Sexual dimorphism of COVID-19 inspires drug repositioning and host-targeting immunotherapy for viral pneumonia

Abstract

Sexual hormones play an important role in modulating disease outcome of COVID-19. The interplay between viral replication, host immune responses, pathology process and sexual hormone levels are complicated, and the underlying mechanisms remain exclusive. Here, we reveal the dose-dependent manner and multi-faceted role of the male hormone testosterone in hamster model of COVID-19 by evaluations of manifestations including survival rate, body weight loss, viral load, immune responses and lung injury. Both low and high doses of testosterone treatment cause more severe illness in male hamsters. Low dose of testosterone is beneficial for female hamsters, but high dose is harmful. Therefore, we evaluate the therapeutic effect of the testosterone inhibitor finasteride in male hamsters and demonstrate that it is sufficient to prevent death and severe pneumonia caused by different SARS-CoV-2 strains. Moreover, pulmonary transcriptome data reveals key clues for the mechanisms of testosterone-mediated disease enhancement and finasteride therapy.

Introduction

Evolutionarily conserved sex differences in several aspects of physiological structure and function result in sexual dimorphism in disease progression and outcome, particularly in the infection and pathology of highly prevalent viruses such as human immunodeficiency virus (HIV), cytomegalovirus (CMV), hepatitis B (HBV), influenza virus, and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).^1–7 In both human patients and animal models, the males tend to have heavier viral loads, more severe symptoms, and higher mortality rates than the females. Sex differences in cellular susceptibility and immune response to certain viruses are thought to be important factors in determining the severity of infectious disease,^8–12 inspiring new approaches to the development of host-target biological therapy. However, the underlying mechanisms and key contributors to the sexual dimorphism of viral infection and the complex virus-host interplay are not fully understood.

Sex hormones, particularly testosterone (TET) and estradiol (E2), are important biological factors that affect sex differences in both innate and adaptive immune responses to virus infection. As the primary androgen, TET is an important biomarker and pathogenic factor to the male-biased course of the infectious disease. For instance, elevated serum TET concentration increases risk of hepatocellular carcinoma in men with chronic HBV infection.¹³ H7N9 influenza virus infection in men is associated with reduced serum TET levels.¹⁴ However, the role of TET in SARS-CoV-2 infection remains unclear. In several small-scale clinical studies, low serum TET concentration is positively associated with a poor prognosis in the coronavirus disease 2019 (COVID-19) patients with SARS-CoV-2 infection.^15–17 In contrary, Samuel et al. provided strong clinical evidence that bio-available TET and androgen signaling serve as deteriorators to increase susceptibility and disease severity of SARS-CoV-2 infection.¹⁸ Mechanistically, TET and its bioavailable form dihydrotestosterone (DHT) bind to androgen receptor (AR) and upregulate transcription of SARS-CoV-2 receptor angiotensin converting enzyme 2 (ACE2) and co-receptor transmembrane serine protease 2 (TMPRSS2), thereby enhancing viral entry and establishing robust infection.^18–21 In addition, similar negative effects of TET on SARS-CoV-2 infection have been confirmed by two other research groups.^20,22 These elusive results highlight the multifaceted roles of TET and androgen signaling in the process of SARS-CoV-2 infection and pathology, and are thought to be a key point to improve our fundamental understanding of the male sex bias of COVID-19.

To clarify this issue, we employed a high-simulation hamster model of COVID-19, which exhibits viral replication kinetics and pathological features in the respiratory tract organs that consistent with human COVID-19 patients,^23–25 and are widely used in the development of vaccines and drugs.^26–28 We have delineated the male sex bias of SARS-CoV-2 infection in hamster model.²⁹ Compared to female controls, male hamsters showed more body weight loss, prolonged viral replication and shedding in the respiratory tract organs, and severe lung injury. In this study, we analyze the relationship between serum TET concentration and disease outcomes in male and female hamsters with SARS-CoV-2 infection. Castration of male hamsters and dose gradient treatment with TET in both sexes of hamsters are performed for a comprehensive function evaluation by detecting disease manifestations, including survival rate, body weight loss, pathological changes in the lungs, viral load in respiratory tract organs, disturbance of inflammation, and innate immune response-associated genes. Since TET has been shown to be a deteriorator of COVID-19 in the male hamster model, we tested the therapeutic efficacy of the FDA-approved TET inhibitor finasteride (FINA) and investigated underlying mechanisms by analysis the bulk and single cell transcriptome data of lung tissues and performed validation experiments. Overall, inspired by the sexual dimorphism of COVID-19, we seek to validate the host-targeting strategy of repositioning anti-androgenic drugs for viral pneumonia therapy.

Results

Serum testosterone concentration is closely associated with the disease outcome of COVID-19 in male and female hamsters

Firstly, 8-week male and female hamsters were intranasally inoculated with 1 × 10⁴ plaque forming unit (PFU) of SARS-CoV-2 prototype, Beta, Delta and XBB.1.9.2.1 variants, respectively. Generally, the male hamsters showed higher mortality, more body weight loss and higher viral RNA load in lung tissues than the female hamsters from 0 to 9 days post infection (dpi) of all the tested SARS-CoV-2 strains (Supplementary Fig. S1). The recovered body weight of the surviving hamsters at 9 dpi indicated an alleviation of acute COVID-19 (Supplementary Fig. S1). In a parallel experiment, six male and female hamsters were intranasally inoculated with 1 × 10⁴ PFU of prototype SARS-CoV-2, and were euthanized at 0 dpi, 3 dpi, 5 dpi, 7 dpi, and 9 dpi, respectively. Serum TET concentration, body weight loss and viral RNA levels in lung tissues were detected at each time point (Fig. 1a). In the male hamsters, approximately 2- to 3-fold increase of serum TET concentration were detected from 0 to 9 dpi, which was negatively related to the percentage of body weight loss (Fig. 1b) and was positively related to the viral RNA levels in lung tissues (Fig. 1c). Although the female hamsters showed approximately 10-fold lower baseline of serum TET concentration than the male hamsters at 0 dpi, a significant increase of serum TET concentration was detectable from 0 to 9 dpi (Fig. 1d). Contrary to male hamsters, serum TET concentration was positively related to the percentage of body weight loss (Fig. 1e), and was negatively related to the viral RNA levels in lung tissues (Fig. 1f) in the female hamsters. Similar to the prototype SARS-CoV-2, the relationship between serum TET concentration and key disease parameters was also observed in the male and female hamsters that infected with the Beta, Delta, and XBB.1.9.2.1 variants, respectively (Supplementary Fig. S2).

