SARS-CoV-2 induces transcription of human endogenous retrovirus RNA followed by type W envelope protein expression in human lymphoid cells.


 Patients with COVID-19 may develop abnormal inflammatory response and lymphopenia, followed in some cases by delayed-onset syndromes, often long-lasting after resolution of the initial SARS-CoV-2 infection. As viral infections may activate human endogenous retroviral elements (HERV), we studied the effect of SARS-CoV-2 on HERV-W and HERV-K envelope (ENV) expression, known to be involved in immunological and neurological pathogenesis of human diseases. We demonstrate here that an initial exposure to SARS-CoV-2 virus activates early HERV-W and K transcription in peripheral blood mononuclear cell (PBMC) cultures from healthy donors. Within a week of primary PBMC culture, only HERV-W ENV protein expression was detected in lymphoid cells of some donors, although SARS-CoV-2 infection of PBMC was not observed. HERV activation was reproduced with UV-inactivated virus and with a recombinant spike protein. Interestingly, exposure to SARS-CoV-2 protein induced a significant production of interleukin 6 in PBMC, independently from detectable HERV expression. Altogether, these results show that SARS-CoV-2 viral protein could induce HERV-W ENV expression in lymphocytes from some individuals, underlying the importance to further address the implicated molecular pathways, to understand patients‘ genetic susceptibility associated to the activation of HERV-W and its possible relevance for targeting therapeutic intervention in COVID-19 associated syndromes.


Introduction
The present COVID-19 pandemic has raised many questions about the underlying biological mechanisms of the many symptoms or syndromes associated with SARS-CoV-2 infection [1][2][3][4][5][6] . Most of them are debated, while others have been explained by, e.g., immune-mediated in ammation and/or activation of blood clotting pathways 7,8 . However, such pathways may themselves be triggered by still unidenti ed factors linking coronavirus infection to a dysregulation of physiological signaling pathways.
Such an indirect pathological activation of receptors and their signaling pathways are known to take place when infectious agents have the potential to lift the epigenetic control and/or to directly activate endogenous retroviral elements present in the human genome 9,10 . The resulting production of an endogenous protein of retroviral origin with pathogenic effects may generate clinical symptoms corresponding to the organ, tissue or cells, in which it is expressed and determined by the tropism of the triggering infectious agent [11][12][13][14][15][16][17][18][19] . Abnormal expression of human endogenous retroviruses (HERV) may become self-sustained, thus creating lifelong chronic expression from host's genome copies in affected tissues 20 , e.g., with cytokine-mediated feedback loops 21 or possibly mediated by their envelope proteins 22 . This has been shown to be involved in brain lesions and in their expansion in patients with multiple sclerosis 20,[23][24][25] .
Abnormally expressed HERV envelope proteins revealed to display major immunopathogenic [26][27][28][29][30][31][32] and/or neuropathogenic 23,28-35 effects in vitro and in vivo, associated with pathognomonic features of human diseases. We therefore studied whether SARS-CoV-2 could activate HERV, considered as 'dormant enemies within' 36 , to evaluate their potential pathogenic contribution in COVID-19 associated syndromes. This question became critical after a recent study revealed the signi cant expression of HERV-W envelope protein (ENV) in lymphoid cells from COVID-19 patients, correlating with disease outcome and markers of lymphocyte exhaustion or senescence 37 .
The present study shows that (i) HERV RNA and known pathogenic HERV-W envelope protein expression can be triggered in naive human peripheral blood mononuclear cells (PBMC) by SARS-CoV-2 virus itself or by its trimeric bioactive spike protein, with analogies to previously known examples of interactions between exogenous viruses and HERV 12 , (ii) only about one-third of healthy blood donors have PBMC responding with HERV activation under SARS-CoV-2 exposure, however, (iii) all donors responded with similar interleukin 6 (IL-6) secretion kinetics after incubation with SARS-CoV-2 protein but without HERV activation in non-responders.

Materials And Methods
Cells PBMC from healthy donors were obtained from the "Etablissement Français du Sang" (EFS) of Lyon (France). PBMC were isolated by Ficoll separation (Ficoll-Plaque PLUS) (GE Healthcare, 17-1440-02) from blood samples and cultured in RPMI-1640 medium (Gibco, 61870-010) completed with 5% of decomplemented Human AB serum (Sigma, H4522). Healthy donors signed a written Informed Consent Form, documented at the EFS, allowing the commercial use of their blood and blood components for medical research after de nite anonymization.
