Autophagy in cells

Sharon Tooze : Autophagy activation through ATG proteins


Autophagy is initiated on intracellular membranes through the activation of a set of proteins known as the Atg proteins.

The initial events in autophagy are the formation of a unique membrane called a phagophore, followed by closure of the phagophore around cytosolic material, which results in the formation of an autophagosome. Next, the autophagosome fuses with the endosomal-lysosomal system to become an autolysosome, which contains proteolytic lysosomal enzymes that degrade sequestered cytosolic material.

This sequence of compartment formation, fusion and maturation occurs in a highly regulated process that is relatively poorly understood. However, through genetic and biochemical screens performed in the yeast Saccharomyces cerevisiae over 38 Atg proteins have been identified and at least 18 of these are conserved in mammalian autophagy. Thus, the focus of the molecular work in mammalian cells has been to understand the function of the mammalian Atg proteins, and extend this work to understand how the process is regulated at every step, and where the membrane for the formation of the phagophore and autophagosome comes from.

Phagophore formation occurs when the ULK (Atg1) kinase complex of four proteins (ULK1 or 2, FIP200, Atg13 and Atg101) becomes activated and membrane associated, best characterised in cells after inactivation of mTORC1 and activation of AMPK.

Given the primacy of the ULK complex in mammalian cells, Harold Jefferies in the lab continues to develop the ULK1/2 transgenic mouse models. The aim is to understand the function of the ULK1/2 complex in both normal processes and under autophagy-inducing conditions.

The ULK kinase complex is a key node in nutrient signalling and a potential target for therapeutic intervention in human diseases. In addition, we have studied the role of ULK in melanogenesis in MNT-1 cells and found ULK1 regulates melanin levels, suggesting an inhibitory function for this protein in melanogenesis. ULK1 function in this context is independent of the canonical ULK1 autophagy partners, ATG13 and FIP200. Furthermore, regulation of melanogenesis by ULK1 is independent of mTORC1 inhibition. Our data thus provide intriguing insights regarding the involvement of the key regulatory autophagy machinery in melanogenesis.

Concomitant with ULK complex activation, the class III PI3 kinase complex or the Beclin 1 (Atg6) complex is mobilised to discrete membrane domains in the cell, found for example on the endoplasmic reticulum (ER), and produces PI3P at these domains. PI3P effectors including the WIPI proteins (Atg18) are then recruited followed by the autophagy-specific ubiquitin-like conjugates Atg12-5-16 and LC3-II.

Phagophore and autophagosome formation both depend on Atg9, a multi-spanning membrane protein my group identified in 2006. During amino acid starvation Atg9 localisation changes, and its trafficking to the phagophore and autophagosome is required for autophagy.

We have shown that in normal conditions Atg9 resides predominately in the Golgi complex, it traffics in small, mobile vesicles between multiple organelles, including recycling endosomes, late endosomes, and the Golgi compartment. Atg9 is also found in a unique conserved vesicular-vacuolar compartment called the 'Atg9 compartment' which is adjacent to forming phagophores.

In collaboration with Lucy Collinson, from the Electron Microscopy Facility, we have studied the morphology of this compartment using cryo-soft x-ray and correlative light microscopy. Under these near-native imaging conditions, we find Atg9 in vacuolar structures adjacent to endosomes and forming phagophores forming along the ER. The control of Atg9 trafficking is likely to involve a complex set of regulatory proteins and machinery, including those responsible for trafficking under normal growth conditions as well as under nutrient-deprivation conditions.

So far our lab has shown control by the p38MAPK pathway, and a conservation of the role of ULK1/2 in regulating Atg9 trafficking. In an effort to expand our knowledge, we have begun to analyze the vesicular pathway taken by Atg9 in mammalian cells at a proteomic level.

As exemplified by Atg9, vesicular trafficking is essential for autophagosome formation. The complexity of the trafficking events so far shown to be required for autophagosome formation underlies the need to understand each contribution at the molecular level. We have shown trafficking of ULK-positive, Atg9 negative vesicles from the recycling endosome to the forming phagophore requires TBC1D14. We are continuing to unravel the function for this Rab GTPase-activating protein and in particular identifying effectors, including both Rab proteins and other trafficking complexes, which are regulated by TBC1D14. Regulation of autophagy by non-Atg proteins has also been a focus of the research in the lab since the completion of our siGenome screen in 2010.

The focus of the work on these novel regulators continues using both structure-function approaches, model organisms and proteomic network analyses. These experiments are done in collaboration with Stephane Mouilleron's Protein Structure Unit and the Protein Analysis and Proteomic Facility headed by Bram Snijders. A molecular understanding of these novel regulators will expand our knowledge of the autophagy pathway and open up new avenues to control the autophagic response in normal and disease affected cells and tissues.