SARS-CoV-2 ORF8 accessory protein is a virulence factor

ABSTRACT Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) encodes six accessory proteins (3a, 6, 7a, 7b, 8, and 9b) for which limited information is available on their role in pathogenesis. We showed that the deletion of open reading frames (ORFs) 6, 7a, or 7b individually did not significantly impact viral pathogenicity in humanized K18-hACE2 transgenic mice. In contrast, the deletion of ORF8 partially attenuated SARS-CoV-2, resulting in reduced lung pathology and 40% less mortality, indicating that ORF8 is a critical determinant of SARS-CoV-2 pathogenesis. Attenuation of SARS-CoV-2-∆8 was not associated with a significant decrease in replication either in the lungs of mice or in organoid-derived human airway cells. An increase in the interferon signaling at early times post-infection (1 dpi) in the lungs of mice and a decrease in the pro-inflammatory and interferon response at late times post-infection, both in the lungs of mice (6 dpi) and in organoid-derived human airway cells [72 hours post-infection (hpi)], were observed. The early, but not prolonged, interferon response along with the lower inflammatory response could explain the partial attenuation of SARS-CoV-∆8. The presence of ORF8 in SARS-CoV-2 was associated with an increase in the number of macrophages in the lungs of mice. In addition, the supernatant of SARS-CoV-2-WT (wild-type)-infected organoid-derived cells enhanced the activation of macrophages as compared to SARS-CoV-2-∆8-infected cells. These results show that ORF8 is a virulence factor involved in inflammation that could be targeted in COVID-19 therapies. IMPORTANCE The relevance of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) ORF8 in the pathogenesis of COVID-19 is unclear. Virus natural isolates with deletions in ORF8 were associated with wild milder disease, suggesting that ORF8 might contribute to SARS-CoV-2 virulence. This manuscript shows that ORF8 is involved in inflammation and in the activation of macrophages in two experimental systems: humanized K18-hACE2 transgenic mice and organoid-derived human airway cells. These results identify ORF8 protein as a potential target for COVID-19 therapies.

The functions of SARS-CoV-2 accessory proteins are not fully understood.Most available studies have been performed in cancerous or transformed cell lines overex pressing individual viral proteins, outside the context of infection, thus providing results in non-physiological conditions.Among them, recent studies describe ORF6, ORF7a, ORF7b (13), ORF8 (14), and ORF9b (15) as antagonists of type I interferon (IFN-I).ORF6 was shown to inhibit STAT nuclear import and antagonize the IFN signaling pathway by its direct binding to nucleoporin Nup98 at the nuclear pore complex (16).ORF7a was described as a potent nuclear factor kappa B (NF-κB) activator of SARS-CoV-2 (17) and also an inhibitor of autophagy (18).ORF8 was shown to promote immune evasion by down-regulating the surface expression of class I major histocompatibility complex molecules (MHC-I) on several cell lines (19).Furthermore, ORF8 mimics interleukin 17A (IL-17A) and interacts with the host IL-17A receptor (IL-17RA) (20,21), inducing a pro-inflammatory response in SARS-CoV-2-infected hamsters (22).Altogether, interac tions of SARS-CoV-2 accessory proteins with mediators of the host immune response suggest a potential contribution to pathogenesis.However, to confirm their relevance in virus virulence and their impact on human disease, infections with deletion mutants of each accessory gene in physiologically relevant experimental systems, such as in vivo lethal animal models, or human airway organoids, are required.
In this manuscript, we engineered deletion mutants of accessory genes using our reverse genetics system for SARS-CoV-2 and evaluated their virulence in vivo in humanized K18-hACE2 transgenic mice.Only the deletion of ORF8 either individually (SARS-CoV-2-Δ8) or in combination with ORF6 (SARS-CoV-2-Δ [6,8]) or ORFs 6 and 7 (SARS-CoV-2-Δ [6,7,8]) partially attenuated the virus leading to 40% or 80% survival, respectively.SARS-CoV-2-Δ8 was further characterized in organoids derived from human airway epithelium.SARS-CoV-2-Δ8 attenuation was not associated with diminished replication either in the lungs of mice or in organoid-derived human airway cells.In the lungs of mice, SARS-CoV-2-∆8 induced a faster IFN response and a reduced proinflammatory response at late times post-infection, which could explain the attenuation.Altogether, this paper shows that SARS-CoV-2 ORF8 is a virulence factor that could be a potential target for antiviral treatment in COVID-19 patients.