Fig. 1.

Serum TET concentration is closely associated with the disease outcome of SARS-CoV-2 infection in hamsters. Male and female hamsters were intranasally inoculated with 1 × 10⁴ PFU of prototype SARS-CoV-2. For sample collection, hamsters were euthanized at 0, 3, 5, 7, and 9 dpi, respectively. TET concentration in the serum samples of (a) male and (b) female hamsters was measured by ELISA (n = 6). Percentage of body weight changes at indicated time points were recorded and viral RNA levels in lung tissues were measured by RT-PCR (n = 6). The linear relationship between testosterone concentration in the serum and (b, e) percentage of body weight changes, and (c, f) viral RNA levels in lung tissues was shown. g TET concentration in the serum samples of castrated hamsters and intact controls was measured by ELISA (n = 6). Castrated and intact male hamsters (controls) were intranasally inoculated with 1 × 10⁴ PFU of prototype SARS-CoV-2. h Percentage of body weight changes were recorded from 0 to 7 dpi (n = 6). i Viral RNA levels in tissue homogenates of lungs collected from prototype SARS-CoV-2 infected hamsters were measured by RT-PCR (n = 6). j Representative H&E staining images of lung tissues collected from castrated hamsters and controls sacrificed at 7 dpi were presented (Bar = 2 mm). Lung images for all of the hamsters were shown in Fig. S4. k Comprehensive pathological scores for lung sections were determined based on the severity and percentage of injured areas of each lung lobe. For each group, 24 lung lobes were collected from six individual hamsters and were scored. Lung pathological scores for all of the hamsters were shown in Table S1. One-way ANOVA followed by Dunnett’s multiple-comparisons test comparing each time point to baseline (0 dpi) for (a, d); simple linear regression for (b, c, e, f); two-way ANOVA with Bonferroni’s post hoc test for planned pairwise comparisons (g) and for between-group comparisons at individual time points across 1–7 dpi for (h); and unpaired two-tailed Student’s t-test for (i, k). P-values < 0.01 were considered significant: ^*P < 0.01, ^**P < 0.001, ^***P < 0.0001, ns indicates no significance

To confirm the causality of male sex bias, serum TET concentration and disease severity in male hamsters of acute COVID-19, six male hamsters were castrated and intranasally inoculated with 1 × 10⁴ PFU of prototype SARS-CoV-2, Beta, Delta and XBB.1.9.2.1 variants, respectively. Compared to the intact controls, the castrated hamsters showed approximately 3-fold lower serum TET concentration (Fig. 1g) and rescued body weight loss (Fig. 1h). Furthermore, we anatomized all of the castrated hamsters and intact controls at 7 dpi and obtained their lungs for analysis of viral load and lung injury. The castrated hamsters showed over 10-fold decrease of viral RNA levels in lung tissues than the intact controls (Fig. 1i). Alleviated disease severity was also observed in the castrated hamsters that intranasally inoculated with 1 × 10⁴ PFU of Beta, Delta, and XBB.1.9.2.1 variants of SARS-CoV-2, respectively (Supplementary Fig. S3). The results of hematoxylin and eosin (H&E) and Masson Trichrome staining for the lung lobes collected from intact controls showed typical features of severe pneumonia, including pulmonary consolidation and alveolar destruction, diffusive inflammation, protein-rich fluid exudate, hyaline membrane formation and severe pulmonary hemorrhage (Fig. 1j and Supplementary Figs. S4 and S5). In contrast, critical lung injury was attenuated in the castrated hamsters (Fig. 1j and Supplementary Figs. S4 and S5). The severity of lung pathogenesis was quantified by comprehensive pathological score based on alveolar septum thickening and consolidation, hemorrhage, exudation, pulmonary edema, and mucous, recruitment and infiltration of inflammatory cells among all the hamster lung lobes. As expected, the castrated hamsters showed a lower lung pathological score than intact controls (Fig. 1k and Supplementary Table S1). In addition, low serum TET concentration is associated with Long COVID at 6 weeks post infection of prototype SARS-CoV-2 in both male and female hamsters, which indicating the multifaceted nature of TET in different course of COVID-19 (Supplementary Fig. S6). Overall, the above results revealed that serum TET concentration is closely associated with diseases outcome of COVID-19 in male and female hamsters.

The disease outcome of extrinsic TET treatment is dose-dependent in male and female hamsters

As castration was able to decrease serum TET concentration and alleviate the severity of COVID-19 in male hamsters, we further investigated the disease outcomes of SARS-CoV-2-infected male and female hamsters with extrinsic TET treatment (Fig. 2a). The prototype SARS-CoV-2-infected hamsters were daily treated with different doses of TET (high dose: TET-H, 1 mg/kg; middle dose: TET-M, 0.2 mg/kg; low dose: TET-L, 0.04 mg/kg per dose) via intraperitoneal injection from 0 to 5 dpi. The prototype SARS-CoV-2-infected hamster without extrinsic TET treatment were set as control group. Physical and health examinations had been undertaken for 7 days by recording their survival rate and percentage of body weight changes. In the male hamsters, SARS-CoV-2 infection caused one death out of 16 hamsters, whereas, two, five and seven out of 16 hamsters deceased in the TET-L, TET-M and TET-H groups respectively (Fig. 2b). The control male hamsters and those in TET-L, TET-M and TET-H groups exhibited the mean body weight loss of 11.2 ± 2.3%, 16.6 ± 2.5%, 21.3 ± 1.7% and 23.4 ± 2.3% at 7 dpi, respectively (Fig. 2c). In the female hamsters, only high dose of extrinsic TET treatment (TET-H) killed three out of 16 hamsters (Fig. 2d). The control female hamsters and those in TET-L, TET-M and TET-H groups exhibited the mean body weight loss of 5.9 ± 3.0%, 3.5 ± 2.4%, 8.7 ± 1.8% and 11.8 ± 2.6% at 7 dpi, respectively (Fig. 2e). To evaluate the severity of lung injury and viral load in respiratory tract organs, all the survived hamsters were euthanized at 7 dpi. The results of H&E and Masson Trichrome staining for lung lobes and comprehensive pathological score of the male hamsters showed that extrinsic TET treatment caused a dose-dependent deterioration of lung pathological changes after SARS-CoV-2 infection (Fig. 2f, g, Supplementary Figs. S7–8 and Supplementary Table S2). In the female hamsters, low dose of extrinsic TET treatment relieved lung pathological changes, whereas middle and high doses of extrinsic TET treatment resulted in more severe lung injury (Fig. 2f, g, Supplementary Figs. S9–10 and Supplementary Table S3).

