Rabouille: Cellular stress in Drosophila

On the other hand, amino-acid starvation leads to the formation of reversible non-membrane bound pro-survival stress assemblies, the Sec bodies, that incorporate the components of the ER exit sites, the Sec16 and the COPII components (7).

We have recently studied whether ADP-ribosylation plays a role in the formation of stress assemblies. We showed that PARP16 MARylates Sec16 on a small region called the SRDC and that this is sufficient to induce Sec body formation (8). We are also working on the role of PARP1 in stress granule formation.

We are in the middle to identifying the signalling pathways required for formation of these stress assemblies as well as their composition.

We are also testing whether Sec bodies form in vivo using the Drosophila fat body, and whether they form in mammalian cells.

Another stress assembly, the stress granules, also forms upon amino-acid starvation. Stress granules form as a result of protein translation inhibition that is imposed by almost any type of stress. This leads in turn to the accumulation of untranslated mRNAs that are bound to specific RNA binding proteins to be stored in stress granules or degraded in P-bodies (7).

Interestingly, we have noticed that the formation of both structures, Sec bodies and stress granules, is linked and we are investigating this. Indeed, we have shown that Sec16, a protein related to ER exit sites and protein transport, is also required in stress granule formation specifically during amino-acid starvation because it binds a modified form of the RNA binding protein Rasputin and stabilises it (9).

GRASP mutant mice

In most cases, newly synthesised proteins destined for the plasma membrane are synthesised in the ER and use the classical secretory pathway (ER>ER exit sites>Golgi>PM) (14). A few years ago, we have shown that in Drosophila, some of these proteins bypass the Golgi in a dGRASP dependent manner (15)(16). Surprisingly, dGRASP is a peripheral protein of the Golgi apparatus and is required for many aspects of its biology (17). To understand the relevance of this pathway in mammals, we have generated a KO mouse for one of the mammalian homologues, GRASP65, that displays no phenotype under state conditions (18). In collaboration with the Malhotra lab in Barcelona (Spain), we are now generating a double KO GRASP65/GRASP55 mouse.