All procedures for animal use and care were approved by the Institutional Animal Care and Use Committees at Institute of Zoology, Chinese Academy of Sciences, and the Ethics and Welfare of Experimental Animals Committee at Institute of Zoology, Chinese Academy of Sciences (No. IOZ20160046). All experiments were performed according to institutional guidelines.

Influenza viruses A/environment/Qinghai/1/2008 (H5N1) (Li et al., 2011) were propagated in 9-day-old embryonated eggs. Viral titer was determined using the plaque assay, and the LD₅₀ was measured in mice administered serial dilutions of the stock.

Specific-pathogen-free (SPF) C57BL/6 mice (6–8-week-old males, n = 90) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., (beijing, China). Mice were randomly assigned to three groups: the H5N1 virus infection group (vehicle, treated with PBS, n = 45), oseltamivir (OS, treated with oseltamivir, n = 45), and geldanamycin (GA, treated with geldanamycin, n = 45). Mice were intranasally infected with 1 × 10⁵ pfu (equivalent to 10 LD₅₀) of the mouse adapted influenza virus A/environment/Qinghai/1/2008 (H5N1) under anesthesia with isoflurane. Two hours after infection, mice in the experimental groups were administered 5 mg/kg of OS (oseltamivir, twice daily for 5 days) (Roche, Switzerland) in PBS by gavage (Sidwell et al., 1998) or 1 mg/kg of GA (geldanamycin, twice daily for 5 days) (NCPC, CHN) in DMSO (dimethyl sulfoxide) (Sigma-Aldrich, USA) by intraperitoneal injection (Chatterjee et al., 2007; Nakano et al., 2007), whereas mice in the control group received PBS only. Body weights and survival of each group were monitored for 12 days or until death. All experiments with influenza virus infections were conducted in a Biological Safety Level-3 (BSL-3) laboratory. Food and water were available ad libitum.

Five mice in each group were sacrificed on days 2, 4, and 7 after lethal H5N1 influenza virus inoculation and treatment. Lung tissues were harvested from sacrificed mice, fixed in 10% formalin buffer, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E). Histopathological changes were observed under a light microscope.

The trachea of euthanized mice were exposed, transected, and intubated with a blunt 18 gauge needle. One milliliter of phosphate-buffered saline supplemented with Complete Mini, EDTA-free Protease Inhibitor Cocktail (Roche) was infused and recovered four times. The recovered bronchoalveolar lavage fluid (BALF) was centrifuged at 3,000 × g for 3 min at 4°C and then stored at −80°C until use. Levels of IL-1α, IL-6, IFN-α, IFN-γ, TNF-α, MIP-1α, MCP-1, CXCL-2, IP-10, and RANTES in BALF were quantified using commercial ELISA kits according to the manufacturer's instructions (R&D Systems).

Cellular analysis was done as described in previous study (Teijaro et al., 2011). Briefly, lungs were harvested from PBS-perfused mice and mechanically diced into small pieces using surgical scissors. Diced lungs were suspended in 4 ml of CDTI buffer (0.5 mg/ml collagenase from Clostridium histolyticum type IV [Sigma], 0.1 mg/ml DNase I from bovine pancreas grade II [Roche], and 1 mg/ml trypsin inhibitor type Ii-s [Sigma] in DMEM) for 1 h at 37°C. Lung tissues were then mechanically disrupted by passing them through a 100 mm filter, and red blood cells were lysed using red blood cell lysis buffer (0.02 Tris-HCl and 0.14 NH₄Cl). Inflammatory cells were purified by centrifugation in 35% PBS-buffered Percoll (GE Healthcare Life Sciences) at 1,500 rpm for 15 min. Cell pellets were re-suspended in staining buffer, and Fc receptors were blocked using 25 mg/ml anti-mouse CD16/32 (BD Biosciences). Cells were stained with the following anti-mouse antibodies: Alexa Fluor 488-conjugated gp38 (eBioscience; clone eBio8.1.1), PE-conjugated (BioLegend, Inc.; clone ME13.3) and APC-conjugated (eBioscience; clone 390) CD31, PE-Cy7-conjugated EpCAM (BioLegend, Inc.; clone 68.8), Pacific blue-conjugated CD45.2 (BioLegend, Inc.; clone 104), PerCP-Cy5.5-conjugated NK1.1 (BD Biosciences; clone PK136), PE-Cy7-conjugated CD3e (eBioscience; clone 145-2C11), e450-conjugated CD4 (eBioscience; clone L3T4), PE-conjugated CD8a (BD Biosciences; clone 53-6.1), Pacific blue-conjugated B220 (BD Biosciences; clone RA3-6B2), PE-conjugated CD19 (BD Biosciences, clone 1D3), PE-Cy7-conjugated CD11b (eBiosciences; clone M1/70), PerCP-Cy5.5-conjugated CD11c (eBiosciences; clone N418), APC-conjugated Gr-1 (BD Biosciences; clone RB6-8C5), Pacific blue- and PE-conjugated Ly6G (BD Biosciences; clone IA8), APC-conjugated F480 (eBioscience; clone BM8), PE-conjugated 7/4 (AbD Serotec; clone 7/4), PE-conjugated I/A-I/E (BD Biosciences; clone M5/114.15.2), PE-Cy7 conjugated CD205 (eBiosciences; clone 205yekta), FITC-conjugated CD69 (BD Biosciences; clone H1.2F3), and APC-conjugated CD25 (eBiosciences; clone PC61.5). The stained inflammatory cells were analyzed using a BD FACSAria III flow cytometer. FlowJo software was used for data analyses.

