Prediction of Quality-control Degradation Signals in Yeast Proteins

Kristoffer E. Johansson, Bayan Mashahreh, Rasmus Hartmann‐Petersen, Tommer Ravid, Kresten Lindorff‐Larsen

TL;DR

This paper introduces QCDPred, a computational model that predicts protein degradation signals (degrons) based solely on amino acid composition. This work sheds light on a fundamental mechanism of cellular quality control, revealing that exposed hydrophobic residues are key to identifying and degrading misfolded proteins, and this predictor can help understand protein abundance and stability.

What Problem Does This Paper Address?

Cells maintain a delicate balance of proteins (proteostasis) to function correctly. A critical aspect of this is the removal of misfolded or damaged proteins, which can be toxic. This "protein quality control" (PQC) system often targets these aberrant proteins for degradation via the ubiquitin-proteasome pathway. This targeting relies on specific "degron" sequences within the proteins. However, the exact nature and sequence properties of these PQC degrons have remained largely elusive. Understanding these degrons is crucial for comprehending how cells maintain proteostasis and how disruptions in this process can lead to disease. This paper aims to bridge this knowledge gap by developing a predictive model for these PQC degrons.

Methods

The core of this research is the development of a predictive model called QCDPred (Quality Control Degron Prediction). The model is built using a logistic regression approach. The key innovation here is that the model relies solely on the amino acid composition of short peptide sequences (17 residues) as input. This is a departure from methods that might require more complex sequence motifs or structural information. The researchers trained this model on experimental data derived from a large-scale yeast screen. They also explored a more complex di-peptide composition model, but found it offered only marginal improvement, reinforcing the power of simple composition-based features.

Datasets

The primary dataset used for training and evaluation was generated from a yeast-based screen of 23,600 17-residue long peptides derived from the yeast proteome. These peptides were fused to GFP in a reporter construct, allowing researchers to quantify their degradation potential via fluorescence-activated cell sorting (FACS) and sequencing. For analysis of protein abundance and half-life, the researchers utilized a dataset of 4533 yeast proteins from a previous study, incorporating cellular abundance data from the Pax database and predicted disordered regions.

Key Insights and Ideas

  • Hydrophobicity is Key: The most striking finding is that exposed hydrophobic amino acids are the primary drivers of degron activity. This aligns with the expectation that these residues are normally buried within the core of folded proteins and become exposed upon misfolding, signaling them for degradation.
  • Negative Charges Stabilize: Conversely, negatively charged amino acids (like Asp and Glu) appear to counteract degron activity, acting as stabilizing elements. This might explain why many intrinsically disordered proteins, which are often rich in charged residues, are less prone to degradation.
  • Composition Over Specific Sequence: The success of a model based purely on amino acid composition suggests that for general PQC, the presence of certain types of residues (hydrophobic) is more critical than a highly specific linear sequence motif. This implies a more promiscuous recognition mechanism by the PQC system.
  • Degrons and Protein Abundance: The study reveals a correlation between the predicted degron potential in disordered regions of proteins and their cellular abundance and half-life. Proteins with higher predicted degron potential tend to be less abundant and have shorter half-lives, suggesting that PQC actively regulates the levels of proteins with potentially destabilizing sequences.

Key Results

  • The QCDPred model, based on amino acid composition, achieved an Area Under the Curve (AUC) of 0.79 in receiver operating characteristic (ROC) curve analysis, indicating good predictive accuracy.
  • Further refinement of the model, excluding peptides with intermediate degradation potential, improved the AUC to 0.88.
  • The model's estimated parameters revealed that hydrophobic amino acids (Met, Val, Leu, Ile, Cys, Phe, Tyr, Trp) promote degron activity, while negatively charged amino acids (Asp, Glu) counteract it. Positively charged amino acids (Lys, Arg) had a more neutral effect.
  • Analysis of the yeast proteome showed that proteins with higher average degron probability in their disordered regions were generally less abundant and had shorter half-lives. This correlation was statistically significant.

Limitations and Future Work

  • Peptide Length Constraint: The current model is trained on and predicts degrons within 17-residue peptides. It does not account for longer degrons or the potential for multiple degrons within a single protein.
  • Sequence Specificity: While composition is powerful, the model doesn't capture potentially specific sequence motifs that might also contribute to degron recognition. The authors suggest that future work could explore this by using larger and more diverse peptide libraries.
  • Linear Degrons: The current focus is on linear degrons. The paper acknowledges that PQC degrons might also exist in three-dimensional structures and not be apparent in the linear sequence, which future computational and experimental approaches could investigate.
  • Disease Variants: The authors highlight that many disease-causing mutations lead to protein degradation, and this work provides a quantitative model that could help understand the relationship between sequence variation, protein stability, and PQC recognition. Future work could integrate this model with studies of disease variants.

Summary updated

Abstract

Effective proteome homeostasis is key to cellular and organismal survival, and cells therefore contain efficient quality control systems to monitor and remove potentially toxic misfolded proteins. Such general protein quality control to a large extent relies on the efficient and robust delivery of misfolded or unfolded proteins to the ubiquitin-proteasome system. This is achieved via recognition of so-called degradation motifs-degrons-that are assumed to become exposed as a result of protein misfolding. Despite their importance, the nature and sequence properties of quality-control degrons remain elusive. Here, we have used data from a yeast-based screen of 23,600 17-residue peptides to build a predictor of quality-control degrons. The resulting model, QCDPred (Quality Control Degron Prediction), achieves good accuracy using only the sequence composition of the peptides as input. Our analysis reveals that strong degrons are enriched in hydrophobic amino acids and depleted in negatively charged amino acids, in line with the expectation that they are buried in natively folded proteins. We applied QCDPred to the yeast proteome, enabling us to analyse more widely the potential effects of degrons. As an example, we show a correlation between cellular abundance and degron potential in disordered regions of proteins. Together with recent results on membrane proteins, our work suggest that the recognition of exposed hydrophobic residues is a key and generic mechanism for proteome homeostasis. QCDPred is freely available as open source code and via a web interface.
BibTeX
@article{johansson2022prediction,
  author = {Kristoffer E. Johansson and Bayan Mashahreh and Rasmus Hartmann‐Petersen and Tommer Ravid and Kresten Lindorff‐Larsen},
  title = {{Prediction of Quality-control Degradation Signals in Yeast Proteins}},
  journal = {Journal of Molecular Biology},
  year = {2022},
  doi = {10.1016/j.jmb.2022.167915},
}
Figure