Fig. 2.

TET treatment affects the disease severity of SARS-CoV-2 infection in male and female hamsters. a Schematic diagram of SARS-CoV-2 infection and animal operations. Male and female hamsters were intranasally inoculated with 1 × 10⁴ PFU of prototype SARS-CoV-2, and then received different doses of intraperitoneal injection of testosterone (TET) from 0 to 5 dpi (once per day; TET-H, 1 mg/kg per dose; TET-M, 0.2 mg/kg per dose; TET-L, 0.04 mg/kg per dose). Age-paired male and female SARS-CoV-2-infected hamsters without treatment were set as control groups. b, d Survival rate (n = 16) and c, e percentage of body weight changes (n = 6) were daily recorded from 0 to 7 dpi. The survived hamsters were euthanized at 7 dpi for pathological, virological, and immunological analysis. f Representative H&E staining images of lung tissues collected from hamsters sacrificed at 7 dpi were presented (Bar = 2 mm). Lung images for all of the hamsters were shown in Figs. S7–10. g Comprehensive pathological scores for lung sections were determined based on the severity and percentage of injured areas of each lung lobe. For each group, 24 lung lobes were collected from six individual hamsters and were scored. Lung pathological scores for all of the hamsters were shown in Tables S2 and S3. Two-way ANOVA with Dunnett’s post hoc test for between-group comparisons at individual time points across 1–7 dpi for (c, e); one-way ANOVA followed by Dunnett’s multiple-comparisons test for (g). P-values < 0.01 were considered significant: ^*P < 0.01, ^**P < 0.001, ^***P < 0.0001, ns indicates no significance

Excessive release of inflammatory cytokines, abnormal perturbation of antiviral innate immune response and high viral load in respiratory tract organs are considered as typical manifestations of critical COVID-19 in both human patients and animal models. Therefore, we evaluated the mRNA levels of inflammatory cytokines, including interleukin-6 (IL-6), interferon gamma (IFN-γ), and tumor necrosis factor alpha (TNF-α), as well as interferon alpha (IFN-α), interferon-stimulated gene 15 (ISG15), and myxovirus resistance 1 (MX1) in lung tissues collected at 7 dpi. In the male hamsters, TET treatment increased the mRNA levels of IFN-γ, IL-6 and TNF-α (Fig. 3a–c). Although the mRNA levels of IFN-α were not significantly changed, middle and high doses of TET treatment down-regulated the mRNA levels of ISG15 and MX1 (Fig. 3d–f). In the female hamsters, low dose of TET treatment decreased the mRNA levels of IL-6, IFN-γ, and TNF-α, whereas high dose TET treatment showed an opposite effect on these inflammatory cytokines (Fig. 3g–i). Furthermore, the female hamsters with high dose TET treatment showed a declined up-regulation of IFN-α, ISG15, and MX1 mRNA levels than those with low dose TET treatment (Fig. 3j–l).

Fig. 3.

Fold changes of inflammation and innate immune response-associated genes in lung tissues of SARS-CoV-2-infected male and female hamsters after TET treatment. The mRNA levels of inflammatory cytokines include a IFN-γ, b IL-6, and c TNF-α in the lung tissues that collected from the male hamsters at 7 dpi and g–i those collected from the female hamsters at 7 dpi were measured by RT-PCR to determine the fold changes (n = 6). The mRNA levels of (d) IFN-α and representative ISGs include (e) ISG15 and (f) MX1 in the lung tissues collected from the male hamsters at 7 dpi and (j-l) those collected from the female hamsters at 7 dpi were also measured by RT-PCR to determine the fold changes (n = 6). These mRNA levels were normalized to the housekeeping gene γ-actin. One-way ANOVA followed by Dunnett’s multiple-comparisons test was used to compare each TET group with the control group. P-values < 0.01 were considered significant: ^*P < 0.01, ^**P < 0.001, ^***P < 0.0001, ns indicates no significance

Afterwards, we asked whether extrinsic TET treatment could affect viral load. To that end, we analyzed viral mRNA levels in respiratory tract organs, including turbinate, trachea, and lung by RT-PCR that amplifies SARS-CoV-2 open reading frame 1ab (ORF1ab) for detection of viral RNA load in the homogenized tissues collected at 7 dpi. Compared to the control group, extrinsic TET treatment caused a dose-dependent increase of viral RNA levels in turbinate, trachea and lung of the male hamsters, respectively (Fig. 4a–c, left panel). The male hamsters in TET-H group showed 5- to 10-fold increase of viral RNA levels than the control group in all of the three detected organs. The female hamsters in TET-L and TET-M groups showed no significant change or decreased viral RNA levels in these respiratory tract organs than the control group (Fig. 4a–c, right panel). Whereas the female hamsters in TET-H group showed over 5-fold increase of viral RNA levels. In addition, the titers of live virus particles in the respiratory tract tissue samples showed a similar trend with the results of viral RNA levels (Fig. 4d–f). The results of immunohistochemistry staining for SARS-CoV-2 nucleocapsid protein in hamster lung tissues confirmed that TET treatment improves replication of viral load (Supplementary Fig. S11). In the primary human alveolar epithelial cell model, TET treatment enhanced the expression of SARS-CoV-2 receptor ACE2 and co-receptor TMRPSS2, which was consistent with the RT-PCR results of hamster lung tissues (Supplementary Fig. S11). These data suggested that TET deteriorates the severity of COVID-19 in the hamster model by arguments of pulmonary viral load. Similar results were also found in human lung epithelial cells, lung organoids, and mouse models.^18,20,22

Fig. 4.

