Intranasal neomycin evokes broad-spectrum antiviral immunity in the upper respiratory tract

Abstract

Respiratory virus infections in humans cause a broad-spectrum of diseases that result in substantial morbidity and mortality annually worldwide. To reduce the global burden of respiratory viral diseases, preventative and therapeutic interventions that are accessible and effective are urgently needed, especially in countries that are disproportionately affected. Repurposing generic medicine has the potential to bring new treatments for infectious diseases to patients efficiently and equitably. In this study, we found that intranasal delivery of neomycin, a generic aminoglycoside antibiotic, induces the expression of interferon-stimulated genes (ISGs) in the nasal mucosa that is independent of the commensal microbiota. Prophylactic or therapeutic administration of neomycin provided significant protection against upper respiratory infection and lethal disease in a mouse model of COVID-19. Furthermore, neomycin treatment protected Mx1 congenic mice from upper and lower respiratory infections with a highly virulent strain of influenza A virus. In Syrian hamsters, neomycin treatment potently mitigated contact transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). In healthy humans, intranasal application of neomycin-containing Neosporin ointment was well tolerated and effective at inducing ISG expression in the nose in a subset of participants. These findings suggest that neomycin has the potential to be harnessed as a host-directed antiviral strategy for the prevention and treatment of respiratory viral infections.

Keywords: antiviral; interferon; mucosal immunity; nasal; transmission.

Fig. 1.

Intranasal application of neomycin induces an upper respiratory ISG response independent of commensal microbiota. (A) Experimental schema. Specific-pathogen-free (SPF) C57BL/6J mice were intranasally treated once with 2 mg neomycin, or vehicle delivered in 10 μL volume per nostril. On days 1, 3, 5, and 7 posttreatment, nasal turbinates were collected for gene expression analysis by RT-qPCR and for RNA FISH analysis (for day 1 tissues only). (B) Expression of ISGs Irf7, Isg15, Usp18, Cxcl10, and Rsad2 in nasal turbinate tissues (Day 1, n = 5; Day 3, n = 4; Day 5, n = 5; Day 7, n = 4). Gene expression was normalized against housekeeping genes Hprt, and then compared against biological controls (vehicle-treated mice). (C) RNA FISH analysis of Cxcl10 expression in paraffin-embedded nasal turbinate tissues. (D) Experimental schema. SPF or GF mice were intranasally treated once with 2 mg neomycin or vehicle. On day 1 posttreatment, nasal turbinates were collected for gene expression analysis by RT-qPCR. (E) Expression of ISGs Irf7, Isg15, Usp18, Cxcl10, and Rsad2 in nasal turbinate tissues (SPF + Vehicle, n = 5; SPF + Neomycin, n = 5; GF + Vehicle, n = 5; GF + Neomycin, n = 5). Gene expression of SPF and GF samples were separately compared to their vehicle controls. To reduce the overall number of experimental animals used, control data points from naïve and neomycin-treated animals housed in the SPF facility are common to B and E. For RNA FISH analyses, representative staining results were shown. Sections are representative of multiple sections from at least five mice per group. Mean ± SEM; statistical significance was calculated by means of two-way ANOVA followed by Tukey’s correction (B and E); *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Individual data points are represented. Samples from different timepoints were collected from different animals, not the same animals longitudinally.

Fig. 2.

Neomycin prophylaxis affords antiviral protection against SARS-CoV-2 infection and disease. (A) Experimental schema. K18-hACE2 mice housed in a SPF facility were intranasally treated once with 0.2 mg neomycin, 2 mg neomycin or vehicle delivered in 10 μL volume per nostril. One day after treatment, mice were intranasally infected with 5 × 10³ PFU SARS-CoV-2 (2019n-CoV/USA_WA1/2020). In one group of mice, weight loss and survival were monitored daily up to 14 DPI. Death was recorded when mice were moribund, or at 80% of original body weight. In a separate cohort, nasal turbinate tissues were collected for viral titer analysis 2 DPI. (B and C) Weight loss and survival of K18-hACE2 mice prophylactically treated with vehicle, 0.2 mg neomycin, or 2 mg neomycin from 1 to 14 DPI (Vehicle, n = 10; 0.2 mg Neomycin, n = 10; 2 mg Neomycin, n = 10). (D) Measurement of infectious virus titer in the nasal turbinate at 2 DPI by plaque assay (Vehicle, n = 10; 0.2 mg Neomycin, n = 9; 2 mg Neomycin, n = 10). Limit of detection (LOD): 10² PFU/mL. (E) Measurement of vRNA in the nasal turbinate at 2 DPI by RT-qPCR against SARS-CoV-2 N gene using the CDCN1 primer-probe set (Vehicle, n = 10; 0.2 mg Neomycin, n = 10; 2 mg Neomycin, n = 10). (F) Representative images of immunohistochemistry staining of SARS-CoV-2 N protein on nasal turbinate sections from vehicle- or neomycin-treated mice. (G–I) Measurement of vRNA in the nasal turbinate at 2 DPI from mice infected with BA.2.12.1, BA.4 or BA.5 by RT-qPCR against SARS-CoV-2 N gene using the CDCN1 primer-probe set (BA2.12.1 + Vehicle, n = 8; BA2.12.1 + Neomycin, n = 9; BA.4 + Vehicle, n = 10; BA.4 + Neomycin, n = 9; BA.5 + Vehicle, n = 10; BA.5 + Neomycin, n = 10). Fold reduction was calculated as average fold reduction in infectious (D) or genomic RNA (E and G–I) virus load between neomycin-treated groups and vehicle controls. For histological analyses, representative H&E staining results were shown. Sections are representative of multiple sections from at least five mice per group. Mean ± SEM; statistical significance was calculated by the log-rank Mantel–Cox test (C), one-way ANOVA followed by Tukey correction (D and E) or Student's t test (G–I). Individual data points are represented. Data are pooled from two independent experiments.

