The Effect of the Novel Macrolide Glasmacinal (EP395) on Allergen-Induced Eosinophil Infiltration into the Lung

Abstract

Purpose

Patients with frequent exacerbations of COPD despite bronchodilator therapy may benefit from inhaled corticosteroids (ICS) if they have eosinophilic inflammation. For those who exacerbate despite bronchodilators plus ICS treatment, or who are not candidates for ICS, use of azithromycin (AZM) off-label may be considered as an add on therapy. Glasmacinal (EP395) is a non-antibiotic macrolide that enhances airway epithelial function and immune defence properties and reduces neutrophilic inflammation in preclinical models. Here, we present the effects of glasmacinal on eosinophilic inflammation in preclinical models of allergic inflammation.

Methods

Glasmacinal was tested in two established models of eosinophil-infiltration airway inflammation: an ovalbumin (OVA)-induced mouse model and a house-dust mite (HDM)-induced guinea pig model. Following sensitisation, animals were pretreated with glasmacinal for 2 weeks prior to challenge. In the OVA model, budesonide- and in the HDM model, AZM- treated animals were also included as comparators. Endpoints included differential cell counts, Th1 and Th2 cytokines in bronchoalveolar lavage fluid (BALF) and/or plasma, and lung histology and vascular permeability.

Results

Glasmacinal significantly reduced allergen challenge-induced eosinophilia in both models. In the OVA model, glasmacinal dose-dependently (0.4, 1, 2 mg/kg) decreased Th1 and Th2 cytokines, with significant reductions at 1 and 2 mg/kg. In the HDM model, glasmacinal preserved lung morphology, reduced mucus accumulation and collagen deposition, and significantly lowered plasma IL-4 and BALF IL-6 levels.

Conclusion

Glasmacinal attenuated eosinophilic airway inflammation in two preclinical models, supporting its potential as a novel therapy for reducing exacerbations in COPD and asthma.

Introduction

Macrolide antibiotics, such as azithromycin (AZM), taken chronically reduce exacerbations of respiratory diseases, including asthma [1] and chronic obstructive pulmonary disease (COPD) [2] likely due to anti-inflammatory and immunomodulatory properties. However, the molecular targets and mechanisms responsible for these effects remain poorly defined, while chronic treatment promotes antibiotic resistance [3]. We developed structural analogues of AZM which retain the anti-inflammatory and enhanced epithelial and immune defence properties of AZM but that lack antibacterial activity. This new class of drug exemplified by glasmacinal, formerly known as EP395, inhibits neutrophilic inflammation in the lung in several murine models [4], impacts neutrophil function in BALF in healthy volunteers exposed to inhaled bacterial lipopolysaccharide (LPS) [5], and reduces neutrophil activation in sputum in stable COPD patients [6].

Airway eosinophilia is a feature of asthma, inhaled corticosteroids (ICS) being the mainstay of treatment to reduce the recruitment and activation of eosinophils. A subgroup of patients with COPD also exhibit eosinophilic inflammation and blood eosinophil levels are used to guide treatment [7]. Patients with COPD exacerbating despite regular bronchodilator therapy may benefit from ICS, if they have evidence of eosinophilic inflammation [7]. For patients who continue to exacerbate despite treatment with bronchodilators plus ICS, or who are not candidates for ICS, off-label use of AZM may be considered as an add-on therapy [7].

Eosinophils release a range of cationic mediators and cytokines that contribute to the tissue damage and bronchial hyperresponsiveness that characterises asthma [8]. While corticosteroids reduce eosinophil recruitment and activation [9, 10], concerns remain about the long-term safety of this class of drug, particularly in children with asthma [11, 12], and in patients with COPD with an increased risk of pneumonia reported [13–15]. Biologics that reduce eosinophil recruitment and activation, including anti-IL-5 (mepolizumab) and anti-IL-4/IL-13 (dupilumab) [16, 17], are effective in both COPD and asthma patients with eosinophilic inflammation, but are administered by injection and expensive.