Viral load in respiratory tract organs of SARS-CoV-2-infected male and female hamsters after TET treatment. Viral RNA levels in respiratory tract organ tissues, including a turbinate, b trachea, and c lung of SARS-CoV-2 infected male and female hamsters with TET treatment were collected from at 7 dpi and measured by RT-PCR (n = 6). The primers were used to amplify SARS-CoV-2 ORF1ab gene. d–f Viral titers of these samples were detected by a titration method of gradient dilution and cytopathic effect observation (n = 6). One-way ANOVA followed by Dunnett’s multiple-comparisons test was used to compare each TET group with the control group. P-values < 0.01 were considered significant: ^*P < 0.01, ^**P < 0.001, ^***P < 0.0001, ns indicates no significance

In addition, we also detected the concentrations for TET, DHT, and E2 in the serum samples collected at 7 dpi, respectively. The hamsters without SARS-CoV-2 infection or extrinsic TET treatment were set as mock group. In both of the male and female hamsters, extrinsic TET treatment corresponded to a dose-dependent increase of TET and DHT levels and a decrease of E2 levels (Fig. 5d–f). Remarkably, the female hamsters with moderate COVID-19 (control and TET-L groups) showed lower TET/DHT levels and/or higher E2 levels than those with middle to critical COVID-19. Taken together, extrinsic TET treatment usually causes deteriorating disease outcome in male hamsters with SARS-CoV-2 infection. Low dose extrinsic TET treatment is beneficial to the female hamsters, however, middle to high doses is harmful.

Fig. 5.

Sex hormone concentration in serum of SARS-CoV-2-infected male and female hamsters after TET treatment. The concentrations of (a, d) TET, (b, e) DHT and (c, f) E2 in the serum samples collected at 7 dpi were measured by ELISA, respectively (n = 6). One-way ANOVA followed by Dunnett’s multiple-comparisons test was used to compare each TET group with the control group. P-values < 0.01 were considered significant: ^*P < 0.01, ^**P < 0.001, ^***P < 0.0001, ns indicates no significance

Pulmonary transcriptome analysis reveals key clues for the pathological mechanisms of TET-associated disease enhancement in hamster model

Furthermore, we performed bulk and single-cell RNA sequencing to delineates the pathological signatures of TET treatment in male hamsters. Lung tissues were collected from mock and prototype SARS-CoV-2-infected male hamsters with or without 6-dose TET-H treatment were collected at 7 dpi, respectively. Principal component analysis (PCA) showed highly similar transcriptional profiles between the mock and mock-TET groups, whereas clear separation was observed between the COVID-19 and COVID-19-TET groups, indicating a TET-dependent transcriptional divergence pattern (Supplementary Fig. S12a). KEGG pathway enrichment revealed distinct functional alterations across groups. Prior to infection, calcium signaling was reduced in the mock-TET group compared with the mock group. Notably, the COVID-19-TET group displayed markedly elevated pathogenic signaling, including the IL-17, HIF-1α, and apoptosis, suggesting that TET might enhance cytokine storm priming and cell death (Supplementary Fig. S12b). To further investigate the alteration at single-cell resolution, we performed scRNA-seq and identified 17 cell subsets based on canonical marker genes (Supplementary Fig. S12c, d). Cell-type compositions differed substantially among the four groups (Supplementary Fig. S12e). In the mock and mock-TET groups, frequencies of the ciliated epithelial cells, monocytes, alveolar macrophages, and B cells were reduced by TET treatment, and these populations declined even further after infection (Supplementary Fig. S12f). In contrast, myeloid dendritic cells (mDCs), plasmacytoid dendritic cells (pDCs), and interstitial macrophages (IMs) exhibited greater post-infection expansion in the COVID-19-TET group (Supplementary Fig. S12f), indicating a TET-driven reshaping of the pulmonary immune landscape in acute COVID-19. Consistent with the results of animal experiments, pulmonary RNA-Seq analysis confirmed that TET exacerbates inflammation and reprograms immune cell dynamics during SARS-CoV-2 infection, provided key clues for the pathological mechanisms of TET-associated disease enhancement. Inhalation delivery of cocktail RNA interference (RNAi) against IL-17 and HIF-1α signaling reduced death, body weight loss, and severe pneumonia in prototype SARS-CoV-2-infected male and female hamsters with or without TET treatment (Supplementary Fig. S13). These validation experiments suggested that TET-associated disease enhancement is reversible.

FINA therapy ameliorated critical COVID-19 in male hamsters

Considering the disease outcomes of SARS-CoV-2 male hamsters with castration or extrinsic TET treatment, we further questioned whether the pharmacological target of TET could counteract the severity of COVID-19. FINA is a widely used TET inhibitor and has been approved by the FDA for the treatment of diseases that are associated with abnormally elevated TET. Therefore, male hamsters were intranasally inoculated with 1 × 10⁴ PFU of prototype SARS-CoV-2 and received sequential FINA therapy. The SARS-CoV-2-infected hamsters were orally administrated with 6-doses of FINA (once per day, 1 mg/kg per dose) from 0 to 5 dpi, or 3-doses of FINA from 3 to 5 dpi (Fig. 6a). The SARS-CoV-2-infected hamsters without FINA therapy were set as controls. Physical and health examinations were undertaken for 7 days to record percentage of body weight changes. All the SARS-CoV-2-infected hamsters with or without FINA therapy survived at 7 dpi. The control hamsters exhibited progressive mean body weight loss of up to 12.1 ± 3.7% from 0 to 7 dpi (Fig. 6b). Whereas, the hamsters with 3-doses and 6-doses of FINA therapy exhibited body weight loss of 8.1 ± 1.5% and 3.4 ± 1.7% at 7 dpi (Fig. 6b).

Fig. 6.

Evaluation for the therapeutic effect of FINA in the male hamsters with SARS-CoV-2 infection. a Schematic diagram of SARS-CoV-2 infection and animal operations. Male hamsters were intranasally inoculated with 1×10⁴ PFU of prototype SARS-CoV-2, and then received 6-dose oral administration of finasteride (FINA) from 0 to 5 dpi, or 3-dose FINA from 3 to 5 dpi. The infected hamsters without FINA therapy were set as controls. b Percentage of body weight changes of the hamsters were recorded from 0 to 7 dpi (n = 6). The survived animals were euthanized at 7 dpi for pathological analysis. c Representative H&E staining images of lung tissues collected from hamsters sacrificed at 7 dpi were presented (Bar = 2 mm). Lung images for all of the hamsters were shown in Figs. S14 and 15. d Comprehensive pathological scores for lung sections were determined based on the severity and percentage of injured areas of each lung lobe. For each group, 24 lung lobes were collected from six individual hamsters and were scored. Lung pathological scores for all of the hamsters were shown in Table S4. Fold changes for the mRNA levels of e inflammatory cytokines, f IFN-α and ISGs in lung tissues were detected by RT-PCR (n = 6). These mRNA levels were normalized to the housekeeping gene γ-actin. g Viral RNA levels in tissue homogenates of turbinate, trachea, and lung collected at 7 dpi were measured by RT-PCR (n = 6). h The concentrations of TET, DHT and E2 in the serum samples collected at 7 dpi were measured by ELISA (n = 6). Two-way ANOVA with Dunnett’s post hoc test for between-group comparisons at individual time points across 1–7 dpi for (b). One-way ANOVA followed by Dunnett’s multiple-comparisons test (each finasteride dose vs Control) for (d–h). P-values < 0.01 were considered significant: ^*P < 0.01, ^**P < 0.001, ^***P < 0.0001, ns indicates no significance