Fig. 3.

Neomycin confers respiratory protection against influenza A virus infection. (A) Experimental schema. Mx1 congenic mice housed in a SPF facility were intranasally treated once with 2 mg neomycin or vehicle delivered in 25 μL volume per nostril. One day after treatment, mice were intranasally infected with 26.5 PFU highly virulent PR8 (hvPR8). On 2 DPI, nasal turbinate and lung tissues were collected for viral titer analysis 2 DPI. (B) Measurement of infectious virus titer in the nasal turbinate at 2 DPI by plaque assay (Naïve, n = 7; hvPR8 + Vehicle, n = 8; hvPR8 + Neomycin, n = 9). (C) Measurement of vRNA in the nasal turbinate at 2 DPI by RT-qPCR against influenza nucleoprotein (NP) or hemagglutinin (HA) genes (Naïve, n = 7; hvPR8 + Vehicle, n = 8; hvPR8 + Neomycin, n = 9). (D) Measurement of infectious virus titer in the lung at 2 DPI by plaque assay (Naïve, n = 7; hvPR8 + Vehicle, n = 8; hvPR8 + Neomycin, n = 9). (E) Measurement of vRNA in the lung at 2 DPI by RT-qPCR against NP or HA genes (Naïve, n = 7; hvPR8 + Vehicle, n = 8; hvPR8 + Neomycin, n = 9). (F and G) Expression of ISGs Irf7, Isg15, Usp18, Cxcl10, Rsad2, and Mx1 in nasal turbinate (F) and lung homogenate (G) tissues at 2 DPI with influenza hvPR8 in Mx1 congenic mice pretreated with 2 mg of neomycin 24 h before infection. Gene expression was normalized against housekeeping genes Hprt, and then compared against biological controls (vehicle-treated mice). Fold reduction was calculated as average fold reduction in infectious (B and D) or genomic RNA (C and E) virus load between neomycin-treated groups and vehicle controls. Mean ± SEM; statistical significance was calculated by one-way ANOVA followed by Tukey correction (B–E) or two-way ANOVA followed by Tukey’s correction (F and G). Individual data points are represented. Data are pooled from two independent experiments.

Fig. 4.

