A SARS-CoV-2-Human Protein-Protein Interaction Map Reveals Drug Targets and Potential Drug-Repurposing

David E. Gordon; Gwendolyn M. Jang; Mehdi Bouhaddou; Jiewei Xu; Kirsten Obernier; Matthew J O'Meara; Jeffrey Z. Guo; Danielle L. Swaney; Tia A. Tummino; Ruth Huttenhain; Robyn Kaake; Alicia L. Richards; Beril Tutuncuoglu; Helene Foussard; Jyoti Batra; Kelsey Haas; Maya Modak; Minkyu Kim; Paige Haas; Benjamin J. Polacco; Hannes Braberg; Jacqueline M. Fabius; Manon Eckhardt; Margaret Soucheray; Melanie Brewer; Merve Cakir; Michael J. McGregor; Qiongyu Li; Zun Zar Chi Naing; Yuan Zhou; Shiming Peng; Ilsa T. Kirby; James E. Melnyk; John S Chorba; Kevin Lou; Shizhong A. Dai; Wenqi Shen; Ying Shi; Ziyang Zhang; Inigo Barrio-Hernandez; Danish Memon; Claudia Hernandez-Armenta; Christopher J.P. Mathy; Tina Perica; Kala B. Pilla; Sai J. Ganesan; Daniel J. Saltzberg; Rakesh Ramachandran; Xi Liu; Sara B. Rosenthal; Lorenzo Calviello; Srivats Venkataramanan; Yizhu Lin; Stephanie A. Wankowicz; Markus Bohn; Phillip P. Sharp; Raphael Trenker; Janet M. Young; Devin A. Cavero; Joseph Hiatt; Theo Roth; Ujjwal Rathore; Advait Subramanian; Julia Noack; Mathieu Hubert; Ferdinand Roesch; Thomas Vallet; Björn Meyer; Kris M. White; Lisa Miorin; Oren S. Rosenberg; Kliment A. Verba; David Agard; Melanie Ott; Michael Emerman; Davide Ruggero; Adolfo Garcí-Sastre; Natalia Jura; Mark von Zastrow; Jack Taunton; Olivier Schwartz; Marco Vignuzzi; Christophe d'Enfert; Shaeri Mukherjee; Matt Jacobson; Harmit S. Malik; Danica G Fujimori; Trey Ideker; Charles S Craik; Stephen Floor; James S. Fraser; John Gross; Andrej Sali; Tanja Kortemme; Pedro Beltrao; Kevan Shokat; Brian K. Shoichet; Nevan J. Krogan

While the cells used for these AP-MS experiments, HEK-293T kidney cells, are permissive to SARS-CoV-2 infection 31 , they do not represent the primary physiological site of infection. Therefore, we asked whether the host proteins bound by SARS-CoV-2 might be specifically relevant to the virus's typical environment, lung tissue. We tested if the interacting human proteins were preferentially highly expressed, at the protein level, in any of 29 human tissues 32 , which identified the lung as the tissue with the highest expression of the prey proteins relative to the average proteome (Fig. 2c) . In accordance to this, when compared to overall RefSeq gene expression in the lung (median=3.198 TPM), the interacting proteins were more highly expressed (median=25.52 TPM, p=0.0007; t-test, Extended Data Fig. 4 ) and were also specifically enriched in lung expression relative to other tissues (Extended Data Fig. 5 ). These data support the hypothesis that SARS-CoV-2 preferentially hijacks proteins available in lung tissue and the degree of tissue specific expression may facilitate the drug targeting of host factors with fewer systemic side-effects; this observation is particularly compelling given its derivation from HEK293 cells, which are themselves not lung derived. To further study the relevance of our map, we analyzed the recently reported protein abundance changes during SARS-CoV-2 infection 33 . We calculated, when possible, the correlation between changes in abundance of viral proteins and their human interaction partners across the 4 timepoints reported. Interacting pairs tended to have stronger correlated changes than other pairs of viralhuman proteins (Fig. 2d , KS test p-value=4.8e−05), arguing that the AP-MS derived interactions are of high relevance for the target tissue and the infection context. We compared our SARS-CoV-2 interaction map with those derived for 10 other pathogens (Fig. 2e) , identifying West Nile Virus (WNV) 23 and Mycobacterium tuberculosis (Mtb) 27 as having the most similar host protein interaction partners. The association with Mtb is particularly interesting considering it also infects lung tissue.