Intestinal epithelium and autophagy: partners in gut homeostasis

Abstract

One of the most significant challenges of cell biology is to understand how each type of cell copes with its specific workload without suffering damage. Among the most intriguing questions concerns intestinal epithelial cells in mammals; these cells act as a barrier between the internally protected region and the external environment that is exposed constantly to food and microbes. A major process involved in the processing of microbes is autophagy. In the intestine, through multiple, complex signaling pathways, autophagy including macroautophagy and xenophagy is pivotal in mounting appropriate intestinal immune responses and anti-microbial protection. Dysfunctional autophagy mechanism leads to chronic intestinal inflammation, such as inflammatory bowel disease (IBD). Studies involving a number of in vitro and in vivo mouse models in addition to human clinical studies have revealed a detailed role for autophagy in the generation of chronic intestinal inflammation. A number of genome-wide association studies identified roles for numerous autophagy genes in IBD, especially in Crohn's disease. In this review, we will explore in detail the latest research linking autophagy to intestinal homeostasis and how alterations in autophagy pathways lead to intestinal inflammation.

Keywords: ATG16L1; IBD; IRGM; autophagy; intestinal epithelium.

Figure 1

Basic steps involved in mammalian macroautophagy.

Figure 2

Initiation via ULK1 complex: the regulatory complex mTORC1 represses autophagy activation in nutrient rich conditions. mTORC1 phosphorylates a serine residue on ULK1 to prevent it interacting with positive regulators of autophagy induction. Atg13 activation is similarly repressed by mTORC1-mediated phosphorylation. Glucose or amino acid starvation results in the repression of mTOR activation. Consequently, ULK1 phosphorylates both FIP200 and Atg13, resulting in the activation of downstream autophagy effector proteins.

Figure 3

PI(3)K-mediated vesicle nucleation: The activated ULK1 complex translocates to the phagophore formation site. In starvation-induced autophagy this is likely to be adjacent to the ER initially but ULK1 also associates with the phagophore. ULK1 complex activation results in the translocation of the Beclin 1 lipid kinase complex to the phagophore assembly site. Beclin 1 acts as a scaffold protein forming a complex with Vps34 PI(3)K, its Vps15 subunit and Atg14. Similarly to orthologs yeast proteins the mammalian Vps15 subunit is predicted to interact with lipid membranes thereby tethering Vps34 to vesicles. Vps34 is the subunit of the Beclin1 lipid kinase complex that catalyzes the phosphorylation of the 3 region of the inositol head group of phosphatidylinositol to generate PI(3)P. The transmembrane Atg9 is found localized to post-Golgi vesicles but is depicted on the same vesicle as the Beclin 1 complex for convenience. The extent to which vesicular Beclin 1 and Atg9 overlap in their distribution is unclear. The inclusion of the Atg14 protein recruits the Beclin 1 complex specifically to the ER. This may serve two functions in assembly of the autophagosome. The first and perhaps most crucial function is the generation of PI(3)P at the omegasome and/or phagophore, which is essential for the recruitment of many effector proteins required for autophagosome assembly. PI(3)P production results in an accumulation of PI(3)P-binding protein DFCP1 at the omegasome membrane, the functional significance of which is unknown. ER-associated WIPI proteins, WIPI1 and WIPI2 also bind PI(3)P and are essential in the transition from omegasome to the double membrane phagophore.

Figure 4

Atg5-Atg12 complex conjugation system. Agt12 is conjugated to Atg5 and Atg16L1 via a ubiquitinylation-like process. Homodimeric Atg7 functions as an ubiquitin-activating (E1)-like enzyme, activating Atg12 through the formation of a thioester bond between the Atg7 active site cysteine and the Atg12 C-terminal glycine. Atg7 transfers the activated Atg12 to the E2-like enzyme Atg10 via the C-terminal thioester linkage. Atg10 catalyzes the formation of an irreversible isopeptide bond between the Atg12 C-terminal glycine and Atg5. After the formation of the Atg5-Atg12 conjugate, Atg16L1 associates with the complex, binding non-covalently with Atg5. Interaction between Atg16 coiled-coil domains on adjacent Atg5-Atg12-Atg16L1 conjugates results in the formation of tetrameric complexes.

Figure 5

LC3-PE conjugation system. Atg4 cleaves the carboxyl terminal region of LC3 immediately after synthesis, generating a soluble LC3-I which now possesses a C-terminal glycine required for further modifications. Homodimeric Atg7 functions similarly to ubiquitin-activating (E1) enzymes and binds the LC3-I C-terminal exposed by Atg4 before recruiting Atg3 via its N-terminal domain. LC3-I is transferred to the Atg3 enzyme during the Atg3-Atg7 interaction. Atg7 dissociates from the LC3-I complex. LC3-II is formed when LC3-I is conjugated to the lipid phosphatidylethanolamine (PE) in a reaction catalyzed by Atg3. Binding of the Atg5-Atg12-Atg16L1 to Atg3 enhances the lipidation of LC3-I. The ability of the Atg5-Atg-12 complex to bind PE already inserted into the isolation membrane ensures that newly converted LC3-II is incorporated into the elongating isolation membrane. Unlike the formation of the Atg5-Atg12 complex, the conversion of LC3-I to LC3-II is reversible. Atg4 cleaves PE from LC3-II and LC3 is recycled.