U87 cells expressing both the human Angiotensin-converting enzyme 2 (ACE2)-receptor and the serine protease TMPRSS2 were generated using a second-generation lentiviral system. A lentivirus producer cell line (HEK293T cells: ATCC® CRL-3216™) was transiently transfected with a transfer plasmid encoding the transgene (ACE-2 or TMPRSS2), simultaneously with two other plasmids encoding either the vesicular stomatitis virus (VSV)-G envelope glycoprotein and a lentiviral packaging plasmid to generate the lentiviral particles 38 . At 48 hours post transduction the cell supernatant was replaced again by fresh medium with Hygromycin at a concentration of 50µg/ml for antibiotic selection. The ACE-2 transduced U87 cells were observed daily and the supernatant replaced every 2-3 days by fresh Hygromycincontaining medium until the generation of the stable U87 ACE-2-expressing cell line. This new cell line was submitted to the same procedure using the TMPRSS-2 encoding lentivirus and then selected with Neomycin G418 at 250µg/ml to generate U87 cell line stably expressing both ACE-2 and TMPRSS2 proteins. The monitoring of ACE2/TMPRSS2 expression by RT-qPCR revealed that both genes were e ciently expressed. SARS-CoV-2 serology of blood donors SARS-CoV-2 serology of blood donors was determined on plasma diluted 10 times using Simple Western technology, an automated capillary-based size sorting and immunolabeling system (ProteinSimple TM ).
The SARS-CoV-2 Multi-Antigen Serology Module (SA-001) was used with Wes device and all procedures were performed according to manufacturer's protocol. Wes device was associated with Compass software for device settings and raw data recording (ProteinSimple/Biotechne).

Quantitative RT PCR (RT-qPCR)
At several time points after infection or recombinant Spike exposure, cells were harvested and total RNA extracted. 200 ng of DNase-treated RNA were reverse-transcribed into cDNA using iScript cDNA Synthesis Kit (Bio-Rad, 1708891) according to the manufacturer's protocol. A control with no-RT was prepared in parallel, to con rm the absence of contaminating DNA in PCR experiments. An amount of 5 ng of initial RNA in RT reaction has been used to quantitatively evaluate the transcriptional levels of HERV-W ENV, HERV-K ENV, N SARS-CoV-2 (Ferren et al. preprint 2021. DOI: 10.21203/rs.3.rs-122126/v1) and ACE2 genes by RT-qPCR (primer sequences are shown in Table 1). The assays were performed in a StepOnePlus instrument (Applied Biosystems) using Platinum SYBR Green (Invitrogen, 11744-500). The housekeeping gene beta-2 microglobulin (B2M) was used to normalize the results in PBMC experiments whereas glyceraldehyde-3-phosphate deshydrogenase (GAPDH) was used in Vero cells experiments. Each experiment was completed with a melting curve analysis to con rm the speci city of ampli cation and the lack of any non-speci c product and primer dimer. Quanti cation was performed using the threshold cycle (Ct) comparative method: the relative expression was calculated as follow:  Quanti cation of IL-6 secretion IL-6 secretion was assessed in PBMC culture supernatant 2, 15 and 24 h after recombinant Spike exposure, by ELISA using BD Opt EIA Set Human IL-6 (BD, 555 220) according to supplier's recommendations.
Cell viability assay PBMC viability was analyzed using CellTiter-Glo 2.0 Assay kit (Promega, G9241) according to the manufacturer's protocol. Viability percentage was calculated using the following equation: Background: wells containing medium without cells; Control: untreated cells at 24 h post treatment.