Engineering deletion mutants of SARS-CoV-2 accessory genes
To investigate the role of accessory genes in SARS-CoV-2 pathogenesis, recombinant viruses with individual or combined deletions of genes 6, 7ab, and 8 were engineered using a reverse genetics system previously established (23) (L.Wang, unpublished data) (Fig. 1A).All mutants were rescued and characterized in Vero E6/TMPRSS2 cells.The absence of expression of proteins encoded by the deleted genes during Vero E6/ TMPRSS2 infection was confirmed by immunoblotting (Fig. 1B).The growth kinetics of mutants was evaluated at 0, 24, 48, and 72 hpi in Vero E6/TMPRSS2 cells infected at a multiplicity of infection (MOI) of 0.001 (Fig. 1C).All mutants reached titers between 10 6 and 10 7 PFU/mL.SARS-CoV-2-∆ [6,7,8] grew to the lowest titer (10 6 PFU/mL), suggesting that accessory genes together, although not critically essential for viral replication, were contributing to some extent to virus growth in cell cultures.

Virulence of deletion mutants of SARS-CoV-2 accessory genes
To evaluate the relevance of each accessory protein in pathogenesis in vivo, 26-weekold female K18-hACE2 mice were either mock-infected, infected with SARS-CoV-2-WT (wild type), or infected with each deletion mutant (SARS- and survival were monitored daily until day 10 pi.Mock-infected mice did not lose any weight, and all of them survived.In contrast, animals infected with the deletion mutants of SARS-CoV-2 accessory genes lost weight to different extents.SARS-CoV-2-Δ6 and SARS-CoV-2-Δ7ab caused initially a severe weight loss similar to that observed in SARS-CoV-WT-infected mice, although mice infected with SARS-CoV-2-Δ7ab started recovering at 7 dpi, while all SARS-CoV-2-Δ6 died.In contrast, SARS-CoV-2-Δ8 and SARS-CoV-2-Δ [6,8] caused less severe weight losses.Initially, the weight loss caused by these viruses was tracked with that induced by SARS-CoV-2-WT, but mice recovered weight faster from 7 dpi.Infection with SARS-CoV-2-Δ [6,7,8] induced the lowest weight loss, delayed in time, starting at 6 dpi (Fig. 2A).
Histopathological evaluation of lungs at 3 dpi (n = 3/group) showed that mice infected with SARS-CoV-2-WT displayed significantly higher lung inflammation scores than mice infected with the deletion mutants (Fig. 3A and B).At 6 dpi, the lung inflammatory pathology increased up to similar levels in mice infected with the virulent viruses SARS-CoV-2-WT, SARS-CoV-2-Δ6, and SARS-CoV-2-Δ7ab.In contrast, no significant increase was observed at 6 dpi in mice infected with partially attenuated viru ses SARS-CoV-Δ8, SARS-CoV-Δ (6,8), or SARS-CoV-Δ (6,7,8).Pulmonary histopathological lesions caused by the virulent viruses were severe and characterized by a diffuse thickening of the alveolar septae, massive presence of mononuclear cell infiltrates within alveolar spaces, and presence of large multifocal perivascular and peribronchiolar mononuclear infiltrates.These lesions were less severe in mice infected with partially attenuated viruses SARS-CoV-Δ8, SARS-CoV-Δ(6,8), and SARS-CoV-Δ(6,7,8) both at 3 and 6 dpi (Fig. 3A and B).As ORF8 was the only individual accessory protein with a significant contribution to SARS-CoV-2 virulence, it was selected for further analysis.
To confirm whether the temporal regulation of interferon and pro-inflammatory responses was associated with SARS-CoV-2-∆8 attenuation, a second experimental infection of mice was performed at earlier time points.Similar levels of infectious SARS-CoV-2-WT and SARS-CoV-2-∆8 were detected in the lungs of mice at 1, 2, and 4 dpi (Fig. 5A).Minimal differences in viral growth were also observed in the nasal turbinates of mice (Fig. 5B).mRNA expression levels of IFN-related genes (Fig. 6A) and pro-inflammatory cytokines (Fig. 6B) were measured in the lungs of mice at 1, 2, and 4 dpi.As compared to SARS-CoV-2-WT, SARS-CoV-2-∆8 induced higher expression levels at 1 or 2 dpi, which decreased at 4 dpi to become lower than those of the SARS-CoV-2-WT (Fig. 6).These results suggested that the combination of early IFN responses and the reduction of an inflammatory state at late times post-infection contributed to the partial attenuation of SARS-CoV-2-∆8.
To confirm that ORF8 was also a virulence factor in human cells, SARS-CoV-2-∆8 mutant was characterized in organoid-derived human airway cultures, previously used to study respiratory virus infections, including SARS-CoV-2 (30,31).In these organoids, basal cells can be differentiated at air-liquid interface (ALI) into a reconstituted airway culture containing basal, goblet, club, and ciliated cells.In this model, ciliated cells are suscepti ble to SARS-CoV-2 infection, just as observed in humans (30,31).Both SARS-CoV-2-WT and SARS-CoV-2-∆8 grew to similar, relatively high titers on these cells (~10 6 PFU/mL), without significant differences in the kinetics, as shown by plaque-assay titration in VeroE6 cells (Fig. 7A) and viral RNA quantification (Fig. 7B).In addition, competition assays by co-infecting organoids with similar proportions (1:1) showed that both viruses had similar replication fitness (Fig. 7C).
To confirm the effect of ORF8 on the innate immune response in a human cell model, organoid-derived human airway cells were infected with SARS-CoV-2-WT or SARS-CoV-2-∆8.Infectious viruses and intracellular RNA were collected at 24 and 72 hpi.In accord ance with the in vivo results, SARS-CoV-2-∆8 infection induced higher expression levels of IFN-β and pro-inflammatory cytokines than SARS-CoV-2-WT at 24 hpi and lower levels of IFN-β, IFN-λ, and IFIT1 at 72 hpi (Fig. 8).These results support a potential mechanism of attenuation of SARS-CoV-2-∆8 by inducing early and time-limited innate immune responses, in contrast to more delayed and sustained responses induced by the virulent SARS-CoV-2-WT.

