ISG15 counteracts Listeria monocytogenes infection.
Bottom Line: ISG15 is an interferon-stimulated, linear di-ubiquitin-like protein, with anti-viral activity.Surprisingly this induction can be type I interferon independent and depends on the cytosolic surveillance pathway, which senses bacterial DNA and signals through STING, TBK1, IRF3 and IRF7.Together, our data reveal a previously uncharacterized ISG15-dependent restriction of Listeria infection, reinforcing the view that ISG15 is a key component of the innate immune response.
Affiliation: Unité, Institut Pasteur, Paris, France.
ISG15 is an interferon-stimulated, linear di-ubiquitin-like protein, with anti-viral activity. The role of ISG15 during bacterial infection remains elusive. We show that ISG15 expression in nonphagocytic cells is dramatically induced upon Listeria infection. Surprisingly this induction can be type I interferon independent and depends on the cytosolic surveillance pathway, which senses bacterial DNA and signals through STING, TBK1, IRF3 and IRF7. Most importantly, we observed that ISG15 expression restricts Listeria infection in vitro and in vivo. We made use of stable isotope labeling in tissue culture (SILAC) to identify ISGylated proteins that could be responsible for the protective effect. Strikingly, infection or overexpression of ISG15 leads to ISGylation of ER and Golgi proteins, which correlates with increased secretion of cytokines known to counteract infection. Together, our data reveal a previously uncharacterized ISG15-dependent restriction of Listeria infection, reinforcing the view that ISG15 is a key component of the innate immune response.
No MeSH data available.
Related in: MedlinePlus
Mentions: Since ISG15 is a ubiquitin-like modifier that is known to be covalently linked to hundreds of cellular and several viral substrates (Giannakopoulos et al., 2005; Zhao et al., 2005), our goal was to identify which substrates following overexpression of ISG15 could account for the protective effect detected in the context of Listeria infection and used a proteomic approach. We made use of SILAC coupled with LC-MS/MS and compared cells that express empty vector, cells that express ISG15 and cells that express ISG15 that were treated with interferon, the primary inducer of ISGylation (Figure 4A,B). We identified thirty ISGylated proteins modified following overexpression of ISG15 (Figure 4C–E, Figure 4—source data 1, depicted in blue). The proteomics data have been deposited to the ProteomeXchange Consortium (Vizcaino et al., 2014) via the PRIDE partner repository with the data set identifier PXD001805 (Radoshevich et al., 2015a). Interestingly, these proteins have not yet been reported to be targets of ISGylation. Following interferon treatment, ISG15 modified twelve additional proteins distinct from those that were ISGylated after overexpression without treatment (Figure 4C–E, Figure 4—source data 1, depicted in red). Three of these proteins are known targets of ISGylation following interferon treatment in the aforementioned screens. The other nine are novel substrates of ISGylation. To gain more insight into the role of ISGylation of the modified substrates following overexpression, we performed a gene ontology (GO) analysis of the thirty ISGylated proteins. To our surprise over 80% of ISG15-target proteins are integral membrane proteins (Figure 5A). Even more intriguingly they are known to be primarily localized to the endoplasmic reticulum and Golgi apparatus and/or are critical for glycosylation, ER morphology and ER to Golgi trafficking (e.g., the oligosaccharyl-transferase (OST) complex, RTN4, ATL3, SEC22B, ERGIC1, and ERGIC3; Figure 4C). One of the proteins enriched following ISG15 overexpression is Magnesium Transporter 1 (MAGT1). MAGT1 is critical for T cell activation and patients with a deletion in the MAGT1 gene are susceptible to viral and certain bacterial infections (Li et al., 2014). The protein is localized to the cell surface or to the ER where it interacts with the OST complex, which is critical for N-glycosylation (Pfeffer et al., 2014). Notably, we also identified multiple proteins from the OST complex as targets of ISGylation (Figure 4C, inset). To validate that MAGT1 was modified by ISG15, we affinity purified ISGylated proteins and performed immunoblot analysis (Figure 5B). Two higher molecular weight ISG15-MAGT1 complexes appear that are absent in control cells (Figure 5B, arrows). To assess whether MAGT1 was ISGylated following infection, we immunoprecipitated ISG15 and could show that there is a higher molecular weight MAGT1 complex that immunoprecipitates with ISG15 (Figure 5—figure supplement 1A). Additionally, non-conjugated MAGT1 was enriched following ISG15 immunoprecipitation as well. Whether the interaction is direct or indirect within a multi-protein complex remains to be determined. Of note, MAGT1 is also induced by Listeria infection (Figure 5—figure supplement 1A). A second target of ISGylation following ISG15 overexpression is Reticulon 4 (RTN4) (Figure 4C,D). It is known that RTN4 helps to determine the morphology of the endoplasmic reticulum (Voeltz et al., 2006). We also validated that endogenous RTN4 was modified by ISG15 by immunoblot (Figure 5B), and we detected an upshifted ladder of modified RTN4 that was also recognized by the ISG15 antibody. ISG15 is not known to make chains so the ladder most likely corresponds to either multiple ISGylations of RTN4 or concomitant modifications of RTN4 by ISG15 together with other modifiers as opposed to a single modification. Here again we could validate the endogenous modification of RTN4 by ISG15 following Listeria infection (Figure 5—figure supplement 1A) by detecting multiple distinct slower migrating bands of RTN4 and ISG15 (Figure 5—figure supplement 1A). Interestingly, Atlastin-3, another protein that affects ER morphology was also identified as an ISG15 target. We attempted to validate this interaction by immunoblot but were unable to identify an upshifted complex, in part due to inadequate antibody sensitivity (data not shown). Atlastin-3 is an ER dynamin-like GTPase that can alter the morphology of both the ER and the Golgi apparatus (Rismanchi et al., 2008; Hu et al., 2009). RTN4 and Atlastin-3 can interact to help shape the tubular ER. When RTN4 is overexpressed it increases protein disulphide-isomerase (PDI) clustering in the cell (Yang et al., 2009; Bernardoni et al., 2013). We thus assessed PDI puncta formation following ISG15 overexpression and in line with our hypothesis found a dramatic increase in puncta formed upon ISG15 expression (Figure 5C). We further assessed the ER morphology of ISG15-overexpressing cells by visualizing SEC61-GFP or KDEL-BFP. Neither of these markers displayed an apparent redistribution in ISG15 overexpressing cells relative to control cells (Figure 5—figure supplement 1B). We also assayed whether there was a difference in ER mass using ER-tracker and FACS analysis (Figure 5—figure supplement 1C). There was no detectable difference between the two cell lines in ER mean fluorescence intensity. Taken together, the most striking effect of ISG15 overexpression on ER morphology is PDI clustering and distribution.10.7554/eLife.06848.011Figure 4.Schematic representation of ISG15 targets identified by proteomics.
No MeSH data available.