deletions | additions       diff --git a/Prophase onset coincide with circadian slowdown.tex b/Prophase onset coincide with circadian slowdown.tex  index 277c492..c3a9a34 100644  --- a/Prophase onset coincide with circadian slowdown.tex  +++ b/Prophase onset coincide with circadian slowdown.tex 
...
\subsection{Prophase onset coincide with circadian slowdown.}  \label{sec:Prophase_onset_coincide_with_circadian_slowdown}  Next, we hypothesized that the observed slow down of the circadian clock just before division, as the \revalphaYFP signal is in it's rising phase, was due to transcriptional shutdown during M phase. In order to temporally resolve the different mitotic phases we generated a NIH-3T3 cell line stably expressing a \histone fusion protein that allow to follow the chromosome condensation that marks the onset of prophase \cite{Sumner1991}, and coincide with the observed shutdown of transcription at division (TODO:ref). \cite{Gottesfeld1997}.  We then manually annotated prophase, metaphase and citokinesis on individual cell traces (Figure \ref{fig:pro_met_cyt}A). We marked the onset of prophase as the first frame where the texture of the \histone reporter became granular, the metaphase as the first frame where the \histone reporter is aligned in the nucleus and the citokinesis as the first frame where two separate objects are clearly visible (which is easier to see in the \revalphaYFP channel). We then inferred the circadian and cell cycle phases (Figure \ref{fig:pro_met_cyt}B) and plotted the phase of the different mitotic phases on top of the phase portrait. As shown in Figure \ref{fig:pro_met_cyt}C the onset of prophase coincide with the deceleration of the circadian phase, although the region extend past the measured times of prophase.  To test whether this wasn't due to a resolution limit of our recordings we imaged a few cells with a $60$x magnification and five minutes time resolution, but the results remained similar.  To better quantify the relationship between circadian phase and prophase onset we computed the mean circadian velocity for all traces passing through the deceleration region (Figure \ref{fig:pro_met_cyt}C, blue rectangle) while taking as a comparison the traces that avoid that region (green rectangle). As shown in Figure \ref{fig:pro_met_cyt}D the minimum velocity of the circadian phase coincide with the mean onset of prophase (red vertical bar). However it seems that the slow down also progressively starts before chromosome condensation, which would exclude a direct causality, and hint that earlier other  processes might influence the circadian phase (TODO:ref?). expression of our \revalphaYFP reporter \cite{Jiang2004}.  This might alternatively be explained by the fact that the inference of the cell cycle phase has a limited precision, and effectively blur, or smear the deceleration region in the phase plane.