Exposure to SARS-CoV-2 virus triggers HERV-W and -K ENV mRNA early transcription from PBMC of heathy donors
We initially analyzed whether infectious SARS-CoV-2 could modulate the expression of HERV W and HERV-K genes in lymphocytes from healthy individuals. RT-qPCR was performed using speci c primers for HERV-W and HERV-K envelope genes, as already validated in patients with HERV-associated diseases 33,39,40 , using B2M mRNA as a suitable reporter gene for PBMC 41 . PBMC of 11 blood donors were cultured or not with infectious SARS-CoV-2 (MOI = 0.1) and RNA was collected at 2H postinoculation ( Figure 1). In PBMC from 3 (Donors # 28, 29 and 40) out of 11 (27%) donors, HERV-W RNA transcription was increased at 2H after exposure to wild type SARS-CoV-2 virus (MOI=0.1); a low response for HERV-W was seen with one donor (#27), as shown in Figure 1 and supplementary Table S1. The same donors also showed relative transcriptional activation of HERV-K ENV ( Figure 1). For further kinetics analysis, RNA from PBMC of 4 representative donors with various RNA results at 2H, was collected at 2H, 19H and 24H and tested in parallel (Supplementary Figure S1) Interestingly, the transcriptional level of both HERV W and HERV-K RNA was decreased at 19H and 24H to a lower level than the baseline of cultures "not inoculated" (NI) with the infectious virus from the same donors at the same time points. A similar decrease of HERV-W and -K ENV transcription, when compared to NI control wells, was observed from the initial time point (2H) in PBMC from of the "non-responding" donors after exposure to SARS-CoV-2 virus (supplementary Figure S1 A, B).
RT-qPCR analysis of the kinetics of SARS-CoV-2 N-RNA only showed abundant RNA load from the inoculum with a signi cant decrease at 19h post-infection (inoculation), which con rmed the absence of viral replication in PBMC (supplementary Figure S1 C).

Exposure to SARS-CoV-2 triggers HERV-W envelope (ENV) production from PBMC of heathy donors
We next analyzed whether the HERV RNA expression is followed by HERV protein production. PBMC cultures of 8 new blood donors, inoculated or not with SARS-CoV-2 at 0.1 MOI, were collected at either 3 or 7 days post-exposure. Cells were stained with speci c monoclonal antibodies raised against envelope proteins of HERV-W, HERV-K and against SARS-CoV-2 N protein, then analyzed by immuno uorescence (Figure 2). Cells were maintained for up to 7 days in culture.
All tested conditions are presented in parallel with an example from donor # 12: HERV-K ENV staining was not detected in the cells after exposure to SARS-CoV-2 ( Figure 2D and Supplementary Figure Figure S6).

Exposure to SARS-CoV-2 triggers HERV-W envelope production in T-cells of heathy individuals
HERV-W ENV protein expression having been observed in CD3 + T-lymphocytes of COVID-19 patients 37 , we next analyzed whether SARS-CoV-2 could induce HERV-W ENV in T lymphocytes. We analyzed HERV-W ENV expression in PBMC cultures from three healthy donors, with or without exposure to SARS-CoV-2, using cyto uorometry analysis. As illustrated in Figure 3, CD3 + T lymphocytes were identi ed with the gating strategy in non infected (NI) cultures, showing an increased detection when inoculated with SARS-CoV-2 at MOI=0.1. However, at 24H post-viral exposure, a CD3-low population originally shown to be characteristic of superantigen activity 42 and tentatively associated to the spike protein of SARS-CoV-2 43 was observed ( Figure 3B), whereas only few such cells were seen in NI cultures ( Figure 3A). When cells were doubled-labeled with anti-HERV-W ENV speci c antibody an important proportion of these CD3-low T-cells was positive for HERV-W ENV ( Figure 3B CD3+ top panel), compared to NI cultures ( Figure 3A CD3+ top panel). Fewer CD3-high were found to express HERV-W ENV after exposure to the virus at this time-point ( Figure 3B, CD3 + bottom panel). The percentage of HERV-W ENV positive cells in each condition is represented in Figure 3E. A signi cant increase in CD3 + low/HERV-W+ lymphocytes exposed to SARS-CoV-2, compared to those from NI cultures was con rmed in at 72H post-inoculation ( Figure 3F) and the same was observed versus all control cells from cultures without virus ( 5. Exposure to SARS-CoV-2 virus recombinant trimeric spike protein triggers HERV-W ENV protein production in PBMC of heathy individuals. We previously observed a rapid response to SARS-CoV-2 virus with an early peak of RNA followed by HERV-W ENV protein expression in cells maintained in culture (≤5 days) (Figures 1-3). This immediate RNA response in the absence of detectable infection of PBMC by SARS-CoV-2 prompted us to investigate a possible direct stimulation by SARS-CoV-2 proteins and, most of all, by its surface spike protein. The same question was also raised following the activation of HERV-W ENV expression in U87-ACE2 + /TMPRSS2 + cells by the UV-inactivated coronavirus, in the absence of any replication and without S-protein detection by double labeling of cells, using cyto uorometry.
