![Figure 6. Computer simulations based on FISH measurements predict findings with live-cell imaging. (A and B) KLacO and miniKLacO signals were followed for 75 generations after removal of selection in two independent clones for each population with FISH. The percentage of signals remaining was measured at several time points. (A) KLacO signals and miniKLacO were lost from SLK or SLK/LANA1 cells at a rate of 3.9% and 4.6% (rate of loss [ROL]), respectively, per population doubling. (B) Although the majority of cells lost KLacO or miniKLacO during 75 doublings in the absence of selection, the percentage of signals remaining in a small fraction of cells were largely unchanged in their distribution. See also Fig. S1. (C) A summary of parameters used in the computer simulations and their predictions shown in the following panels, including the mean of the initial population, the probability of cluster breakup, and the size distribution of dissociated clusters. One parameter, the probability of cluster formation, was found to be optional. A final parameter, the probability of equal distribution between daughter plasmids, was varied only for the fusion protein shown in Fig. 8 D. (D) The distributions of miniKLacO signals after 30 generations without selection (light blue line) were predicted by computer simulations (dark blue line). Two clones of SLK/LacI-tdTomato cells expressing wild-type LANA1 and carrying miniKLacO were cultured without selection for 60 generations, and the distributions of miniKLacO signals were measured each 10 generations by FISH. (E) Varying the parameters in the simulations by removing clustering (orange line) or decreasing the efficiency of synthesis (red line) impaired the simulations. (F) Computer simulations based on FISH measurements predicted the number of miniKLacO plasmids per cluster observed in SLK/wild-type LANA1 cells using live-cell imaging. (G) Computer simulations predicted that clustering accelerates the accumulation of plasmids in a subset of newly infected/transfected cells. See also Materials and methods.](https://cdn.rupress.org/rup/content_public/journal/jcb/216/9/10.1083_jcb.201702013/8/jcb_201702013_fig6.jpeg?Expires=1773459616&Signature=jAea5kTiZiBAG~LX6zqXpWGcAYEcmkLE1mLnzfwiqtGUBnf2nm3Xy31zaNduRsKA9Xn25YyStUOLpIGdodyBYv-fgoE816MRXpgibN95YT2o6kNucJhsA083rkOXNeit~o1EcjxILheACJOoNPCD0trVdU-v6iMncu0vF2zxpUovl7U47E6jAI2AN5PAtHJ9heSrzlVNWpBdcB8bdEy1gchXE2ECJd-YyHCJxZmQ~GlSgUZ-ZqXv9Kz9eEzP7cXZIVTjqjvi5zzxx1p~FTKHGUD0c5rv5xT3KbJ22iVkmJuAXDmTR6GDo6FA1ekySCcf1E6940mGTrbHHty22~kv9g__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Computer simulations based on FISH measurements predict findings with live-cell imaging. (A and B) KLacO and miniKLacO signals were followed for 75 generations after removal of selection in two independent clones for each population with FISH. The percentage of signals remaining was measured at several time points. (A) KLacO signals and miniKLacO were lost from SLK or SLK/LANA1 cells at a rate of 3.9% and 4.6% (rate of loss [ROL]), respectively, per population doubling. (B) Although the majority of cells lost KLacO or miniKLacO during 75 doublings in the absence of selection, the percentage of signals remaining in a small fraction of cells were largely unchanged in their distribution. See also Fig. S1. (C) A summary of parameters used in the computer simulations and their predictions shown in the following panels, including the mean of the initial population, the probability of cluster breakup, and the size distribution of dissociated clusters. One parameter, the probability of cluster formation, was found to be optional. A final parameter, the probability of equal distribution between daughter plasmids, was varied only for the fusion protein shown in Fig. 8 D. (D) The distributions of miniKLacO signals after 30 generations without selection (light blue line) were predicted by computer simulations (dark blue line). Two clones of SLK/LacI-tdTomato cells expressing wild-type LANA1 and carrying miniKLacO were cultured without selection for 60 generations, and the distributions of miniKLacO signals were measured each 10 generations by FISH. (E) Varying the parameters in the simulations by removing clustering (orange line) or decreasing the efficiency of synthesis (red line) impaired the simulations. (F) Computer simulations based on FISH measurements predicted the number of miniKLacO plasmids per cluster observed in SLK/wild-type LANA1 cells using live-cell imaging. (G) Computer simulations predicted that clustering accelerates the accumulation of plasmids in a subset of newly infected/transfected cells. See also Materials and methods.
