Protein Degradation Pathways Associated with Nonsense-Mediated mRNA Decay

Abstract

The presence of a premature termination codon (PTC) in an mRNA results in the generation of a truncated protein which may aggregate or perform dominant negative functions in the cell. To limit the production of these truncated proteins and their deleterious effects, PTC-containing mRNAs are targeted for degradation via the nonsense-mediated decay (NMD) pathway. However, NMD requires the translation of the mRNA before degradation can be triggered, necessarily resulting in the generation of truncated proteins. The fate of the truncated proteins that are synthesized before NMD can act is not well understood. In Chapters 2 and 3, we develop reporter systems in mammalian cells that encode the same protein before a normal or premature termination codon. We show that termination at a PTC is sufficient to target the nascent polypeptide for proteasomal degradation. Proteasome inhibition specifically increases the levels of newly synthesized proteins terminating at a PTC within an hour, but also affects mRNA homeostasis within a few hours. The degradation of proteins that terminate at a PTC requires UPF1, the central NMD factor, and SMG1, the kinase that phosphorylates UPF1 to activate NMD, but not SMG1 kinase activity. These studies reveal that PTC recognition by NMD factors triggers the degradation of both the nascent polypeptide and the mRNA and provide a framework for further mechanistic studies. In Chapter 4, we investigate how the NMD factor UPF3B promotes the protein degradation of its paralog UPF3A. The regulation of UPF3A protein levels by UPF3B is thought to act as a feedback mechanism to maintain proper levels of NMD within the cell, and mutations in UPF3B that result in an increase in UPF3A protein levels are linked to intellectual disability in humans. Consistent with previously published results, we find that UPF2 is required for the upregulation of UPF3A protein levels when UPF3B is depleted, but UPF2 binding to UPF3B is required for downregulation of UPF3A protein levels by UPF3B. The current model reconciles this dual role of UPF2 in regulating UPF3A protein levels by proposing that UPF3A competes with UPF3B for binding to UPF2 and is stabilized upon binding to UPF2. We challenge this model by showing that UPF3A does not need to bind to UPF2 to be upregulated by UPF3B depletion. We also identify the N-terminal 65 residues of UPF3A as the region required for UPF3A to respond to UPF3B levels and show that UPF3A can be monoubiquitinated in human cell lysate. Our results provide the starting point for a revised model of how UPF3B regulates UPF3A protein levels.