Differences in the cellular immune response induced by SARS-CoV-2-WT and SARS-CoV-2-∆8
Macrophages are immune cells with relevant functions in host defense.Since their dysregulated hyperinflammation has also been proposed to contribute to SARS-CoV-2 pathogenesis (32), we studied macrophage responses in the lungs of K18-hACE2 mice and in the organoid model after SARS-CoV-2-WT and SARS-CoV-2-∆8 infection.Notably, the number of macrophages was significantly increased in infected mice at 6 dpi.However, this increase was higher in mice infected with SARS-CoV-2-WT than in those infected with SARS-CoV-∆8 at 6 dpi (Fig. 9A and B).Moreover, in the lesions of SARS-CoV-2-WT-infected mice, macrophages aggregated into small clusters at 6 dpi.In contrast, a more diffuse distribution of macrophages was observed in SARS-CoV-2-∆8infected mice (Fig. 9A).
The activation of human macrophages was evaluated using the organoid-derived human airway cells.Human macrophages were incubated with conditioned media collected from the basal compartment of infected organoids at 72 hpi, which contains activating cytokines.Macrophage activation was observed as a morphological transfor mation from round-shaped cells to flattened spreading cells with pseudopodium-like protrusions (Fig. 10A).A higher number of activated cells was observed when macro phages were incubated with conditioned media obtained after 72 hpi from organoids infected with SARS-CoV-2-WT, as compared to the media from SARS-CoV-2-∆8-infected organoids.Accordingly, mRNA levels of IL-6, IL-8, and TNF-α were significantly higher when macrophages were activated by the conditioned media from organoids infected with SARS-CoV-2-WT (Fig. 10B).These results suggest that macrophages could be crucial to promote the hyperinflammation observed in patients with critical disease manifesta tions since over-activation of these cells could result in a cytokine storm.

