The proteins of nontypable and type b Haemophilus influenzae isolates were characterised using two-dimensional polyacrylamide gel electrophoresis (2-D PAGE). Coomassie Brilliant. Blue R-250 was used for protein detection. Two hundred and twenty eight proteins were resolved from whole cell lysates prepared from a standard nontypable H. influenzae strain (designated HI-64443) when isoelectric focusing was used for the first-dimensional separation of 2-D PAGE. When nonequilibrium pH gel electrophoresis (NEPHGE) was used to separate basic proteins in the first dimension, 50 proteins were detected for HI-64443; 20 of the basic proteins detected were considered to be unique for this separation protocol. The apparent molecular weights and isoelectric points were determined for 82 of the proteins resolved for HI-64443. The variation of the proteins from the standard bacterial strain (HI-64443) was determined for nontypable H. influenzae isolates. On the basis of their electrophoretic mobilities, 17.5% of the proteins of HI-64443 were shared by four other nontypable H. influenzae strains analysed. These data identified both conserved and variable proteins among the nontypable H. influenzae isolates analysed. The results obtained indicated that 2-D PAGE was able to discriminate nontypable H. influenzae into population clones identified by other procedures. The 2-D protein profiles obtained for type b H. influenzae strains were similar to those obtained for nontypable H. influenzae strains. The extent of the protein variation observed between type b and nontypable H. influenzae strain was similar to that observed among nontypable strains alone. These data are discussed in relation to the application of 2-D PAGE as a tool for studies on bacterial epidemiology and for the analysis of the genome structure and gene expression of Haemophilus influenzae.

Nippostrongylus brasiliensis induces a biphasic anorexia in laboratory rats, the first phase coincident with lung invasion (ca day 2) and the second when the worms mature in the intestine (ca day 8). Using the anthelminthic, mebendazole (MBZ), N. brasiliensis infections of the rat were eliminated between the first and second anorexic episodes. This intervention prevented the expression of the second phase of anorexia. Rats exposed to a second infection with N. brasiliensis, 3 weeks after the primary infection, exhibited only a first phase anorexic response which was not influenced by MBZ termination of the primary infection. The lower cumulative food intake and weight gain of all infected rats after 8 days of infection were accompanied by elevated plasma insulin and, in some individuals, by elevated plasma leptin, compared with uninfected controls and previously-infected MBZ-treated rats. Messenger RNA levels for neuropeptide Y were higher in the hypothalamic arcuate nucleus of 8-day infected rats than in recovering MBZ-treated animals. Inoculation of rats with heat-killed N. brasiliensis larvae failed to induce anorexia and did not alter the severity of biphasic anorexia on subsequent injection of viable larvae. The first anorexic episode is therefore dependent upon viable migrating larvae. The second phase of anorexia clearly requires the continuing presence of the parasite beyond the lung phase. Viable migrating larvae are also required to confer 'resistance' to reinfection.

Length variations of Haemophilus influenzae outer membrane porin protein P2 were found at the DNA and protein levels, notably in non-capsulate strains. Protein length, measured by SDS-polyacrylamide gel electrophoresis, was found to correlate with the length of the gene, measured by polymerase chain reaction amplification, and ranged from 35-42 kDa and 970-1090 nucleotides, respectively. This represents a length variation of some 15%. The genetic location of these variations was studied by restriction enzyme mapping 10 of the non-capsulate strains revealing further polymorphisms at the DNA level. All 10 strains were distinct and differed from a type b strain. The conservation and assortment of the different restriction sites in the alleles is discussed in relation to the very great diversity previously described for this protein and of the whole genome itself in non-capsulate strains. The roles of selection, horizontal gene transfer, and transformation in generating this diversity are discussed.

Thirty-four clinical isolates of noncapsulate Haemophilus influenzae representing isolates with either related or dissimilar patterns of whole-cell polypeptide profiles on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) were further characterized by restriction enzyme analysis (REA) and rRNA gene restriction patterns. Total cellular DNA was extracted by a rapid, microcentrifuge-scale method and digested with BamHI, which gave a pattern of about 18 discrete bands. This confirmed the five closely related groupings suggested by SDS-PAGE. Isolates dissimilar by SDS-PAGE were also distinguishable by REA. However, there was no correlation between the degrees of similarity estimated from whole-cell polypeptide profiles and those obtained from REA for the dissimilar isolates. Therefore, inferences of genetic relatedness made on the basis of these data should be interpreted with caution. rRNA gene restriction patterns also confirmed the groupings suggested by the other two techniques. We conclude that the three methods were highly discriminatory and that whole-cell polypeptide patterns or REA with BamHI would be appropriate techniques for epidemiological studies of noncapsulate H. influenzae.

The characterization and comparison of isolates of bacterial species by comparing restriction enzyme digests of their chromosomal DNA (fingerprints) is highly discriminatory for different strains and allows similarities between them to be readily determined. However, the utility of the technique is dependent on the selection of appropriate restriction enzyme(s) and on the method of determining the similarities between the fingerprints generated. We report here a system which circumvents these two problems. The restriction enzyme is selected from amongst those which have a suitable frequency of restriction for given enzymegenome combinations. The frequencies of restriction enzyme recognition sites are calculated from the frequencies of di-and trinucleotides in sequenced genes from the species of interest using Markov chain analysis. Fingerprints are compared by dividing them up into sections with DNA size standards, scoring the number of bands in a few of these sections, and comparing these scores (numerical profiles) to establish similarities. In this way a single electrophoretic gel yields easily analysable data which can be compared with data from other gels. The time from the acquisition of bacterial isolates to their final characterization is much reduced in comparison to existing methods.