SARS-CoV-2 proteases PLpro and 3CLpro cleave IRF3 and critical modulators of inflammatory pathways (NLRP12 and TAB1): implications for disease presentation across species

Abstract

The genome of SARS-CoV-2 encodes two viral proteases (NSP3/papain-like protease and NSP5/3C-like protease) that are responsible for cleaving viral polyproteins during replication. Here, we discovered new functions of the NSP3 and NSP5 proteases of SARS-CoV-2, demonstrating that they could directly cleave proteins involved in the host innate immune response. We identified 3 proteins that were specifically and selectively cleaved by NSP3 or NSP5: IRF-3, and NLRP12 and TAB1, respectively. Direct cleavage of IRF3 by NSP3 could explain the blunted Type-I IFN response seen during SARS-CoV-2 infections while NSP5 mediated cleavage of NLRP12 and TAB1 point to a molecular mechanism for enhanced production of cytokines and inflammatory responThe genome of SARS-CoV-2 encodes two viral proteases (NSP3/papain-like protease and NSP5/3C-like protease) that are responsible for cleaving viral polyproteins during replication. Here, we discovered new functions of the NSP3 and NSP5 proteases of SARS-CoV-2, demonstrating that they could directly cleave proteins involved in the host innate immune response. We identified 3 proteins that were specifically and selectively cleaved by NSP3 or NSP5: IRF-3, and NLRP12 and TAB1, respectively. Direct cleavage of IRF3 by NSP3 could explain the blunted Type-I IFN response seen during SARS-CoV-2 infections while NSP5 mediated cleavage of NLRP12 and TAB1 point to a molecular mechanism for enhanced production of cytokines and inflammatory response observed in COVID-19 patients. We demonstrate that in the mouse NLRP12 protein, one of the recognition site is not cleaved in our in-vitro assay. We pushed this comparative alignment of IRF-3 and NLRP12 homologs and show that the lack or presence of cognate cleavage motifs in IRF-3 and NLRP12 could contribute to the presentation of disease in cats and tigers, for example. Our findings provide an explanatory framework for indepth studies into the pathophysiology of COVID-19.

Keywords: IRF3; NLRP12; NSP3 (PLpro); NSP5 (3CLpro); SARS-CoV-2; TAB1; innate immunity; protease activity.

Figure 1.

Principle of the screen of protease activity of SARS-CoV-2 PLpro and 3CLpro. (A) Schematic of the organization of the genome of SARS-CoV-2, focusing on the non-structural proteins Nsp1-16. As depicted, two proteases are encoded in ORF1a: NSP3 or papain-like protease (PLpro) and NSP5, or 3C-like protease (3CLpro). PLpro is responsible for three proteolytic cleavages, while 3CLpro cuts the large polyprotein at eleven different sites. (B) Results obtained for the family of IRF proteins ; the additional band obtained for IRF3 in the presence of PLpro indicates cleavage (C) Overview of the proteins tested in this study and the proteolytic events detected: out of the 71 proteins tested, PLpro cleaves only IRF3 (indicated in blue) and 3CLpro cleaves NLRP12 and TAB1 (as shown in red).

Figure 2.

Cleavage of IRF3 by SARS-CoV-2 PLpro. (A): SDS-page analysis of the cleavage of human IRF3 protein, with a N-terminal GFP tag. The protein was expressed alone or in the presence of increasing concentrations of the SARS-CoV-2 protease PLpro. (B) Logo analysis of the cleavage site predicted from the polyprotein cleavage of SARS-CoV-2. (C) Representation of the domains found in IRF3 and the position of the cleavage site. (D) Representation of IRF3 structure (from PDB 1J2F). The cleavage sequence LGGG is highlighted in blue (E) Alignment of the amino acids for human IRF1-IRF9, demonstrating that only IRF3 would be predicted to be cleaved as observed. (F) Structure of the IRF3 homodimer (from PDB 1QWT), showing the two fragments (green, N terminal, green, and blue for C-terminal).

Figure 3.

