Harvard University COVID-19 updates

Department News



(l to r) Vlad Denic, Roarke Kamber, and Chris Shoemaker

The modern urbanite doesn’t bother with the bus – she taps her phone and meets a personal driver at her door. Proteins and organelles seeking transport in the cell are similarly impatient – late deliveries can cause death and disease – and frequently use the protein equivalents of Uber apps to exert control over their transporters.

Selective autophagy, a specialized transport process that captures unwanted intracellular “target” structures and shuttles them to the cell’s recycling bins, had been thought to function more like a bus line, operating at a steady rate despite spatial and temporal variabilities in demand. Under this model, proteins called autophagy receptors recognize targets, such as protein aggregates and damaged organelles, and then await pickup by transport vesicles called autophagosomes.

But while inefficiencies in public transit systems are largely tolerable (at worst, some gas is wasted when buses run empty or some commuters are delayed when buses fill up), the consequences of allowing toxic selective autophagy targets to linger in the cytoplasm could be lethal. Thus, a constant rate of autophagosome formation would have to be set very high, at great energetic cost to the cell, to ensure that targets are always eliminated promptly.

In our paper, we tested the alternative hypothesis: that autophagosomes are produced on demand. A mechanism linking target detection by receptors with the formation of autophagosomes would guarantee rapid target destruction while minimizing overall costs.

To investigate this possibility, we monitored the activity of a protein, Atg1, that serves as the master overseer of autophagosome formation. Atg1 is a kinase, an enzyme that changes the behavior of other proteins by putting electric charges on them. When Atg1 is active, it instructs a set of proteins to begin construction on a new autophagosome. Using our assay for Atg1 activity, we found that Atg1 was only active in cells that contained a target – in this case, a protein aggregate that is constantly produced and captured by autophagosomes – and its dedicated receptor protein.

How do receptor-bound targets control the activity of Atg1? One hint came from our examination of the proteins that Atg1 binds in the cell. We developed a gentle procedure for purifying Atg1 and found that it co-purified with the same receptor-aggregate complex that controls its activity, suggesting that Atg1 activation occurs by a direct mechanism rather than a more circuitous pathway. The identification of a third protein present in the Atg1 complex, Atg11, which was known to bind receptors but had no clear function, led us to a more satisfying explanation. In cells lacking Atg11, Atg1 was both inactive and unable to interact with receptor-bound targets – suggesting that Atg11 allows receptor-bound targets to dictate Atg1’s activity.

To rigorously test the idea that Atg11 enables receptor-bound targets to activate Atg1, and thus drive local autophagosome formation, we developed a novel tool for monitoring Atg1 kinase activity in a test tube, and mixed cell extracts containing Atg1 with purified Atg11 and receptor-bound aggregates. The results of these in vitro experiments were unambiguous: receptor-bound aggregates activated Atg1, but only when Atg11 was also included in the mixture. When we instead used versions of Atg11 or the receptor with mutations designed to disrupt their interaction, Atg1 activation by targets was abolished, providing further evidence that a binding mechanism drives Atg1 activation.

Finally, we asked whether this mechanism might be generalizable to other types of target structures. We looked into the possibility that peroxisomes, organelles that become autophagy targets when damaged, also activate Atg1. Indeed, when we combined damaged peroxisomes with Atg1 in a test tube, Atg1 became active, dependent on the presence of the damaged-peroxisome receptor.

Overall, our work revealed a key piece of molecular logic that explains how cells are able to rapidly eliminate targets whose appearance in the cytoplasm can be unpredictable. Going forward, figuring out how receptor-bound targets physically flip the Atg1 switch, via structural characterization of the protein complex we identified, is an exciting prospect. It will also be important to determine whether failures in the mechanism we uncovered contribute to the toxic accumulation of autophagy targets commonly seen in diseased and aging organisms. It may turn out that relying on public transit is not only unbearable for technophiles, but fatal for the cell. 

Read more in Molecular Cell or download PDF

View Vlad Denic’s Faculty Profile