Regnase-1-mediated regulation of neutrophils modulates SARS-CoV-2 pneumonia

TL;DR

Reduced Regnase-1 in neutrophils improves resistance to SARS-CoV-2 infection by dampening harmful immune responses.

Contribution

Regnase-1 is identified as a novel regulator of neutrophil function and antiviral immunity during SARS-CoV-2 infection.

Findings

Reduced Regnase-1 expression in neutrophils enhances resistance to SARS-CoV-2 MA10 infection.

Regnase-1 suppresses Tsc22d3, a negative regulator of interferon signaling, contributing to immune dysregulation.

A specific neutrophil subset with interferon-stimulated gene signatures is decreased in Regnase-1+/– mice.

Abstract

The innate immune response to viral infection needs to be tightly regulated to ensure effective pathogen clearance while avoiding excessive immune activation. During SARS-CoV-2 infection, however, the immune system often fails to elicit appropriate responses, resulting in cytokine-release syndrome in patients with COVID-19. In this study, we show that reduced expression of Regnase-1, an RNase that negatively regulates immune cell activation, confers resistance to infection with the mouse-adapted SARS-CoV-2 MA10 strain. In Regnase-1+/– mice, altered neutrophil function contributed to the amelioration of MA10-induced pneumonia. Single-cell RNA sequencing of lung tissue during MA10 infection revealed four distinct neutrophil subsets, and among these, a subset characterized by an interferon-stimulated gene (ISG) signature was decreased in Regnase-1+/– mice. Furthermore, Regnase-1+/– neutrophils exhibited reduced ISG expression without corresponding changes in proinflammatory gene expression. Regnase-1 was found to repress the expression of Tsc22d3, a gene involved in the negative regulation of interferon responses, through its 3′ untranslated region. Collectively, these findings suggest that Regnase-1 attenuates resistance to SARS-CoV-2 MA10 infection by promoting excessive interferon responses in neutrophils.

Significant progress has been made in understanding the pathophysiology of SARS-CoV-2 infection. However, the mechanisms by which innate immune cells are activated to eliminate viruses while restraining hyperinflammation remain unclear. Antiviral innate immune responses are regulated by multiple checkpoints that balance effective inflammatory activity with the prevention of cytokine release syndrome. In this study, we demonstrate that the expression level of Regnase-1, a posttranscriptional regulator of immune cells, determines resistance to infection with the mouse-adapted SARS-CoV-2 MA10 strain. Reduced Regnase-1 expression in neutrophils significantly enhances resistance to viral infection. A decreased representation of a neutrophil subset characterized by a robust interferon-stimulated gene signature is associated with the attenuation of excessive inflammation. These excessive interferon responses may contribute to increased mortality from a lethal dose of the MA10 strain. Regnase-1 suppresses the expression of Tsc22d3, a negative regulator of IFN signaling. This implicates Regnase-1 in the dysregulation of immune reactions during infection. Our findings identify Regnase-1 as a novel regulator of neutrophil function and antiviral immunity, which modulates disease severity and pulmonary inflammation.

Figures

Ameliorated pneumonia and increased pulmonary neutrophil infiltration in MA10-infected Reg1+/– mice.A) Female BALB/c WT and Reg1+/– mice (12-weeks old, n = 4 per group) were intranasally infected with 102 TCID₅₀ MA10. Body weight was monitored daily from -2 to 14 days dpi. The average weight from day -2 to 0 for each group was defined as 100%, and subsequent changes are shown as mean ± s.d. Significant differences was determined by two-way ANOVA with Bonferroni’s post-test (*P < 0.05, P < 0.01). B) Representative H&E-stained lung sections (200 × magnification) from MA10-infected mice collected at 3 and 6 dpi. Lung injury was evaluated in a blinded manner using the DAD scoring at these time points. C) Female BALB/c WT and Reg1+/– mice (12-weeks old, n = 4 per group) were intranasally infected with 103 TCID₅₀ MA10. Body weight was monitored daily from -2 to 6 dpi. The average weight from day -2 to 0 was defined as 100%, and changes were shown as the mean ± s.d. Significant difference was assessed by two-way ANOVA with Bonferroni’s post-test (*P < 0.05, P < 0.0001). D) Representative H&E-stained lung sections (200 × magnification) from MA10-infected mice collected at 3 and 6 dpi. Lung injury was evaluated using the DAD scoring at 3 dpi (n = 3 per group) and 6 dpi (n = 3 for WT, n = 4 for Reg1+/–). E) Absolute numbers of the indicated cell populations infiltrating lungs at 3 and 6 dpi. Cell numbers were calculated by multiplying the total number of cells from lung single cell suspension by the proportion of each population determined by flow cytometry. F) qPCR analysis of Ifna, Il1b and Il6 mRNA expression in RNA extracted from infected lungs at 3 and 6 dpi. G) qPCR analysis of MA10 viral RNA, including nucleocapsid phosphoprotein (N1, N2) and ORF1b, in the lungs of infected mice at 3 and 6 dpi. Bar graphs (B, D-G) show the mean ± s.d. (n = 4). P-values are shown. (two-tailed unpaired Student’s t-test).

Neutrophils mediate resistance to MA10 infection in Reg1+/– mice.A) qPCR analysis of Reg1 mRNA expression in the indicated lung cell types under steady state conditions (n = 3). B) Schematic representation of antibody treatment and MA10 infection. Female BALB/c WT mice were divided into two groups: no antibody treatment (PBS) group and Ly6G antibody treatment group. Female BALB/c Reg1+/– mice were also treated with Ly6G antibody. Antibody-treated groups daily received 100 μg/day of Ly6G antibody intraperitoneally (i.p.) until 4 dpi. All mice were intranasally infected with 103 TCID₅₀ MA10. Lung-infiltrating cells were collected at 5 dpi for flow cytometry analysis and a portion of the lung tissue was used for histopathological evaluation. C) Flow cytometry analysis of neutrophils in CD45+ live cells isolated from infected lung at 5 dpi. D) Body weight was monitored from day -2 to 5 dpi. The average weight from day -2 to 0 was defined as 100%, and changes were shown as the mean ± s.d. Significant difference was assessed by two-way ANOVA with Bonferroni’s post-test (*P < 0.05). E) Representative H&E-stained lung sections (200 × magnification) from MA10-infected mice collected at 5 dpi. Lung injury was evaluated using the DAD scoring at 5 dpi (n = 3 for no untreated WT mice, n = 4 for anti Ly6G treated WT and anti-Ly6G treated Reg1+/–). F) qPCR analysis of MA10 viral RNA, including nucleocapsid phosphoprotein (N1, N2) and ORF1b, in the lungs of infected mice at 5 dpi. (n = 3 for no untreated WT mice, n = 4 for anti Ly6G treated WT and anti-Ly6G treated Reg1+/–). The bar graphs (A, E, F) show the mean ± s.d. (n = 3 or 4). P-values are shown. (two-tailed unpaired Student’s t-test).

scRNA-seq reveals the subdivision of lung-infiltrating neutrophils into four clusters following 103 TCID₅₀ MA10 infection.A) UMAP representation and clustering of lung infiltrating CD45+ cells (0, 3 and 5 dpi, WT mice and Reg1+/– mice). Data-driven clustering and cell type based on canonical marker genes are shown. B) UMAP representation of lung CD45+ cells at the indicated time points and genotypes. For each group, the CD45+ cells from three mice were combined and processed as a single sample (n = 1) for further analysis. C) UMAP representation of lung infiltrating neutrophils, corresponding to clusters 9, 13 and 21 in (A). These neutrophils were further subdivided into four clusters based on gene expression profiles. D) Absolute numbers of neutrophils at 0, 3 and 5 dpi in WT mice or Reg1+/– mice.

