Treatment with inhaled aerosolised ethanol reduces viral load and potentiates macrophage responses in an established influenza mouse model

Abstract

Aim

Treatment options for viral lung infections are currently limited. We aimed to explore the safety and efficacy of inhaled ethanol in an influenza-infection mouse model.

Materials and methods

In a safety and tolerability experiment, 80 healthy female BALB/c mice (20 per group) were exposed to nebulized saline (control) or three concentrations of ethanol (40/60/80% ethanol v/v in water) for 3x30-minute periods, with a two-hour break between exposures. In a separate subsequent experiment, 40 Female BALB/c mice were nasally inoculated with 10^4.5 plaque-forming units of immediate virulence “Mem71” influenza. Infection was established for 48-h before commencing treatment in 4 groups of 10 mice with either nebulized saline (control) or one of 3 different concentrations of ethanol (40/60/80% ethanol v/v in water) for 3x30-minute periods daily over three consecutive days. In both experiments, mouse behavior, clinical scores, weight change, bronchoalveolar lavage cell viability, cellular composition, and cytokine levels, were assessed 24-h following the final exposure, with viral load also assessed after the second experiment.

Results

In uninfected BALB/c mice, 3x30-minute exposures to nebulized 40%, 60%, and 80% ethanol resulted in no significant differences in mouse weights, cell counts/viability, cytokines, or morphometry measures. In Mem71-influenza infected mice, we observed a dose-dependent reduction in viral load in the 80%-treated group and potentiation of macrophage numbers in the 60%- and 80%-treated groups, with no safety concerns.

Conclusions

Our data provides support for inhaled ethanol as a candidate treatment for respiratory infections.

Keywords:

• Ethanol
• influenza
• inhalation treatment
• respiratory infection
• toxicology

Introduction

The current management of respiratory viral infections is usually restricted to supportive care, with limited available antiviral and immunomodulatory treatment options.^1–5 Ethanol, delivered by inhalation, has several properties that make it a promising candidate treatment for acute respiratory infections.⁶

Ethanol has well-characterised antimicrobial activity, with concentration-dependent activity against most viruses.⁷ Ethanol also has potent immunomodulatory properties, described across different contexts in infection, inflammation, allergy, and autoimmunity.⁶ While negative effects following chronic alcohol misuse are well described,⁶ short-duration, inhaled ethanol treatment may prove efficacious in the context of the dysregulated immune response seen in severe infections, comparable to other immunomodulatory treatments (e.g. corticosteroids/biologics).³ Finally, ethanol has surfactant-like properties, with demonstrated efficacy in cardiogenic pulmonary edema in humans,⁸ with this property potentially useful for treating the acute respiratory distress syndrome-related pulmonary edema seen in critical infection.

In general, the primary benefit of the inhaled route of administration is the ability to deliver large doses directly to the target site of infection, while potentially reducing systemic absorption and side effects. Indeed, ethanol delivered by inhalation has been shown in humans to result in significantly lower blood alcohol levels than an equivalent dose delivered orally or intravenously.^9,10

While previously published work has supported the potential safety/efficacy of inhaled ethanol in influenza infection when delivered prophylactically before infection,¹¹ ethanol utility as a treatment in established infection has not been explored. We hypothesized that inhaled ethanol would be safe and reduce viral replication in mild established influenza infection in mice.

Materials and methods

Animals

Eight-week-old female BALB/c mice (Animal Resources Center; Murdoch, Western Australia) were housed in individually ventilated cages at the Telethon Kids Institute on non-allergic, dust free bedding (ALPHA-dri TM, Shepherd Specialty Papers, MI, USA). Water and food (Rat and Mouse Cubes, Specialty Feeds, Glen Forrest, Western Australia) were supplied ad libitum. Daily weights and behavior/severity monitoring were performed according to standard operating procedures. All procedures were approved by the Institutional Animal Ethics Committee (AEC # 364) and conformed to National Health and Medical Research Council of Australia guidelines. Mice were monitored during all exposures for signs of distress and clinically scored after each exposure.

Ethanol exposure

For all nebulized exposure studies, individual mice, in groups of ten, were placed in separate compartments of a “pie-box” Perspex exposure chamber. Solutions of 0.9% saline (control) or ethanol at 40%, 60%, and 80% concentrations v/v in water were then nebulized into each chamber using an ultrasonic nebulizer (UltraNeb®, DeVilbiss, Somerset, Pennsylvania). Lignocaine 1% was added to all saline and ethanol solutions to minimize any potential for discomfort during inhalation.