To evaluate the severity of lung injury and viral load in respiratory tract organs, all of the hamsters were euthanized at 7 dpi. The results of H&E and Masson Trichrome staining of lung lobes and comprehensive pathological scores showed that SARS-CoV-2-induced lung pathological changes were largely suppressed by 6-dose FINA therapy (Fig. 6c, d, Supplementary Figs. S14, 15 and Supplementary Table S4). Partial relief of lung injury was observed in the male hamsters with 3-dose FINA therapy. In a parallel experiment, reduced dosage of FINA (0.2 or 0.04 mg/kg per dose) cannot fully protect against the body weight loss and lung injury caused by prototype SARS-CoV-2 infection (Supplementary Fig. S16). Of note, FINA therapy resulted in a dose-dependent down-regulation of IL-6, IFN-γ, and TNF-α (Fig. 6e), as well as up-regulation of IFN-α, ISG15, and MX1 (Fig. 6f) mRNA levels in lung tissues. We further analyzed viral replication in respiratory tract organs, including turbinate, trachea, and lungs by RT-PCR that amplifies SARS-CoV-2 ORF1ab for the detection of viral RNA loads in the homogenized tissues collected at 7 dpi. In contrast of the hamsters without therapy, lower levels of viral RNA in turbinate, trachea, and lung were detected in the hamsters with 3-doses and 6-doses FINA therapy, but only the decreases in lungs were statistically significant (Fig. 6g). Although FINA therapy slightly increased serum TET levels, remarkable decreases of serum DHT levels were coupled with increase of E2 levels at 7 dpi (Fig. 6h). In addition, 6-doses FINA therapy (1 mg/kg per dose) can rescue male hamsters from lethal COVID-19 caused by the highly pathogenic Beta, Delta and XBB.1.9.2.1 variants of SARS-CoV-2 (Supplementary Figs. S17 and 18), suggesting that host-target therapy can largely overcome the challenge of rapid viral mutation and immune escape. Overall, these data demonstrated the potent and broad-spectrum therapeutic efficacy of FINA in male hamsters with SARS-CoV-2 infection.

Pulmonary transcriptome analysis of FINA therapy reveals key clues for the mechanisms of FINA therapy in hamster model

To investigate the underlying therapeutic mechanisms of FINA, we performed bulk and single-cell RNA sequencing on lung tissues of male hamsters with or without 6-dose FINA therapy (COVID-19 and FINA groups). Age-paired mock male hamsters without SARS-CoV-2 infection were set as health controls. PCA showed that samples under each group were well separated and the FINA group was closer to the mock group (Fig. 7a). SARS-CoV-2 infection resulted in down-regulation of 3090 genes and up-regulation of 3087 genes, however, 2127 and 2462 of them were reserved by FINA therapy, respectively (Fig. 7b). The results of Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Set Enrichment Analysis (GSEA) revealed that signaling pathways include neutrophil extracellular trap (NET) formation, p53, IL-17 and HIF-1 were up-regulated in the COVID-19 group and down-regulated by FINA therapy, whereas Wnt, Rap1 and calcium signaling pathways etc. were down-regulated in the COVID-19 group and up-regulated by FINA therapy (Fig. 7c). In addition, the regulatory network and heat map of differentially expressed genes (DEGs) in five representative biological process and signaling pathways were displayed (Fig. 7d and Supplementary Fig. S19).

Fig. 7.

Pulmonary transcriptome analysis of FINA therapy in male hamsters with SARS-CoV-2 infection. a Principal component analysis (PCA) for the transcriptome analysis of lung tissues. Health controls, SARS-CoV-2 infected hamsters with or without FINA therapy were colored in blue (Mock), red (COVID-19) and green (FINA), respectively (n = 5). b Number of DEGs in comparisons between the mock and COVID-19 groups, and between the COVID-19 and FINA groups. The pie chart displays the percentage of significantly up-regulated (in red) or down-regulated (in blue) genes in the overlapping DEGs. c Enrichment analysis based on KEGG database. The first column shows the up-regulated and down-regulated pathways that were significantly altered in overlapping DEGs based on enriched KEGG pathways, with the color intensity on behalf of the counts of genes. The second column shows the results of the GSEA. Pathways in which gene expression was increased or decreased are shown in red and blue, respectively. The color intensity is proportional to the NES calculated using the GSEA. d The diagram shows the changes and interactions of genes in the key signaling pathways after SARS-CoV-2 infection and FINA therapy. The intensity of the colors indicates average changes in genetic levels. In two columns, dot size is used to presented p value of the pathways, with only p value < 0.05 depicted here. DEGs differentially expressed genes, KEGG Kyoto Encyclopedia of Genes and Genomes, GSEA gene-set enrichment analysis, NES normalized enrichment score

Afterwards, the protein-protein interaction (PPI) network was performed to explore interactions between proteins regulated by the core genes. Molecular complex detection (MCODE) component identified 20 genes enriched in the Wnt and calcium signaling pathways within the network (Supplementary Fig. S20a). Ten hub genes were subsequently defined by overlapping six cytoHubba algorithms (Supplementary Fig. S20b, c), six of which were confirmed by scRNA-seq to be reversed by FINA treatment (Supplementary Fig. S20d). Notably, these hub genes, especially Fzd2, Wnt2, and Ntrk3, are predominantly expressed in fibroblasts, with Fzd2 additionally showing high expression in epithelial cells, indicating that FINA may counteract disease progression by modulating proliferation and growth programs primarily in fibrotic and epithelial compartments (Supplementary Fig. S20e).

