Variability of cell cycle times measured in vivo in embryonic chick retina by continuous labelling with BUdR

Abstract

IN this report, cell cycle times were measured in the embryonic chick retina by continuous labelling with 5-bromodeoxyuridine (BUdR) and the fluorescence plus Giemsa¹ (FPG) technique. The measurement of cell cycle times in embryonic chick tissues in vivo is preferably carried out by continuous rather than pulse labelling because a single injection of a nucleotide precursor into an egg can form a long-lasting pool which remains available for at least 24 h (ref. 2). Continuous labelling with BUdR provides information about the number of S phases a metaphase cell has encountered since the start of labelling whereas the previous continuous labelling procedure³ using tritiated thymidine indicates that a mitotic cell has encountered an S phase but does not show how many. BUdR labelling thus gives more detailed data for estimating the means, variances and distributions of the durations of the cell cycle and its phases. Analysis of the data by statistical modelling has shown that the cell cycle is made up of a phase of constant duration and one of variable duration and we were able to test the appropriateness of different models for the variable phase. The variable phase was described well by a gamma distribution which was clearly not exponential in form and, somewhat less well, by an inverse gaussian distribution. A random transition in the cell cycle would be indicated by an exponentially distributed period; therefore, for the random transition model^4,5 of the cell cycle to hold, there would have to be an exponentially distributed component within the variable phase. Our results do not exclude this but they do show that such a component could only be a minor part of the variable phase.