SARS-CoV-2 awakens ancient retroviral genes and the expression of proinflammatory HERV-W envelope protein in COVID-19 patients

Figures

• 

  Figure 1. Expression of HERV-W ENV or HERV-K ENV RNA and HERV-W ENV protein in PBMC from healthy blood donors (HBDs) exposed to SARS-CoV-2 in vitro (A) The level of HERV-W ENV and HERV-K ENV mRNA in PBMC cultures from HBD, exposed to SARS-CoV-2 (MOI:0.1) or mock-treated (culture medium), was analyzed by RT-qPCR. The graph presents the mean results from triplicate (7 out 11 donors) or single cultures (4 out 11 donors) at 2h post-exposure. For each healthy blood donor, a color was assigned in order to be able to compare the induction of HERV-W ENV and HERV-K ENV according to the cultures of PBMC from the same individual. The details of the RT-qPCR results are available on the following dataset: https://doi.org/10.17632/rc74sdgksk.1. (B and C) PBMC cultures of 8 HBD, inoculated or not with SARS-CoV-2 at 0.1 MOI, were collected either 3 or 7 days post-exposure and stained for HERV-W ENV (green) and anti-SARS-CoV-2 nucleocapsid (red). (B) PBMCs from HBD #12 were inoculated or not with SARS-CoV-2 (MOI: 0.1). Bi: HERV-W ENV labeling, Bii: HERV-K ENV labeling (green staining). (C) PBMC cultures from 3 representative donors with variable number of HERV-W ENV-positive cells (green staining): donor…

• 

  Figure 2. Induction of HERV-W ENV and HERV-K ENV mRNA after exposure to SARS-CoV-2 spike trimer, followed by HERV-W ENV protein expression in PBMC from healthy blood donors (A) PBMCs from 4 HBDs were exposed to 0.5 μg/mL of active trimer spike recombinant protein in parallel to mock control (buffer), and HERV-W ENV (Ai) and HERV-K ENV (Aii) mRNA levels were assessed at 2, 15, and 24h post-inoculation (pi) by RT-qPCR. (B) PBMCs isolated from HBD were incubated or not with recombinant spike trimer. IL-6 secretion was monitored at 24h by ELISA in cultures inoculated with increasing concentrations of non-stabilized SARS-COV-2 spike protein. (C) Kinetics of IL-6 release at 2, 15, and 24h post treatment (hpt) with 500 ng of spike protein or without was performed with 3 donors (#30–32). (D) To exclude an effect of endotoxin possibly present in the recombinant protein buffer, the PBMCs of 3 HBDs were exposed either to 10 ng/mL of LPS or to 2 μg/mL of recombinant spike, in combination or not with 12.5 μg/mL of polymyxin B (PMB). IL-6 was then quantified in the culture supernatants by ELISA. The volume of buffer equivalent to 2 μg/mL of spike protein was used as a negative control. (E) PBMCs from…

• 

  Figure 3. COVID-19 patients produce HERV-W ENV in the plasma and on the membrane of their PBMC (A) The percentage of HERV-W ENV-positive cells was analyzed by cytofluorometry in PBMC from available fresh PBMC of 33 healthy blood donors (HBDs) (blue dots) and 57 COVID-19 patients (red dots) hospitalized in Lyon. The percentage of HERV-W ENV-positive cells has been evaluated in CD3high T cells, CD3low T cells, CD14+ monocytes, and CD19+ B cells. The dotted lines correspond to the positivity threshold calculated from the mean of HBD values +3SD. (B) Distribution of patients of the same cohort infected with the different SARS-CoV-2 variants. The dotted lines correspond to the positivity threshold calculated from the mean of HBD values +3SD. (C) Electrophoregrams (Ci) and digital western blot representations -WBs- (Cii) from healthy blood donors (HBDs, blue panel) and COVID-19 patients (red panel). Dotted line in Ci represents the upper limit of non-specific chemiluminescence due to the post-migration deposit of high molecular weight protein trapping detection reagents. This background signal is also seen on “HBD” run of the digital WB representation. (Ciii) Quantification of the soluble HERV-W…

