suggests that plastids may play more of a genetic role in their inheritance than is usually assumed. Their prevailing non-Mendelian mode of inheritance thus remains to be convincingly explained.

SummaryThe successful exploitation of natural genetic diversity requires a basic knowledge of the extent of the variation present in a species. To study natural variation in Arabidopsis thaliana, we de®ned nested core collections maximizing the diversity present among a worldwide set of 265 accessions. The core collections were generated based on DNA sequence data from a limited number of fragments evenly distributed in the genome and were shown to successfully capture the molecular diversity in other loci as well as the morphological diversity. The core collections are available to the scienti®c community and thus provide an important resource for the study of genetic variation and its functional consequences in Arabidopsis. Moreover, this strategy can be used in other species to provide a rational framework for undertaking diversity surveys, including single nucleotide polymorphism (SNP) discovery and phenotyping, allowing the utilization of genetic variation for the study of complex traits.

FRIGIDA (FRI) is a major gene involved in the regulation of flowering time in Arabidopsis thaliana. Nucleotide variation at this gene was investigated by sequencing 25 field ecotypes collected from western Europe. Genetic diversity at FRI was characterized by a high number of haplotypes and an excess of low-frequency polymorphisms. A large excess of intraspecific nonsynonymous variation associated with low synonymous variation was detected along the first exon in the FRI gene. In contrast, no excess of nonsynonymous divergence was detected between A. thaliana and A. lyrata. The Tajima and McDonald and Kreitman tests, however, suggested that this gene has evolved in a nonneutral fashion. Nonsynonymous variation included eight loss-of-function mutations that have probably arisen recently and independently in several locations. A phenotypic evaluation of the sequenced ecotypes confirmed that these loss-of-function mutations were associated with an early-flowering phenotype. Taken together, our results suggest that DNA polymorphism at the FRI gene in A. thaliana from western Europe has been shaped by recent positive selection for earliness in a set of isolated populations.

Lines of Chiamydomonas were selected for growth either in Light or in Dark conditions for several hundred generations. Evolved lines that grew well in the environment of selection grew less well in the other environment, so that negative genetic correlation between Light and Dark growth was created by selection. The existence of a cost of adaptation was confirmed by reverse selection. The lines were also exposed to environments that varied either in space or in time with respect to Light and Dark conditions. Specialization (genetic variation) was retained in spatially variable environments, whereas generalization (phenotypic plasticity) evolved in temporally varying environments. The original negative correlation between adaptation to Light and Dark conditions seemed to be caused primarily by mutation accumulation rather than by antagonistic pleiotropy. It was thereby possible to select a generalist type nearly as well adapted in each environment as a specialist line.