The BPAN and intellectual disability disease proteins WDR45 and WDR45B modulate autophagosome-lysosome fusion

ABSTRACT

WDR45 and WDR45B are β-propeller proteins belonging to the WIPI (WD repeat domain, phosphoinositide interacting) family. Mutations in WDR45 and WDR45B are genetically linked with beta-propeller protein-associated neurodegeneration (BPAN) and intellectual disability (ID), respectively. WDR45 and WDR45B are homologs of yeast Atg18. Atg18 forms a complex with Atg2 for autophagosome biogenesis, probably by transferring lipids from the ER to phagophores. We revealed that WDR45 and WDR45B are critical for autophagosome-lysosome fusion in neural cells. WDR45 and WDR45B, but not their disease-related mutants, bind to the tether protein EPG5 and facilitate its targeting to late endosomes/lysosomes. In Wdr45 Wdr45b-deficient cells, the formation of tether-SNARE fusion machinery is compromised. The macroautophagy/autophagy deficiency in wdr45 wdr45b DKO cells is ameliorated by suppression of O-GlcNAcylation, which promotes autophagosome maturation. Thus, our results provide insights into the pathogenesis of WDR45- and WDR45B-related neurological diseases.

KEYWORDS:

• Autophagy
• autophagosome maturation
• WDR45
• WDR45B
• BPAN
• ID

Human genetic studies revealed that de novo mutations in WDR45 are causatively linked with a broad spectrum of neurological diseases, including beta-propeller protein-associated neurodegeneration (BPAN), Rett-like syndrome, intellectual disability (ID), West syndrome, and epileptic encephalopathy. WDR45B has been identified as a causative gene for ID. WDR45 and WDR45B share extensive sequence similarities and are both human homologs of the yeast autophagy protein Atg18. Our previous studies showed that depletion of Wdr45 or Wdr45b in mice causes impaired learning and memory capability and accumulation of swollen axons, thus recapitulating some key features of BPAN and ID patients. wdr45 and wdr45b single knockout (KO) mice show weak autophagy defects, whereas double KO (DKO) mice exhibit strong autophagy defects in the brain, but not in other tissues, which supports the notion that Wdr45 and Wdr45b play partially redundant roles in neural autophagy to maintain CNS homeostasis. However, the molecular functions of WDR45 and WDR45B in the autophagy pathway remain unclear.

Autophagy is an essential intracellular degradation mechanism, conserved from yeast to mammals. The autophagic process starts from the formation of a double-membrane phagophore that expands into an autophagosome, which eventually fuses with the vacuole (yeast and plants)/lysosomes (metazoans) for degradation of its contents. Extensive study of the molecular mechanism of autophagy originated from the identification of multiple Atg (autophagy-related) proteins from yeast genetic screens. Among these Atg proteins, the Vps34 phosphatidylinositol (PtdIns) 3-kinase complex catalyzes local production of PtdIns3P, which triggers the subsequent recruitment of PtdIns3P effectors, such as Atg18 and Atg21. Atg18 forms a complex with Atg2. Recently, Atg2 was shown to possess lipid transfer activity in vitro, which is facilitated by Atg18. This indicates that Atg18-Atg2 may be involved in transferring lipid from the ER for phagophore membrane expansion.

Mammalian cells have four WD40-repeat-containing PtdIns3P-interacting proteins, WIPI1, WIPI2, WDR45B/WIPI3 and WDR45/WIPI4, which all belong to the Atg18 family. They can be further divided into two subgroups, WIPI1/2 and WDR45B/WDR45. WIPI2 interacts with ULK1-RB1CC1/FIP200 and VAPA-VAPB to tether the ER with phagophores and recruits ATG12–ATG5-ATG16L1 for LC3 lipidation on PtdIns3P-positive phagophore membranes. WDR45 and WDR45B were previously reported to modulate autophagy through upstream AMPK and MTOR pathways, and may also regulate the size of autophagosomes downstream of WIPI2. WDR45 promotes the lipid transfer function of mammalian ATG2s in vitro. These observations suggest that the complex may transfer phospholipids from the ER to phagophores. However, the in vivo evidence for lipid transfer is still lacking.

