Articles
CLONING OF THE GENES FROM ERWINIA HERBICOLA EH252 NECESSARY FOR PRODUCTION OF AN ANTIBACTERIAL COMPOUND WHICH INHIBITS ERWINIA AMYLOVORA
Transposon induced Ant mutants do not protect immature pear fruit against fire blight as well as the wild type strain, indicating that this antibacterial compound is one of the factors involved in the inhibition of E. amylovora (Vanneste et al. 1990, 1992). In this study, we cloned the genes necessary for production of this compound, and we compared the ability to protect immature pear fruit from fire blight of Ant mutants complemented for antibiotic production with that of the wild-type strain.
A gene library of Eh252 constructed into E. coli Dh5
yielded 5 clones that inhibited the growth of E. amylovora on minimal medium.
The inhibitory compound produced by these clones of E. coli was found to be, as that produced by Eh252, inactive in the presence of proteases or histidine.
Plasmids isolated from these clones restored antibiotic production to all Ant mutants of Eh252, and conferred antibiotic production to different strains of E. coli. They carried inserts ranging from 23 to 12 Kb.
Two of these plasmids (pANT3 and pANT5) were used to complement 10:12, a transposon induced Ant mutant.
The complemented mutants (10:12 pANT3 and 10:12 pANT5) protected immature pear fruit as well as the wild type strain Eh252, and significantly better (P<0.001) than the mutant (10:12) or the mutant carrying the plasmid vector (10:12 pCPP25) (Fig. 1), confirming the role of this compound in the biological control of fire blight.
A 3.2 Kb fragment of DNA from pANT3 was cloned in both orientations into pUC19, yielding the plasmids pANT23 and pANT29. These two plasmids conferred antibiotic production to E. coli DH5∞, indicating that this 3.2 Kb fragment most probably contains the gene(s) and the promoter necessary for production of the inhibitory compound.
Analysis of unidirectional nested deletions of pANT23 and pANT29 revealed that only 2.4 Kb of the 3.2 Kb fragment were necessary for antibiotic production.
This was confirmed by transposon mutagenesis of the plasmids pANT23 and pANT29, using Tn5tac1 (Chow and Berg, 1989) and TnnphoA (Taylor et al., 1988). Of the 1000 TnphoA mutants analysed, the only 6 which expressed phosphatase alkaline had the transposon in the vector, suggesting that the inhibitory compound is not secreted in the periplasm via a leader sequence.
Over 90% of the insert from pANT23 and pANT29 have been sequenced.
Comparison of this DNA sequence with those in international databases might indicate the likely nature of this peptide and its mode of action.