Quantitative Ubiquitylome Analysis Reveals the Specificity of RNF111/Arkadia E3 Ubiquitin Ligase for its Degradative Substrates SKI and SKIL/SnoN in TGF-β Signaling Pathway

Victor Laigle,Florent Dingli,Sadek Amhaz,Tiphaine Perron,Mouna Chouchène,Sabrina Colasse,Isabelle Petit,Patrick Poullet,Damarys Loew,Céline Prunier,Laurence Levy

doi:10.1016/j.mcpro.2021.100173

Abstract

RNF111/Arkadia is an E3 ubiquitin ligase that activates the transforming growth factor-β (TGF-β) pathway by degrading transcriptional repressors SKIL/SnoN and SKI. Truncations of the RING C-terminal domain of RNF111 that abolish its E3 function and subsequently activate TGF-β signaling are observed in some cancers. In the present study, we sought to perform a comprehensive analysis of RNF111 endogenous substrates upon TGF-β signaling activation using an integrative proteomic approach. In that aim, we carried out label-free quantitative proteomics after the enrichment of ubiquitylated proteins (ubiquitylome) in parental U2OS cell line compared with U2OS CRISPR engineered clones expressing a truncated form of RNF111 devoid of its C-terminal RING domain. We compared two methods of enrichment for ubiquitylated proteins before proteomics analysis by mass spectrometry, the diGlycine (diGly) remnant peptide immunoprecipitation with a K-ε-GG antibody, and a novel approach using protein immunoprecipitation with a ubiquitin pan nanobody that recognizes all ubiquitin chains and monoubiquitylation. Although we detected SKIL ubiquitylation among 108 potential RNF111 substrates with the diGly method, we found that the ubiquitin pan nanobody method also constitutes a powerful approach because it enabled the detection of 52 potential RNF111 substrates including SKI, SKIL, and RNF111. Integrative comparison of the RNF111-dependent proteome and ubiquitylomes enabled the identification of SKI and SKIL as the only targets ubiquitylated and degraded by RNF111 E3 ligase function in the presence of TGF-β. Our results indicate that lysine 343 localized in the SAND domain of SKIL constitutes a target for RNF111 ubiquitylation and demonstrate that RNF111 E3 ubiquitin ligase function specifically targets SKI and SKIL ubiquitylation and degradation upon TGF-β pathway activation.

Highlights

Identification of the endogenous ring finger protein 111 (RNF111) ubiquitylated substrates. A new powerful method to profile protein ubiquitylation using pan UB nanobody. Lysine K343 as a target for RNF111 ubiquitylation. SKI and SKIL are the only identified RNF111 degradative targets in transforming growth factor-β (TGF-β) signaling
Further sequencing of the genomic region of exon 5 indicates that RNF111-RING-KO clone #1 carries two alleles with a frameshift mutation at position 434 followed by 15 extra amino acids before stop mutation and 1 allele with a frameshift mutation at position 437 followed by 46 extra amino acids; whereas RNF111-RING-KO clone #2 carries the same frameshift mutation at position 426 followed by three extra amino acids on all alleles
We further validated by Western-blot that Kynureninase (KYNU), growth differenciation factor 15 (GDF15), and fatty acid binding protein 3 (FABP3), some of the strongest candidates with commercially available antibodies, are increased in both RNF111-RING-KO clones #1 and #2 as compared with parental U2OS cells, but unlike SKI and SKIL, we found that this increase is independent of TGF-β (Fig. 2C)

Summary

Introduction

RNF111/Arkadia is an E3 ubiquitin ligase that activates the transforming growth factor-β (TGF-β) pathway by degrading transcriptional repressors SKIL/SnoN and SKI. Polyubiquitylation can occur by polymerization of the ubiquitin molecules via one of its seven lysine residues (K6, K11, K27, K29, K33, K48, and K63) or its N-terminal methionine (M1), which generates as many different polyubiquitin linkages [1] This ubiquitylation code leads to distinct biochemical outcomes for the substrate, and it is admitted that only K48 polyubiquitylation, and to a lesser extent K11 polyubiquitylation, drive substrates toward degradation by the proteasome. We and others have found that RNF111 harbors a C-terminal RING domain required for its E3 ubiquitin ligase function that activates SMAD-dependent transcription in response to TGF-β by inducing degradation of SKI and SKIL ( named SnoN) transcriptional repressors [4,5,6].

Objectives

Results

Conclusion