Next, single-cell RNA sequencing was further performed to determine how FINA modulates cell fate alterations in response to SARS-CoV-2 infection. FINA treatment reversed the SARS-CoV-2-induced decrease in alveolar macrophages and monocytes, as well as the increase in IMs and dendritic cell populations (Supplementary Fig. S21a). At the level of cell-cell communication, SARS-CoV-2 infection primarily reshaped macrophage-centered interaction networks, particularly those involving IMs and their crosstalk with other pulmonary cell types (Supplementary Fig. S21b). These alterations were largely reversed by FINA treatment (Supplementary Fig. S21b). These findings suggest that FINA may counteract disease progression by modulating macrophage fate and their interactions with other cell types, prompting further investigation into the functional heterogeneity of macrophage subpopulations. Enrichment analysis of marker genes from the three macrophage subclusters revealed that IMs were uniquely characterized by prominent activation of inflammatory pathways, including the IL-17 signaling pathway and Chemokine signaling pathway, in contrast to the other subsets (Supplementary Fig. S21c). Collectively, these findings demonstrate that FINA alleviates SARS-CoV-2-induced lung injury by reshaping IM fate and dampening macrophage-centered inflammatory signaling. Overall, these findings indicate that FINA may alleviate SARS-CoV-2-induced lung injury mainly through reprogramming macrophage fate and attenuating inflammatory signaling associated with IMs.

Discussion

In the past decades, infectious diseases are one of the major threats to public health and human lifespan, which necessity the search for prophylactic and therapeutic countermeasures. More than 260 million chronic HBV carriers are at high risk of cirrhosis and liver cancer.³⁰ Seasonal flu causes millions of hospitalizations and severe pneumonia cases every year.³¹ The SARS-CoV-2 pandemic results in over 700 million cases of COVD-19 and global average mortality of 1%.³² However, the efficiency of current vaccines and antiviral drugs are challenged by both viral factors and host factors. In terms of virus factors, the emerging SARS-CoV-2 variants usually obtain increased transmissibility, cellular entry, and immune escape capacity, and thereby cause breakthrough infection among the vaccinated populations. In terms of host factors, the sexual dimorphism of the immune system might affect the host response after vaccination and long-term protection efficiency.³³ Furthermore, the therapeutic effects of the antiviral agent may also be influenced by sex differences.³⁴ In addition to the virus-target strategy, the development of a host-target strategy, especially for the use of sexual dimorphism, has become a new approach to combat infectious diseases.

The sexual hormones have been considered as the main biological basis of sexual dimorphism in the progression of infectious diseases. Here, we revealed the male sexual hormone TET is associated with a poor disease outcome of SARS-CoV-2 infection in male hamsters. However, the interplay between sex hormones and viral pathology is complicated and varies between viruses and hosts of different sexes. For example, H1N1 influenza virus infection reduces TET concentration in male mice, but not in female mice.³⁵ It is demonstrated that TET treatment can reduce the severity of H1N1 influenza virus infection in adult and aged male mice by mitigating pulmonary inflammation rather than by affecting viral replication.^36–38 TET suppressed the increase of pulmonary inflammatory monocytes, CD8-positive T cells, and eosinophils in H1N1-infected mice.³⁹ Neither depletion of eosinophils nor adoptive transfer of CD8-positive T cells could reverse the ability of TET to protect males against H1N1, suggesting these were secondary immunologic effects.³⁹ Conversely, SARS-CoV-2 infection resulted in increase of serum TET concentration in both male and female hamsters in our present study. In male hamsters with SARS-CoV-2 infection, serum TET concentrations are positively correlated with body weight loss and lung viral load, whereas female hamsters show the opposite linear relationship. Exogenous TET treatment showed dose-dependent disease exacerbation in the male hamsters with SARS-CoV-2 infection and castrated hamsters presenting milder symptoms, particular in the representative manifestations such as survival rate, body weight changes, viral load and lung pathogenic changes. Similar results were observed in a mouse model of Coxsackievirus B3 (CVB3) infection.⁴⁰ TET treatment enhanced viral dissemination in both male and female mice, and broadly increased proinflammatory cytokines and chemokines. The TET-treated CVB3-infected male mice with TET treatment had a higher mortality rate than those with TET depletion, suggesting TET and androgen signaling pathway might play roles of survivor or deteriorator in different infectious diseases. Interestingly, low-dose exogenous TET treatment can relieve the severity of COVID-19 in female hamsters, whereas high-dose exogenous TET treatment is harmful. Compared to the male hamsters, dose-specific disease outcomes and host responses of TET treatment in female hamsters highlighted the sexual dimorphism of COVID-19 and suggested that TET may play opposite roles under different host environments and concentrations. Therefore, modulation of serum TET concentration might determine the disease outcome of COVID-19 in the hamster model and other infectious diseases.

Drug repositioning has been demonstrated as an important strategy to combat the emerging outbreak of SARS-CoV-2.^41–45 Several studies indicated that anti-androgenic drugs, including Apalutamide, Darolutamide, ARD-61, Enzalutamide, Dutasteride, and FINA, can down-regulate the levels of SARS-CoV-2 receptor ACE2 and co-receptor TMPRSS2 in indicated cell types and suppress viral load of SARS-CoV-2.^18,20,22 However, the diverse antiviral effects of drugs in different in vitro models might hinder their further application. For instance, Enzalutamide can inhibit SARS-CoV-2 infection in LNCaP cells and human prostate organoids, but was ineffective in human lung organoids. Modulation of circulating androgen concentration and androgen signaling can alter cell susceptibility to virus and pulmonary immune microenvironment, but the in vivo therapeutic effects of these anti-androgenic drugs remain unclear.

In this study, FINA was shown to reduce the severity of the prototype SARS-CoV-2 infection, as well as the beta, delta, and XBB variants in the hamster model, highlighting the broad spectrum of host-targeting therapeutic strategies. Despite its antiviral effects in respiratory tract organs, our pulmonary transcriptome analysis delineated the multi-layered immune regulatory function of FINA. Remarkably, NET formation,^46–48 IL-17,^49–51 HIF-1,⁵² Wnt⁵³, and calcium^54–56 signaling pathways were demonstrated as important therapeutic targets of COVID-19, which were significantly changed in the lung tissues of SARS-CoV-2-infected hamsters and were reserved by FINA therapy. In addition, FINA therapy can reserve the abnormal disturbance of B cells, eosinophils, and neutrophils, but has little effect on the changes in pulmonary immune cell populations such as DCs, Tfh, Treg, and memory resting CD4 T cells.