Attenuated type I IFN signature in lung neutrophils of MA10-infected Reg1+/– mice.A) Heatmap showing differences in gene expression among the four neutrophil subclusters (cluster 2, 3, 4 and 5 from Fig 3C). B) Heatmap of the top marker genes ranked by log fold change which are associated with the neutrophil clusters indicated in (A), segmented by time points (0, 3 and 5 dpi). C) Heatmap of the top marker genes ranked by log fold change which are associated with the neutrophil clusters indicated in (A), segmented by time points (0, 3 and 5 dpi) and genotype of mice. D) Heatmap of genes upregulated in cluster 5 by comparing 0 vs. 3 dpi and 0 vs. 5dpi in WT mice. Gene expression is displayed across time points and genotypes. E) GSEA comparing transcriptome profiles of lung-infiltrating neutrophils at 5 dpi (103 TCID₅₀ MA10 infection) between WT and Reg1+/– mice using HALLMARK_INTERFERON_ALPHA_RESPONSE gene sets (M5911; genes upregulated in response to IFNα).

Reg1 suppresses the expression of genes including Tsc22d3 via its RNase activity.A) List of differentially expressed genes upregulated in Reg1+/– mice, identified by RNA-seq analysis of lung-infiltrating neutrophils from WT and Reg1+/– mice at 0 dpi (uninfected). The top 135 genes are listed, ranked by the genes with the smallest log2 fold change values (-9.26 ~ -0.34). Five candidate genes potentially related to neutrophil regulation (Tsc22d3, Nr2f2, Axl, Muc4 and Fgfr1) are highlighted in yellow. Two biological replicates were analyzed. B) qPCR of Tsc22d3 mRNA expression in lung-infiltrating neutrophils at 0 dpi from WT and Reg1+/– mice (n = 5 per group). C) qPCR analysis of Tsc22d3 mRNA in the indicated cell types from uninfected lung (n = 3). D) Luciferase activity in HEK293T cells transfected with reporter plasmids expressing murine Tsc22d3 3’ UTR sequence and expression plasmids for Reg1 (WT or D141N) (n = 3 per group). The bar graphs (B, C, D) show the mean ± s.d. (n = 3 or 5). P-values are shown. (two-tailed unpaired Student’s t-test).

Full text

Introduction

The clinical course of COVID-19 exhibits considerable variability. While most patients experience mild to moderate symptoms, 10% to 20% develop pneumonia and severe, life-threatening complications [1]. A hallmark of severe COVID-19 is the excessive production of proinflammatory cytokines, triggered by the recognition of SARS-CoV-2 by pattern recognition receptors [2–7] along with hypercoagulation [8–10]. In addition, accumulating evidence suggests that multiple factors contribute to disease severity, including defects in the type Ⅰ interferon (IFN) response [11,12]. In contrast, delayed and excessive type I IFN responses may exacerbate inflammation and contribute to the severe progression of COVID-19 [13].

Both innate and adaptive immune cells play critical roles in mediating protective immune responses while also contributing to detrimental inflammation during viral infection. Among these, neutrophils, constituting the most abundant nucleated cells in the blood, serve as a first line of defense against invading microorganisms [14]. They exert antimicrobial functions through mechanisms such as degranulation, production of reactive oxygen species, phagocytosis, and the formation of neutrophil extracellular traps (NETs) [15,16]. In patients with severe COVID-19, elevated neutrophil counts and increased neutrophil-to-lymphocyte ratios in peripheral blood are frequently observed [17–19]. Additionally, dysfunctional low-density neutrophils have been reported in the peripheral blood of severe COVID-19 patients [20,21], suggesting that both neutrophil increase and dysfunction associated to disease severity. The accumulation of neutrophils in the lungs, as documented in autopsy cases of life-threatening COVID-19, along with NET formation [22], provides compelling evidence for the critical role of neutrophils in disease pathogenesis. Nevertheless, the molecular mechanisms regulating neutrophil function during SARS-CoV-2 infection remain poorly understood.

Regnase-1 (Reg1) functions as an RNase that suppresses inflammation by degrading mRNAs encoding inflammation-related genes such as Il6, Il12b and Il1b in immune and non-immune cells [23,24]. Reg1 recognizes stem-loop structures containing a pyrimidine-purine-pyrimidine loop sequence in the 3′ untranslated regions (UTRs) of target mRNAs and degrades them in a translation-dependent manner. Mice deficient in Reg1 (Reg1^–/–^) develop severe inflammatory and autoimmune diseases [23]. Moreover, the Reg1 haploinsufficiency exacerbates autoimmune disease mouse models such as experimental autoimmune encephalomyelitis and imiquimod-induced psoriasis [25–27]. However, the in vivo role of Reg1 during viral infection, particularly in the context of SARS-CoV-2 infection, remains uncharacterized.

In this study, we investigated the role of Reg1 on SARS-CoV-2 infection by utilizing a mouse-adapted SARS-CoV-2 strain (MA10), which replicates severe pneumonia observed in life-threatening COVID-19 (ref.[28]). Reg1^+^^/–^ mice exhibited increased resistance to MA10 infection compared to wild-type (WT) mice, depending on the activity of neutrophils. Furthermore, single-cell RNA sequencing analysis suggests that Reg1 modulates both infiltration and functional characteristics of neutrophil subsets in the lung during MA10 infection. Collectively, these findings uncover the roles of Reg1-mediated control of neutrophils in regulating immune responses against SARS-CoV-2 pneumonia.

Results

Amelioration of SARS-CoV-2 MA10-induced pneumonia under Reg1 haploinsufficiency in mice

We investigated the in vivo role of Reg1 in mice intranasally infected with the mouse-adapted SARS-CoV-2 (MA10) strain. Whereas Reg1 haploinsufficient (^+^^/–^) mice exhibited reduced levels of Reg1 expression [29], immune cell proportions in the lungs remained comparable between Reg1^+^^/–^ and WT mice under steady state conditions (S1A Fig). Following infection with 10² TCID₅₀ of MA10, Reg1^+^^/–^ mice showed significantly less weight loss compared to WT mice which lost approximately 10% of their body weight and required 10 days for full recovery (Fig 1A). Histopathological analysis revealed that bronchial damage, alveolar obstruction, and airway lumen retention were less severe in Reg1^+^^/–^ mice compared to WT mice at 3 and 6 days post infection (dpi) (Fig 1B). By contrast, intranasal PBS treatment did not induce pathological lesions in the lungs from either Reg1^+^^/–^ or WT mice (S1B Fig), indicating that observed changes were attributable to MA10 infection. Moreover, diffuse alveolar damage (DAD) scores, assessed based on cellular sloughing, hyalinization, and necrosis, were significantly lower in Reg1^+^^/–^ mice (Fig 1B). Furthermore, infection with higher dose (10^3^ TCID₅₀) of MA10, which resulted in 100% mortality in WT mice by 6 dpi, did not lead to any fatalities in Reg1^+^^/–^ mice. Instead, these mice began to gain body weight by 5 dpi after an initial approximately 10% weight loss (Fig 1C). Consistently, immune cell infiltration and DAD scores were significantly lower in Reg1^+^^/–^ mice at 3 and 6 days following infection with 10^3^ TCID₅₀ of MA10 (Fig 1D).