• 

  Figure 4. Prevalence of soluble HERV-W ENV in serum from COVID-19 patients correlates with blood biomarkers and disease severity (A) Scatterplots comparing COVID-19 patients either positive for HERV-W ENV in plasma (11 patients) or 10 negative patients with their corresponding percentage of polymorphonuclear neutrophils (Ai), lymphocytes (Aii), and polymorphonuclear basophils (Aiii) in the blood. Yellow areas represent the reference range for blood tests in healthy adults (QC-validated normal values provided with results). (B) Comparison of neutrophil/lymphocyte ratio between patients positive for HERV-W ENV from two different COVID-19 cohorts, early and late consisting in (i) a cohort of early post-PCR COVID-19 diagnosed outpatients from Zaragoza and in (ii) in patients (hospitalized) from “Lyon” cohort (cf. clinical data of Zaragoza and Lyon cohorts; https://doi.org/10.17632/3v4hfxv4w8.1) with symptomatic COVID-19 evolution. The normal interval for the neutrophil/lymphocyte ratio is represented by the colored area above the X axis. HBD: healthy blood donor. (C) Detection of soluble HERV-W ENV in COVID-19 patients’ serum at the day of RT-qPCR diagnosis (Zaragoza cohort,…

• 

  Figure 5. Immunohistology detection of HERV-W ENV and SARS-CoV-2 antigens in postmortem lung tissue from acute COVID-19 patients (A) N-SARS-CoV-2 (Ai, Aiv) and HERV-W ENV (Aii, Av) (brown-red staining) were immunodetected in lung tissue sections from necropsies of COVID-19 postmortem patients (Ai-Aiii) but not in non-COVID-19 (Aiv and Av), with representative examples of lung tissue from non-COVID-19 patients, comprising normal tissue surrounding lung tumors (presented: non-COVID-19 lung necropsy donor D1). To further confirm the phenotype of HERV-W ENV-producing macrophage-like cells seen in lung alveoli, CD68 immunostaining targeting a specific marker of macrophages was performed on sections from COVID-19 lungs (Aiii). Higher magnifications of pulmonary cells (Ai-Av, red squares) and of alveolar macrophages (Ai-Av, black squares) are presented on COVID-19 and non-COVID-19 tissues for comparison. (B) The specificity of HERV-W ENV immunodetection was also assessed on COVID-19 tissues by comparing anti-HERV-W ENV murine IgG1 GN_mAb_Env01 (BiBiii) and murine IgG1 isotype control (Biv-Bvi) at the same concentration (10 μg/mL) on adjacent tissue sections. (C) In addition to macrophage cells,…

• 

  Figure 6. Immunohistology detection of SARS-CoV-2 and HERV-W ENV in postmortem brain parenchyma of the olfactory bulb and in nasal mucosa tissues from acute COVID-19 patients (A) Schematic representation of nasal cavity and location of tissue sampling. b.: brain; o.b.: olfactory bulb; n.p.: naso-pharynx; p.: palate; n.: nostrils; o.m.: olfactory mucosa; o.f.: olfactory fibers; c.p.: cribriform plate (ethmoid bone); cf.: cribriform foramina. (B) SARS-CoV-2 virus was searched for in brain parenchyma (Bi and Bii) and nasal mucosa (Bi) using anti-N SARS-CoV-2 immunodetection (brown-red staining). The black dotted line on Bi represents the boundary between brain parenchyma and nasal mucosa (position of the cribriform plate). SARS-CoV-2 nucleocapsid immunostaining was only observed in nasal mucosa and not in neighboring brain parenchyma (Bi). Bars = 100 μm. (C) HERV-W ENV immunodetection using the murine IgG1 antibody GN_mAb_Env01 (brown-red staining) revealed that HERV-W ENV IHC staining was detected in nasal epithelium (Ci), nasal mucosa (Cii), and brain parenchyma (Ciii). Bars = 100 μm. (D) Murine IgG1 isotype control (Di-Diii) was used at the same concentration (10 μg/mL) than anti-HERV-W…

• 

  Figure 7. Summary of immunohistology detection of HERV-W ENV and N-SARS-CoV-2 proteins in all studied tissues from COVID-19 patients (A and B) Detection of HERV-W ENV and SARS-CoV-2 N is presented for each donor, for all analyzed organs. (A) lungs (B) brain and nasal mucosa/epithelium. (C and D) (C) heart and (D) digestive tract. Donors (D) 1–3 of lung tissue are obtained from non-covid patients from CRB Lyon, while donors D 4–24 are from postmortem COVID-19 patients from Mexico City hospital. Representative immunohistology results are presented in Figures 5, 6, and S4 (n.d. not detected; n.a. not available). The analysis of potential tissue cross-reactivity in healthy human organs, obtained in similar postmortem conditions and including the presently studied tissues, showed no significant staining and, thereby, no expression of HERV-W ENV protein (described in STAR Methods). Isotype controls have been used on all tissues, which constantly resulted with negative staining, and the negative results in staining obtained with anti-HERV-K monoclonal also provided another “isotype control” in the corresponding Figures.

•