It is an open question how genetic switches that control cell fate remain in the correct state for the entire lifetime of an animal, despite fluctuations in transcription factor levels. The transcription factor CHE-1 acts as a genetic cell fate switch by activating its own expression and that of >100 genes that define ASE neuron fate. Hence, CHE-1 must simultaneously induce its target genes and maintain its own expression at high enough levels to prevent the switch from reverting from the 'on' (high CHE-1) to the 'off' state (no CHE-1). It is thought essential for the stability of genetic switches that its core transcription factors are present at high mRNA and protein levels and have long lifetimes. However, this has never been tested experimentally in any developmental system. Here, we used a quantitative, biophysical approach to measure, for the first time, the key parameters governing the stability of the ASE cell fate switch, namely the absolute che-1 mRNA and protein numbers and lifetimes. Surprisingly, we found relatively short lifetimes (~2 hours) and low molecule numbers (~700 CHE-1 proteins and ~5 mRNA molecules per cell). We used simulations to predict the stability of the CHE-1 induced switch for the measured parameters and found spontaneous collapses to the 'off' state occurring with a rate as high as once per day, indicating that low-copy number fluctuations can significantly limit the stability of the ASE-cell fate switch. The model produced exceedingly stable switches (>10 year lifetime) when CHE-1 bound its own promoter very strongly. In this case, the simulations reliably returned to the state with high CHE-1 level, even when all but ~50 CHE-1 proteins were removed. We tested this prediction using the auxin inducible degradation (AID) system to deplete CHE-1::GFP::degron in ASE neurons in larvae. In agreement with our model predictions, we found that the ASE neurons recovered to express high CHE-1 levels after ~24 hours of CHE-1::GFP::degron depletion. Strikingly, after CHE-1::GFP::degron depletion, che-1 mRNA levels were unchanged, even while its target genes gcy-14 and gcy-22 were no longer expressed. Overall, this suggests a mechanism where strong, preferential binding of CHE-1 to its own promoter allows the ASE-cell fate switch to recover from transient CHE-1 depletion by recruiting CHE-1 away from its other targets, resulting in a highly stable genetic switch. We are currently testing this model by swapping CHE-1 binding sites.

Affiliations:
- AMOLF, Amsterdam, NL
- Dept. of Cell Biology, Erasmus MC, Rotterdam, NL