Characterization of the rfc region of Shigella flexneri.

The O antigen of the Shigella flexneri lipopolysaccharide (LPS) is an important virulence determinant and immunogen. We have isolated S. flexneri mutants which produce a semi-rough LPS by using an O-antigen-specific phage, Sf6c. Western immunoblotting was used to show that the LPS produced by the semi-rough mutants contained only one O-antigen repeat unit. Thus, the mutants are deficient in production of the O-antigen polymerase and were termed rfc mutants. Complementation experiments were used to locate the rfc adjacent to the rfb genes on plasmid clones previously isolated and containing this region (D. F. Macpherson, R. Morona, D. W. Beger, K.-C. Cheah, and P. A. Manning, Mol. Microbiol 5:1491-1499, 1991). A combination of deletions and subcloning analysis located the rfc gene as spanning a 2-kb region. Insertion of a kanamycin resistance cartridge into a SalI site in this region inactivated the rfc gene. The DNA sequence of the rfc region was determined. An open reading frame spanning the SalI site was identified and encodes a protein with a predicted molecular mass of 43.7 kDa. The predicted protein is highly hydrophobic and showed little sequence homology with any other protein. Comparison of its hydropathy plot with that of other Rfc proteins from Salmonella enterica (typhimurium) and Salmonella enterica (muenchen) revealed that the profiles were similar and that the proteins have 12 or more potential membrane-spanning segments. A comparison of the S. flexneri rfc gene and protein product with other rfc and rfc-like proteins revealed that they have a similarly low percentage of G + C content and have similar codon usage, and all have a high percentage of rare codons. An attempt to identify the S. flexneri Rfc protein was unsuccessful, although proteins encoded upstream and downstream of the rfc gene could be identified. Examination of the distribution of rare or minor codons in the rfc gene revealed that it has several minor codons within the first 25 amino acids. This is in contrast to the upstream gene rfbG, which also has a high percentage of rare codons but whose gene product could be detected. The positioning of the rare codons in the rfc gene may restrict translation and suggests that minor isoaccepting tRNA species may be involved in translational regulation of rfc expression. The low percentage of G + C content of rfc genes may be a consequence of the selection pressure to maintain this form of control.