DISCUSSION
CoV accessory genes play important roles in pathogenesis (4,33,34).However, the relevance of SARS-CoV-2 accessory genes in virulence still remains unclear, since limited information is available from studies performed in the context of SARS-CoV-2 infection in animal models (9).Using a SARS-CoV-2 reverse-genetics system to engineer infectious cDNAs with deletions of genes, we studied the relevance of accessory genes to the virus pathogenesis in the K18-hACE2 mouse model.The individual deletion of ORF6 and weight loss, lethality, and lung inflammatory pathology of infected mice, indicating that ORF8 is a virulence factor.The highest attenuation was observed by the combined deletion of the four accessory ORFs 6, 7a, 7b, and 8, suggesting that the deletion of ORF7ab in the absence of ORFs 6 and 8 further attenuates SARS-CoV-2.
The most characteristic pulmonary histopathological lesions induced by WT and virulent mutants (SARS-CoV-2-Δ6 and SARS-CoV-2-Δ7ab) included severe diffuse thickening of the alveolar septae, severe diffuse mononuclear cell infiltrates within alveolar spaces as well as the presence of large multifocal perivascular and peribron chiolar mononuclear infiltrates (35,36).In contrast, the lungs infected with partially attenuated SARS-CoV-2-Δ8, SARS-CoV-2-Δ [6,8], and SARS-CoV-Δ [6,7,8] showed only mild to moderate lung damage.In addition, a reduction in the number of macrophages was observed in lung lesions in mice infected with SARS-CoV-2-Δ8 compared to lungs infected with SARS-CoV-2-WT at late times post-infection, previously associated with mild SARS disease in BALB/c mice (37).An excessive macrophage response could induce abnormal secretion of pro-inflammatory factors causing apoptosis of endothelial and epithelial cells of the lungs, resulting in hypoxia, alveolar edema, and vascular leakage (38).
Natural variants with mutations in ORF6 (39,40), ORF7a (41,42), ORF7b (43), or ORF8 (43-45) have been described.ORF6 variants showed no differences in clinical presentations as compared to hospitalized patients infected with SARS-CoV-2-WT (40).However, variants with mutations in ORF8, either a large 382-nt deletion comprising part of 7b and ORF8 transcription-regulating and coding sequences (45) or the alpha variant (B1.1.7),with a stop mutation at Q27 (46), were associated with a milder COVID-19 disease, suggesting that ORF8 might contribute to SARS-CoV-2 virulence.Since these natural variants include a variety of mutations in addition to those in ORF8, it is not possible to establish a causal relationship between the expression of ORF8 and the severity of the disease.However, these findings support a potential contribution of ORF8 to the virulence of SARS-CoV-2, as proposed by one study showing that ORF8 mimics the host IL-17 cytokine and contributes to COVID-19 severe inflammation (21) and the observation that SARS-CoV-2-∆8 induces less inflammation in the lungs of hamsters (22).In addition, a histone mimic motif was described in ORF8 (47).Chroma tin remodeling induced by viral infections can disrupt epigenetic regulation of gene expression, including the immune response, which might contribute to pathogenicity.However, the implications of this host response in acute infection or post-acute sequelae of SARS-CoV-2 are not well understood yet.
In contrast, another study postulated that ORF8 plays a minor role in the disease outcome since the deletion of ORF8 from the SARS-CoV-2 infectious cDNA induced 100% of mortality in K18-hACE2 mice (9).Experimental differences in the age of mice between that study (4-8 weeks old) (48) and our work (26 weeks old) might explain the differences observed, as it is well known that SARS-CoV-2 virulence is age dependent.Moreover, using the hamster sublethal model, one study described that infection with SARS-CoV-2 deficient in ORF6 showed significantly reduced weight loss as compared to WT-infected animals.However, both viruses caused in the lungs similar histopathological lesions, including necrosuppurative bronchiolitis at 2 dpi that progressed to bronchointerstitial pneumonia with edema and hemorrhage at 4 dpi, suggesting that ORF6 was not a major determinant of virulence (49).
