The data indicate that FA1090(M1) possessed a small insertion of 7 nucleotides about midway through the coding sequence, producing a frame shift mutation in nfsB. This genetic data supported the hypothesis that the nitrofurantoin resistant phenotype is due to the loss of nitroreductase activity. Conclusive evidence that this gene was responsible for nitrofurantoin resistance was obtained by deleting the coding

sequence for this gene from FA1090 and then demonstrating that FA1090NfsB-BsmIS lacked nitroreductase CT99021 molecular weight activity (data not shown). Identification of the genetic basis of spontaneous nitrofurantoin resistant mutants We isolated numerous independent spontaneous nitrofurantoin resistant mutants and determined the DNA sequence of the ABT-737 mw nfsB gene in these strains. Most of these mutants (90%) possessed the insertion of an adenine in a run of 5 adenines near the beginning of the gene, suggesting a bias for base insertion during

DNA replication at this position. This gene contains three other polynucleotide runs of five nucleotides distal to the start codon; 2 poly adenines and one polythymine. Interestingly, even though we were able to isolate base insertions at each of these three clusters, none of the clusters showed the elevated propensity for generating spontaneous mutations. To eliminate the bias introduced by the 5 bp polyadenine run at the 5′ end of the all gene, this DNA sequence was altered to remove the poly-A tract, while preserving the corresponding amino acid sequence. The plasmid, pEC3 was constructed as described in figure 4. Plasmid DNA was isolated from E. coli and DNA used to transform N. gonorrhoeae. Spectinomycin resistant transformants were identified, and DNA sequence analysis of a PCR amplicon derived

from the constructed strain indicated that the derivative of FA1090, FA1090-NfsB(mod) contained the desired sequence modification. Nitroreductase assays of this strain indicated that it possessed wild-type NfsB activity (data not shown). Figure 4 Schematic illustrating the strategy used to modify the nfsB coding region. Each numbered arrow corresponds to the procedures summarized below: 1: PCR using primers NfsBsmI-3F and NfsBsmI-2R to introduce a BsmI recognition sequence and to alter a poly-A tract. 2: Treatment with S1 nuclease to create blunt ends, polynucleotide kinase to phosphorylate 5′ ends, and T4 DNA ligase. E. coli were transformed using this construct (pEC2). Plasmid DNA was isolated by alkaline lysis. 3: Treatment with BsmI to generate pEC1. Digestion product was ligated with T4 DNA ligase. The construct was transformed into E. coli. 4: pEC1 was amplified with primers dwnstrm-F and dwnstrm-R.