*Ameliorated pneumonia and increased pulmonary neutrophil infiltration in MA10-infected Reg1+/– mice.A) Female BALB/c WT and Reg1+/– mice (12-weeks old, n = 4 per group) were intranasally infected with 102 TCID₅₀ MA10. Body weight was monitored daily from -2 to 14 days dpi. The average weight from day -2 to 0 for each group was defined as 100%, and subsequent changes are shown as mean ± s.d. Significant differences was determined by two-way ANOVA with Bonferroni’s post-test (*P < 0.05, **P < 0.01). B) Representative H&E-stained lung sections (200 × magnification) from MA10-infected mice collected at 3 and 6 dpi. Lung injury was evaluated in a blinded manner using the DAD scoring at these time points. C) Female BALB/c WT and Reg1+/– mice (12-weeks old, n = 4 per group) were intranasally infected with 103 TCID₅₀ MA10. Body weight was monitored daily from -2 to 6 dpi. The average weight from day -2 to 0 was defined as 100%, and changes were shown as the mean ± s.d. Significant difference was assessed by two-way ANOVA with Bonferroni’s post-test (*P < 0.05, ***P < 0.0001). D) Representative H&E-stained lung sections (200 × magnification) from MA10-infected mice collected at 3 and 6 dpi. Lung injury was evaluated using the DAD scoring at 3 dpi (n = 3 per group) and 6 dpi (n = 3 for WT, n = 4 for Reg1+/–). E) Absolute numbers of the indicated cell populations infiltrating lungs at 3 and 6 dpi. Cell numbers were calculated by multiplying the total number of cells from lung single cell suspension by the proportion of each population determined by flow cytometry. F) qPCR analysis of Ifna, Il1b and Il6 mRNA expression in RNA extracted from infected lungs at 3 and 6 dpi. G) qPCR analysis of MA10 viral RNA, including nucleocapsid phosphoprotein (N1, N2) and ORF1b, in the lungs of infected mice at 3 and 6 dpi. Bar graphs (B, D-G) show the mean ± s.d. (n = 4). P-values are shown. (two-tailed unpaired Student’s t-test).

Examination of the immune cell populations in the lungs revealed increased numbers of neutrophils and alveolar macrophages in Reg1^+^^/–^ mice compared with WT mice at 3 dpi with 10^2^ TCID₅₀ of MA10 (Fig 1E). However, at 6 dpi, the number of alveolar macrophages had decreased in both WT and Reg1^+^^/–^ mice. Nevertheless, Reg1 haploinsufficiency failed to alter the numbers of other pulmonary innate and acquired immune cells examined (Fig 1E).

Then we examined if Reg1 expression alters cytokine expression during MA10 infection. We observed that Reg1^+^^/–^ mice exhibited increased expression of Il6 and Cxcl1 at 6 dpi without altering the expression of other cytokines, such as Il1b, Tnf, Il12b, Ifnb1 and Ifna (Figs 1F and S2A). Nevertheless, the levels of MA10 viral RNA corresponding to the nucleocapsid (N1, N2) and exonuclease (ORF1b) regions were comparable between lungs from WT and Reg1^+^^/–^ mice at both 3 dpi and 6 dpi (Fig 1G), suggesting that Reg1 haploinsufficiency confers protection against SARS-CoV-2 MA10 infection through mechanisms other than direct viral clearance, potentially involving enhanced accumulation of myeloid cells in the lung.

Contribution of neutrophils to resistance in Reg1+/– mice against MA10 infection

The observation that neutrophil numbers tend to be elevated in Reg1^+^^/–^ mice than in WT mice during MA10 infection, coupled with the higher expression of Reg1 mRNA in neutrophils among lung-resident immune cells under steady state conditions (Fig 2A), led us to hypothesize that neutrophils contribute to the increased resistance of Reg1^+^^/–^ mice to MA10 infection. To elucidate the role of neutrophils in mediating resistance to MA10 infection under Reg1 haploinsufficiency, we depleted neutrophils in mice using an anti-Ly6G antibody prior to 10^3^ TCID₅₀ MA10 infection (Fig 2B and 2C). The Ly6G antibody efficiently depleted neutrophils, but not monocytes, during MA10 infection (Figs 2C and S3A). Whereas both neutrophil-depleted WT and Reg1^+^^/–^ mice exhibited continued weight loss by 5 dpi, there was a significant difference in body weight compared with untreated WT mice at 5 dpi (Fig 2D). Consistently, neutrophil depletion was associated with severe lung inflammation in both WT and Reg1^+^^/–^ mice, although DAD scores did not reach statistically significant compared with WT mice without neutrophil depletion (Fig 2E). DAD scores were similar between neutrophil-depleted WT and Reg1^+^^/–^ mice during MA10 infection (Fig 2E). Notably, neutrophil depletion did not significantly alter the levels of MA10 viral RNAs (N1, N2 and Orf1b) in lungs at 5 dpi in WT and Reg1^+^^/–^ mice (Fig 2F). These results imply that neutrophil depletion does not alter viral clearance. These results indicate that Reg1^+^^/–^ mice no longer displayed enhanced resistance to MA10 infection compared to WT mice in the absence of neutrophils. Collectively, these results suggest that Reg1 expressed in neutrophils contributes to the regulation of resistance against MA10 infection.

*Neutrophils mediate resistance to MA10 infection in Reg1+/– mice.A) qPCR analysis of Reg1 mRNA expression in the indicated lung cell types under steady state conditions (n = 3). B) Schematic representation of antibody treatment and MA10 infection. Female BALB/c WT mice were divided into two groups: no antibody treatment (PBS) group and Ly6G antibody treatment group. Female BALB/c Reg1+/– mice were also treated with Ly6G antibody. Antibody-treated groups daily received 100 μg/day of Ly6G antibody intraperitoneally (i.p.) until 4 dpi. All mice were intranasally infected with 103 TCID₅₀ MA10. Lung-infiltrating cells were collected at 5 dpi for flow cytometry analysis and a portion of the lung tissue was used for histopathological evaluation. C) Flow cytometry analysis of neutrophils in CD45+ live cells isolated from infected lung at 5 dpi. D) Body weight was monitored from day -2 to 5 dpi. The average weight from day -2 to 0 was defined as 100%, and changes were shown as the mean ± s.d. Significant difference was assessed by two-way ANOVA with Bonferroni’s post-test (P < 0.05). E) Representative H&E-stained lung sections (200 × magnification) from MA10-infected mice collected at 5 dpi. Lung injury was evaluated using the DAD scoring at 5 dpi (n = 3 for no untreated WT mice, n = 4 for anti Ly6G treated WT and anti-Ly6G treated Reg1+/–). F) qPCR analysis of MA10 viral RNA, including nucleocapsid phosphoprotein (N1, N2) and ORF1b, in the lungs of infected mice at 5 dpi. (n = 3 for no untreated WT mice, n = 4 for anti Ly6G treated WT and anti-Ly6G treated Reg1+/–). The bar graphs (A, E, F) show the mean ± s.d. (n = 3 or 4). P-values are shown. (two-tailed unpaired Student’s t-test).