The individual deletion of SARS-CoV-2 accessory genes did not affect viral titers in cell cultures, in the organoid model, or in susceptible mice, supporting the idea that they are not essential for replication, as demonstrated for other coronaviruses, SARS-CoV and MERS-CoV (4,5,9).The combined deletion of the four accessory genes (6, 7a, 7b, and 8) of SARS-CoV-2 led to a slight reduction in viral titers both in VeroE6/TMPRSS2 cells and in vivo.Similarly, the attenuated deletion mutant MERS-CoV-Δ [3,4a,4b,5] showed a reduction in viral replication (4,50).Conversely, attenuation of MERS-CoV by the deletion of single accessory genes is mainly caused by the induction of less inflammation and not by lower viral titers in the lungs (4,12).
It is well-documented that CoVs encode multiple interferon antagonists to evade the host innate immune response (51)(52)(53).Accessory proteins are not conserved among CoVs and have different functions, mostly related to the interference with the innate immune response.Highly pathogenic human CoVs induce a modest IFN response (26), as confirmed in this study (Fig. 3A and Fig. 4A), which is associated with disease severity and depends on the expression of viral genes that interfere with the host antiviral response.Deletion of ORF8 increased the early expression of MX1 and ISG15 at 1 dpi in the lungs of mice and IFNβ in human organoids at 24 hpi, suggesting that ORF8 is an IFN antagonist at early times post-infection.Overexpression of SARS-CoV-2 ORF8 in HEK-293 T cells led to controversial results.Studies described that ORF8 downregulated the expression of IFN-β and ISGs induced by Sendai virus infection (14) or poly I:C (54) by decreasing nuclear import of IRF3, while in another study ORF8 did not suppress IFN production or signaling (55).These different results might be explained by the overexpression of a single protein at non-physiological levels in the absence of other viral proteins.Moreover, preferably all the results obtained relating one protein with one viral function should be validated in physiologically relevant infection models, such as human organoids, as this work reported.In fact, the organoid model used here is useful to study the virulence of SARS-CoV-2 as it recapitulates the lung inflammatory disease observed in transgenic mice and humans (56,57).
The early increase in the production of IFN and the down-regulation of pro-inflammatory cytokines at late times post-infection with SARS-CoV-2-Δ8 led to a partial attenua tion (Fig. 2, 3, 4 and 6), suggesting a protective effect, as previously described for the early administration of IFN in SARS-CoV- (37) and MERS-CoV-( 58) infected mice.The increase of IFN at late times post-infection has been associated with excessive recruit ment of monocytes and with a more severe infection in mice in the case of SARS-CoV (37).In the same line, COVID-19 patients with milder symptoms induced higher levels of interferon early in infection and lower levels of pro-inflammatory cytokines than severe patients (56,57).In contrast, prolonged IFN responses resulted in a delayed adaptive immune response and continued upregulation of inflammatory chemokines (59).In accordance with this observation, the SARS-CoV-2-Δ382 isolate lacking ORF8 expression induced higher levels of IFN, but lower levels of pro-inflammatory responses, resulting in a milder infection (45).
Together, these results support that ORF8 promotes an innate immune response to a viral infection that contributes to immune-mediated pathology.A slight increase in IFN at early times post-infection, not prolonged in time, and a decrease in the pro-inflammatory responses during infection likely protected 40% of mice from death and reduced macrophage recruitment and activation.These results provide evidence for the relevance of innate immune timing in inflammatory pathology and disease outcome.Moreover, since ORF8 is a virulence factor during SARS-CoV-2 infection, therapies targeting ORF8 might contribute to limit immunopathology and improve the disease outcome.