To further address the activation of HERV-W in the absence of infection, a recombinant trimeric spike protein was added into the culture medium of PBMC from 4 new healthy donors, in parallel to mockcontrol, corresponding to the buffer only. Cell viability was measured and did not vary signi cantly within the culture period of experiments (≤5days; Cf. Supplementary Material Figure S7).
PBMC RNA from 4 new donors were collected at 2H, 15H and 24H post-inoculation and analyzed by RT-qPCR for HERV-W and HERV-K envelope gene expression. Donor #31 showed transcriptional increase for both HERV-W and HERV-K ENV at 2H post-inoculation, donor #30 showed a peak of HERV-W and HERV-K ENV RNA at 15h, while both RNA levels remained rather stable or decreased in donors #32 and #33 ( Figure 6 A-B). RNA levels for both HERV-W and HERV-K also decreased below the baseline of identical non-exposed cells after having shown a peak of transcription. This was again seen in the absence of signi cant cell death that would anyhow have been compensated by the differential quanti cation with B2M RNA. Of interest, donor #30 was the only one tested positive for anti-SARS CoV-2 antibodies (supplementary Figure S6), which did not prevent HERV transcriptional activation by this recombinant spike trimer but coincides with slightly delayed peak of HERV RNA.
IL-6 secretion was signi cantly increased at 15H and 24H PE (p<0.01 and p<0.05) in PBMC from all donors, including non-responding donors (without HERV activation) after exposure to SARS-CoV-2 S antigen (Figure 6 C).. HERV activation independently from IL-6 production was con rmed by immuno uorescence analysis at 72H, which showed HERV-W ENV positive lymphoid cells in cultured PBMC of 2 out the four donors inoculated with spike trimers (0.5 µg/mL) but not in mock-control cultures (Figure 6 D).
Of note, as seen with the infectious virus, few HERV-W ENV positive cells were detected but an increase in HERV RNA was seen despite this low proportion of activated cells within the cultures. Also, even in the presence of IL-6 within the culture media of all cultures (Figure 6 C), HERV-W ENV was expressed in few cells only within PBMC cultures from responders, but not detected in PBMC from non-responders at both RNA and protein levels.

Discussion
Unlike physiological effectors, the transcription of which is normally upregulated to produce active molecules when needed, HERV activation by environmental factors does not re ect a physiological response and seldom results in protein expression 44 . Exceptions are known with a few "domesticated" HERV genes, known to encode physiological proteins 45 or at the transcriptional level only, with the high number of non-coding RNA types involved in major post-genomic regulations 46 . Thus, though representing about 8% of the human genome (whereas physiological genes represent much less), HERV are mostly defective and non-coding remnants of ancestral integrations of retroviral genomes in the germ line of individuals infected by exogenous retroviruses. Non-physiological HERV envelope protein expression has been shown to be detrimental in certain human diseases and cannot be considered as just another response to environmental triggers like, e.g., an immune response with cytokines release, though a parallel modulation mediated by certain HERV RNA is conceivable. The expression of HERV-K or HERV-W envelope proteins may have trophic roles in the placenta 45 or in stem cells 47 , but envelope proteins from the same HERV families were rather shown to be involved in disease pathogenesis when expressed in adult cells 34,39 .
This study shows that SARS-CoV-2 is able to trigger both HERV-W and HERV-K RNA transcription, while only HERV-W ENV protein was detected during these short-term primary cultures of non-stimulated PBMC from about 30% of healthy donors. However, a quite potent expression is made consistent from the observed ENV-protein immuno uorescent labeling in producing cells, which lasted for several days after exposure to SASR-CoV-2, and from the detectable HERV-W ENV RNA increase above the baseline, though expressed in few cells only. but.
Most importantly, after addition of SARS-CoV-2 spike protein trimer in PBMC cultures, the dosage of IL-6 showed that its secretion occurred in the supernatants from all donors, either responding or nonresponding with HERV activation. The fact that this IL-6 production was induced by SARS-CoV-2 in the absence of detectable HERV RNA or protein expression provides a supplementary evidence that the presently observed HERV activation is not induced by cytokines, or by in ammation due to viral infection, but rather by SARS-CoV-2 spike protein itself.
