Differentiation of Salmonella Serotypes by pyrolysis-gas-liquid chromatography of cell fragments

Abstract

Pyrolysis-gas-liquid chromatography (PGLC) of whole cells and cell fragments was used to differentiate 10 Salmonella serotypes. Lyophilized samples (200 ,ug) of whole cells, cell walls, flagella, and deoxyribonucleic acid from each serotype were analyzed in duplicate by PGLC. Pyrochromatograms recorded as pyrolytic elution patterns represented thermal fragmentation products of the samples. Mathematical expressions of percent similarity and percent conformity were calculated for all possible pair combinations of the 10 serotypes. Stepwise discriminant analysis of the PGLC data showed that 100% correct classification of the 10 serotypes was possible from the flagella or deoxyribonucleic acid pyrochromatograms. The classification matrix of the whole-cell data showed a 90% correct classification. PGLC of cell fragments may provide useful information for taxonomic studies of Salmonella and other microorganisms.

Pyrolysis Gas Chromatography of Pseudomonas and Acinetobacter Species

“…These were all clinical or environmental isolates from St. Thomas's Hospital, and were identified by orthodox methods described previously (King & Phillips 1978 For pyrolysis, all organisms were grown overnight at 37°C or room temperature on bacterial filters (Millipore type HA, pore size 0.45 pm) placed on the surface of 6% lysed Horse Blood Agar Plates (Columbia Blood Agar, Oxoid CM 331) (Oxborrow et al 1976(Oxborrow et al , 1977a.…”

“…The id~nti~ca~jon of P s e~~o~~~~s species from clinical specimens is difficult and time consuming (King & Phillips 1978). We have, therefore, investigated the possibility of using Py-GC as a simple and rapid method for the ~d~ntification of pseudomonads of medical importance.…”

The Application of Pyrolysis Techniques to the Identification of Micro‐organisms

“…For pyrolysis using low resolution packed columns the problem of recognizing unique retention times is compounded by the difficulty in ensuring that changes in the resolution of components are not providing false information. It is often not clear just how authors are comparing pyrograms but Garner & Chi Yuan (1964), Cone & Lechowich (1970), Vincent & Kulik (1970, , Sekhon & Carmichael (1972, Taylor (1976) and Emswiler & Kotula (1978) all imply qualitative differences between micro-organisms of varying taxonomic relatedness. It may be significant, however, that, with the exception of Cone & Lechowich (1970) and Emswiler & Kotula (1978), all the studies have been on fungi where one might expect to find more chemical differences, but by the same token, more problems with standardizing the preparation of samples.…”

“…peak height, or peak area data. Emswiler & Kotula (1978) used the %S formula of Vincent & Kulik (1970) but also developed a percentage conformity (%C) estimate to include peak areas: where R, = retention time of a peak %A = standardized peak area.…”

“…When the modified formula was applied to the data it clarified the picture of similarity. Emswiler & Kotula (1978) also applied stepwise discriminant analysis to their data, by dividing the area under each pyrogram into 1 min intervals. This analysis was able to select variables for which the ratio of the variances between and within samples was maximal.…”