Reduced number of neutrophils with a high IFN signature in the lungs of MA10-infected Reg1+/– mice

To further characterize the phenotypic changes in immune cells during infection with SARS-CoV-2 MA10 strain, we isolated lung CD45^+^ cells from mice at 0, 3 and 5 days after infection with 10^3^ TCID₅₀ MA10 and subjected them to single-cell RNA sequencing (scRNA-seq). We analyzed 7507, 8180 and 12466 cells from WT mice and 4630, 8526 and 11485 cells from Reg1^+^^/–^ mice at 0, 3 and 5 dpi, respectively. Transcriptomic analysis identified 22 distinct cell clusters, allowing for the characterization of lung immune cell populations based on their signature gene expression profiles (Fig 3A). The lethal dose (10^3^ TCID₅₀) MA10 infection induced dynamic alterations in lung immune cell composition, marked by increases in monocytes, macrophages, T and B cells, along with a decrease in alveolar macrophages in both WT and Reg1^+^^/–^ mice (Fig 3B). However, we failed to observe significant differences in the proportions of immune cell subsets including T cells, B cells, NK cells, monocytes/macrophages, alveolar macrophages, neutrophils and plasmacytoid dendritic cells (pDCs) between WT and Reg1^+^^/–^ mice during lethal dose of MA10 infection (Fig 3B). Flow cytometry analysis confirmed that the numbers of immune cells were comparable between WT and Reg1^+^^/–^ mice, except NK cells showing an increase in Reg1^+^^/–^ mice at 6 dpi (S4A Fig). The discrepancy between scRNA-seq and flow cytometry results in NK cells is thought to be due to the difference in sample collection timing at 5 dpi and 6 dpi. Further studies will be required to elucidate the role of Reg1 in NK cells during SARS-CoV-2 MA10 infection. Furthermore, viral RNA levels and cytokine gene expression in the whole lung as well as NK cells, monocytes/macrophages and pDCs were comparable between WT and Reg1^+^^/–^ mice (S4B and S4C Fig and S1 Table). These results suggest that the enhanced resistance of Reg1^+^^/–^ mice to viral infection is not attributable to altered proinflammatory cytokine production. Furthermore, prior to infection, neutrophils exhibited higher Reg1 mRNA expression than other lung-resident cell types (Fig 2A). Following infection, neutrophils also maintained higher Reg1 mRNA expression among innate immune cells (S5A Fig).

scRNA-seq reveals the subdivision of lung-infiltrating neutrophils into four clusters following 103 TCID₅₀ MA10 infection.A) UMAP representation and clustering of lung infiltrating CD45+ cells (0, 3 and 5 dpi, WT mice and Reg1+/– mice). Data-driven clustering and cell type based on canonical marker genes are shown. B) UMAP representation of lung CD45+ cells at the indicated time points and genotypes. For each group, the CD45+ cells from three mice were combined and processed as a single sample (n = 1) for further analysis. C) UMAP representation of lung infiltrating neutrophils, corresponding to clusters 9, 13 and 21 in (A). These neutrophils were further subdivided into four clusters based on gene expression profiles. D) Absolute numbers of neutrophils at 0, 3 and 5 dpi in WT mice or Reg1+/– mice.

Given that neutrophils were shown to contribute to increased resistance to MA10 infection in Reg1^+^^/–^ mice (Fig 2D and 2E), we investigated their properties in more detail. scRNA-seq analysis identified eight neutrophil clusters in the lung (S6A Fig). However, based on their gene expression profiles, clusters 1, 6, 7, and 8 likely represented doublets composed of neutrophils and other cells such as T cells or B cells (S6B Fig), and were excluded from further analysis and we focused on clusters 2–5. Under steady state conditions, a majority of neutrophils belonged to clusters 2, with a smaller population in cluster 3 (Fig 3C and 3D). Following MA10 infection, the proportion of cluster 2 neutrophils initially diminished but reemerged at 5 dpi. In contrast, other neutrophil clusters exhibited distinct kinetics: cluster 5 expanded rapidly but transiently at 3 dpi, whereas clusters 3 and 4 increased notably at 5 dpi. Reg1 haploinsufficiency had minimal impact on the overall distribution of neutrophil clusters, with the exception of cluster 5, which was significantly reduced at both 3 and 5 dpi (Fig 3C and 3D).

We next characterized the transcriptional features of these neutrophil clusters. Although severe COVID-19 has been associated with emergency myelopoiesis leading to the release of immature neutrophils into circulation [20], neutrophils in clusters 2–5 from both WT and Reg1^+^^/–^ mice expressed genes characteristic of mature neutrophils (e.g., S100a8, S100a9 and Junb) and lacked expression of immature markers including Mpo, Elane and Gata1 [30–34] (S6C and S6D Fig). These findings indicate that the identified clusters represent functional heterogeneity among mature neutrophils rather than distinct stages of maturation.

Each neutrophil cluster exhibited a distinct gene expression profile (Fig 4A). Clusters 3 and 5 were characterized by a robust, transient induction of ISGs, including Irf7, Ifit1, Ifit2, Ifit3, Rsad2, and Isg15, at 3dpi (Fig 4B). Notably, cluster 3, unlike 5, showed elevated expression of pro-inflammatory cytokine genes and their regulatory elements, including Il1a, Tnf, Nfkbia and Tnfaip3, in addition to ISGs. In contrast, clusters 2 and 4 exhibited only modest ISG expression. Cluster 4 was enriched for genes involved in neutrophil migration, protease inhibition and genes with unknown functions (e.g., Cd177, Cstdc4, Cstdc6, Stfa2 and Stfa3), whereas cluster 2 lacked prominent transcriptomic features.

Attenuated type I IFN signature in lung neutrophils of MA10-infected Reg1+/– mice.A) Heatmap showing differences in gene expression among the four neutrophil subclusters (cluster 2, 3, 4 and 5 from Fig 3C). B) Heatmap of the top marker genes ranked by log fold change which are associated with the neutrophil clusters indicated in (A), segmented by time points (0, 3 and 5 dpi). C) Heatmap of the top marker genes ranked by log fold change which are associated with the neutrophil clusters indicated in (A), segmented by time points (0, 3 and 5 dpi) and genotype of mice. D) Heatmap of genes upregulated in cluster 5 by comparing 0 vs. 3 dpi and 0 vs. 5dpi in WT mice. Gene expression is displayed across time points and genotypes. E) GSEA comparing transcriptome profiles of lung-infiltrating neutrophils at 5 dpi (103 TCID₅₀ MA10 infection) between WT and Reg1+/– mice using HALLMARK_INTERFERON_ALPHA_RESPONSE gene sets (M5911; genes upregulated in response to IFNα).

The expression of proinflammatory cytokine-associated genes, Il1a, Tnf, Nfkbia and Tnfaip3, was comparable between WT and Reg1^+^^/–^ neutrophils, or was slightly elevated in Reg1^+^^/–^ mice (Figs 4C and S7A). In contrast, the expression of ISGs including Ifit1, Ifit3, Ifitm3, Oasl2 and Usp18 was reduced particularly in clusters 3 and 5 in Reg1^+^^/–^ mice compared to WT, while the expression of Ifitm3 showed slight increase in cluster 5 of Reg1^+^^/–^ mice at 5 dpi (Figs 4C, 4D, and S7A). Consistently, transcriptome analysis of neutrophils sorted from MA10-infected WT and Reg1^+^^/–^ mice confirmed a downregulation of a gene set associated with type I IFN responses in Reg1^+^^/–^ neutrophils (Fig 4E). Collectively, these findings suggest that Reg1 haploinsufficiency attenuates type I IFN responses in neutrophils during MA10 infection, which may contribute to the reduced disease severity observed in Reg1^+^^/–^ mice.