Neomycin mitigates disease progression and curbs infection of SARS-CoV-2. (A) Experimental schema. K18-hACE2 mice housed in a SPF facility were intranasally infected with 5 × 10³ PFU SARS-CoV-2 (2019n-CoV/USA_WA1/2020). Four hours after infection, mice were intranasally treated once with 0.2 mg neomycin, 2 mg neomycin or vehicle delivered in 10 μL volume per nostril. Weight loss and survival were monitored daily up to 14 DPI. Death was recorded when mice were moribund, or at 80% of original body weight. (B and C) Weight loss and survival of K18-hACE2 mice therapeutically treated with vehicle, 0.2 mg neomycin, or 2 mg neomycin from 1 to 14 DPI (Vehicle, n = 10; 0.2 mg Neomycin, n = 10; 2 mg Neomycin, n = 11). (D) Experimental schema. K18-hACE2 mice housed in a SPF facility were intranasally infected with 5 × 10³ PFU SARS-CoV-2 (2019n-CoV/USA_WA1/2020). Four hours after infection, mice were intranasally treated once with 0.2 mg neomycin, 2 mg neomycin or vehicle delivered in 10 μL volume per nostril. Nasal turbinate tissues were collected for viral titer analysis 2 DPI. (E) Measurement of infectious virus titer in the nasal turbinate at 2 DPI by plaque assay (Vehicle, n = 10; 0.2 mg Neomycin, n = 10; 2 mg Neomycin, n = 9). Limit of detection (LOD): 10² PFU/mL. (F) Measurement of vRNA in the nasal turbinate at 2 DPI by RT-qPCR against SARS-CoV-2 N gene using the CDCN1 primer-probe set (Vehicle, n = 10; 0.2 mg Neomycin, n = 9; 2 mg Neomycin, n = 10). (G) Expression of Irf7, Isg15, Usp18, Cxcl10, and Rsad2 in nasal turbinate tissues at 2 DPI from mice infected with 5 × 10³ PFU SARS-CoV-2 (2019n-CoV/USA_WA1/2020) and treated with Neomycin at 4 h postinfection. Gene expression was normalized against housekeeping genes Hprt, and then compared against biological controls (vehicle-treated mice) (Vehicle, n = 10; 0.2 mg Neomycin, n = 9; 2 mg Neomycin, n = 9), (H) Experimental schema. K18-hACE2 mice housed in a SPF facility were intranasally infected with 5 × 10³ PFU SARS-CoV-2 (2019n-CoV/USA_WA1/2020). Four HPI, mice were intranasally treated once with 2 mg neomycin or vehicle. Two subsets of neomycin-treated mice were provided with a second dose of 2 mg neomycin on 1 and 3 DPI, respectively. Weight loss and survival were monitored daily up to 14 DPI. Death was recorded when mice were moribund, or at 80% of original body weight. (I and J) Weight loss and survival of K18-hACE2 mice therapeutically treated with vehicle, neomycin dosed at 4 HPI, 2 mg neomycin dosed at 4 HPI and 1 DPI or neomycin dosed at 4 HPI and 3 DPI from 1 to 14 DPI (Vehicle, n = 9; Neomycin [+4 h], n = 10; Neomycin [+4 h & +24 h], n = 10; Neomycin [+4 h & +72 h], n = 10). Fold reduction was calculated as average fold reduction in infectious (D) or genomic RNA (E) virus load between neomycin-treated groups and vehicle controls. Mean ± SEM; statistical significance was calculated by the log-rank Mantel–Cox test (C and J), one-way ANOVA followed by Tukey correction (E and F) or two-way ANOVA followed by Tukey’s correction (G). In weight loss curves, error bars for timepoints with less than 3 alive animals were not shown. Individual data points are represented. Data are pooled from two independent experiments.

Fig. 5.

Neomycin reduces contact transmission in a hamster model of SARS-CoV-2. (A) Transmission experimental schema. Syrian hamsters were intranasally treated with 5 mg of neomycin or vehicle delivered in 25 μL volume per nostril. Twenty-four hours after neomycin treatment, recipient hamsters were cohoused for 4 h with naïve donor hamsters that had been infected 24 h earlier with 10⁴ PFU SCV2. Daily OP swabs from 0 to 2 DPI as well as lung tissues on 2 DPI were collected. (B) Measurement of infectious virus titer in OP swabs on 0 to 2 DPI by plaque assay (Vehicle, n = 10; Neomycin, n = 10). (C) Measurement of infectious virus titer in lung tissues on 2 DPI by plaque assay (Vehicle, n = 10; Neomycin, n = 10). Fold reduction was calculated as average fold reduction in infectious (B and C) virus load between neomycin-treated groups and vehicle controls and indicated above all bar plots. Mean ± SEM; statistical significance was calculated by means of one-way ANOVA followed by Tukey’s correction (B) or Student's t test (C). *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Individual data points are represented. Data are pooled from two independent experiments.

Fig. 6.

Neosporin application induces ISG expression in the nasal mucosa in healthy humans. (A) Experimental schema. Healthy human participants were randomized at a 2:1 ratio into an experimental arm and a placebo arm to receive Neosporin (n = 12) or Vaseline (n = 7) treatment, respectively. Drug application was performed twice-daily for 7 d. Participants had a total of 4 in-person meetings. On days 1, 4, 8, and 12, a nasopharyngeal swab and a nasal brush collection were performed for viral testing and ISG assessment, respectively. (B–D) Response rate defined as a participant with at least 2 ISGs undergoing at least 5-, 10-, or 20-fold induction at any timepoint compared to day 0 measurements. (E–J) Expression of RSAD2, IRF7, CXCL9, CXCL10, IL10, and USP18 in nasal swab samples. Gene expression was normalized against housekeeping genes HPRT, and then compared against day 0 samples. Mean ± SEM; statistical significance was calculated by means of two-way ANOVA followed by Bonferroni’s correction (E–J); *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.