In the baseline safety experiments, 20 healthy naïve mice per treatment group (saline, 40%, 60%, and 80% ethanol) were whole-body exposed for three, 30-minute periods, with a two-hour break in their home cage in between exposures. In the influenza infection model experiments, 10 mice per treatment group were exposed daily for three, 30-minute periods over three consecutive days in the influenza infection model experiments described below.

Influenza infection model

40 mice (n = 10 per exposure group) were inoculated under isoflurane anesthesia by pipetting drops of inoculum onto the nostrils. Mice received 10^4.5 plaque-forming units (pfu) of immediate virulence “Mem71” (H3N1) diluted in 50 μl phosphate buffered saline. Mem 71 is a reassortment strain of the Influenza A family, consisting of neuraminidase of A/Bellamy/42 and hemagglutinin of A/Memphis/1/71, as previously described.^12,13 Mem71 was chosen to assess the impact of inhaled ethanol on established influenza infection in a model of moderate severity where infected mice do not show dramatic weight loss or overt signs of infection.^12,13 Infection was established for 48-h (days 1-2) prior to commencing the three days of saline or ethanol exposure (days 3-5) to coincide with the known peak of Mem71 viral replication and cellular immune responses occurring between infection days 4-7.^12–14

Sample processing

24-h following their final exposure (day 6), mice were terminally anesthetized via intraperitoneal injection of a solution containing ketamine (40 mg/mL; Troy Laboratories, New South Wales, Australia) and xylazine (2 mg/mL; Troy Laboratories, New South Wales, Australia) at a dose of ∼0.10 mL/10 g body weight. Serum ethanol was not measured as all the inhaled ethanol would have been metabolized and cleared by this timepoint.¹⁵

Bronchoalveolar lavage cell and mediator measurements were performed on a random subsample of mice (n = 10 per treatment) in the baseline safety exposure experiment and all mice (n = 10 per treatment) in the influenza infection experiment. Airway morphometry measurements were performed on the other half of mice (n = 10 per treatment) from control and 80% ethanol treatments in the baseline exposure experiment.

Bronchoalveolar lavage fluid (BAL) was collected by washing 0.5 mL of chilled saline (Pfizer Pty Ltd., WA, Australia) in and out of the lungs 3 times. To separate the cells and supernatant, BAL was centrifuged at 350 g for 4 min. A proportion of re-suspended cell pellet was stained with trypan blue and total cell counts obtained using a hemocytometer. Remaining cells were cytospun, stained with Rapid-Stain (Amber Scientific, Traralgon, Victoria, AUS) and differential cell counts conducted using light microscopy. The supernatant was analyzed for interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-a) by ELISA as per the manufacturers’ instructions (BD Biosciences, North Ryde, NSW, Australia and R&D Systems, Gymea, NSW, Australia).

For airway morphometry measurements, lungs were inflation fixed at a trans-respiratory pressure of 10 cm H₂O by intra-tracheal instillation of 4% formaldehyde prior to removal en bloc. The left lung was embedded in paraffin and 5 µm thick sections were taken from the proximal region. These were stained with Masson’s trichrome, and then analyzed for epithelial thickness, airway smooth muscle mass, and airway lumen diameter using the stereological software (newCAST; Visiopharm, Hørsholm, Denmark) as previously described.¹⁶ Airway morphometry measurements were only performed on mice in the control and 80% ethanol treated groups to assess the maximal effect of ethanol treatment.

Viral load

For the viral load assessments, lungs were removed en bloc from mice euthanised as described above. Lungs were immediately flash-frozen in liquid nitrogen and stored at −80 °C until shipped to the School of Chemistry and Molecular Biosciences, University of Queensland for analysis. RNA was extracted and qPCR was performed as previously described.¹⁷ Viral copy number was determined using 10-fold dilutions of a plasmid containing the MEM Matrix gene.

Statistics

All statistical analyses were performed in R. Differences between groups were assessed using pairwise t-tests (normal distribution) or Mann-Whitney U-tests (non-normal), with the control group as reference for pairwise group comparisons, or ANOVA for multiple group comparisons. The relationship between variables was assessed using Pearson correlation. A p-value <0.05 was considered statistically significant. P-values were adjusted for multiple comparisons using the Holm method.