Reg1 suppresses genes including Tsc22d3 in neutrophils

To investigate the molecular mechanisms by which Reg1 modulates neutrophil functions, we assessed gene expression profiles in lung neutrophils from WT and Reg1^+^^/–^ mice. Given that excessive type I IFN responses in neutrophils may contribute to tissue damage and impair resistance to viral infection, we hypothesized that Reg1 suppresses genes involved in the negative regulation of the IFN responses. RNA-seq analysis of lung neutrophils revealed that Reg1 haploinsufficiency led to the upregulation of 135 genes compared to WT neutrophils (Fig 5A). Among these, the expression of 5 genes reported to be associated with IFN response signature (Tsc22d3, Nr2f2, Axl, Muc4 and Fgfr1) [35–39] were validated by qPCR. Noteworthy, Tsc22d3 exhibited significant upregulation in Reg1-haploinsufficient neutrophils, whereas the other 4 genes did not show significant upregulation (Figs 5B and S8A). Additionally, the expression of Tsc22d3 in the lung was higher in neutrophils than in other immune cell types in uninfected mice (Fig 5C). To determine whether Reg1 directly regulates Tsc22d3 mRNA via the 3′ UTR, we conducted the luciferase reporter assays. Overexpression of wild-type Reg1 significantly suppressed the activity of the reporter construct harboring the Tsc22d3 3′ UTR (Fig 5D). In contrast, a nuclease-inactive Reg1 mutant (D141N) failed to inhibit the reporter activity (Fig 5D), indicating that Reg1 acts as an RNase to control Tsc22d3 mRNA expression. These findings suggest that Reg1 modulates the expression of neutrophil genes, including ISG signatures, by degrading mRNA such as Tsc22d3.

Reg1 suppresses the expression of genes including Tsc22d3 via its RNase activity.A) List of differentially expressed genes upregulated in Reg1+/– mice, identified by RNA-seq analysis of lung-infiltrating neutrophils from WT and Reg1+/– mice at 0 dpi (uninfected). The top 135 genes are listed, ranked by the genes with the smallest log2 fold change values (-9.26 ~ -0.34). Five candidate genes potentially related to neutrophil regulation (Tsc22d3, Nr2f2, Axl, Muc4 and Fgfr1) are highlighted in yellow. Two biological replicates were analyzed. B) qPCR of Tsc22d3 mRNA expression in lung-infiltrating neutrophils at 0 dpi from WT and Reg1+/– mice (n = 5 per group). C) qPCR analysis of Tsc22d3 mRNA in the indicated cell types from uninfected lung (n = 3). D) Luciferase activity in HEK293T cells transfected with reporter plasmids expressing murine Tsc22d3 3’ UTR sequence and expression plasmids for Reg1 (WT or D141N) (n = 3 per group). The bar graphs (B, C, D) show the mean ± s.d. (n = 3 or 5). P-values are shown. (two-tailed unpaired Student’s t-test).

Discussion

In the present study, we found that Reg1^+^^/–^ mice exhibited ameliorated pneumonia following infection with the mouse-adapted SARS-CoV-2 MA10 strain compared to WT mice. Neutrophil depletion experiments demonstrated that the reduced Reg1 expression in neutrophils contributes to the enhanced resistance of Reg1^+^^/–^ mice to infection. scRNA-seq analysis further revealed a decrease of a neutrophil subset with robust IFN signatures under Reg1 haploinsufficiency during MA10 infection, implicating that excessive type I IFN activation in WT mice is involved in disease pathogenesis. Further, we identified a set of Reg1-suppressed genes in neutrophils including Tsc22d3, that may regulate type I IFN signaling. These findings suggest that decreased expression of Reg1 expression in lung neutrophils provide protective effects against SARS-CoV-2 MA10 infection.

Our experimental model offers several advantages in elucidating the pathogenesis contributing to the severity of SARS-CoV-2 infection. Clinical specimens from COVID-19 patients are typically collected at medical facilities after symptom onset and definitive diagnosis, making it challenging to collect samples from the early stages of infection and track disease progression. In contrast, the SARS-CoV-2 MA10 mouse model employed in this study provided robust evidence for the contribution of neutrophils in resistance to SARS-CoV-2. This model allowed targeted neutrophil depletion and precise temporal control over analysis, enabling a detailed examination of immune cell dynamics in the lung following MA10 infection. An additional advantage of the MA10 mouse model is its suitability for investigating the role of specific genes in COVID-19 resistance using genetically modified mice. In this study, we demonstrated that reduced Reg1 expression due to haploinsufficiency enhances resistance to MA10 infection, further highlighting the utility of this model for studying host-pathogen interactions.

It should be noted, however, that naturally occurring SARS-CoV-2 variants such as Omicron lineages are also capable of infecting laboratory mice, but generally induce milder disease with limited lung pathology compared with MA10. While this limits their use for modeling life-threatening pneumonia, Omicron-based models may be particularly valuable for investigating host responses associated with upper airway infection, immune evasion, viral persistence, and post-acute or subclinical disease states. Future studies comparing MA10 with Omicron strains in genetically modified mice may therefore help to delineate shared versus strain-specific mechanisms governing neutrophil function, tissue damage, and immune regulation during SARS-CoV-2 infection.

Our neutrophil depletion studies further underscored the critical role of these cells in mediating antiviral resistance. In influenza infection, it has been reported that pneumonia severity worsens in the absence of neutrophils at the time of infection, based on similar neutrophil depletion experiments [40]. We demonstrated that neutrophils are also critical for host defense against MA10 infection. Numerous clinical studies have highlighted the detrimental roles of excessive neutrophil activation in severe COVID-19, including increased peripheral neutrophil counts [1] and elevated neutrophil-to-lymphocyte ratios [41], as well as the formation of NETs, which may induce secondary organ damage, such as thrombus formation [22]. Together with our findings, these observations emphasize the need for tight control of neutrophil activation during viral infection. While neutrophil deficiency impairs antiviral defense, excessive neutrophil activation is a major driver of severe disease and life-threatening complications. Our scRNA-seq analysis further revealed dynamic changes in neutrophil populations during viral infection, providing a detailed characterization of lung immune cells responses. These findings complement prior studies analyzing peripheral blood and bronchoalveolar lavage samples from COVID-19 patients [1,17–21,41–48], offering additional insights into the immunological landscape of SARS-CoV-2 infection.

An adequate IFN response is essential for antiviral host defense. In one hand, both insufficient type I IFN production [42] and the presence of antibodies to type I and type III IFNs [11,12] exacerbate COVID-19. Conversely, delayed yet excessive IFN responses can exacerbate inflammation, contributing to severe disease progression [13]. This dysregulated IFN response, in combination with excessive inflammatory cytokine production, can lead to a “cytokine storm” causing severe lung damage and multiple organ failure [49]. The timing and location of IFN responses are therefore crucial determinants of disease outcomes [13,43–46]. This notion is further supported by studies demonstrating that dexamethasone, which dampens the ISG response, provides protection against SARS-CoV-2 infection [47,48]. Accordingly, dexamethasone is recommended for treating hospitalized patients suffering from life-threatening COVID-19 requiring oxygen supplementation or mechanical ventilation [50]. Given that Reg1 haploinsufficiency confers resistance to mouse SARS-CoV-2 infection while attenuating IFN responses, targeted suppression of Reg1 expression may represent a novel therapeutic strategy for COVID-19 and other severe viral infection.

