VA Wadsworth Anaerobe Laboratory and UCLA School of Medicine, Los Angeles, California, USA
Bacteroides fragilis group organisms are among the anaerobes most frequently isolated from clinical infections, and are among those most resistant to antimicrobial agents. In aerobes (E. coli, Salmonella, and Pseudomonas), the contribution of porin molecules to antimicrobial resistance is well documented. Similar research in Gram-negative anaerobes has been hampered by the lack of knowledge about the structure-function characteristics of their outer membrane proteins. We are analyzing the structure function relationships of outer membrane proteins of B. fragilis. One of these proteins, HMP-1 (heat-modifiable protein 1), shares some characteristics with aerobic OmpA proteins and is discussed here.
BLAST, BLITZ and several other data base searches were used to see whether the sequences derived from the amino acid sequencing of peptide fragments from HMP-1 yielded any reasonable matches. No likely matching proteins were found. PROPSEARCH (EMBL) was used to analyze the amino acid analysis results for an outer membrane protein from Bacteroides fragilis. PROPSEARCH neglects the order of amino acid residues in a sequence, but includes consideration of amino acid composition, bulkiness, molecular weight, content of small residues, and average hydrophobicity and charge. Membrane proteins from Neisseria meninigitidis (e.g. porA) were listed as the likeliest matches, but several other membrane proteins (Psuedomonas aeruginosa porin, E. coli ompA precursor) were included as well. AG SSCP analysis ( prediction of secondary structural content from amino acid composition) (EMBL) predicted that HMP-1 is likely composed of 68.5-80.5% coil contents and 19.5-31.5% alpha helical structure (similar to the aerobic OmpA proteins.).
An oligo probe probe was derived from an amino acid sequence an SDS-PAGE purified fragment of HMP-1 according to accepted codon use tables. After negative results on southern blot analysis, we investigated Bacteroides gene and amino acid sequences in GenBank and other databases and found that the codon usage was quite different than that assumed in these common protocols. This is in accord with the phylogenetic distance between E. coli (for which many of these were derived) and Bacteroides, which is as great as that between gram-positive and gram-negative organisms. Using this information, we developed a probe sequence for a probe which differed significantly from the first. We also used the CODON USAGE TABLE for B. fragilis using the codon usage database from GenBank to confirm our codon choices.
PROSITE (EXPASY) was used to find a consensus sequence for the aerobic OmpA molecules, which is found at the C-terminal end of the protein and is probably peptidoglycan associated. The B. fragilis HMP-1 sequence, when identified, will be examined to see if it contains this sequence or a homolog of this sequence. Other predictor-based programs will be used to compare HMP-1 to aerobic OmpA proteins.