NCOA4 drives ferritin phase separation to facilitate macroferritinophagy and microferritinophagy

Abstract

A ferritin particle consists of 24 ferritin proteins (FTH1 and FTL) and stores iron ions within it. During iron deficiency, ferritin particles are transported to lysosomes to release iron ions. Two transport pathways have been reported: macroautophagy and ESCRT-dependent endosomal microautophagy. Although the membrane dynamics of these pathways differ, both require NCOA4, which is thought to be an autophagy receptor for ferritin. However, it is unclear whether NCOA4 only acts as an autophagy receptor in ferritin degradation. Here, we found that ferritin particles form liquid-like condensates in a NCOA4-dependent manner. Homodimerization of NCOA4 and interaction between FTH1 and NCOA4 (i.e., multivalent interactions between ferritin particles and NCOA4) were required for the formation of ferritin condensates. Disruption of these interactions impaired ferritin degradation. Time-lapse imaging and three-dimensional correlative light and electron microscopy revealed that these ferritin–NCOA4 condensates were directly engulfed by autophagosomes and endosomes. In contrast, TAX1BP1 was not required for the formation of ferritin–NCOA4 condensates but was required for their incorporation into autophagosomes and endosomes. These results suggest that NCOA4 acts not only as a canonical autophagy receptor but also as a driver to form ferritin condensates to facilitate the degradation of these condensates by macroautophagy (i.e., macroferritinophagy) and endosomal microautophagy (i.e., microferritinophagy).

Digest it all: the lysosomal turnover of cytoplasmic aggregates

“…So far, only a very recent report could show that specific aggregates (i.e., ferritin conglomerates) can be degraded by eMI. This process depends on the LLPS of these conglomerates and on the autophagy receptors NCOA4 and TAX1BP1 [38]. This is in line with studies showing that eMI can engulf other large cargoes such as fragments of the ER, mitochondria, peroxisomes, or nuclei [81,82].…”

“…Several soluble SARs, namely TAX1BP1, p62/SQSTM1, NBR1, OPTN, and perhaps TOLLIP, play a central role not only in aggrephagy [14,37,38] but also in other selective types of autophagy where they facilitate the removal of different unwanted intracellular material, including mitochondria (mitophagy), peroxisomes (pexophagy), invading pathogens (xenophagy), lysosomes ) and other crucial adaptor proteins that facilitate and assist SARs (e.g., ALFY) and the chaperone-assisted selective autophagy/BAG-instructed proteasomal to autophagosomal switch and sorting (CASA/BIPASS) system) have been described to mediate the sequestration of these cargoes into autophagosomes. Within the CASA/BIPASS system, it remains unknown in which form the cargo proteins are being degraded (i.e., soluble, p62 bodies, or solid aggregates).…”

“…Although p62 bodies have frequently been used as a bona fide model structure to study the mechanism of aggrephagy (Figure 1i-iii), other LLPS structures such as Ape1 complexes, stress granules, ferritin particles, and P granules are also autophagy targets [38,44,[52][53][54][55]. In yeast, the Ape1 protease undergoes LLPS to form so-called semi-lipid droplets, which are subsequently recognized by Atg19, a SAR, in an Ub-independent manner.…”

Autophagy genes in biology and disease

“…In mammalian cells, fluid-like ferritin–NCOA4 condensates are subjected to macroautophagy and endosomal microautophagy, both of which require TAX1BP1 for incorporation 112 , 113 . Because TAX1BP1 interacts with NCOA4, TAX1BP1 can bridge autophagosomal ATG8 and ferritin–NCOA4 condensates in macroautophagy.…”

Availability of Ferritin-Bound Iron to Enterobacteriaceae

“…On the other hand, facultatively intracellular microbes may trigger autophagy to access ferritin-stored iron. This idea is supported by a recent study demonstrating ferritin to be compartmentalized in the cytosol as a liquid phase condensate with NCOA4 [ 83 ]. This may alter the way previous and present studies have to be interpreted, as the availability of ferritin-bound iron present in this condensate has not been studied thoroughly.…”