We observed that the expression of Tsc22d3 was upregulated, whereas ISG signature genes were suppressed in Reg1^+^^/–^ neutrophils. Furthermore, we demonstrated that Reg1 directly destabilizes Tsc22d3 via its 3’ UTR in an RNase activity-dependent manner. Tsc22d3 (also known as GILZ: glucocorticoid-induced leucine zipper) was originally identified as an anti-inflammatory molecule induced by glucocorticoid [51–53]. A previous study showed that Tsc22d3 suppress ISG upregulation by binding to STAT1 and preventing its nuclear translocation [35]. In addition, Tsc22d3 expression in neutrophils has been shown to exert protective effects in an acute colitis mouse model by inhibiting the MAPK pathway, and thereby preventing excessive neutrophil activation [54]. Thus, increased expression of Tsc22d3 in Reg1^+^^/–^ neutrophils may confer protection against MA10 infection by damping ISG expression and excessive neutrophil activation. However, further investigations are needed to validate these hypotheses. Moreover, in addition to Tsc22d3, Reg1 may target additional mRNAs involved in the regulation of ISG expression in neutrophils. mRNAs upregulated in Reg1^+^^/–^ neutrophils represent the candidates target, and future studies will be required to determine which of these mRNAs are directly degraded by Reg1 and contribute to the control of IFN responses during SARS-CoV-2 MA10 infection.

In this study, we primarily focused on the role of Reg1 expressed in neutrophils in shaping host responses against MA10 infection. However, Reg1 expressed in other cell types may also contribute to disease outcomes. For instance, flow cytometry analysis revealed an increased number of alveolar macrophages in Reg1^+^^/–^ mice compared with WT mice at 3 dpi following low-dose MA10 infection. Future investigations will be required to elucidate the roles of Reg1 in other immune and non-immune cell types during MA10 infection.

It was reported that Regnase-1 can degrade viral RNA in cell systems overexpressing Reg1 [55,56]. However, these studies did not demonstrate a contribution of Reg1-mediated viral RNA degradation to antiviral defense in vivo. If Reg1-mediated viral RNA degradation plays a significant role during SARS-CoV-2 infection in vivo, one would expect Reg1^+^^/–^ mice to exhibit increased viral loads and exacerbated disease. In contrast, our study showed that Reg1^+^^/–^ mice developed ameliorated SARS-CoV-2 pneumonia, and we did not detect significant differences in viral loads in the lungs. These findings indicate that Reg1 contributes to the exacerbation of SARS-CoV-2 pneumonia in vivo, even though we cannot exclude the possibility that Reg1 may degrade SARS-CoV-2 RNA in certain cell types, including airway epithelial cells.

A limitation of this study is the reliance on a mouse model to assess the role of Reg1 in resistance against mouse-adapted SARS-CoV-2 MA10 infection, and its relevance to human COVID-19 remains untested. In addition, in our scRNA-seq analysis, three mice were used per time point and pooled to generate a biological replicate. Owing to this experimental design, statistical comparisons between groups could not be performed, and formal statistical analyses were therefore not feasible for these datasets. Nonetheless, a previous study reported that Reg1 expression is upregulated in dysfunctional neutrophils in the peripheral blood of patients with severe COVID-19 [20], suggesting that maintaining optimal Reg1 levels in neutrophils may be protective, consistent with our findings. However, further studies are required to establish the causal relationship between Reg1 expression and neutrophil functions.

Materials and methods

Mice and infections

Ethics statement.

Regnase-1^–/–^ mice on C57BL/6 background [23] were backcrossed more than six times to BALB/c background [29], and Regnase-1^+^^/–^ (BALB/c) mice were used in the present study. WT (BALB/c) mice were purchased from CLEA Japan and bred at the Kyoto University animal facility for at least two weeks before being used in the experiments. All the mice used were maintained under SPF conditions at the animal facility of the Institute for Life and Medical Sciences, Kyoto University. Twelve-week-old female mice were used for MA10 infection experiments. Mice were infected intranasally with 10^2^ TCID_50_/50μl or 10^3^ TCID_50_/50μl/ MA10 in an approved animal biosafety level 3 facility in Osaka University following regulations. Mouse experiments were approved by the animal experimentation committees of Graduate School of Medicine, Kyoto University and Research Institute for Microbial Diseases, Osaka University, and conducted according to institutional guidelines. Virus-infected mice were euthanized if they lost more than 25% of their initial body weight. Lung tissue and peripheral blood (serum or cells) were collected at the indicated time points for analysis.

Viruses.

SARS-CoV-2 MA10 was established by a reverse genetics method called CPER (circular polymerase extension reaction) [57] using SARS-CoV-2 strain 2019-nCoV_Japan_TY_WK-5212020 (isolated by NIID) as the backbone. MA10 contains seven mutations, including nsp4: T295I, nsp7: K2R, nsp8: E23G, S: Q493K/Q498Y, P499T, and orf6: F7S, which have been reported as adaptive mutations introduced in SARS-CoV-2 during serial passages in BALB/c mice by Leist et al [28]. All viruses were amplified in VeroE6/TMPRSS2 cells, and the culture supernatants were harvested and titers were measured using the TCID_50_ assay and stored at -80℃ until use. MA10 preparation was done in the biosafety level 3 facility in Research Institute for Microbial Diseases, Osaka University.

Viral titration of SARS-CoV-2 MA10.

Viral titers were measured by median tissue culture infectious dose (TCID_50_) assays. VeroE6/TMPRSS2 cells (Japanese Collection of Research Bioresources, JCRB1819) were cultured in MEM supplemented with 5% FBS and 1% penicillin/streptomycin and seeded into 96-well cell plates. Samples were serially diluted 10-fold 10^-1^ to 10^-8^ in the cell culture medium. Dilutions were plated in triplicate on VeroE6/TMPRSS2 cells and incubated at 37℃ for 72 hours. Cytopathic effects were evaluated by microscopy. TCID_50_/ml was calculated by the Reed-Muench method.

Plasmid construction

Expression plasmids for murine Reg1 and the D141N mutant were as described previously [23]. The 3’ UTR sequence of murine Tsc22d3 (796–2135 of NM_001077364.2) was purchased from Thermo Fisher Scientific GeneArt, and was inserted in the pGL3-promoter vector (Promega).

Luciferase assay

HEK293T cells were transfected with luciferase reporter plasmid pGL3 containing the 3’ UTR sequence of murine Tsc22d3 together with expression plasmids for murine Reg1 (WT or D141N) or empty (control) plasmid, and the Relilla luciferase plasmid. Cells were lysed and luciferase activity in the lysates was determined using the Dual-Luciferase Reporter Assay System (Promega) 24 hours after transfection.

Cell preparation

Lungs were harvested, minced with scissors, and digested in 50μg/ml Liberase (Roche) and 100μg/ml DNase Ⅰ (Roche) containing RPMI medium for 40 minutes at 37℃. Then EDTA was added to a final concentration of 10mM and incubated at 37℃ for 10 minutes. After digestion, the digested lung tissue was homogenized using a gentle MACS (Miltenyi Biotec) and filtered through a 40 μm cell strainer, and the cells were collected.

