Annals of Botany

ALS well in mutant medic

‘Test the trait, not the breeding method’ has almost become a mantra in my own contributions to the UK’s debate on GM crops. For me, this point is well illustrated in the paper by Oldach et al., Urrbrae, South Australia (pp. 997–1005). The authors describe the use of sulfonylurea (SU) herbicides with cereal crops. Although these herbicides are regarded as safe, nevertheless their rate of degradation in alkaline soils is slow; slow enough to affect the annual medics (Medicago spp.) grown in rotation with the cereals. An SU-tolerant cultivar of M. littoralis (‘Angel’) has been developed by mutagenesis-based breeding and the main aim of the authors has been to identify the molecular basis of this mutation. Analysis of segregation ratios of the F2 populations from crosses between ‘Angel’ and intolerant M. trunculata indicated a single dominant gene. Based on the mode of action of SU herbicides, this was likely to be a mutant form of ALS, the gene that encodes acetolactate synthase, an enzyme involved in synthesis of branched-chain amino acids. The sequence of the Arabidopsis thaliana ALS gene was used to interrogate the Medicago database, revealing two homologues located respectively on chromosomes 2 and 3. Linkage analysis using known markers then revealed that the herbicide-tolerance trait was associated with a region of the chromosome 3 containing the ALS locus. Sequencing of the wild-type and mutant ALS genes revealed a single amino-acid change from proline to leucine. These data facilitated the development of a diagnostic marker for SU tolerance while RT–PCR showed that ‘Angel’ does indeed express the mutant ALS. Theway is thus open for use of the mutant allele in Medicago breeding programmes. However, a similar mutant ALS from Arabidopsis has been used to transform Nicotiana and Brassica napus by GM techniques. To return to my opening comment: which is more important here, the genetic trait or the breeding method?