Cyto uorometry analysis from these PBMC cultures also con rmed that HERV-W ENV protein expression was effectively induced in T-lymphocytes. A noticeable sub-group of CD3-low T-lymphocytes appeared in which HERV-W ENV protein was predominantly expressed, whereas it remained minimal and HERV-W negative in control cultures not exposed to the virus. These observations may corroborate previous works describing superantigen motifs of SARS-CoV-2 spike protein 43 , and may involve cellular mechanisms associated to lymphopenia and hyperin ammation already characterized for other emerging viruses (e.g., Ebola or Lassa) 48,49 . However, because HERV-W ENV was also shown to display superantigen-like effect 28 , the origin of this effect on T-lymphocytes or potentially combined effects are questioned. Indeed, this work now calls for further analyses of possible markers of exhaustion and for studying the fate of these cells in kinetics analyses with, e.g., markers of apoptosis versus markers of early activation. The present in vitro observation is consistent with the co-stained HERV-W ENV positive CD3 lymphocytes shown to express either exhaustion or terminal differentiation markers in leukocytes from COVID-19 patients with severe clinical evolution 37 . Thus, despite an expected parallel stimulation of T-cell antiviral reactivity, the present experimental results consistently raise the question of a possible role of this HERV protein expression in the induction of anergy or in the depletion of lymphocytes. The frequent lymphopenia in COVID-19 patients corroborates a defect of adaptive immunity and an imbalance with innate immune reactivity, compared to usual viral infections 50 . This question may now be addressed under the new angle of HERV expression in lymphocytes, in parallel with a mechanism of induction in Tlymphocytes that remains to be characterized. Moreover, since HERV-W ENV (previously named MSRV-ENV), is already known to be a potent TLR4 agonist and itself induces in ammation via TLR4 signaling pathway 30,51 , its superimposed or synergistic role in the hyperactivation of innate immunity should also be considered in parallel with a potential contribution to the impairment of adaptive immunity as seen in COVID-19.
However, the presently observed HERV activation occurs without signs of infection of lymphocytes or monocytes by SARS-CoV-2. Furthermore, UV-inactivated virus and a recombinant trimer of its spike protein appeared su cient to reproduce similar HERV-W and -K ENV RNA stimulation as well as HERV-W ENV protein production in lymphoid cells. This type of HERV activation mediated by an interaction between a triggering virus and a speci c receptor on certain cells has already been described for HHV-6A and U87 cell-line 12  SARS-CoV-2 virus, unlike HHV-6A, did not activate HERV-W expression in U87 cells naturally devoid of ACE2/TMPRSS2 receptors, but did so after ACE2/TMPRSS2 receptors expression in transduced U87 cell membrane. Because SARS-CoV-2 induced HERV-W expression in human lymphoid cells that neither express ACE2 (con rmed with speci c RT-qPCR) nor TMPRSS2 receptor and were not infected, these cannot be expected to be the receptors through which SARS-CoV-2 would trigger a signaling pathway resulting in HERV activation. This activation was also obtained with a recombinant trimer tested for an effective binding to ACE2 receptors, but not stabilized in a speci c conformation. Pre-and post-fusion variable conformations of this spike protein may confer divergent properties 52 , whereas vaccine preparations are likely to stabilize its structure 53-55 with a good safety pro le 54,56 . Here again, further studies are needed to de ne sequence-related and protein conformational properties involved in a proteinreceptor interaction leading to HERV activation in susceptible individuals. This may further provide relevant information for viral constructs or preparation envisaged for SARS-CoV-2 vaccines.
At the transcriptional level, within the present short-term PBMC cultures after a single exposure to SARS-CoV-2 or to its active protein, a single peak of HERV RNA is seen when occurring in responding donors.
The absence of recurrent stimulation and/or of renewed PBMC comprising susceptible cell-type(s) may explain this single peak nonetheless followed by active HERV-W protein synthesis. Thus, an increase in the number and percentage of lymphoid cells with HERV-W activation strongly expressing the envelope protein under persisting SARS-CoV-2 virion release would be consistent with observations from COVID-19 patients 37 .