Results

Baseline safety

No significant differences in clinical scores or pre-/post-exposure weights were observed between BALB/c mice exposed to saline (control) or three different concentrations (40%,60%,80%) of ethanol (Table 1). Ethanol exposed mice displayed unsteady gait for approximately 5 min after exposure, with complete normalization of behavior within 10 min in all animals. BAL and lung samples for morphometric analyses were collected 24-h after exposure and compared between groups. No significant differences in total BAL cell viability (p = 0.71) or lung epithelial layer thickness (p = 0.23), airway smooth muscle mass (p = 0.056), or airway lumen diameter (p = 0.27) were observed (Table 1). There was a statistically significant reduction in the number of neutrophils in mice exposed to 80% ethanol as compared to control (median cells/ml 0 vs. 1105, p = 0.044; Table 1). Ethanol exposure had no other significant effects on BAL immune cell composition or inflammatory cytokine levels (all p > 0.05; Table 1).

Table 1. Baseline exposure measurements.

	Saline Control	40% Ethanol	60% Ethanol	80% Ethanol
Pre-Exposure Weight (g)	18.5 ± 0.9	18.6 ± 1.0	18.1 ± 1.3	18.3 ± 0.9
Post-Exposure Weight (g)	18.8 ± 1.2	18.8 ± 1.0	18.4 ± 1.3	18.0 ± 0.9
Total Cells (x10^4 cells/mL)	7.0(1.9-15.3)	6.1(3.4-11.3)	3.8(1.2-5.5)	3.9(2.8-7.1)
Macrophages (x10^4 cells/mL)	5.5(1.9-10.3)	5.6(3.4-11.3)	3.8(1.2-5.5)	3.9(2.8-7.1)
Neutrophils (x10^3 cells/mL)	1.1(0-55.1)	0.0(0.0-5.0)	0.0(0.0-0.4)	0.0(0.0-0.1)*
Lymphocytes (cells/mL)	0.0 (0.0-0.0)	0.0 (0.0-313.3)	0.0 (0.0-0.0)	0.0(0.0-0.0)
BAL cell viability (%)	91.8 ± 3.9	92.5 ± 4.5	93.0 ± 7.2	92.9 ± 8.6
Epithelial thickness (√area/Pbm)	0.035 ± 0.006	–	–	0.031 ± 0.005
Airway smooth muscle mass (√area/Pbm)	0.200 ± 0.036	–	–	0.184 ± 0.024
Lumen (√area/Pbm)	0.042 ± 0.005	–	–	0.039 ± 0.005
IL-6	2.7(2.7-4.1)	2.7(2.7-4.3)	2.7(2.7-2.7)	2.7(2.7-9.8)
TNFα	5.0(5.0-34.6)	5.0(5.0-5.0)	5.0(5.0-5.0)	5.0(5.0-5.0)

Baseline assessments from naïve BALB/c mice (n = 10/group) exposed to 3x30minute exposures with saline (control) or three different concentrations of ethanol (40%,60%,80%). data are presented as mean ± standard deviation or median (range). *statistically significant difference (p < 0.05) relative to control. IL-6 = interleukin-6, TNFα = tumour necrosis factor alpha, p_bm = perimeter of the basement membrane.

Viral load and disease severity

In mice with established “Mem71” influenza infection, exposure to 3 × 30-minute aerosols of 80% ethanol daily for three days resulted in a significant reduction in influenza viral load compared to controls (mean copy number 1225 vs. 1702, p = 0.042) (Figure 1A). No significant change in viral load was observed after exposures to 40% (mean 1905, p = 0.70) or 60% (mean 1697, p = 0.98) ethanol (Figure 1A). Consistent with this model of mild influenza infection,¹² mice in the control group had only a minor reduction in weight after infection with a peak mean weight loss of 6.89% on infection day 6. No significant differences in mouse weight were observed at any timepoint between ethanol- and control-exposed mice (Figure 1B). No significant differences in mouse behavior or clinical scores were observed.

Figure 1. Efficacy of inhaled ethanol in treating established influenza infection. Female BALB/c mice received 10^4.5 plaque-forming units of immediate virulence “Mem71” influenza. 48-h after infection, mice were exposed to 3 days of 3x30minute exposures to saline (control) or 40%, 60%, or 80% ethanol and analyzed 24-h after the last exposure. (A) Viral load in whole lung tissue was measured by quantitative polymerase chain reaction (n = 10/group). (B) Mice were weighed daily, with weights normalized relative to starting weight (100%). bronchoalveolar lavage samples were collected by washing with chilled saline and then centrifuged to separate the cell pellet and supernatant. (C) Differential counts in the cell pellet were obtained using light microscopy after Rapid-Stain. (D) Interleukin-6 (IL-6) and tumour necrosis factor alpha (TNFα) levels in the supernatant were measured by ELISA. * = holm adjusted p < 0.05 (calculated with t-test in a and Mann-Whitney U-tests in C,D with the control group as reference).