Recently, Stanelle-Bertram et al. reported that clinically approved CYP19A1 inhibitor letrozole can reduce SARS-CoV-2 infection caused lung injury and support recovery of imbalanced sex hormones in male hamsters.⁵⁷ However, increased CYP19A1 mRNA levels were only found in SARS-CoV-2-infected cells, but not in those infected with SARS-CoV and influenza virus H1N1,⁵⁷ which largely limits the application scope of letrozole. FINA inhibits the conversion of TET to DHT through specific antagonism of type II 5α-reductase, and has been widely used for clinical therapy of prostate cancer and baldness. In SARS-CoV-2-infected male hamsters with FINA therapy, serum TET concatenation was slightly increased, serum DHT concatenation was significantly reduced and serum E2 concatenation was significantly increased, indicating indirectly regulation of sex hormones to achieve a re-balance (Fig. 6h). Although the overarching results of our animal experiments revealed that balance of sex hormones is important to alleviate the severity of COVID-19 in hamsters, the detailed regulatory mechanisms and transformation pathways between TET/DHT and E2 was not fully understood. This needs further investigation by metabolomics analysis and molecular biology validations. Besides its therapeutic effects in acute phase, long-term use of FINA is associated with side effects include persistent sexual dysfunction, including erectile dysfunction, decreased libido, impaired orgasm, and potential risk of psychiatric adverse events.^58–60 Therefore, short-term use of FINA with careful monitoring of sexual function and mental health is recommended in clinical practise. Collectively, directly targeting the conversion from TET to DHT and indirectly regulating the balance of TET and E2 are both effective in reducing the severity of COVID-19 in the hamster model, providing a new therapeutic strategy for individualized patient management and treatment of emerging viral pneumonia.

There are several limitations and gaps that should be considered and investigated in our follow-up studies. Although the hamster model was widely used for the studies of COVID-19, the lack of species-specific immunological reagents largely limits the performance of immunological analysis, perturbation of immune cells and associated signaling observed in the pulmonary transcriptome data need to be validated in hACE2-transgenic mouse model, humanized mouse model, and clinical cohorts in future studies. Because of the experimental scale limitation in the animal biosafety laboratory, we did not explore the relationship between inoculation viral doses and sexual hormone levels after infection. Robust viral shedding and broad organ tropism of SARS-CoV-2 determine its strong immune stimulation capacity, TET production in testis, muscle, fat, adrenal gland, and other organs might be affected by the virus component, inflammatory cytokines, and chemokines, which made it difficult to clarify how SARS-CoV-2 directly affects the dynamic changes of serum TET levels. Moreover, the relationships between androgen receptors and the downstream inflammatory signaling, such as HIF-1α and IL-17, and the underlying regulation mechanisms need to be clarified and validated. We observed that low serum TET concentration is associated with Long COVID in both male and female hamsters; the molecular consequences and regulation mechanisms need further investigation.

Taken together, our study investigated the male bias in COVID-19, unveiled a unique strategy for targeting the male sex hormone TET, and delineated the multidimensional therapeutic mechanisms of the TET inhibitor FINA in the hamster model. As sexual dimorphism is prevalent in virus infection and pathology, the repositioning of approved sex hormone regulating drugs provides new options for the in-house treatment of infectious diseases, reducing rates of severity, hospitalization, and death. Furthermore, host-targeting agents tend to have a broad therapeutic effect against viruses with high mutation rates and immune escape capacity, which effectively compensates for the limitations of vaccines and virus-targeting agents.

Materials and methods

Experimental animal study and sample size

Hamsters (LVG strain) used in this study were raised in the specific pathogen-free animal feeding facilities. To determine the sexual dimorphism of prototype SARS-CoV-2, Beta, Delta and XBB.1.9.2.1 variant infection, 16 male or female hamsters were used for survival observation in each group, and the percentage of body weight changes of six survived hamsters were recorded. To analyze the relationship between lung viral load and serum TET concentrations, six male and female hamsters were euthanized at 0, 3, 5, 7, and 9 dpi of prototype SARS-CoV-2, respectively. To investigate the impact of castration on the severity of COVID-19, six castrated male hamsters and six age-paired male controls were infected with prototype SARS-CoV-2, Beta, Delta, and XBB.1.9.2.1 and euthanized at 7 dpi, respectively. To investigate the relationship between serum TET concentration and the disease outcomes of long COVID, 30 male and 30 female hamsters survived from prototype SARS-CoV-2 infection were employed for evaluation. Percentage of body weight changes and serum TET concentration were detected at 6 weeks post infection. To investigate the impact of exogenous TET on the severity of COVID-19, 16 male or female hamsters were treated with low, middle, and high doses of TET after infection of prototype SARS-CoV-2, respectively. Age-paired male and female hamsters without TET treatment after infection were set as controls. Survival rates of all the hamsters and percentage of body weight changes of six survived hamsters in each group were recorded from 0 to 7 dpi. The survived hamsters were euthanized at 7 dpi for sample collection. For the validation of TET-caused disease enhancement, 12 male and 12 female hamsters were infected with prototype SARS-CoV-2 and received high dose of TET treatment, half of them received RNAi therapy.

For the study of FINA therapy, six male hamsters received 3-doses FINA therapy from 3 to 5 dpi, another six male hamsters received 6-doses FINA therapy from 0 to 5 dpi of prototype SARS-CoV-2. Age-paired male hamsters without TET treatment after infection were set as controls. In the follow-up evaluation, 12 male hamsters were infected with Beta, Delta, and XBB variants of SARS-CoV-2, respectively. Half of them received 6-doses FINA therapy from 0 to 5 dpi, and the other half were set as controls. Survival rates and percentage of body weight changes of all the hamsters in each group were recorded from 0 to 7 dpi. The survived hamsters were euthanized at 7 dpi for sample collection. The lung tissues collected from the prototype SARS-CoV-2-infected male hamsters with or without 6-doses FINA therapy at 7 dpi were used for pulmonary transcriptome analysis. For the dose exploration experiment of FINA, 24 male hamsters were infected with prototype SARS-CoV-2 and received 3-dose or 6-dose of FINA therapy with different dosages. All the animal experiments were approved by the Medical Ethics Committee of Guangdong-Hong Kong Joint Laboratory of Emerging Infectious Diseases (SUCM2021-112).

System evaluation of disease outcome

Detailed information of biosafety operations, preparation of virus stocks, castration operation, virus inoculation and sample collection, administration of TET and FINA, transcriptome analysis of lung tissue samples, detection of viral RNA, viral titer and cytokine mRNA levels, measurement of serum sexual hormone concentration, histopathological studies and statistical analysis were shown in the supporting information file.