3CLpro (Nsp5) cleaves TAB1 at two separate sites. A): SDS-page analysis of the cleavage of human TAB1 protein, with a C-terminal GFP tag. The protein was expressed alone or in the presence of increasing concentrations of the SARS-CoV-2 protease 3CLpro. The gel shows two additional bands upon cleavage, corresponding to the fragment 132-504 and to the fragment 444 to 504.. (B) same, but for the N-terminal GFP. In this case, the fragments 1-132 and 1-144 are fluorescent and can be detected on the gel, while the fragments 132-444, 444-504 and 132-504 are not fluorescent. (C) Schematic representation of TAB1 protein structure with the location of the identified cleavage sites on the primary sequence. (D) Representation of TAB1 structure (from PDB 2POM). The cleavage sequence ASLQS is highlighted in red. (E) full sequence of amino acids for human TAB1, showing the two putative cleavage sites ASLQS and LTLQS.

Figure 4.

Cleavage of NLRP12 by 3CLpro of SARS-CoV-2. (A): SDS-page analysis of the cleavage of NLRP12 protein, with a C-terminal GFP tag. The protein was expressed alone or in the presence of increasing concentrations of the SARS-CoV-2 protease 3CLpro. (B) Logo analysis of the cleavage site predicted for 3CLpro, from the polyprotein cleavage of SARS-CoV-2. (C) Representation of the domains found in NLRP12 and the position of the cleavage sites.(D) Alignment of the amino acids for human NALPs (NLRPs), demonstrating that only NLRP12 would be predicted to be cleaved as observed. Below, alignment of mouse NLRP12, showing that the first cleavage site is conserved, but the second site presents an A→V mutation that would disrupt cleavage. (E) SDS-page analysis of human and mouse NLRP12, with different tag orientations, to demonstrate the differences between species. The banding patterns obtained in the presence of 3CLpro are consistent with the predicted sizes. Using the mouse NLRP12 constructs, only one cleaved fragment is observed in the N-term and C-term constructs. (F) Representation of NLRP12 structure (derived from the structure of NLRP3 from PDB 6NPY). (top): The cleavage sequence LFQG (site 1, at residue 238) is highlighted in red; the site is presented in a flexible loop and seems fully exposed for cleavage by 3CLpro. (middle): The second cleavage site, (LQA) found at residue 938, it presented in a flexible loop and would be accessible at the tip of an alpha helix in the LR repeats. (bottom): in this view, both cleavage sites are visible.

Figure 5.

IRF3, TAB1 and NLRP12 are decreased upon infection with SARS-CoV-2 in 293T-ACE2 cells, in two independent laboratories. (A–C) 293T-ACE2 cells were infected with SARS-CoV-2 and analysed for viral and host protein levels 72 h post-infection. (D–I) Independently, stable 293T-ACE2 were infected with icSARS-CoV-2 mNeonGreen) and the levels of IRF3, TAB1 And NLRP12 were visualized by Western Blotting at 6 h (D), 24 h (E) and 48 h (F).(G–I): Protein levels measured by densitometry for IRF3 (G), TAB1 (H) and NLRP12 (I) and plotted for the Mock, poly(I:C) and SARS-CoV-2 infection at 6 h, 24 h and 48 h.

Figure 6.

Analysis of the protein sequences across species for IRF3 and NLRP12 cleavage sites. (*) see Materials and Methods for protein sequences for Cotton rats (Sigmodon hispidus) and Minks (Neovison vison).

Figure 7.

PLpro and 3CLpro of SARS-CoV-2 interfere with innate immune response by directly cleaving IRF3, TAB1 and NLRP12. (1) In blue: PLpro (Nsp3) inhibits IFNβ production by cleaving IRF3. (2) in red: 3CLpro could interfere with production of pro-inflammatory cytokines at two levels: cleavage of TAB1 would inhibit activation of NF-κB via TAK1 (2a), while cleavage of NLRP12 could release its inhibitory effect on NF-κB (2b). In addition, the cleavage of NLRP12 by 3CLpro could perturb the NLRP3 inflammasome assembly (2c), especially as one of the cleavage sites would release a PYD domain from NLRP12. This could trigger the cleavage of pro-Caspase-1 and enhance the release of IL-1β.

Figure 8.

Analysis of cleavage sites in potential host species. Most “exotic” species that would be relevant for SARS-CoV, MERS or SARS-CoV-2 present the correct cleavage site for IRF3. This (limited) analysis identified only one species (Myotis davidii) that possess all 5 similar cleavage sites compared to human.