Immune modulation

As expected, influenza infection resulted in large relative increases in measured cell and cytokine levels in all group as compared to their levels measured in uninfected mice (Table 1; Figure 1C,D). Comparing exposure groups, macrophage numbers were significantly higher in infected mice exposed to 60% (median 4.0x10⁵cells/mL, p = 0.037) and 80% (median 4.3x10⁵cells/mL, p = 0.002) ethanol as compared to infected control mice (median 2.4x10⁵cells/mL). A smaller relative increase in macrophage numbers was also observed in mice exposed to 40% ethanol (median 3.0x10⁵cells/mL, p = 0.165), but this was not statistically significant. No significant differences in neutrophil or lymphocyte numbers were observed between groups (Figure 1C).

We also observed a statistically significant increase in IL-6 in mice exposed to 60% ethanol (median 10.2 pg/mL, p = 0.034) relative to control (median level 6.73 pg/mL) (Figure 1D). Non-statistically significant increases in IL-6 were measured in 40% (median 8.17 pg/mL) and 80% (median 7.70 pg/mL) ethanol-exposed mice relative to control (p > 0.05). The increase in IL-6 was also statistically significant when all ethanol-exposed mice were pooled and compared as a single group to control (median 8.73 pg/mL, p = 0.042). No significant differences in TNFα were observed between ethanol and control mice (Figure 1D).

There was a statistically significant inverse correlation between viral load and macrophage count across all mice, with higher macrophage counts associated with lower viral loads (Pearsons R=-0.31, p = 0.049). No correlations between viral load and IL-6, TNFα, neutrophil, or lymphocyte levels were observed (p > 0.05).

Discussion

Ethanol’s properties make it an interesting candidate therapy for respiratory infections if it can be delivered safely to the airways via inhalation. In this study, we observed no major adverse effects from ethanol exposure across all tested concentrations, with the only statistically significant difference being a reduction in BAL neutrophils after 80% exposure, although in infected mice, neutrophils were present in all groups with no significant differences between ethanol and control groups. These results are consistent with previously published studies modeling the effects of short-term ethanol exposures on the rodent lung, which have also demonstrated no tissue damage on histopathology or large alterations in lung cellularity measured both in BAL or on histology, even after weeks of intermittent exposure.^11,18–23 In one study using a rat model of inhaled ethanol exposure with a similar experimental set-up to ours, three fifteen-minute exposures to 65% ethanol vapor repeated for five days resulted in no significant tissue damage as assessed by immunohistochemistry.¹⁸ Similar results were observed in C57BL/6J mice exposed to 50% ethanol vapor twice daily for three weeks.¹¹ Other studies of short-term ethanol exposures in rats have also shown no significant changes in BAL cellularity.^19,20,23 A lack of effects on airway histology and BAL cellularity has been observed in longer term inhalation studies of rats exposed to 3000 ppm ethanol for six hours per day over two to thirteen weeks.^21,22 However, in direct contrast to these results, studies investigating pulmonary effects after chronic oral ethanol exposure have typically reported significant alterations in BAL cellularity with mixed reports of both increases and decreases in macrophage, lymphocyte, and neutrophil cell counts in different studies.^24–26 Models of acute intoxication where a single high dose of intravenous/intraperitoneal ethanol is administered have also reported significant alterations in BAL cellularity.²⁷ Collectively, these results highlight both route of administration and duration of exposure as critical factors when determining the safety of ethanol exposure, with our study lending support to the concept that short-term inhaled ethanol may be safe.

In our model of mild established influenza infection, we observed a statistically significant reduction in viral load following three days of exposure to 80% ethanol. This result is consistent with a recent report demonstrating a protective effect in lethal influenza A infection following prophylactic treatment with inhaled 50% ethanol.¹¹ However, the beneficial effect after short-duration treatment in both studies, directly contrasts with the higher viral loads and more severe disease observed in mouse models of chronic oral ethanol misuse,²⁸ highlighting once more the importance of route and duration of exposure in mediating ethanol’s beneficial versus negative effects.