A feed-back inhibitory loop triggered by abnormal HERV activation is probably limiting the yield of HERV expression in targeted PBMC in the present cultures. When HERV activation was not seen in PBMC cultures from 7 out of 11 healthy blood donors exposed to the virus and from 1 out of 4 exposed to the spike trimers, this coincided, with constantly decreased RNA levels in PBMC exposed to the virus or to the recombinant spike trimer, compared to non-exposed PBMC. Thus, an inter-individual variability in the rapidity and potency of expression of HERV inhibitors, with a genetic or epigenetic origin, may offer an avenue of research to understand why some individuals do not activate this HERV expression in PBMC.
Symptoms potentially superimposed or synergized by the pathogenic effects of HERV envelope protein expression may consequently occur in certain individuals only. This should also be investigated in COVID-19 patients for it may provide possible prognostic information on individuals who may present an enhanced reaction to the infection with severe evolution or long-lasting post-infectious syndromes.
Finally, since the HERV-W ENV protein was shown to have multifaceted pathogenic effects with a common TLR4 activation 51 [57][58][59][60] . Altogether, these data call for evaluating HERV-W ENV as a potential target for therapeutic intervention in COVID-19 associated syndromes.
In conclusion, this study aimed at elucidating the possible origin of HERV-W ENV expression in lymphocytes of COVID-19 patients and revealed that SARS-CoV-2 via its spike protein may trigger HERV expression in human PBMC and, in particular, in T-lymphocytes, which could play a role in the immunopathogenesis of COVID-19.  Quanti cation of HERV-W ENV and HERV-K ENV RNA in PBMC from healthy blood donors exposed to SARS-CoV-2. The level of mRNA of HERV-W ENV and HERV-K ENV of PBMCs cultures from 11 healthy blood donors, exposed or not to infectious SARS-CoV-2 (MOI:0.1), was analyzed by RT-qPCR. The graph presents mean results from triplicate cultures of 11 donors at 2H post-exposure. Numbers of donors with increased expression of HERV-W and HERV-K are indicated and values of each sample is presented on the Supplementary table S1.  SARS-CoV-2 induces the expression of HERV-W ENV in CD3low T cells. PBMCs from 3 healthy donors were either incubated with SARS-CoV-2 (MOI=0.1) for 24H (B) or 72H (D), or remained non exposed to the virus (A, C). Cells were stained using anti-CD3 and GN_mAb_Env01 antibodies and analyzed by ow cytometry. A-D, Gating strategy for the separation of CD3 high and CD3 low expressing T cells is presented. E,F The percentage of HERV-W ENV positive cells from subpopulations within CD3 T lymphocytes at 24Hpi and 72Hpi (average from 3 donors +SD) is presented with histograms. Statistical analysis was performed using Tukey's multiple comparisons test.   Induction of HERV-W ENV and HERV-K ENV mRNA expression and IL-6 secretion in PBMCs cultures from healthy donorsafter exposure to the recombinant Spike trimer PBMCs from 4 healthy blood donors were exposed during 2, 15 and 24 h to 0.5µg/mL of active trimer Spike recombinant protein. (A) HERV-W ENV and (B) HERV-K ENV mRNA levels were assessed by RT-qPCR using speci c primers. Results were presented as fold change versus the corresponding non infected condition. Hatched histograms correspond to the non-exposed conditions and plain histograms to conditions with inoculation of spike protein in the cultures. pi: post-inoculation. (C) PBMCs isolated from 3 healthy blood donors (donors #30 to 32) were incubated, or not, with 0.5µg/mL of recombinant spike trimer. IL-6 secretion was monitored in culture at 2, 15 and 24 h post treatment (hpt) using the BD® Opt EIA Set Human IL-6® ELISA dosage.
Statistical analysis: Sidak's multiple comparison. Here, not treated and treated condition were compared at each time point 2hpt, 15 and 24hpt. (n.s.: p>0.5; *: 0.5<p<0.01; **: 0.01<p<0.005; ***: 0.005<p<0.0001; ****: p<0.0001). (D) PBMCs from healthy blood donors were exposed during 24 h with 0.5 µg/mL of active trimer Spike recombinant protein. Here, results obtained on 2 responding donors (donor # 18 and donor # 21) are presented. The recombinant Spike protein was not detected following the staining with anti-S-SARS-CoV-2 antibody (red staining). When expressed, HERV-W ENV was detected in few cells of cultures exposed to the spike trimer (green labeling). Higher magni cation of HERV-W ENV positive cells is presented in white squares. Non exposed cells do not present any HERV-W ENV labelling using GN_mAb_Env01 antibody. DAPI was used to stained nuclei (blue staining).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.