Preparation of virus stock

In this study, different SARS-CoV-2 strains include prototype (EPI_ISL_1655937), Beta variant (EPI_ISL_2779638), Delta variant (EPI_ISL_2385091), and XBB.1.9.2.1 variant (EPI_ISL_17660518) were cultured on Vero cells (#CCL-81, ATCC). Viral stocks were prepared in Vero cells with Dulbecco’s Modified Eagle Medium, DMEM (#11995-040, GIBCO) containing 2% Fetal Bovine Serum, FBS (#10270106, GIBCO), 5ug/mL of TPCK-trypsin (#25200-114, GIBCO), 100 U/mL Penicillin Streptomycin (#15140-122, Invitrogen) and 30 mmol/L MgCl2 (#AM9530G, Thermo-Fisher). Virus stocks were harvested and stored in ultra-low temperature refrigerator. The titers were determined by means of plaque assay in Vero cells.

Castration operation

Male hamsters were anesthetized by isoflurane (#R510-22, RWD Life Science). After that, bilateral testis and epididymis were excited by surgical knife within three minutes. The wound was well closed by absorbent sutures and completely healed within 2 weeks.

Virus inoculation and sample collection

The hamsters were anesthetized by isoflurane (#R510-22, RWD Life Science) and nasally inoculated with 1 × 10⁴ PFU dose of SARS-CoV-2 diluted in 200 μL Phosphate Buffer Solution, PBS (#10010031, GIBCO). The body weights of these hamsters were measured by electronic balance. Hamsters were euthanized and sacrificed at indicated time point for detection of viral load in respiratory tract organ tissues, analysis of pathogenesis in lung lobes and pulmonary transcriptome analysis. Before the anatomy operation, blood was collected for measurement of serum sexual hormone concentration.

Detection of viral RNA and viral titer

For the solid organ samples, we collect 1 mg turbinate, 0.1 mg trachea and 0.1 mg lung in 1 mL PBS for homogenate and detection of viral RNA and viral titer. Viral RNA was extracted by using a QIAamp Viral RNA Mini Kit (#52906, Qiagen) according to the manufacturer’s instructions. The RT-PCR was conducted by using the SLAN-96S Real-Time System (Hongshi, Shanghai, China) with a SARS-CoV-2 RT qPCR Kit (WS-1248, Wantai). Relative viral RNA of SARS-CoV-2 ORF1ab and NP genes were determined using primer pairs and probes shown in the kit instruction. Viral RNA copies were expressed on a log10 scale after normalized to the standard curve obtained by using ten-fold dilutions of a SARS-CoV-2 stock, which was calibrated using the SARS-CoV-2 RNA standard sample (GBW(E)091133, CNRM). The titers of viral stocks and homogenized tissues were measured by plaque assay and half tissue culture infective dose (TCID50) titration method in Vero cells seeded in 6-well and 96-well plates, respectively.

Measurement of serum sexual hormone concentration

The concentrations of TET (#ml059793, Shanghai Meilian Biotech), DHT (#ml059794, Shanghai Meilian Biotech), and E2 (#ml059792, Shanghai Meilian Biotech) in hamster serum samples were measured by the indicated ELISA Kit, respectively. It is a cornerstone of comparative endocrinology that the core sex steroid hormones TET, DHT, and estradiol (E2) are identical in their molecular structure across all mammalian species, including humans, mice, and hamsters. Their synthesis follows a deeply conserved steroidogenic pathway. Therefore, the analyte detected by a specific ELISA antibody is the exact same molecule, regardless of the species of origin.

Animal and sample size justification

Sample sizes maximized considering limits in BSL-3 working capacity, numbers of animals that can be handled under ABSL-3 conditions, and availability of well-trained staff.

Statistical Analysis

Data are presented as the mean ± SD and analyzed in R (version 4.2.3). One-way ANOVA, followed by Dunnett’s multiple-comparisons test, was used for Fig. 1a, d, 2g, 3–5, 6d–h, and S11b. Two-way ANOVA with Bonferroni’s post hoc test was used for Figs. 1g, h, S3c–e, S13c and S17. Two-way ANOVA with Dunnett’s post hoc test for between-group comparisons at individual time points across 1–7 dpi for Figs. 2c, e and 6b. Unpaired two-tailed Student’s t test was used for Figs. 1i, k and S3f. P-values < 0.01 were considered significant: ^*P < 0.01, ^**P < 0.001, ^***P < 0.0001, ns indicates no significance.

Author contributions

Y.L.Z., X.H.Q. L.X. and Z.M. contributed equally to this work. Y.L.Z., Z.M., W.K., Y.J.H., M.J., Z.L.M., L.X.L., Z.C. and C.R.R. performed animal experiments and sample measurements. X.H.Q. and L.X. performed transcriptome analysis and draw the figures. F.M.J., Z.H.C., Z.Y.L., and Y.Q. participated in project design and provided proposals. Y.L.Z., X.H.Q., and L.X. designed this study and wrote the manuscript. Y.L.Z., C.T., G.Y., and X.N.S. supervised this study. All authors have read and approved the article.

Funding

This study was supported by National Science and Technology Major Project (#2025ZD01900504, Recipient: C.T.); National Natural Science Foundation of China (#32201152, Recipient: L.X.; #82272310, Recipient: C.T.); National Natural Science Foundation of Xiamen City (#3502Z202471022, Recipient: Y.L.Z.); Funding supports from Guangdong Province Government (#HZQB-KCZYZ-2021014, Recipient: G.Y.); Research Grants Council of Hong Kong (#T11-705/21-N, Recipient: Z.H.C.); Scientific Research Foundation of State Key Laboratory of Vaccines for Infectious Diseases (#2024SKLVDzy009, Recipient: Y.L.Z.) and Fundamental Research Funds for the Central Universities (#20720250004, Recipient: X.N.S.; #20720250177, Recipient: Y.L.Z.).

Data availability

The sequencing data generated in this study have been deposited in the ArrayExpress database at EMBL-EBI under accession number E-MTAB-16610. All other data supporting the findings of this study are available within the article and its Supplementary Information files. Detailed information of biosafety operations, administration of TET and FINA, transcriptome analysis of lung tissue samples, detection of cytokine mRNA levels, and histopathological studies were shown in the supporting information file.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41392-026-02636-1.