A major drawback to study gene functions in plant systems is the lack
of an effective gene knockout strategy. With a large number of plant g
enes isolated and the accelerating pace by which this collection is gr
owing, the need for their functional analyses at the whole plant level
has become increasingly urgent. Here evidence is reported for the fir
st successful disruption of a non-selectable gene in Arabidopsis thali
ana by creating a mutant of the TGA3 locus via targeted insertion of t
he bacterial neo gene conferring kanamycin (Km) resistance. A beta-glu
curonidase (GUS) expression unit outside the region of homology was us
ed as a screenable marker to distinguish homologous recombination even
ts from those of ectopic insertions. PCR amplification coupled with So
uthern blot screening identified two putative homologous recombination
events among 2580 Km(r) calli. One callus line was subsequently isola
ted and the structure of the targeted TGA3 allele confirmed by Souther
n blot analyses. This study demonstrates the feasibility of targeting
a non-selectable locus in Arabidopsis. Combined with future improvemen
ts in negative selection strategies and efficient transformation metho
dologies, gene replacement studies in plants could become a routine te
chnique.