Lung-infiltrating neutrophils were isolated using the Neutrophil Isolation Kit (Miltenyi Biotec) with a modification to enhance the depletion of non-neutrophil populations. Briefly, cell suspensions were incubated with biotin-conjugated anti-CD3ε (145-2C11) and anti-CD19 (6D5) antibodies prior to the labeling with anti-biotin Microbeads followed by depletion of non-neutrophils according to the manufacturer’s protocol. The purity of isolated neutrophils was confirmed by flow cytometry and consistently exceeded 95%.

Flow cytometric analysis

Cells were incubated with anti-CD16/32 antibody (Biolegend) for 15 minutes on ice to block non-specific Fc receptor binding, and then incubated with fluorochrome-conjugated following antibodies for 20 minutes at 4℃. CD11b (M1/70), CD19 (6D5), CX3CR1 (SA011F11), Ly6C (HK1.4), Ly6G (1A8), TCRβ (H57-597), TCRγδ (GL3), B220 (RA3-6B2), CD8 (53-6.7), CD49b (DX5), and NK1.1 (PK136) antibodies purchased from Biolegend, and CCR2 (475301), CD11c (HL3), CD40 (3/23), CD45 (30-F11), CD64 (X54-5/7.1), MHC class Ⅱ (I-A/I-E) (M5/114.15.2), SiglecF (E50-2440), CD4 (RM4–5), and CD19 (1D3) antibodies purchased from BD Biosciences were used. Dead cells were excluded from the analysis by using a Live/Dead cell stain kit purchased from Invitrogen (eBioscience). Cells were then fixed with fixation buffer (Biolegend) for 20 minutes at room temperature and subsequently transferred out of the BSL3 facility. Data were acquired on a FACSAria Fusion (BD Biosciences) or LSRFortessa (BD Biosciences), and analyzed using FlowJo software (BD Biosciences). The gating strategy for each cell fraction is shown in S9 Fig.

Histological analysis

Lung tissue was collected and fixed in 10% neutral buffered formalin. Fixed tissues were embedded in paraffin and sectioned at 4μm thickness, and stained with hematoxylin and eosin. Pathological evaluation was performed in a blind manner using three randomly selected 400 × fields per tissue, and scored according to the DAD histologic scoring systems. The DAD scores [28] were determined as follows: 1; absence of cellular sloughing and necrosis, 2; uncommon solitary cellular sloughing and necrosis, 3; multifocal cellular sloughing and necrosis with uncommon septal wall hyalinization, 4; multifocal (>75% of field) cellular sloughing and necrosis with common and/or prominent hyaline membranes. The average score across three fields was calculated for each sample. All samples were imaged using an all-in-one fluorescence microscope (BZ-X 710) (Keyence) and analyzed with NDP Viewer software (NDP.view2) (Hamamatsu).

RNA isolation and Quantitative RT-PCR

Total RNA was extracted using TRIzol Reagent (Life technologies) and Directzol (Zymo Research) according to the manufacturer’s protocol. cDNA was synthesized using SuperScript VILO (Invitrogen). Quantitative RT-PCR was performed using PowerUp SYBR Green Master Mix (Applied Biosystems) for the following genes: SARS-CoV-2 Nucloecapsid phosphoprotein (N1, N2 region selected by Centers for Disease Control and Prevention) and ORF1b, Ifna, Ifnb1, Il1b, Il12b, Il6, Tnf, Ccl5, Cxcl1, Cxcl2, Zc3h12a (Reg1), Tsc22d3, Nr2f2, Ax1, Muc4, and Fgfr1, and Hprt as a control gene. Target gene expression was calculated by the comparative method for relative quantification after normalization to Hprt expression. Primer sequences for the quantitative RT-PCR are listed in the S2 Table.

RNA-sequencing

Lung infiltrating neutrophils were isolated using the Neutrophil Isolation Kit (Miltenyi Biotec) as described above, and total RNA was isolated from neutrophils using TRIzol reagent (Life technologies) and Directzol (Zymo Research). For the library preparation, NEBNext Ultra Ⅱ Directional RNA Library Prep Kit for Illumina (NEB) was used and RNA-sequencing was performed on an Illumina NextSeq 500 system. Bioinformatic analysis of RNA-sequencing data was performed as described previously [58]. Briefly, the sequences were mapped to the mouse genome version GRCm39 using Hisat2 (version 2.1.0) [59]. Differential gene expression analyses were performed using DESeq2 package in R (version 1.20.0) [60] on read counts obtained by FeatureCounts (version 1.6.4) [61] were used for GSEA [62]. GSEA was performed with GSEA software (v.4.3.3) using the response to type I interferon mouse gene set (GO:0034340), the ranked list file, and the enrichment plot, normalized enrichment score (NES), and false discovery rate (FDR) were generated.

ScRNA-sequencing

Lung infiltrating CD45^+^ cells were isolated using the CD45 microbeads (Miltenyi Biotec). Samples collected from three mice for both WT or Reg1^+^^/–^ mice were combined into one sample for subsequent processing in each time point after MA10 infection. The scRNA-seq analysis was performed using the Chromium Fixed RNA Profiling Kit (10X genomics) according to the manufacturer’s instructions. Briefly, cells were fixed with 1 ml of fixation buffer and stored at 4℃ for 23.5 h. After removing the supernatant and adding 1 ml of quenching buffer, 100 μl of enhancer, and 275 μl of 50% glycerol, the Chromium Fixed RNA Profiling protocols were followed. For library preparation, Chromium Fixed RNA Kit, Mouse Transcriptome (10X genomics) was used and sequencing was performed on an Illumina NovaSeq 6000 system. Analysis of the resulting scRNA-seq data was conducted using the Seurat R package (version 5.0.1) [63]. Data of each sample was read in and 52,794 cells with more than 200 detected genes and less than 10% mitochondrial reads were retained. Cells were processed using a default workflow, including normalization (NormalizeData function), prediction of variable genes (FindVariableFeatures function), scaling (ScaleData), followed by dimensionality reduction using principal component analysis (PCA) and UMAP using the first 10 principal components (PCs). This revealed the presence of batch effects, which we removed using the Harmony R package (version 1.2.0), using the first 10 PCs as input [64]. The first 10 dimensions returned by Harmony were used for dimensionality reduction using UMAP, and clustering using functions FindNeighbors and FindClusters (using resolution = 1.2). This resulted in 22 clusters, which we annotated manually using the expression of known marker genes of several immune cell types. Marker genes used are listed in S3 Table.

For the further analysis of neutrophils, we first selected cells of clusters 9, 13, and 21 (Fig 3A). On these cells, we recalculated highly variable genes, ran PCA followed by Harmony and UMAP, and obtained 8 clusters of cells (S6A Fig). However, inspection of marker genes suggested that clusters 1, 6, 7, and 8 consisted of doublets of neutrophils and B, T, or NK cells. To confirm this, we generated 60 artificial doublet cells by merging the transcription data of 60 randomly sampled neutrophil cells (from clusters 9, 13, or 21) with 20 randomly sampled B cells (from cluster 2), 20 randomly sampled T cells (from clusters 1, 6, or 7), and 20 randomly sampled NK cells (from cluster 3). We added these 60 artificial doublets to the neutrophil cells and repeating the original processing steps, resulting in the UMAP plot of S6B Fig. Artificial doublets accumulated clearly inside 3 clusters here, while a large cluster contained relatively few artificial doublets, corresponding to the neutrophil clusters 2, 3, 4, and 5. Therefore, for the remainder of the neutrophil analysis, we focused on neutrophil clusters 2, 3, 4, and 5. Genes with differential expression between genotypes, clusters, or dpi were predicted using the FindMarkers function using default parameters.

