All cells have ways of conserving energy during times of stress. One strategy is to reuse larger or older cellular structures by encasing them in membranes and breaking them back down into building blocks. This process is termed “autophagy,” which is Greek for “self-eating.”
Autophagy occurs in every organism from yeast to humans and is crucial for energy conservation as well as protection from diseases, such as cancer, diabetes, and neurodegeneration, that are caused by aberrant signaling or other molecules within cells. It turns out that many infectious microorganisms can be destroyed by the autophagy pathway as well. However, many viruses have evolved ways of not only evading the pathway but of actually using it for purposes of replication and spread from cell to cell. This is especially true of the simplest kinds of RNA viruses, which take advantage of the extensive cellular membrane surfaces that occur during the induction of the autophagy pathway to cluster their proteins and RNA. Then, they use the resulting membrane vesicles to traffic to the surface of the cell and exit.
We are interested in how viruses subvert the cellular process of autophagy because it contains many proteins that could serve as targets for inhibitors to block. Targeting parts of the human host that the virus relies on is one strategy to address the problem of drug resistance. When drugs target viral proteins, the rapid mutation rate of the virus can facilitate the formation of drug-resistant variants. However, human DNA mutates much more slowly and it is unlikely that a host protein would change its susceptibility to an inhibitory molecule. Furthermore, the autophagy pathway is a stress-induced pathway; it is dispensable for humans in the short term and, therefore, could be targeted to inhibit viruses without harm to the human host.
We were interested in taking a closer look into how different viruses use this pathway to see if there are any common features that multiple viruses use and would thus make for a good antiviral drug target. We also wanted to get a big-picture view of what parts of the autophagy pathway are used by each virus, since there are many contradictory reports about the use or not of the autophagy pathway for many viruses. To accomplish these goals, we used the popular and effective CRISPR technology to make a panel of cell lines, each of which lacks a different important component of the autophagy pathway of human cells. This allowed us to probe the entire pathway and do a broad comparative study between viruses in different families. We chose three different RNA viruses: poliovirus, a well-studied and notorious human pathogen, and both dengue and Zika virus, two closely related mosquito-borne viruses that cause considerable human disease.
We found that while all three viruses subvert the autophagy pathway, they utilize different parts of the pathway for efficient viral growth. The cellular autophagy pathway requires a series of protein complexes that initiate and form the membrane vesicles that enwrap cellular structures to prepare them for degradation. There are two important initiation complexes and a protein, LC3 that becomes embedded in autophagic membranes after it is covalently attached to a lipid. While all three viruses used different initiation complexes, they all required the LC3 protein. However, they bring LC3 to membranes without requiring it to be linked to the lipid, thus bypassing the extensive machinery required for lipidation that is the hallmark of the “canonical” cellular autophagy pathway. This finding was especially intriguing because it suggests a convergent strategy by different viruses to manipulate the autophagy pathway downstream of its initiation.
We found that during viral infection, LC3 that is not lipidated is still associated with membrane structures. We further discovered that poliovirus proteins are binding LC3 and likely mediating their recruitment to membranes in the absence of the lipidation machinery. There are several reasons viruses would want to bind LC3, including using this abundant protein to localize the viral proteins and to stabilize the curvature of the bound membranes, both for vesicle formation and viral exit from the cell.
We also showed that two specific autophagy-inhibiting drugs that target either the first or second initiation complexes are capable of inhibiting virus growth consistent with our CRISPR cell line data. These findings are revealing and promising — the more we understand which components of cellular processes such as autophagy are exploited by viruses, the better targeted our approach will be for finding ways to inhibit their growth.
These findings are described in the article entitled Differential and convergent utilization of autophagy components by positive-strand RNA viruses, recently published in the journal PLOS Biology.