Research Article: Multiple Roles of the Cytoskeleton in Bacterial Autophagy

Date Published: November 20, 2014

Publisher: Public Library of Science

Author(s): Serge Mostowy, Virginia Miller.


Partial Text

Several bacterial pathogens are targeted by macroautophagy (referred to as autophagy hereafter), and autophagy is widely recognized as an important antibacterial defence mechanism [1]. On the other hand, new work has shown that some bacterial pathogens have mechanisms to avoid or exploit the autophagy machinery for intracellular survival [2]–[4]. Strikingly, the host cytoskeleton plays a crucial role in autophagy and its ability to restrict or promote bacterial replication. A complete understanding of autophagy-cytoskeleton interactions is therefore needed to enable the manipulation of autophagy for therapeutic purposes. Actin, microtubules, intermediate filaments, and septins are four main cytoskeletal components of vertebrate cells (Box 1), yet their roles in autophagy are not fully understood. This Pearl revisits our current understanding of autophagy-cytoskeleton interactions and highlights new concepts and emerging roles for the cytoskeleton in bacterial autophagy.

Autophagy is a membrane trafficking process delivering cytoplasmic material to the lysosome for degradation (Figure 1A). The different steps of canonical autophagy have been well characterized and involve the assembly of at least 36 autophagy-related (ATG) proteins into distinct complexes [5]. The ATG1-UNC-51-like kinase (ULK) complex initiates formation of the isolation membrane (also called a phagophore), the class III phosphatidylinositol 3 (PI3) kinase complex generates PI3-phosphate (PI3P) phospholipid for membrane biogenesis, the ATG12–ATG5–ATG16L1 complex mediates autophagosome formation and elongation, and the ATG8 lipidation system mediates closure of the autophagosomal membrane (Figure 1B). In humans there are six ATG8 orthologues belonging to the light chain 3 (LC3) or γ-aminobutyric acid receptor-associated protein (GABARAP) subfamilies, and by interacting with an extensive repertoire of proteins, they have important roles in mediating membrane-remodelling processes.

During starvation-induced autophagy, actin has been shown to promote the generation of PI3P for autophagosome formation [8]. In agreement with this, Wiskott-Aldrich syndrome protein (WASP) family member WASH is a nucleation-promoting factor (NPF) for the actin-related protein 2/3 (ARP2/3) complex necessary for the trafficking of ATG9 and autophagosome formation [9]. By contrast, during nutrient-independent, ubiquitin-selective autophagy, actin has been shown to promote the fusion of autophagosomes to lysosomes [10]. Taken together, these observations suggest that actin can link membrane acquisition to autophagosome biogenesis or regulate the fusion of autophagosomes to lysosomes, depending on the cargo targeted for degradation (Figure 2AI, 2AII).

Microtubules facilitate autophagosome trafficking and are viewed to promote autophagosome biogenesis and the fusion of autophagosomes with lysosomes [18]. Dynein, a motor protein, is required for autophagosome trafficking along microtubules. To interact with autophagosomes, dynein can bind to ATG8 family proteins and their interaction partners, microtubule-associated protein 1A and 1B (MAP1A and MAP1B). In this way, autophagosomes can move along microtubule tracks toward lysosomes concentrated near the microtubule organizing centre (MTOC) (Figure 2BI) [19].

Intermediate filaments are more stable and exhibit less dynamic behaviour than actin filaments or microtubules. In agreement with this, vimentin intermediate filaments have recently been shown to suppress autophagy dynamics by interacting with Beclin 1/ATG6, a protein crucial for autophagy initiation [32]. The phosphorylation of Beclin 1 by the protein kinase Akt promotes a Beclin 1/14-3-3/vimentin complex and inhibits the role of Beclin 1 in autophagy, highlighting a fundamental link between autophagy and intermediate filaments (Figure 2CI).

Septins have been mostly implicated in cytokinesis, a process in which autophagy is tightly regulated, and their role in nondividing cells is poorly understood [37]. The recruitment of autophagy critical components, e.g., p62 and LC3B, to their sites of function can be reduced by septin depletion [14], suggesting a general role for septins in autophagic activity (Figure 2DI). Septins and the septin-like GTPase of immunity-associated proteins (GIMAPs) function as nucleotide-regulated scaffolds on intracellular membranes, and GIMAP6 interacts with GABARAPL2 for recruitment to autophagosomes [38]. The specific interaction between GIMAP6 and GABARAPL2, similarly to what has been described between NDP52 and LC3C [39], suggests different roles for different ATG8 family proteins during autophagy and may highlight another level of specificity underlying selective autophagy.

Studies using intracellular bacterial pathogens have shown great potential to significantly advance our understanding of autophagy-cytoskeleton interactions. Despite new insights, many outstanding issues remain. For example, the precise signals targeting bacteria to autophagy and the source of membrane for autophagosome biogenesis (endoplasmic reticulum, mitochondria, and/or plasma membrane) are poorly understood. Work has shown that intracellular bacteria can trigger multiple canonical and noncanonical (i.e., alternative autophagy pathways independent of some core machinery components) autophagy pathways [2]. However, what determines the pathway of autophagy responding to bacterial infection is not yet clear. To help resolve these issues, future studies investigating intracellular pathogens and how they modulate the cytoskeleton during infection shall be crucial to illuminate structural determinants of autophagy and their roles in mediating disease outcome. They can also reveal new links between autophagy and the cytoskeleton.