WIPI4

Genomic variants outside of the canonical splicing site (±2) may generate abnormal mRNA splicing, which are defined as non-canonical splicing variants (NCSVs). However, the clinical interpretation of NCSVs in neurodevelopmental disorders (NDDs) is largely unknown.

We investigated the contribution of NCSVs to NDDs from 345,787 de novo variants (DNVs) in 47,574 patients with NDDs. We performed functional enrichment and protein-protein interaction analysis to assess the association between genes carrying prioritised NCSVs and NDDs. Minigene was used to validate the impact of NCSVs on mRNA splicing.

We observed significantly more NCSVs (p = 0.02, odds ratio [OR] = 2.05) among patients with NDD than in controls. Both canonical splicing variants (CSVs) and NCSVs contributed to an equal proportion of patients with NDD (0.76% vs. 0.82%). The candidate genes carrying NCSVs were associated with glutamatergic synapse and chromatin remodelling. Minigene successfully validated 59 of 79 (74.68%) NCSVs that led to abnormal splicing in 40 candidate genes, and 9 of the genes (ARID1B, KAT6B, TCF4, SMARCA2, SHANK3, PDHA1, WDR45, SCN2A, SYNGAP1) harboured recurrent NCSVs with the same variant present in more than two unrelated patients with NDD. Moreover, 36 of 59 (61.02%) NCSVs are novel clinically relevant variants, including 34 unreported and 2 clinically conflicting interpretations or of uncertain significance NCSVs in the ClinVar database.

This study highlights the common pathology and clinical importance of NCSVs in unsolved patients with NDD.

The present study was funded by grants from the National Natural Science Foundation of China, China Postdoctoral Science Foundation, the Hunan Youth Science and Technology Innovation Talent Project, the Provincial Natural Science Foundation of Hunan, The Scientific Research Program of FuRong laboratory, and the Natural Science Project of the University of Anhui Province.

The four human WIPI β-propellers, WIPI1 through WIPI4, belong to the ancient PROPPIN family and fulfill scaffold functions in the control of autophagy. In this context, WIPI β-propellers function as PI3P effectors during autophagosome formation and loss of WIPI function negatively impacts autophagy and contributes to neurodegeneration. Of particular interest are mutations in WDR45, the human gene that encodes WIPI4. Sporadic WDR45 mutations are the cause of a rare human neurodegenerative disease called BPAN, hallmarked by high brain iron accumulation. Here we discuss the current understanding of the functions of human WIPI β-propellers and address unanswered questions with a particular focus on the role of WIPI4 in autophagy and BPAN.

PROPPINs/WIPIs are β-propeller proteins that bind phosphoinositides and contribute to the recruitment of protein complexes involved in membrane remodelling processes such as autophagosome formation and endosomal trafficking. Yeast Atg21 and mammalian WIPI2 interact with Atg16/ATG16L1 to mediate recruitment of the lipidation machinery to the autophagosomal membrane. Here, we used the reverse double two-hybrid method (RD2H) to identify residues in Atg21 and Atg16 critical for protein-protein binding. Although our results are generally consistent with the crystal structure of the Atg21-Atg16 complex reported previously, they also reveal that dimerization of the Atg16 coiled-coil domain is required for Atg21 binding. Furthermore, most of the residues identified in Atg21 are conserved in WIPI2 and we showed that these residues also mediate ATG16L1 binding. Strikingly, these residues occupy the same position in the β-propeller structure as residues in PROPPINs/WIPIs Hsv2 and WIPI4 that mediate Atg2/ATG2A binding, supporting the idea that these proteins use different amino acids at the same position to interact with different autophagic proteins. Finally, our findings demonstrate the effectiveness of the RD2H system to identify critical residues for protein-protein interactions and the utility of this method to generate combinatory mutants with a complete loss of binding capacity.