1. It has been possible to transduce the property of fimbriation to a number of permanently non-fimbriate strains of Salmonella typhimurium, using pellicle formation, which occurs on serial subculture in aerobic, static broth, as a selection mechanism. The transduction rate, using lysates prepared on wild -type fimbriate S.typhimurium strains, was calculated as 1 per 5 x 10 - 10 infected cells. It was shown that before successful transduction was effected, the recipient strain was required to be motile. This requirement was not satisfied by the presence of paralysed flagella, and is a reflection of the need for positive aerotactic movements by the recipient strain, which will result in its transport to the broth surface.
2. Among 140 naturally -isolated non-fimbriate strains of S.typhimurium, evidence of non -identical allelism in the fim segment of the chromosome was shown. Reciprocal transduction among the mutants suggested the existence of five transductional groups. The order of the mutations involved in these different groups was not ascertained because of limitations in the selection method and physiological differences inherent in the mutants.
3. Reciprocal transduction among 132 non -fimbriate strains of Salmonella typhimurium classified as FIRN, i.e. fin - inl - rha - , did not reveal recombination either to fin+ or rha+. These strains, although epidemiologically independent, are thought to have identical or overlapping mutations in fin and rha. The hypothesis was presented that these arose from a common fin - rha - parent, and that present-day FIRN strains represent a world -wide, successfully distributed typhimurium type.
4. There was no evidence of a close linkage between fin and any other markers tested, so that the initial hypothesis of a close genetic linkage between fin and rha, to explain the almost absolute correlation of these characters in S.typhimurium strains, was not valid. With the non-fimbriate, non-flagellate derivatives of S.typhimurium LT -2 by ethyl methane sulphonate treatment, called Dubnau strains, indications of cotransduction of fin and fla were presented, but this phenomenon was not examined in detail.
5. One non-fimbriate strain, S.typhimurium Sa 519, produced a small colony fimbriate revertant. Genetic analysis suggested that this was due to suppressor mutation.
6. Induced reversion to a fimbriate state was achieved wi the mutagens ultra- violet light and manganous chloride. Limited tests with ethyl methane sulphonate did not induce fimbriate revertants. Of the seven fin - rha + strains of S.typhimurium, three strains reverted to a fimbriate state - Sa 749, 1436 and 1566b. Six of the 132 FIRN strains revert to fimbriate state, and were characterised as FIRN derivat- ives. The value of these findings is discussed. Spon- taneous mutation to fimbriation was never detected with any non -fimbriate S.typhimurium mutants.
7. A number of fimbriate, non- haemagglutinating strains of S.paratyphi B was transduced to a haemagglutinating state with traneducing lysates from wild -type fimbriate strains o Salmonella tyohimurium. This was judged indicative of a gene Ha controlling fimbrial function, but alternate inter- pretations were discussed.
8. Attempts to demonstrate the presence of a fin episome controlling phase variation in Salmonella typhimurium were without success. Another theory to explain the fin phase variation was presented, and discussed also in relation to the larger number of fimbriate and non-fimbriate types Escherichia coli.
9. Competition experiments were devised to investigate the transduction procedure for fimbriate cells on a quantitative basis. Small numbers of fimbriate cells were used to challenge large numbers of non-fimbriate cells, both strains being flagellate, in mixed culture under a range of environmental conditions - static broths incubated (a) aerobically, (b) anaerobically, (c) microaerophilically, artificially aerated broths (d) on a reciprocating shaker at 100 oscillations per min. and (e) on a rotator at 12 rotations per min., and, finally, (f) on agar plates. Buffered and unbuffered media were used. The strains used were naturally isolated fimbriate and non-fimbriate strains of Salmonella typhimurium, and pairs of fimbriate and non-fimbriate strains isolated from a transduction mixture. Under conditions favourable to the formation of a fimbrial pellicle, the fimbriate cells outgrew the non -fimbriate cells and formed as much as 80% of the final population after 48 hr. A small outgrowth by naturally isolated fimbriate strains was sometimes detected, not associated with the formation of a fimbrial pellicle, and due to physiological strain differences. The possible mechanisms of this out- growth are discussed.
10. Similar competition experiments in which the challenger fimbriate cells were non -motile showed that, if the challenger cell was phenotypically fimbriate, it could outgrow large numbers of non -fimbriate, non- flagellated challenged cells even in the absence of motility. This outgrowth was accompanied by the formation of a fimbrial pellicle in the later stages of growth.
11. Competition experiments also revealed the importance of flagella alone. Small numbers of flagellate, motile cells outgrew large numbers of non -motile cells under aerobic static conditions by their ability to form a flagellar pellicle. The outgrowth, however, was less significant than that associated with the formation of a fimbrial pellicle.
12. From these competition experiments and the parallel growth experiments between S.paratyphi B strains with functional and non -functional fimbriae, the advantage bestowed on a cell by its ability to form a pellicle was seen. It is considered that these experiments provide quantitative experimental data supporting the hypothesis that fimbriae are organs of survival.
13. Fimbriae from Escherichia coli were isolated in a pure form by differential centrifugation and by chromatography on DEAE-cellulose exchanger at a neutral pH, and elution with a salt gradient. The fimbriae were judged pure by five criteria - electron microscopy, ultraviolet spectrophotometry, chemical analysis, agar gel diffusion and biological activity. The relative values of these criteria are discussed.
14. Chemical analysis reîealed that fimbriae were protein structures, unrelated to cell wall, capsule or flagella.
15. The range of amino acids detected on two -dimensional chromatography was not suggestive of an atypical protein. The adsorption of fimbriae to DEAE - but not to CM- cellulose suggests a preponderance of acidic amino acids in their structure. The relevance of this finding is discussed in relation to their agglutinating properties and to their resistance to proteolytic enzymes.
16. Chromatography of fimbriae from Shigella flexneri FIaI showed a behaviour different from E.coli 2v,. The possible relevance of this to sharing of fimbrial antigens is discussed.