SORTING OUT THE STOCHASTIC COMPONENTS OF TENASCIN-C GENE REGULATION

 

Daniel R. Sisan, Michael Halter, Anne L. Plant

 

A population of clonal cells in homogeneous culture often displays a considerable range of phenotypes.  To better understand this heterogeneity, we used fluorescence activated cell sorting (FACS) to sort a clonal population of GFP reporter cells based on promoter activity and then to observe the relaxation of each subpopulation back to steady state.  We also collected time-lapse images, and segmented and tracked hundreds of individual cells.  The cells are NIH-3T3 fibroblasts stably expressing destabilized eGFP under the control of the promoter for tenascin-C (TN-C), an extracellular matrix protein that has a complex transcriptional control and poorly understood biological function.  We find that the distribution of activities is unusually broad (CV~2) and bimodal (though only when log-transformed), making it an interesting system for studying expression noise.  The sorted subpopulations, we observe, relax over long time scales (up to several months), while the low GFP expressing subpopulation returns to a steady state much slower than the high expressing subpopulation.  We also observe that the high expressing cells proliferate faster, likely due to the effect of secreted tenascin-C, which contains mitotis-inducing EGF-like repeats.  The steady state TN-C distribution is thus determined by a combination of intracellular TN-C expression noise and TN-C state-dependent population dynamics.