My project was divided into two main working packages. The first consisted in developing new tools to identify ISGylated substrates while the second consisted in using these strategies for identifying the function of ISG15 in macrophages.
(1) Developing tools to identify ISGylated substrates
Five strategies were proposed for identifying ISGylated proteins.
The most ambitious and promising strategy relied generating a stable cell line expressing a mutant of ISG15 (WP1.4). After genome edition using CRISPR CAS9, I have obtained and isolated this mutant. Unfortunately, this mutant has loss ISGylation capabilities and couldn’t be used for the intended purpose (Figure 1).
One strategy was based using transgenic mice lacking ISG15 (ISG15 KO) and comparing them to mice expressing the unaltered ISG15 protein (ISG15 WT) (WP1.5). Unfortunately, these transgenic mice present different expression levels of the protein ubiquitin. This change in expression rendered the proposed approach impractical at best and was considered unfeasible.
The two antibody-based strategies (WP1.1 and WP1.3) were attempted and ended up being inconclusive as ISG15 is only present at traces amount in cells.
A spin-off of WP1.1 was attempted and was based on tools available for studying ubiquitin and SUMO, two proteins closely related to ISG15. In brief, I tried to identify a protein binding strongly to ISG15 and use it to produce a synthetic protein containing multiple ISG15 binding motifs. I have identified more than 1000 proteins (Figure 2) and tried validate the interaction between ISG15 and the strongest candidate identified: USP14. After several months of trying to produce USP14 in bacterial and mammalian cells, USP14 was deemed to toxic for the cells used. The Trost lab is now focusing on other candidates identified from this list of 1000 proteins.
The last strategy was considered too unspecific by the lab and was rolled out immediately (WP1.2). However, all the other strategies attempted during my fellowship being inconclusive, this last strategy was reconsidered and is now implemented in the lab.
(2) Identifying the function of ISG15 in macrophages
The original goal was to combine the previously proposed approaches to perform a dual affinity purification of ISGylated peptides in mouse cell line and/or tissues. As all attempts proposed in the first working packages have been inconclusive, I have decided to use the previously obtained ISG15 KO mouse line to determine the effect of ISG15 during phagocytosis of latex beads in macrophages using mass spectrometry based proteomics.
Macrophages were derived from ISG15 WT and ISG15 KO mouse bone marrow. After differentiation, cells were incubated with latex beads and phagosome were enriched by ultracentrifugation. A small fraction of the cells was also kept for total cell extract analysis. Proteins from both phagosomal fraction analysed by mass spectrometry based proteomics a state of the art Lumos Tribrid mass spectrometer (Figure 3).
Up and down-regulated proteins were assessed using statistical analysis (Figure 4). Both total cell extract and phagosome fraction show a reduction in lysosomal proteins as well as an increase in early endosome in KO cells. This suggest that phagosome maturation is slower in ISG15 KO mouse macrophages. Mitochondrial proteins were also upregulated in the phagosome fraction while they were downregulated in the total cell extract. This indicates a shift in degradation pathways used in ISG15 KO macrophages.
To confirm these observations, I have then assessed the biochemical functions of the phagosome by flow cytometry using the ISG15 KO cell line. Overall, I observed lower phagocytosis efficiency as well as slower digestion and acidification in ISG15 KO phagosomes. The results of these assays support the LCMS data obtained with the bone marrow derived macrophages and confirm ISG15 is a key modulator of the phagocytic activity and phagosome maturation.
A similar phenotype is usually observed after interferon treatment and is also correlated to a higher antigen presentation rate. In order to assess the antigen presentation efficiency of ISG15 KO cells, bones dissected from ISG15 KO and WT mice and sent to Sophia Maschalidi from Kodi Ravichandran laboratory (VIB-UGent Center for Inflammation Research, Ghent University, Belgium). After bone marrow extraction, cells were differentiated into dendritic cells which are more efficient antigen presenter than macrophages. And antigen presentation was assessed by flow cytometry (Figure 7). Here, we observed higher antigen presentation in ISG15 KO cells after incubation with ovalbumin coated beads and in solution ovalbumin, but not with the already processed SIINFEKL peptide. This suggest ISG15 is not required for antigen presentation, however it is a key factor of antigen processing for later presentation.
