LC3-independent autophagy is vital to prevent TNF cytotoxicity

Abstract

The (macro)autophagy field is facing a paradigm shift after the recent discovery that cytosolic cargoes can still be selectively targeted to phagophores (the precursors to autophagosomes) even in the absence of LC3 or other Atg8-protein family members. Several in vitro studies have indeed reported on the existence of an unconventional selective autophagic pathway that involves the in-situ formation of an autophagosome around the cargo through the direct selective autophagy receptor-mediated recruitment of RB1CC1/FIP200, thereby bypassing the requirement of LC3. In an article recently published in Science, we demonstrate the physiological importance of this unconventional autophagic pathway in the context of TNF (tumor necrosis factor) signaling. We show that it promotes the degradation of the cytotoxic TNFRSF1A/TNFR1 (TNF receptor superfamily member 1A) complex II that assembles upon TNF sensing and thereby protects mice from TNFRSF1A-driven embryonic lethality and skin inflammation.Abbreviations: ATG: autophagy related; CASP: caspase; FIR: RB1CC1/FIP200-interacting region; LIR: LC3-interacting region; M1: linear; PAS: phagophore assembly site; PtdIns3K: phosphatidylinositol 3-kinase; TNF: tumor necrosis factor; TNFRSF1A: TNF receptor superfamily member 1A.

Keywords: ATG9A; Apoptosis; LC3-independent autophagy; cell death; embryogenesis; skin disease.

Figure 1.

Schematic overview of the role of LC3-independent autophagy in counteracting TNF cytotoxicity. (A) TNF sensing triggers the assembly of a first membrane-associated multiprotein complex called TNFRSF1A complex I. Complex I subsequently dissociates from the receptor and associates with CASP8 to form the cytosolic TNFRSF1A complex II. Complex II has the ability to activate CASP8 and promote cell death induction. Within complex II, RIPK1 is decorated with M1-linked ubiquitin chains, which are recognized by the selective autophagy receptor TAX1BP1 as a marker for lysosomal degradation. TAX1BP1 directly binds RB1CC1 through its N-terminal SKICH-domain, allowing the direct recruitment of the autophagy initiation machinery to form an in-situ autophagosome around complex II, thereby bypassing the need for LC3. The formation of this autophagosome depends on the PtdIns3K complex, and on the lipid transfer proteins ATG9A and ATG2A/ATG2B. Once encapsulated, complex II can no longer transmit its lethal signal and is ultimately destroyed by lysosomal degradation. (B) When the unconventional autophagy pathway is blocked, complex II is no longer targeted for lysosomal degradation and accumulates in the cytosol, leading to cell death induction.

Figure 2.

Schematic overview of the consequence of constitutive deletion of Atg genes on mice viability. Mice lacking the upstream ATG proteins PIK3C3, RB1CC1, ATG101, ATG13 and ATG9A – identified to be involved in the new ATG9A-dependent cell death checkpoint – die during embryogenesis, which is likely caused by dysregulated TNF-induced cell death. The embryonic lethality of atg9a^-/- and rb1cc1^-/- mice is indeed fully rescued by the additional deletion of Tnfrsf1a. Embryonic lethality of the RB1CC1-deficient mice is not caused by a defect in canonical autophagy, because autophagy-deficient Rb1cc1 knock-in mice (rb1cc1^MUT) develop normally and die at birth. Mice deficient for TAX1BP1 also die in utero, consistent with the pro-survival role of TAX1BP1 in the TNF pathway. In contrast, mice deficient for the other autophagy receptors (SQSTM1, OPTN and NBR1) develop normally. ULK1 and ULK2 are dispensable for the ATG9A-dependent cell death checkpoint, and accordingly, ulk1^-/- and ulk1^-/- ulk2^-/- mice do not die during embryogenesis. Mice deficient in the LC3-conjugation machinery (ATG5, ATG7, ATG16L1 and LC3) only die at, or after, birth. Original studies describing these mouse strains are referred to in the main text. *pro-survival role in the TNF pathway addressed in vitro.