In vivo neutrophil depletion experiments

Ultra-LEAF purified anti Ly6G (clone 1A8) was purchased from Biolegend. The procedure was performed as previously reported [65]. In brief, Anti-Ly6G (100 μg in 100 μl PBS) was injected intraperitoneally from 2 days prior to MA10 infection to 4 dpi daily. Alternatively, an equivalent volume of PBS was injected as the control.

Statistical analysis

Two-tailed Student’s t-test was used for most statistical analyses (GraphPad Prism), and a P value < 0.05 was considered statistically significant. Body weight after MA10 infection between each group was analyzed by two-way ANOVA with Bonferroni’s post-test.

Supporting information

S1 FigReg1^+/–^ mice show no significant alteration in the absolute number of lung immune cells.A) The absolute numbers of the indicated cell types obtained from the lungs of 12-week-old female BALB/c WT and Reg1^+^^/–^ mice (n = 3 per group) at a steady state. Absolute numbers were calculated from the number of cells counted after the single cell suspension from the lung and the percentage of each cell type observed by flow cytometry. The bar graphs show the mean ± s.d. P-values are shown. (two-tailed unpaired Student’s t-test) B) Representative H&E-stained lung sections (200 × magnification) from intranasally PBS administrated mice collected at 28 days after administration.(EPS)

S2 FigCytokine and chemokine levels in WT and Reg1^+/–^ mice following 10^2^ TCID₅₀ MA10 infection.A) qPCR of Ifnb1, Il12b, Tnf, Ccl5, Cxcl1 and Cxcl2 mRNA expression in RNA extracts obtained from infected lungs at 3 and 6 dpi. The bar graphs show the mean ± s.d. (n = 4). P-values are shown. (two-tailed unpaired Student’s t-test).(EPS)

S3 FigUnder MA10 infection, the Ly6G antibody increased the proportion of Ly6C^high^CCR2^+^ monocytes.A) Flow cytometry analysis of Ly6C^high^CCR2^+^ monocytes in CD45^+^ live cells isolated from infected lung at 5 dpi.(EPS)

S4 FigReg1^+/–^ mice exhibit significant amelioration of pneumonia in 10^3^TCID₅₀ MA10 infection.A) The absolute numbers of the indicated cell types infiltrating infected lungs on 3 dpi (n = 3 per group) and 6 dpi (n = 3 for WT, n = 4 for Reg1^+^^/–^) by flow cytometry. B) qPCR of MA10 viral nucleocapsid phosphoprotein (N1, N2 region) and ORF1b RNA expression in lungs of infected mice collected 3 dpi (n = 3 per group) and 6 dpi (n = 3 for WT, n = 4 for Reg1 ^+/–^). C) qPCR of Ifna, Il1b and Il6 mRNA expression in RNA extracts obtained from infected lungs at 3 dpi (n = 3 per group) and 6 dpi (n = 3 for WT, n = 4 for Reg1^+^^/–^). The bar graphs (A-C) show the mean ± s.d. (n = 3 or 4). P-values are shown. (two-tailed unpaired Student’s t-test).(EPS)

S5 FigscRNA-seq revealed that, among innate immune cells in the lungs following infection with 10^3^ TCID_50_ MA10, neutrophils highly express *Reg1.*A) The expression of Reg1 in the indicated lung cell types on 3 and 5 dpi in WT mice, as shown in Fig 3B, is presented.(EPS)

S6 FigscRNA-seq analysis identified doublets of neutrophils and other cell types.A) UMAP representation of lung infiltrating neutrophil clusters (cluster 9, 13 and 21 in Fig 3A). B) Neutrophils plotted in Fig 3A as clusters 9, 13, and 21 in UMAP were extracted and further classified into 8 clusters as shown in (A). Artificial doublets of 60 T cells (or B cells, or NK cells) and neutrophils were created and plotted in UMAP. The results revealed that clusters 2, 3, 4, and 5 consisted of true neutrophils, while the remaining clusters consisted of doublets. C) Dot plot of neutrophil developmental status enhancing genes in the indicated neutrophil clusters (UMAP from Fig 3C). D) Dot plot of neutrophil developmental status related transcriptional factors in the indicated neutrophil clusters (UMAP from Fig 3C).(EPS)

S7 FigscRNA-seq reveals that, at 3 dpi, cluster 3 neutrophils in Reg1^+/–^ mice express lower level of ISGs than WT mice.A) Heatmap of the gene sets whose expression was found to be upregulated after MA10 infection in Cluster 3 neutrophils. Gene expression is further segmented at the indicated time points (0dpi, 3dpi and 5dpi) obtained from WT and Reg1^+^^/–^ mice.(EPS)

S8 FigExpression levels of candidate Reg1-target genes in Reg1^+/–^ lung neutrophils.A) qPCR of Nr2f2, Axl, Muc4, and Fgfr1 mRNA expression in RNA extracts of lung infiltrating neutrophils at 0 dpi obtained from WT and Reg1^+^^/–^ mice (n = 3 per group). The bar graphs show the mean ± s.d. P-values are shown. (two-tailed unpaired Student’s t-test).(EPS)

S9 FigGating strategies for flow cytometry analysis.A, B) Gating strategies used to analyze alveolar macrophages, dendritic cells, Ly6C^-^CCR2^-^ monocytes, Ly6C^+^CCR2^-^ monocytes, Ly6C^high^CCR2^+^ monocytes, interstitial macrophages, neutrophils, eosinophils, B cells, CD4^+^ T cells, CD8^+^ T cells and NK cells from BALB/c WT or Reg1^+^^/–^ mice presented on Fig 1E. The same strategy was used to analyze those cells in MA10 infected BALB/c WT or Reg1^+^^/–^ mice presented on Figs 2A, 2C, 5C, S1A and S4A.(EPS)

S1 TableList of differentially expressed genes in the Reg1^+/–^ vs. WT comparison across various cell types identified by scRNA-seq.Based on the scRNA-seq results, we summed the total reads for each cell type across 0, 3 and 5 dpi and compared gene expression between WT and Reg1^+^^/–^ mice. The table lists genes that were upregulated in Reg1^+^^/–^ mice, ranked by fold change. T cells represent the sum of clusters 1, 6, 7, and 12 in Fig 3A. B cells represent the sum of clusters 2 and 22 in Fig 3A. NK cells represent cluster 3 in Fig 3A. Monocytes/macrophages are the sum of clusters 4, 8, 11, 15 and 20 in Fig 3A. Alveolar macrophages represent the sum of clusters 5 and 14 in Fig 3A. pDCs represent cluster 10 in Fig 3A.(XLSX)

S2 TableList of primers used for qPCR.List of primers for qPCR in Figs 1F, 1G, 2A, 5B, 5C, S2A, S4B, S4C and S8A.(EPS)

S3 TableList of marker genes used for the immune cell type annotation.List of marker genes for the immune cell type annotation in Figs 3A, 3B, 3C, 3D, 4A, 4B, 4C, 4D, S5A, S5B, S6A, S6B, S6C, S6D and S7A.(EPS)

S1 DataAll raw values for all graphs represented in figures.(XLSX)