Proximity labeling approaches to study protein complexes during virus infection

Profiling phagosome proteins identifies PD-L1 as a fungal-binding receptor

Phagocytosis is the process by which myeloid phagocytes bind to and internalize potentially dangerous microorganisms1. During phagocytosis, innate immune receptors and associated signalling proteins are localized to the maturing phagosome compartment, forming an immune information processing hub brimming with microorganism-sensing features2-8. Here we developed proximity labelling of phagosomal contents (PhagoPL) to identify proteins localizing to phagosomes containing model yeast and bacteria. By comparing the protein composition of phagosomes containing evolutionarily and biochemically distinct microorganisms, we unexpectedly identified programmed death-ligand 1 (PD-L1) as a protein that specifically enriches in phagosomes containing yeast. We found that PD-L1 directly binds to yeast upon processing in phagosomes. By surface display library screening, we identified the ribosomal protein Rpl20b as a fungal protein ligand for PD-L1. Using an auxin-inducible depletion system, we found that detection of Rpl20b by macrophages cross-regulates production of distinct cytokines including interleukin-10 (IL-10) induced by the activation of other innate immune receptors. Thus, this study establishes PhagoPL as a useful approach to quantifying the collection of proteins enriched in phagosomes during host-microorganism interactions, exemplified by identifying PD-L1 as a receptor that binds to fungi.

A multitask transfer learning framework for the prediction of virus-human protein-protein interactions

BACKGROUND

Viral infections are causing significant morbidity and mortality worldwide. Understanding the interaction patterns between a particular virus and human proteins plays a crucial role in unveiling the underlying mechanism of viral infection and pathogenesis. This could further help in prevention and treatment of virus-related diseases. However, the task of predicting protein-protein interactions between a new virus and human cells is extremely challenging due to scarce data on virus-human interactions and fast mutation rates of most viruses.

RESULTS

We developed a multitask transfer learning approach that exploits the information of around 24 million protein sequences and the interaction patterns from the human interactome to counter the problem of small training datasets. Instead of using hand-crafted protein features, we utilize statistically rich protein representations learned by a deep language modeling approach from a massive source of protein sequences. Additionally, we employ an additional objective which aims to maximize the probability of observing human protein-protein interactions. This additional task objective acts as a regularizer and also allows to incorporate domain knowledge to inform the virus-human protein-protein interaction prediction model.

CONCLUSIONS

Our approach achieved competitive results on 13 benchmark datasets and the case study for the SARS-COV-2 virus receptor. Experimental results show that our proposed model works effectively for both virus-human and bacteria-human protein-protein interaction prediction tasks. We share our code for reproducibility and future research at https://git.l3s.uni-hannover.de/dong/multitask-transfer .