The titration of vaccinia virus on the chorio-allantoic membrane of the developing chick embryo

Autor: P. H. Phipps, E. A. Boulter, J. C. N. Westwood
Jazyk: angličtina
Rok vydání: 1957
Předmět:
Popis: The technique of infecting the developing chick embryo by the chorio-allantoic route, introduced by Woodruff & Goodpasture (1931), was developed by Burnet (1936a, b) into a general method for the titration of pock-producing viruses, and was applied to vaccinia virus by Keogh (1936). The method is now widely used and is the most accurate available for the titration of vaccinia. Nevertheless, owing to the wide scatter in individual counts from a single virus dilution, its accuracy is far below that of the analogous plate count titration of bacterial suspensions. This scatter was recognized by Burnet & Faris (1942), and is well illustrated by the figures published by Fenner & McIntyre (1956). Burnet & Faris (1942) attempted to overcome it by suitably weighting their results. On the grounds that marked departure from the expected scatter was always towards a low, rather than a high, count, they doubled the value given to all those membranes which 'could reasonably be regarded as giving counts in the normal range'. This they regarded as the 'least objectionable' way of overcoming the difficulty, but, while it would tend to give a truer estimate of the virus concentration, the procedure is highly undesirable. Fenner & McIntyre (1956), working with myxoma virus, accepted the wide scatter as in intrinsic defect in the method due to variation in the susceptibility of egg membranes. They exercised no selection over their data, and their figures show coefficients of variation varying from 28 to 110%, with the majority lying above 60 %. In a series of thirty-six titrations of the same material the betweenegg coefficient of variation was 63%. These results are far outside Poissonian expectation. Reid, Crawley & Rhodes (1949), using fowl pox virus, are the only workers who have obtained a good fit with Poisson distribution, but their findings were unfortunately restricted to mean counts of less than 15, below which level the expected coefficient of variation is, in any case, high. The recent work of Overman & Tamm (1956) does not materially contribute to the precision of the technique since they failed to investigate many of the factors affecting the virus count, and their use of a 3-day period of incubation for inoculated
Databáze: OpenAIRE