Organoid culture and differentiation
Human adult airway stem cells were isolated from lung parenchyma (n = 2) and grown as described previously (30).Adult human lung tissue was obtained from non-tumor lung tissue from patients undergoing lung resection.The Medical Ethical Committee of the Erasmus MC granted permission for this study (METC 2012-512).Stem cells were isolated from adult lung tissue, maintained, and passaged in Basement Membrane Extract (BME) in the airway organoid medium described previously (30).To obtain cultures for infection experiments, stem cells were dissociated into single cells and seeded on Transwell membranes.Cells were differentiated at ALI in Pneumacult-ALI medium (StemCell Technologies) for at least 4 weeks before performing infection experiments, and the medium was replaced every 5 days.

Plasmids and bacterial strains
Bacterial artificial chromosome (BAC) pBeloBAC1 was used to assemble recombinant SARS-CoV-2 infectious cDNA clones.BAC plasmids were purified using a large-con struct kit (Qiagen), following the manufacturer's specifications.Escherichia coli DH10B (Invitrogen, Thermo Fisher Scientific) cells were transformed by electroporation using a MicroPulser unit (Bio-Rad) according to the manufacturer's instructions.

In vivo infections
Female K18-hACE2 C57BL/6J transgenic mice (strain 2B6.Cg-Tg(K18-ACE2)2Prlmn/J) were purchased from the Jackson Laboratory (Bar Harbor, ME).Mice were housed under pathogen-free conditions and acclimatized to the BSL3 animal facilities in the Animal Health Research Center (CISA/INIA-CSIC) for 7 days prior to infection.Twenty-six-weekold mice were anesthetized with isoflurane and intranasally inoculated with 10 5 PFU/ animal of each recombinant virus diluted in 50 µL of DMEM.Body weight loss and mortality were monitored daily until 10 dpi (n = 5 mice per recombinant virus).Mice showing >20% reduction of their initial body weight were euthanized.Twenty-week-old mice were euthanized and necropsied at 1, 2, 3, 4, and 6 dpi (n = 4).
For 2D ALI cultures, cells were apically washed three times with AdDF+++ medium prior to infection.Then, cells were apically inoculated at an MOI of 0.1 and incubated for viral adsorption at 37°C.After 2 hours, the inoculum was removed and cells were washed.Then, cells were incubated at 37°C and 5% CO.Samples for RNA extraction was taken at 24 and 72 hpi.

Competition assay by virus co-infection
Air-liquid airway organoids differentiated in Pneumacult were infected at a 1:1 ratio with SARS-CoV-2-WT and SARS-CoV-2-∆8, at a final MOI of 0.1 in triplicate.Apical washes were performed daily for 4 days and stored at −80°C until further processing.RNA was extracted from the washes and subjected to cDNA synthesis using Superscript IV (Invitrogen).Next, a 1,506-or 1,153-bp region including or not ORF8 was amplified by PCR using Pfu UltraII (Agilent) using the following primers: Forward: 5′-ACCAAGAGTGTGTTAGAGG-3′; reverse: 5′-GTTCAATCTGTCAAGCAGC-3′.
Conditions for the PCR were as follows: (i) denaturation for 3 minutes at 94°C; (ii) denaturation for 20 seconds at 94°C; (iii) denaturation for 20 seconds at 94°C; (iv) annealing for 20 seconds at 56°C; (v) extension for 90 seconds at 72°C; (vi) steps ii-iv 30 times; and (vii) final extension for 5 minutes at 72°C.PCR products were purified using PCR purification kit (Qiagen) before Sanger sequencing using the forward primer.Sanger sequences were analyzed using QSVana lyser.

Processing of mice samples
Lungs and nasal turbinates from euthanized mice were collected to determine viral titers for RNA extraction and histopathological analysis.The left lung lobes were collected for histopathological analysis, while the right lungs were divided into two longitudinal sections (including the three lobes) for viral titration and RNA analysis, respectively.Samples were homogenized in 1 mL of phosphate-buffered saline (PBS) (Sigma-Aldrich) with 100 IU/mL penicillin (Sigma-Aldrich), 0.1 mg/mL streptomycin (Sigma-Aldrich), 50 µg/mL gentamicin (Sigma-Aldrich), and 0.5 µg/mL amphotericin B (Sigma-Aldrich) using a gentleMACS dissociator (Miltenyi Biotec, USA) for lung tissues, or the Bead Bug 6 microtube homogenizer (Sigma-Aldrich) and metal beads lysing matrix S tubes (MP biomedicals, USA) for nasal turbinates.RNA from homogenized mouse lungs was purified using the RNeasy kit (Qiagen, Germany) following the manufacturer's specifications.

Histopathological analysis
Left lung lobes from mice were fixed in 10% zinc formalin (Sigma-Aldrich) for 48 hours.After the fixation period, samples were processed and embedded in paraffin blocks that were then sectioned at 4-µm thickness on a microtome, mounted onto glass slides, and stained with hematoxylin and eosin (H&E).Lung sections were evaluated using an Olympus BX43 microscope.To assess the character and severity of histopathologi cal lesions, lung inflammation scoring parameters based on the previous reports on SARS-CoV-2 infection in mouse models were used (36).The histopathological parame ters evaluated were hemorrhages, edema, septal thickening, alveolar damage, hyaline membranes, and inflammatory cell infiltration in alveoli.The histopathological parame ters were graded following a semi-quantitative scoring system as follows: (0) no lesion; (1) minimal lesion; (2) mild lesion; (3) moderate lesion; and (4) severe lesion.The cumulative scores of histopathological lesions provided the total score per animal.In each experimental group, the individual scores were used to calculate the group average.