Total RNA was extracted from mouse lung tissues and reverse-transcribed into cDNAs using the M-MLV reverse transcriptase (Promega). The expression of the influenza viral np gene (Wang et al., 2017) was measured using the Stratagene MX3000P real-time PCR system with the following primers: NP forward, 5′-AGT CCT GCT TGC CTG CTT G-3′, and NP reverse, 5′-CTG CAG AGT GGC ATG CCA T-3′. Reactions were performed in triplicate. Ct values were normalized to β-actin levels, and fold changes were calculated (ΔΔCt) with respect to the vehicle group.

Serum samples were harvested from mice on day 14 after infection with the lethal H5N1 influenza virus, and then heat-inactivated for 30 min at 56°C. Levels of neutralizing antibodies in mouse sera were determined by the neutralization assay as described by a protocol (WHO Global Influenza Surveillance Network, 2011). Briefly, Serum samples were first heat inactivated, and then serial two-fold dilutions were made starting at an initial 1:10 dilution. Influenza viruses (100 50% tissue culture infectious doses [TCID50]) were added to serum dilutions, incubated at 37°C with 5% CO₂ for 1 h, and used to infect 1.5 × 10⁵ MDCK cells per ml. After overnight incubation, the presence of the viral proteins was detected by an enzyme-linked immunosorbent assay (ELISA) using monoclonal antibodies (colon name:GT1236, GeneTex, Inc. USA) specific to the influenza A virus nucleoprotein. MN titers were defined as the reciprocal of the highest dilution of serum that yielded at least 50% neutralization.

Data are presented as means ± SD. The log rank Kaplan-Meier or χ2 tests were used to analyze the survival times, survival rates, and neutralizing antibody levels, respectively. All other analyses were performed using unpaired two-tailed Student's t-tests, with a 95% confidence level representing a significant difference. All statistical analyses were performed using GraphPad software (GraphPad Prism Version 6.0, GraphPad Software, San Diego, CA).

In vivo experiments were performed in the mouse model to analyze the anti-influenza virus potential of geldanamycin. We first intranasally infected anesthetized mice with 1 × 10⁵ pfu of influenza virus, and they were treated with two drugs, geldanamycin or oseltamivir, 2 h later. The control mice were treated with vehicle (PBS) only. We then analyzed the survival rate and weight loss of the treated mice. As shown in Figure 1A, the onset of death in vehicle-treated mice was observed earlier (at day 3 post-infection) than in mice treated with either oseltamivir (at day 4 post-infection) or geldanamycin (at day 4 post-infection). In addition, the survival rate of drug-treated mice was significantly increased. Geldanamycin showed a more marked protection of survival (93.3%) than oseltamivir (53.3%). Meanwhile, we also monitored the animals' body weights for 12 days after virus infection. As shown in Figure 1B, mice from each group started losing weight beginning 1 day after infection. The body weights of the control mice continued to decrease until they all died at 6 day post-infection; mice treated with geldanamycin stopped losing weight after 4 days and the surviving mice started gaining weight to levels similar to the baseline weight. On the other hand, the oseltamivir-treated mice recovered their body weights beginning on day 9, which represented a greater delay than the geldanamycin treatment. Mice treated with geldanamycin lost 15.7% of their body weights, which was significantly less than the loss observed in mice treated with vehicle or oseltamivir (24.9%, p < 0.05). Based on these results, geldanamycin was more effective at protecting mice against influenza virus infection and significantly reduced mortality compared with oseltamivir.

Avian and mammalian hosts, including humans, will exhibit strong lung immunopathology upon infection with lethal influenza viruses. We further investigated the efficacy of geldanamycin in treating mice infected with lethal H5N1 influenza viruses by histopathology. Lung sections were collected from mice at different time points after infection. On days 4 and 7 after infection, the vehicle-treated mice displayed severe bronchopneumonia, interstitial pneumonitis, and the presence of necrotic debris in the bronchioles and alveoli (Figure 2). In comparison, treatment with either geldanamycin or oseltamivir markedly reduced tissue injury, mononuclear cell accumulation, hemorrhage, pulmonary edema, and tissue inflammation (Figure 2). However, increased tissue consolidation and vascular hemorrhaging were detected in the oseltamivir-treated group (Figure 2). These findings were consistent with the survival results observed for each group (Figure 1A).

