Chronic stimulation with SARS-CoV-2 spike protein does not trigger autoimmunity

Abstract

Autoimmune manifestations were reported in people infected with SARS-CoV-2. Repetitive exposure of mice to foreign antigen may lead to the onset of autoimmunity. We therefore investigated whether repetitive exposure to the SARS-CoV-2 spike protein could result in autoimmunity. To address this hypothesis, we repeatedly immunized C57Bl/6 mice with spike protein injected intraperitoneally. At the end of the immunization, mice which received spike protein produced anti-spike IgG but none of them developed anti-dsDNA antibodies or proteinuria. In conclusion, repetitive immunization with SARS-CoV-2 spike protein does not induce autoimmunity in the present mice model. Albeit reassuring, these results need to be confirmed by large epidemiological study evaluating the incidence of autoimmune diseases in individuals with repetitive SARS-CoV-2 antigen exposure.

Systemic autoimmunity is a complex and multifactorial process characterized by the loss of tolerance toward self-antigens and the development of a cellular and humoral immune response against multiple tissues/organs. Several theories have been formulated to address the development of autoimmunity in humans, including excessive self-immunosurveillance [1], danger [2], and antigen discontinuity [3]. The antigen discontinuity (and self-criticality) theory postulates that repetitive antigen exposure leads to a continuous immune response. In line with this concept, Tsumiyama and Shiozawa have demonstrated that repetitive immunization with an exogenous antigen such as ovalbumin leads to the appearance of auto-antibodies and a lupus-like disease with glomerulonephritis in BALB/c and C57Bl/6 mice [4,5].

These data led us to consider the potential role of the SARS-CoV-2 pandemic in the onset of human autoimmune diseases. In fact, since the beginning of the pandemic in 2020, a significant proportion of the population became infected, sometimes repeatedly, and received multiple immunizations either through mRNA or vector-based vaccines. Furthermore, autoimmune conditions have been reported after COVID-19 [6], and it has been proposed that they are responsible for the post-acute sequelae of COVID-19 [7]. We investigated whether multiple exposure to the SARS-CoV-2 spike protein could trigger autoimmunity.

To test this hypothesis, we conducted repetitive immunization of C57Bl/6 mice (n = 5/group) with recombinant SARS-CoV-2 spike protein (1 μg per i.p. injection; R&D systems) or vehicle (phosphate buffered saline [PBS]); experimental settings shown in Fig. 1A). The first i.p. injection was mixed with Alum (Thermofisher), and the subsequent injections were diluted in PBS, administered every 5 days for a total of 16 i.p. injections. At the end of the immunization protocol, the mice were sacrificed, and serum was collected for further analysis. Anti-spike and Anti-dsDNA IgG were evaluated using a homemade ELISA kit as previously described [8], and the serum from 34-week-old lupus-prone C57Bl/6.lpr mice was used as a positive control for anti-dsDNA IgG. In brief, spike protein (0.2 μg/mL) or calf thymus DNA (0.1 mg/mL, Sigma; after precoating with poly-L-lysin 0.05 mg/mL, Sigma) were coated on a 96-well plate overnight. The plates were blocked and subsequently incubated with diluted serum samples for 2 h at room temperature. After washing, an alkaline phosphatase-conjugated anti-mouse IgG was added (1:5000; Jackson Immunoresearch) to measure plate-bound IgG. For anti-dsDNA IgG levels, the serum of an autoimmune mice was used as a standard and expressed in units per volume.

Fig. 1.

Repetitive immunization with SARS-CoV-2 spike protein does not induce autoimmunity in mice.

(A) Overview of the experimental setting. 8-week-old C57Bl/6 mice (n = 5/group) received every 5 days 1 μg spike protein or vehicle (PBS) intraperitoneally at a total of 16 injections. At the end of the immunization protocol, the mice were euthanized and the serum was collected. 34-week-old C57Bl/6.lpr (B6.lpr) mice were used as a positive control.

(B) ELISA results for anti-spike IgG at the end of the immunization phase. The results are expressed in relative optic density (O·D).

