Fig. 5.

Kinetics analysis of SGs reveals altered formation and dispersal of SGs in an inducible cell model of DM1, and increased variance in MBNL1 dynamics. (A) Selected time-points from representative time-lapse images showing the formation (top) and dispersal during recovery (bottom) of SGs (arrows) in doxocycline (Dox)-induced CTG12 (left) and CTG960 (right) HeLa cell lines treated with NaAsO₂ to induce SGs. The ‘formation time’ of SGs was recorded as the time-point at which at least five SGs were first detected. The ‘dispersal time’ of SGs was recorded as the first time-point at which SGs could no longer be clearly detected. Scale bars: 10 µm. See also Movies 1-4. (B) CTG12 control and CTG960 DM1 model cells were induced with Dox (+DOX), and time of SG formation (left) and SG loss (right) were analysed after treatment with NaAsO₂ (analysed immediately for SG formation, and after 90 minutes of pre-treatment and drug removal for SG dispersal). Bar graphs show the time of SG formation after induction for 24 h (left) and time of SG loss after induction for 24 h or 72 h (right) in CTG960 cells normalised to the mean values measured in control (CTG12) cells (set as 100%). CTG12, n=19 cells; CTG960, n=27 cells for SG formation; ****P<0.0001. CTG12, n=35 cells; CTG960, n=25 cells for SG loss at 24 h; *P<0.05. CTG12 n=33; CTG960 n=18 cells for SG loss at 72 h; ****P<0.0001 unpaired t-tests. Each graph is representative of two independent experiments. (C) The proportion of cells containing SGs was plotted against time for the full time-course of treatment with or recovery from treatment with NaAsO₂. This emphasises the difference in dynamics of SG formation (top) between cells expressing CTGexp RNA (DM1) and control RNA, but also the differences regarding SG loss when cells were Dox-induced for 24 h (centre) or 72 h (bottom). The increased difference seen with RNA induction for 24 h or 72 h confirms that these changes are due to CUGexp expression, rather than random variation between the cell lines. (D) The mobile fraction (left) and half-time of fluorescence recovery (right) of GFP-tagged MBNL1 (GFPMBNL1) in SGs is similar in CTG960 and CTG12 HeLa cells (Mann–Whitney U test; ns, not significant). (E) Plotted is the relative fluorescence recovery of GFPMBNL1 in SGs over time in response to photobleaching (488 nm laser, 2.5 s pulse), see also Fig. S7. Closed circles, SGs in HeLa_CTG₁₂_GFPMBNL1 cells (n=41); open circles, SGs in HeLa_CTG₉₆₀_GFPMBNL1 cells (n=25). All data are displayed as the mean±s.d.