Immunohistochemistry
The detection of macrophages was assessed in paraffin sections.Antigen retrieval was done by microwaving sections in citrate buffer (10 mM, pH 6.0, Sigma-Aldrich) for 5 minutes at 100°C.After endogenous peroxidase blocking with 3% H 2 O 2 for 10 minutes, slices were blocked again with 10% Normal Goat Serum for 1 hour at RT.Primary antibodies [1:100, Rabbit-anti-F4/80 (no.D2S9R) XP, Cell Signaling Technology or Normal Rabbit IgG Control (AB-105-C; R&D systems)] were incubated in incubation buffer (0.1% BSA in PBS) for 1 hour at RT.Then, sections were incubated with a secondary antibody conjugated with HRP (1:100, Goat-anti-rabbit IgG-HRP, P0448; DAKO) for 1 hour at RT. Finally, HRP activity was revealed by incubating the slides for 20 minutes at RT in a substrate containing 1.5 mL of 3-amino-9-ethyl-carbazole (AEC; Sigma-Aldrich) dissolved in N, N-dimethylformamide (DMF; Sigma-Aldrich) (17 mg/mL) and 30 µL of H 2 O 2 30% in 60 mL of NaAc buffer, pH 5.0, followed by washing step in PBS.The use of AEC results in a bright red precipitate.
Staining on two sections of each tissue was performed.The evaluation was per formed by light microscopy by two independent observers blinded to the fixation procedure.The evaluation included the macrophage count in lesions, including three mice per conditions.

Macrophage incubation with conditioned media from infected organoids
Conditioned media from organoid-derived human airway cells either mock-infected or infected with viruses at an MOI 0.1 were collected at 24 and 72 hpi from the bottom part of the insert.After confirming that the infectious virus was not detected in the conditioned media, MDMs were incubated with 100 µL of conditioned media from each condition for 24 hours.Images were taken with a Zeiss Primovert microscope using Zen software (Zeiss) before and after MDM incubation.The percentage of activated macrophages after incubation with the supernatant from infected cells was quantified with the ImageJ v2.1.0software, using three independent images per condition.
Survival analysis was interpreted by log-rank (Mantel-Cox) test.Survival of all groups were compared with the survival of mock-infected group.Significant values were indicated as asterisks (*P < 0.05; **P < 0.01).

FIG 1
FIG 1 Generation and characterization of SARS-CoV-2 deletion mutants of accessory genes.(A) Diagram of deletions of accessory genes engineered in each mutant virus.Letters above or below boxes indicate viral genes.L, leader sequence; An, poly-A tail.Red dashed line indicates deleted regions.The shadowed area shows the 3′-end viral genes.(B) Immunoblot analysis of expression of accessory proteins 6, 7a, 7b, and 8 with specific antibodies in lysates from Vero E6 cells infected at an MOI 1 for 24 hours with the indicated mutants.βActin was used as a loading control.(C) Growth kinetics of SARS-CoV-2 deletion mutants.Vero E6/TMPRSS2 cells were infected at an MOI of 0.001, and the infections followed for 72 hours.Results are represented as a mean ± SEM.

FIG 2
FIG 2 Evaluation of virulence of SARS-CoV-2 deletion mutants of accessory genes in K18-hACE2 mice.Twenty-six-week-old female K18-hACE2 mice were intranasally mock-infected or infected with 10 5 PFU/animal of each virus.Weight loss (A) and survival (B) were monitored daily for 10 days.Weight loss is represented as the mean ± SDs of the mean (n = 5 mice per group).Lung samples were obtained at 3 and 6 dpi.(C) Viral titers were determined by plaque assays in Vero E6 cells at 4 and 6 dpi.Mean ± SDs of the mean are represented (n = 3).**P < 0.01; *P < 0.05.

FIG 7 10 FIG 8 FIG 9 12 FIG 10
FIG 7 Characterization of SARS-CoV-2-WT and SARS-CoV-2-∆8 in two-dimensional (2D) human airway organoids.Representative data of replication kinetics in terms of viral titers (A) and RNA (B) studied in organoids from three different donors.(C) Proportion of replication of each virus in 2D airway organoids infected with a 1:1 ratio.
Statistical analysis was performed by GraphPad Prism, version 9. Author(s) MEC | Spanish National Plan for Scientific and Technical Research and Innovation (Plan Estatal de Investigación Científica y Técnica y de Innovación)