(C) ELISA results for anti-dsDNA IgG at the end of the immunization protocol. Serum from lupus-prone mice was used as a reference and the results are expressed as units per ml.

(D) Urine-strip results for proteinuria at the end of the immunization protocol.

Each point represents the results from an individual mouse, bars indicate mean ± s.e.m. **, p < 0.01; ****, p < 0.0001 using One-way ANOVA with Holm-Sidak's correction.

At the end of the immunization phase, the clinical phenotype of control and spike-immunized mice was unremarkable. All the mice immunized with the spike protein produced measurable amounts of anti-spike IgG determined by ELISA, while the control and the B6.lpr mice did not (Fig. 1B). In contrast to autoimmune B6.lpr mice, none of the mice in the immunization group developed anti-dsDNA IgG (Fig. 1C), or proteinuria (Fig. 1D).

Overall, we found that repetitive exposure to SARS-CoV-2 spike protein does not induce autoimmunity in C57Bl/6 mice. Several explanations may account for the different results observed compared to the ones from Tsumiyama. First, the genetic background of the mice may explain discrepancies. Indeed, Tsumiyama used the BALB/c mice which is more susceptible to pristane-induced glomerulonephritis compared to C57Bl/6 mice that we used [9]. However, the same team previously reported that C57Bl/6 mice repeatedly immunized with ovalbumine developed anti-dsDNA antibody mice, suggesting that this phenomenon is not restricted to BALB/c [5]. The present results do not rule out that repetitive immunization in genetically predisposed mice (or individuals), may foster autoimmunity or that repetitive immunization may induce flares in mice (or individuals) with autoimmune diseases. Follow-up studies which evaluate repeated immunization in mice with preexisting autoimmunity (ie, MRL/lpr or NZBxNZW (F1)) would be greatly informative. Second, differences in the dose and the frequency of administration of the antigen or of the adjuvant, may direct immune cells toward tolerance rather than activation. To circumvent this risk, we used the same frequency of administration as used by Tsumiyama [4], but we did not evaluate different dose regimen of the spike protein. Follow-up studies with higher number of mice and different dose regimen should be conducted. We cannot rule out that difference between immunization groups may appear later on, but the Tsumiyama and coauthors observed autoantibodies as soon as 6–8 immunizations [4]. Furthermore, we only screened for anti-dsDNA IgG and not for anti-histone, anti-Sm antibodies or rheumatoid factors. Finally, differences in the antigen itself, its processing and presentation to the immune system may impact the development of autoantibodies. Autoimmunity may be induced only when the major histocompatibility complex (MHC) molecules can present the antigen to the immune cells [4]. Since all the mice which were immunized with spike protein developed anti-spike IgG (Fig. 1C) and that the C57Bl/6 strain has been used in other SARS-CoV-2 related research [10], it is unlikely that the absence of autoantibody is due to the lack of immune stimulation by the spike protein.

In conclusion, repetitive exposure to SARS-CoV-2 spike antigen does not induce autoimmunity in mice. Albeit reassuring, these results need to be confirmed in follow-up studies using different mouse genotypes and immunization regimen, and in human by large epidemiological study evaluating the incidence of autoimmune diseases in individuals with repetitive SARS-CoV-2 antigen exposure.

Disclosures

MS received consulting fees for Sandoz, Nordic Pharma and Amgen.

JEG received research subsidies from BMS (unrelated to this work); consulting fees from AbbVie, BMS, CSL Behring, Galapagos, Gilead, Lilly, Roche-Chugai, Pfizer, Sanofi, and UCB.

JS received research subsidies (unrelated to this work) from Pfizer, BMS, Roche, MSD. JS received speaker honoraria from Roche, Chugai, Bristol-Myers Squibb, UCB, GSK, LFB, Actelion, Pfizer, MSD, Novartis, Amgen, Abbvie, Sandoz, Gilead, Lilly, Sanofi Genzyme, Janssen, Mylan, Galapagos, Sobi.

Declaration of Competing Interest

GCT reports no competing interest.

Data availability

Data will be made available on request.