In a recent study [1], we have investigated the role of WDR45 and WDR45B in the autophagy pathway. We generated wdr45 and wdr45b single and double KOs in multiple cell lines, and found that KO of Wdr45 or Wdr45b individually causes no obvious autophagy defect. Depletion of both genes causes dramatically increased levels of SQSTM1/p62 and LC3-II in neuroblast-derived Neuro-2a (N2a) cells, but not in cell types with other tissue origins. These results indicate a partial functional redundancy of WDR45 and WDR45B specifically in neural autophagy. Furthermore, BPAN- and ID-related mutations of WDR45 or WDR45B fail to suppress the increased number of LC3 puncta in wdr45 wdr45b DKO N2a cells, which suggests that autophagy defects may contribute to the pathogenesis of the diseases. Transmission electron microscopy (TEM) analyses revealed that closed autophagosomes are formed in wdr45 wdr45b DKO cells, although the size of autophagosomes is reduced compared to control cells. Overexpression of ATG2A partially reverses the smaller autophagosome phenotype in wdr45 wdr45b DKO cells, providing in vivo evidence that ATG2s supply phospholipids for phagophore expansion with the assistance of WDR45 and WDR45B. The complete closure of autophagosomes in wdr45 wdr45b DKO cells was further verified by fluorescence protease protection/FPP and Halo-LC3 assays.

Yeast Atg18 regulates the morphology and function of the vacuole by PtdIns(3,5)P₂-dependent targeting to the vacuolar membrane. We found that Wdr45 Wdr45b deficiency does not affect lysosomal function, which prompted us to examine the autophagosome-lysosome fusion step. wdr45 wdr45b DKO causes dramatically reduced colocalization of autophagosomes with lysosomes. TEM images demonstrated increased accumulation of autophagosomes and reduced numbers of autolysosomes in wdr45 wdr45b-depleted cells. The fusion between autophagosomes and late endosomes/lysosomes requires dedicated cooperation among tether proteins, SNAREs and RAB proteins. Interestingly, WDR45 and WDR45B interact with the tether protein EPG5 and facilitate its late endosome/lysosome localization. EPG5 mistargets to autophagosomes in the absence of Wdr45 and Wdr45b, which compromises its interaction with the SNARE complex and impairs assembly of the fusion machinery. Disease-causing mutations significantly decrease the binding of WDR45 and WDR45B to EPG5, which suggests the importance of these mutated residues in mediating this interaction. Enhancing fusion machinery assembly by inhibiting O-GlcNAcylation rescues the autophagy defects in wdr45 wdr45b-deficient cells. This observation provides potential targets for treating BPAN and ID patients.

This study and our previous findings in mice indicate that WDR45 and WDR45B are specifically required for autophagy in neural cells. But why do neural cells require more factors to regulate autophagosome-lysosome fusion? This is probably due to the special nature of neuronal autophagy. In neurons, autophagosomes are generated at the distal tip of the axons, followed by transport to the soma of the cells for formation of degradative autolysosomes. During this long-distance transport, newly formed autophagosomes mature by sequential fusion with endolysosomal compartments, which gradually increase their degradative activity from the distal regions to the soma. We speculate that more factors (e.g., WDR45 and WDR45B) are required in neurons to ensure that autophagosomes fuse correctly with diverse endolysosomal compartments. In the future, more studies are needed to figure out how WDR45 and WDR45B are differentially regulated in non-neural and neural cells.

In conclusion, our study uncovered that WDR45 and WDR45B are critical for autophagosome maturation into autolysosomes in neural cells, and their disease-related mutations disrupt their autophagic function. WDR45 and WDR45B bind to the tether protein EPG5 and assist its targeting to late endosomes/lysosomes, and therefore facilitate the formation of tether-SNARE complexes that mediate autophagosome-lysosome fusion. This mechanism provides novel insights into the pathogenesis of the human diseases BPAN and ID and potential therapeutic targets.

Acknowledgments

We are grateful to Dr. Isabel Hanson for editing work.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Natural Science Foundation of China (NSFC) [31671430 to Y.G.Z.].