This antiviral effect was dose-dependent with no significant change in viral load following exposures to 40% or 60% ethanol, which differs from the previously described efficacy in vitro at 40% and in vivo at 50% concentrations.^7,11 However, direct comparison is challenging given that the experimental set-up, duration of exposure, airway/respiratory physiology, and distribution/deposition dynamics all mean that the concentration of ethanol achieved in the lower airways will be lower than that administered into a dosing chamber and difficult to compare/estimate.²⁹ In our study, we also deliberately chose a model of moderate influenza infection to allow characterization of safety and efficacy outcomes without mice rapidly succumbing to illness. This likely contributed to the relative magnitude of ethanol’s effect in the 80% group, given the modest viral load elevation in the control group.^12,13 Future studies aimed at identifying the optimal dosing regimen and minimal effective concentration in more severe infection remain a priority for clinical translation.

One important consideration when interpreting the viral load results is our use of lignocaine to reduce inhalation discomfort, given that lignocaine has previously been shown to have direct anti-viral effects.³⁰ However, lignocaine was included across both ethanol and saline control exposures, making it unlikely to be the primary driver of the observed effect.

Across all tested concentrations, ethanol’s primary immunomodulatory effect appeared to be in potentiating the macrophage-dominant host response seen in control animals. Interestingly, we observed a relative increase in macrophage across all ethanol-exposure groups, but a viral load reduction only in the 80% group, suggesting the concentration required for immunomodulation may be lower than that for antiviral activity. Macrophages have previously been shown to play an essential protective role in the host immune response to influenza,³¹ and in this model, we observed a weak inverse relationship between higher macrophage numbers and lower viral load. The role of IL-6 in influenza infection is less clear with studies showing mixed associations with both improved and worse outcomes, depending on the model.³² In contrast to the observed dose-dependent relationship between increasing ethanol concentration and increasing macrophage numbers, a more modest effect on cytokine expression was observed with no clear dose-dependent relationship, suggesting ethanol may not be directly regulating their expression in this model. However, while we observed a relative potentiation of macrophages and no-inhibition of TNFα or IL-6 expression, previously-published work has frequently observed suppression of all these factors in different models of ethanol exposure,^11,33 which again highlights the complexity of ethanol’s properties likely related to timing, dose, and duration of exposure. Future studies aimed at clarifying the role of other key immune players, including chemokines responsible for monocyte/macrophage recruitment (e.g. monocyte chemoattractant protein-1),^34,35 will be important to characterize the full spectrum of ethanol’s antiviral and immunomodulatory properties. However, ethanol has previously been shown to have broad-spectrum immunomodulatory properties and to interact with multiple key signaling molecules/pathways (e.g. nuclear factor kappa B, toll-like receptor, and inflammasome pathways)^6,36,37 that are critically involved in the influenza-immune response³⁸ so clarifying simple mechanistic relationships will be challenging.

One important consideration when interpreting our results, and those previously published, are the limitations of animal models of human inhalation dynamics.³⁹ Significant species differences in airway anatomy and breathing pattern influence aerosol distribution, deposition, and systemic absorption.³⁹ One critical difference is that inhaled ethanol exposure in rodent models tends to lead to high blood levels (0.05-0.5%),^11,40 while human studies have consistently shown much lower elevations (<0.01%) and a plateauing effect,^9,10 which may influence the translation of safety/efficacy assessments. For example, the transient features of intoxication observed in this model and similar rodent models,^11,40 would be of major concern if reproduced in humans, given the significant deleterious biological effects of acute alcohol intoxication.⁴¹ However, the reduced blood ethanol levels and lack of intoxication after high-dose ethanol inhalation in humans is reassuring.¹⁰

In summary, we have demonstrated the potential efficacy of short duration, high-dose exposures to inhaled aerosolised ethanol on established influenza infection. No significant safety concerns were identified. Our results complement a recent study showing safety and a protective effect of prophylactic inhaled aerosolised ethanol on lethal influenza infection,¹¹ and lend support to inhaled ethanol as a promising and safe candidate treatment. Further studies in other infection models and with optimized dosing regimens would be helpful in extending these results toward potential clinical translation.

Acknowledgments

We would like to thank Drs Kirsty Short and Keng Chew at University of Queensland for measurement of the influenza viral titres.

Disclosure statement

The authors report there are no competing interests to declare.

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article.

Additional information

Funding

This work was supported by a grant from the Telethon Kids Institute.