Cellular responses to environmental cues

Whereas our biochemical understanding of RNAi mechanisms is rather advanced, very little is known about its spatial and temporal regulation. By addressing these aspects we found that the RNAi protein Dicer is a predominantly nuclear protein in S. pombe, but has the potential to shuttle between the nucleus and cytoplasm. Counter-intuitively, we found that RNA binding of Dicer’s double-stranded RNA-binding domain (dsRBD) is dispensable for the generation of siRNAs in S. pombe.  Instead, the dsRBD functions to control the nucleo-cytoplasmic distribution of Dicer. We showed that the fold of this domain is rather unusual and includes a novel zinc-binding motif. If the coordination of zinc is disrupted, nuclear retention of Dicer is lost. As a result, centromeric heterochromatin is lost and tight repression of BANC genes via CTGS is disrupted. Interestingly, similar to disrupting zinc-coordination genetically, Dicer is deported to the cytoplasm under chronic heat shock conditions. Here, it transiently accumulates in cytoplasmic aggregates, most likely due to temperature-induced structural changes in its C terminus. Thus, the dsRBD of Dicer constitutes a thermo-switch that controls the conditional relocalization of Dicer to the cytoplasm. This provides an elegant mechanism for the release of BANC genes from RNAi repression during heat shock, but also poses a potential threat to the integrity of heterochromatin.

Intriguingly, one of the BANC genes released from RNAi repression during heat shock is hsp104+. We have found that the Hsp104 protein dissolves the cytoplasmic Dicer aggregates that form under acute heat stress. These data suggest that this negative feedback loop involving the RNase Dicer and the protein disaggregase Hsp104 is crucial to confer epigenetic robustness to the organism. In the absence of Hsp104, Dicer accumulates in cytoplasmic inclusions and heterochromatin becomes unstable at elevated temperatures, an epigenetic state that is inherited for many cell divisions after the heat stress. On the other hand, we found that Dicer averts the toxic aggregation of a prionogenic protein.

These results highlight the importance of feedback regulation in building epigenetic memory. They furthermore uncovered Hsp104 and Dicer as homeostatic controllers that buffer environmentally induced stochastic epigenetic variation and toxic aggregation of prionogenic proteins. These data also show that relocalization of RNAi factors can be actively exploited by a cell to regulate the activity of RNAi. Finally, our results suggest that RNAi might be linked to prion biology in S. pombe. It is well known that prions form in S. cerevisiae, which has no RNAi pathway. Thus, we are tempted to speculate that evolution of prions as bet-hedging devices able to respond to environmental stimuli may have been one of the negative forces that led to the loss of RNAi in S. cerevisiae.


Original publications:

Oberti et al, Cell Rep, 2015

Woolcock et al, G&D, 2012

Barraud & Emmerth et al, 2011, EMBO J

Emmerth et al, Dev. Cell, 2010


Further reading:

Holoch and Moazed, Genes Dev. 2012; 26(8): 741-5.