Infection with the HPAIV H5N1 influenza virus often correlates with an increase in virus replication in the lungs; therefore, we investigated whether geldanamycin affected viral NP levels in the lungs. Expression of the viral NP gene in the lungs of geldanamycin-treated mice was significantly reduced compared with the vehicle-treated group on days 2, 4, or 7 after infection (p < 0.01), and a similar inhibitory effect on virus NP levels was observed in mice treated with oseltamivir (Figure 3A). However, oseltamivir administration reduced the viral NP levels to a greater extent than the geldanamycin treatment on post-infection days 4 and 7, and difference was statistically significant (p < 0.05) (Figures 3B,C). Interestingly, oseltamivir administration did not significantly prolong survival compared with geldanamycin administration, although it reduced the viral NP levels in the lungs to a greater extent (Figure 1A). Thus, a high viral NP levels might not be the direct cause of high mortality following H5N1 influenza virus infection.

Influenza virus infection promotes the excessive production of proinflammatory cytokines and chemokines in the lung and serum, resulting in massive acute pulmonary hemorrhage and edema. Here, we further assayed whether geldanamycin administration reduced cytokine responses induced by the influenza virus. BALF was harvested from mice in the vehicle, geldanamycin and oseltamivir groups on days 2, 4, and 7 after infection with the lethal H5N1 influenza virus. We measured the levels of proinflammatory cytokines, chemokines, and antiviral cytokines using enzyme immunoassays. The analysis of the levels of proinflammatory cytokines in geldanamycin-treated mice revealed a significant reduction in TNF-α, IL-6, and IL-1α levels on days 2, 4, or 7 after infection compared with the vehicle-treated and oseltamivir-treated mice (Figures 4A–C). Meanwhile, the chemokine levels were similar to the proinflammatory cytokine levels in BALF; geldanamycin-treated mice produced significantly less MIP-1α, MCP-1, CXCL-2, IP-10, and RANTES on days 2, 4, and 7 after infection than the vehicle-treated mice and oseltamivir-treated mice (Figures 4D–H). Finally, the levels of the antiviral cytokines IFN-α and IFN-γ were also determined, and geldanamycin administration decreased the levels of IFN-α and IFN-γ compared with the vehicle-treated mice and oseltamivir-treated mice on days 2, 4, and 7 (Figures 4I,J). Based on these results, the geldanamycin treatment significantly limited the cytokine responses induced by infection with the lethal H5N1 influenza virus.

The immune system is a feedback loop, and the infiltration and activation of immune cells is affected by cytokines and chemokines in the lung and respiratory tract. On the other hand, the secretion of cytokines and chemokines is also affected by innate immune cells. Therefore, we tested the effect of geldanamycin on the inflammatory cell response using flow cytometry. As shown in Figures 5–7, the geldanamycin treatment markedly inhibited the inflammatory cell response and protected mice from the lethal H5N1 influenza virus. Geldanamycin-treated mice showed a significant reduction in the number of macrophages/monocytes compared with the vehicle-treated group (p < 0.01) and oseltamivir-treated group (p < 0.05) on days 2 and 4 post-infection (Figure 5A). In addition, the mean fluorescence intensity (MFI) of the early activation marker CD69 and MHC class II on macrophages/monocytes was diminished, indicating the attenuation of cell activation (p < 0.01) (Figures 5B,C). Natural killer (NK) cells exert well-known, potent cytotoxic activities and robustly produce inflammatory cytokines, including IFN-α, TNF-α, and MIP-1α; the number of NK cells and the MFI of CD69 on NK cells in geldanamycin-treated mice were significantly reduced compared with the vehicle- or oseltamivir- treated mice on days 2 and 4 (Figures 5D,E). Pulmonary dendritic cells play a key role in bridging innate and adaptive immune responses following HPAIV H5N1 infection, and their numbers were also measured on days 2 and 4 after infection. Geldanamycin administration significantly inhibited the MFI of MHC class I, MHC class II, CD80 and CD86 on DC cells compared with vehicle- or oseltamivir-treated mice, indicating that geldanamycin decreased the number of antigen-presenting cells (Figures 6A–H). We also simultaneously measured the accumulation of CD4⁺ and CD8⁺ T cells in lungs of vehicle-or oseltamivir-treated mice. Geldanamycin administration significantly reduced the number of CD4⁺ and CD8⁺ T cells on days 4 and 7 compared with vehicle- or oseltamivir- treated mice (Figures 7A,B). Because CD4⁺ and CD8⁺ T cells are sources of MIP-1α and TNF-α, these results are consistent with the levels of MIP-1α and TNF-α (Figures 4A,D).

In our study, geldanamycin administration in mice infected with the lethal H5N1 influenza virus not only inhibited viral NP levels but also reduced the inflammatory response in the lung tissue compared with the vehicle-treated mice, and had similar effect on the inhibition of viral NP levels in mice treated with oseltamivir. However, we have not clearly determined whether geldanamycin administration affects the host's ability to control the infection. We analyzed the serum levels of neutralizing antibodies after influenza virus infection. No obvious differences were observed among mice treated with geldanamycin, oseltamivir or vehicle (Figure 8). Thus, geldanamycin blocked certain proinflammatory mediators to inhibit the cytokine response, but did not alter the ability of hosts to clear pathogens.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.