The SARS-CoV-2 accessory protein Orf3a is not an ion channel, but does interact with trafficking proteins

Read the full article See related articles


The severe acute respiratory syndrome associated coronavirus 2 (SARS-CoV-2) and SARS-CoV-1 accessory protein Orf3a colocalizes with markers of the plasma membrane, endocytic pathway, and Golgi apparatus. Some reports have led to annotation of both Orf3a proteins as a viroporin. Here we show that neither SARS-CoV-2 nor SARS-CoV-1 form functional ion conducting pores and that the conductances measured are common contaminants in overexpression and with high levels of protein in reconstitution studies. Cryo-EM structures of both SARS-CoV-2 and SARS-CoV-1 Orf3a display a narrow constriction and the presence of a basic aqueous vestibule, which would not favor cation permeation. We observe enrichment of the late endosomal marker Rab7 upon SARS-CoV-2 Orf3a overexpression, and co-immunoprecipitation with VPS39. Interestingly, SARS-CoV-1 Orf3a does not cause the same cellular phenotype as SARS-CoV-2 Orf3a and does not interact with VPS39. To explain this difference, we find that a divergent, unstructured loop of SARS-CoV-2 Orf3a facilitates its binding with VPS39, a HOPS complex tethering protein involved in late endosome and autophagosome fusion with lysosomes. We suggest that the added loop enhances SARS-CoV-2 Orf3a ability to co-opt host cellular trafficking mechanisms for viral exit or host immune evasion.

Article activity feed

  1. eLife assessment

    The function of specific proteins made by SARS-CoV-1 and SARS-CoV-2 is under debate, with diverging claims previously published regarding the ability of Orf3a proteins from either virus to form ion channels. The authors undertook a thorough characterization of Orf3a from CoV-1 and CoV-2 by combining data from a range of different structural and functional experiments, arguably providing the most compelling evidence to date that Orf3a from viruses is not an ion channel. Instead, the orthologue-specific interaction with a component of a larger protein complex suggests a role of one of the two membrane proteins in the endo-lysosomal pathway. The work is significant from a fundamental science perspective, for its implications for COVID antiviral development strategies, and also for establishing guidelines for future identification of true viral ion channels.

  2. Reviewer #1 (Public Review):

    The accessory protein Orf3a from severe acute respiratory syndrome coronavirus (SARS-CoV-1 or SARS-CoV-2) was initially suggested to be a viroporin and function as a cation channel. In this study, Miller performed a comprehensive structural and functional investigation of SARS-CoV-2 Orf3a utilizing a multidisciplinary approach, including extensive electrophysiological analysis using different systems and determination of multiple single-particle EM structures of the protein under different conditions. Their findings demonstrated that Orf3a has no channel function and is unlikely to be a viroporin. In addition, they tried, but failed to record any channel activity of Orf3a claimed in other studies. They demonstrated that large single-channel currents measured from vesicle-reconstituted Orf3a are due to transient membrane leakiness caused by high protein/lipid ratio and/or channel contamination. Furthermore, they found that SARS-CoV-2 Orf3a, but not SARS-CoV-1 Orf3a, interacts with VPS39, a host HOPS protein involved in autophagosome/late endosome fusion with the lysosome. They proposed that the interaction between SARS-CoV-2 Orf3a and VPS39 may function to assist with SARS-CoV-2 exit and host intracellular immune evasion. This is a meticulously executed research work. I appreciate the tremendous effort the authors spent in the study to clarify some misconceptions related to the role and function of Orfsa from coronavirus.

  3. Reviewer #2 (Public Review):

    There is currently much discussion about the function of several viral proteins hypothesized to be "viroporins", especially specific proteins within SARS-CoV-1 and CoV-2, such as Orf3a. While some prior studies suggest that Orf3a exhibits ion channel activity, others disagree on this important topic. In the present study, compelling evidence is presented that Orf3a does not function as an ion channel, and suggestions are made as to its actual function. The study combines imaging to delineate Orf3a location in the cell, extensive functional analyses that demonstrate a lack of ion channel activity beyond endogenous currents, and compelling structural evidence that Orf3a does not take the form of an ion channel - lacking a clear conduction pathway and also having a basic aqueous vestibule that would not be predicted to support cation channel activity. Finally, co-assembly with trafficking proteins suggest, instead, functioning of Orf3a as a host cell trafficking disruptor that could contribute to immune cell evasion or even viral exit.

    The authors present exhaustive, high-quality data to support their conclusions that Orf3a proteins from SARS-CoV-1 and SARS-CoV-2 do not exhibit ion channel activity. They clearly show Orf3a at the cell membrane and fail to detect ion channel activity using multiple modalities. I believe this work closes the book on the question of Orf3a as a viroporin. It is difficult to find any deficiencies in the experimental work. The parts about a role disrupting trafficking are a little more speculative but nevertheless appropriate and serve as a guidepost for future studies to fully elucidate the true role of Orf3a.

  4. Reviewer #3 (Public Review):

    This study combines data from cryo-electron microscopy, electrophysiology and cellular localization studies to provide insight into the structure and potential function of two orthologues of the membrane protein Orf3a from the corona viruses SARS-CoV-1 and SARS-CoV-2. The work follows up on previous studies, which assigned these proteins as viral ion channels (viroporins). By using patch-clamp electrophysiology in different cellular systems and from reconstituted protein, the authors provide convincing evidence that these proteins do likely not function as ion channels and that previous conclusions in this direction were presumably based on experimental artifacts. The lack of functional evidence is supported by structures of both proteins in different lipid environments, which concur with previous structures of the same system, and which do not show characteristic features of an ion channel. Instead, the authors describe the localization of both proteins on the plasma membrane and endo-lysosomal compartments, and they show specific interactions of the orthologue from SARS-CoV2 but not SARS-CoV1 with the protein VPS39, which as part of the HOPS complex is involved in the fusion of late endosomes and autophagosomes with lysosomes.

    The strength of this manuscript relies on the wealth of high-quality data and its careful analysis, which refutes the presumed function of the viral membrane protein Orf3a as viroporin. Instead, the work provides conclusive evidence for its involvement in a different process. The electrophysiology data is very well carried out and the authors make a convincing case that the observed lack of specific currents renders a role of Orf3a as ion channel as highly unlikely. Similarly, the structural data and the cellular studies are of high quality.

    The main weakness of the study, which should be considered minor in light of the strong results, relates to the unclear relevance of structural features of Orf3a to the still poorly defined function of the protein. In this respect, I regard the discussion of potential lipid density at the cytoplasmic side as exaggerated. The only region that was assigned a functional importance in mediating interactions with the protein VPS39 is unstructured and only found in one of the two orthologs. Although the data describing the interaction between SARS-CoV-2 Orf3a and VPS39 is conclusive, a function of Orf3a that is common to both viral orthologs is still missing. These weaknesses can be addressed by some revision of the text whereas the clarification of the role of Orf3a is beyond the scope of the current study and should be addressed in future work.