Biochemical and structural studies of the midnolin Catch domain bound with both wild-type and mutant IRF4 peptides reveal the molecular basis for its broad substrate specificity

Abstract

The newly discovered midnolin-proteasome pathway is a unique ubiquitin-independent mechanism for degrading nuclear proteins, which is crucial for maintaining cellular protein homeostasis. The Catch domain of midnolin is essential for substrate recognition and binding, yet the underlying mechanism for its broad substrate specificity remains elusive. Transcription factor IRF4, essential for the functions of B and T cells, is a substrate of midnolin. This study presents comprehensive biochemical and structural analyses of the human midnolin Catch domain in complexes with both wild-type and mutant IRF4 peptides. The crystal structure of the Catch-IRF4 complex reveals that the Catch domain creates a substrate-binding groove at the interface of the Catch1 and Catch2 subdomains, recognizing and binding to the 215-QVTGTFYAC-223 sequence motif of IRF4. The binding motif of IRF4 forms a β-strand that is embedded into the substrate-binding groove, resulting in an antiparallel five-stranded β-sheet. The interactions between the IRF4 peptide and the Catch domain are predominantly hydrophobic and exhibit high spatial complementarity. Additionally, the biochemical, modeling and structural data indicate that the V2 and A8 positions of the IRF4 sequence motif can be substituted with other hydrophobic or small polar residues (G/A/V/L/I/M/P/F/Y/C/S/T), but not with large polar and charged residues (D/N/E/Q/H/K/R). The G4 position can be replaced by Ser, while the F6 position can be substituted with Tyr. These results suggest that the Catch domain can recognize and bind to a variety of substrates containing the sequence motif x[G/A/V/L/I/M/P/F/Y/C/S/T]x[G/S]x[F/Y]x[G/A/V/L/I/M/P/F/Y/C/S/T]x or briefly the G/SxF/Y motif (where x represents polar residues) located in an unstructured or loop region on the protein surface, and the hydrophobic interactions and spatial complementarity between the binding motifs of substrates and the Catch domain govern the substrate specificity. Collectively, these findings elucidate the molecular basis for midnolin’s broad substrate specificity.

1. Introduction

Protein homeostasis is a vital biological process that regulates intracellular protein synthesis, folding, and degradation. This process relies on a complex regulatory network that includes molecular chaperones, endoplasmic reticulum protein synthesis, protein degradation systems, and their regulatory factors [1–3]. Dysfunction in protein homeostasis can lead to the accumulation of misfolded or damaged proteins, impairing cellular function and contributing to various diseases, including neurodegenerative disorders and cancers [4,5].

A key aspect of maintaining protein homeostasis is the regulation of protein degradation. In the human body, most proteins are regulated through two well-established degradation pathways: the ubiquitin-proteasome pathway and the lysosomal autophagy pathway. The ubiquitin-proteasome pathway primarily is responsible for degrading most proteins, particularly short-lived cellular regulatory factors [6–8]. In this pathway, ubiquitin is covalently attached to the target protein, which is then recognized by ubiquitin ligases and subsequently presented to the proteasome for degradation. In contrast, the lysosomal-autophagy pathway primarily targets the degradation of long-lived proteins as well as damaged organelles and other cellular components [9,10]. This pathway involves the formation of autophagosomes that engulf target cellular components and their proteins, which then fuse with lysosomes where their contents are degraded by lysosomal enzymes.

Additionally, some proteins undergo degradation through pathways that are neither ubiquitin- nor lysosome-dependent [11,12]. Very recently, a novel non-ubiquitin-dependent proteasomal degradation pathway mediated by midnolin (MIDN) was identified through whole-genome CRISPR-Cas9 screening [13]. This study demonstrated that MIDN can mediate the proteasomal degradation of specific proteins encoded by immediate-early genes (IEGs), such as early growth response 1 (EGR1), FBJ murine osteosarcoma viral oncogene homolog B (FOSB), interferon regulatory factor 4 (IRF4), and several other cell type-specific transcriptional regulators. These IEGs primarily encode rapidly responsive transcription factors, cytoplasmic enzymes, and secretory proteins, which are expressed transiently in response to cellular stimulation and function within minutes to hours before being degraded [14,15]. Cells lacking MIDN exhibit significantly increased stability of EGR1 and FOSB. Conversely, overexpression of MIDN enhances the degradation of these proteins.