Poster CC8 - Mechanistic Insights into SARS-CoV-2 Spike Protein Induction of the Chemokine CXCL10

Introduction: In many instances of severe COVID-19, there is an overexpression of pro-inflammatory cytokines leading to a cytokine storm that causes tissue damage and, in certain cases, organ failure or death [1]. The expression of these cytokines is due to an overstimulated innate immune response caused by the viral infection. Innate immune cells express pattern recognition receptors (PRRs) that detect pathogens via pathogen-associated molecular patterns (PAMPs) shed from pathogens such as viruses [2]. Improper regulation of cytokine expression can create a cytokine storm.

In addition to the obvious response to the genomic material of the SARS-CoV-2 virus, an emerging, but controversial, hypothesis is that the innate immune system responds to the protein components of the virus, in particular the spike protein (e.g. [3-5]). The spike protein is made up of two subunits, S1 which binds to the host cell receptor ACE2, and S2 whose function is to facilitate fusion with the cellular membrane [6]. We have previously reported that both S and S2 induce the production of CXCL10 which is a chemokine upregulated in severe COVID-19 [7]. Further, we have shown that some, but not all, inhibitors of glycogen synthase kinase-3 (GSK-3), attenuate S and S2 induction of CXCL10 [7]. In the present study, we have further probed (i) the ability of a highly specific inhibitor for GSK-3, termed COB-187, to inhibit spike protein-induced CXCL10 production and (ii) the mechanism by which the spike protein induces CXCL10.

Materials and

Methods: Reagents
SARS-CoV-2 spike proteins from: (i) ACROBiosystems were full-length extracellular S protein (AS) (cat# SPN-C52H8), mutated AS (cat# SPN-C52H9), the S2 subunit (AS2) (cat# S2N–C52H5; lot# 3621-2042F1-X5), or mutated AS2 (cat# S2N–C52H5; lot# P3840b-2087F1-T7) - the mutated AS and AS2 proteins had the following substitutions: F817P, A892P, A899P, A942P, K986P, and V987P and both versions of the S protein had the following substitutions R683A and R685A.; (ii) RayBiotech was S2 (RS2); and (iii) Biotechne was S2 (BS2). Unless otherwise noted, all spike protein treatments were done in the presence of Polymyxin B (PB; Sigma). Pam3CSK4 and LPS were from (InvivoGen) and SC-514 was from Santa Cruz Biotechnology.

Cell Culture and PBMC preparation
The THP-1 human monocytic cell line (ATCC), dual reporter THP-1, and a TLR2 knockout version of the dual cells (InvivoGen; referred to herein as DRTHP1 and KO-DRTHP1) were cultured according to the manufacturers’ protocols, and differentiated into the macrophage phenotype using PMA (MilliporeSigma). The dual reporter cells report NF-kB activation via SEAP and IRF via QUANTI-Blue. Peripheral blood mononuclear cells (PBMCs; ImmuoSpot) were thawed, resuspended, and used immediately in the experiments.

Analytical Techniques
The proteome profiler cytokine array was from R&D and the CXCL10/IP-10 ELISA was from ThermoFisher. RT-qPCR was used to quantify mRNA expression and flow cytometric analysis was used to probe expression of TLR2 on the cell surface of THP-1, DRTHP1, and KO-DRTHP1 macrophages. Western blotting was used to detect NF-κB and activated NF-κB.

Results, Conclusions, and Discussions: Results
We found that: (i) AS and the AS2 subunit induce CXCL10, a cytokine implicated in severe COVID-19, gene expression by human macrophage cells (THP-1); (ii) a highly specific GSK-3 inhibitor (COB-187) attenuates this induction; (iii) AS and AS2 do not activate NF-κB but do activate IRF; (iv) AS and AS2 do not require TLR2 to elicit CXCL10 production or activate IRF; and (v) AS and AS2 elicit CXCL10 production by PBMCs. We also discovered that the cellular response, or lack thereof, to S and S2, is a function of the recombinant/source of S and S2 used. While such a finding raises the possibility of confounding LPS contamination, we found evidence that potentially contaminating LPS does not appear to underlie induction of CXCL10. Most notably, the spike protein induction experiments were conducted in the presence of PB (PB neutralizes the activity of LPS [8] and has been used by others to “rule out” or “rule in” a role for possible LPS contamination in cytokine induction [8-10]) and induction of CXCL10 was clearly observed.

Discussion
Our data support an emerging hypothesis that the spike protein of SARS-CoV-2 can induce cytokine expression. While contaminating LPS, or any other contaminant(s), cannot be entirely ruled out, the evidence in totality indicates that (i) a simple explanation that contaminating LPS underlies all observed spike-protein induced THP-1 and PBMC CXCL10 production is not consistent with all of the data and (ii) the S and S2 spike protein themselves can cause cytokine induction. In probing S and S2 induction of cytokines in general, it is important to consider that the recombinant proteins discussed herein and used by others have small differences in their sequences which could influence their ability to induce cytokine production.

Conclusion
In summary, AS and AS2 proteins robustly induce the expression of CXCL10 and increase CXCL10 protein levels both of which are attenuated by a specific GSK-3 inhibitor termed COB-187. AS and AS2 activate IRF but not NF-κB. AS and AS2 do not require TLR2 to increase CXCL10 levels or activate IRF.

Acknowledgements (Optional): This work was funded by the National Institutes of Health (R15GM110602), the National Science Foundation (CDS&E Grant 1953311) and a Fast Grant from Emergent Ventures at the Mercatus Center, George Mason University.