Although orally administered therapies are desirable from a patient perspective, licensed oral medications for respiratory indications are limited to montelukast (asthma) and roflumilast (severe COPD). Chronic off-label AZM use tends to be limited to patients without eosinophilic inflammation, due to the limited treatment options for that patient subtype, although those with eosinophilic inflammation also benefit [18, 19]. However, chronic use of AZM leads to antimicrobial resistance in both patients with asthma and COPD which is not desirable. We have therefore, investigated glasmacinal’s ability to reduce allergen-induced eosinophil recruitment to the lung as an AZM analogue lacking significant antimicrobial activity which may provide a viable alternative to macrolides for chronic future treatment of patients with respiratory diseases.

Materials and Methods

Materials

EpiEndo synthesised glasmacinal in-house. AZM was from Chemtronica, China and budesonide from MCE, Medchem, UK.

Animals

Mice

Male BALB/c mice (20–30 g, Charles River Laboratories, UK) were acclimatised for 7 days in cages of 4–5 animals based on weight, under a 12-h light dark cycle at 17–24 °C and 40–70% humidity. Environmental enrichment with ad libitum chow and water were provided. Group size was based on power analyses from similar studies. Cages were randomly assigned to treatment groups using Excel RAND functionality and positioned randomly. Blinded initial analysis was conducted based on group number and animal ID and unblinded after analysis completion.

Guinea Pigs

Male Dunkin Hartley guinea pigs (250–450 g; Marshall Bioresources, Hull, UK) were acclimatised for one week, housed ≤ 5 per cage at 17–24 °C and 40–70% humidity. Environmental enrichment and ad libitum chow and water were provided. Animals were randomly assigned to four groups: (n = 5–7): sham-sensitised, HDM-sensitised/vehicle treated (citrate buffer), HDM-sensitised/AZM treated and HDM-sensitised/glasmacinal treated.

Ovalbumin (OVA) Challenge in Mice

Mice were sensitised with subcutaneous injections of a 25 µL solution containing OVA (15 µg; Grade V, Sigma Aldrich) and 1 mg of aluminium hydroxide on days 1 and 7. To elicit an inflammatory response, animals were challenged on days 15, 16, and 17 with an aerosol of either 1% w/v OVA in PBS, or PBS alone, generated with an ultrasonic nebuliser (Aerogen) for 20 min. Animals (n = 10/treatment) in 6 groups were pretreated via oral gavage (10 mL/kg) once a day for 2 weeks (14 doses) with glasmacinal (0.4, 1.0, 2.0 mg/kg/day) or vehicle (citrate buffer, pH 4) or budesonide (10 mg/kg, 3 doses, given 1 h prior to challenge on days 15, 16 and 17). Administration started 3 days after first OVA sensitisation and ended 1 h before the final OVA challenge. Control groups were vehicle with and without OVA challenge.

House Dust Mite (HDM) Sensitisation and Challenge in Guinea Pigs

Ten milligrams of lyophilised HDM extract containing 430 µg of Dermatophagoides pteronyssinus extract (Citeq Biologics, NL) was diluted in 5 mL sterile saline to give 2 mg/mL solution of total protein and 86 µg/mL of the specific antigen. Guinea pigs were immunised subcutaneously in the dorsal neck region with 200 µL of a solution containing 1:1 HDM extract (2 mg/mL) and aluminium hydroxide (2.5 mg/mL; Sanofi, Brazil). Sham animals received aluminium hydroxide. The animals were injected on days 1, 2, 3, and 8, 9, 10 (six injections). On day 15, animals were anaesthetised with isoflurane (5%; 2 L/min) and challenged intranasally with 100 µL of HDM extract diluted in 100 µL sterile saline. Measurements were performed 48 h later. Two mg/kg glasmacinal or AZM diluted in citrate buffer, pH 4.0, or vehicle (citrate buffer) was administered by once daily oral gavage from day 3 to day 17.

Bronchoalveolar Lavage and Serum Collection

OVA Model

Twenty-four hours after the final OVA challenge, animals received an overdose of pentobarbitone. The airways were lavaged via a tracheal cannula and lungs flushed using 1 mL of PBS, repeated until the recovered volume was 1.6 mL bronchoalveolar fluid (BALF). BALF was centrifuged (1500 rpm, 10 min, 4 °C) and supernatant aliquots (400 µL) stored at -80 °C for cytokine analysis. Cell pellets were resuspended in 0.8 mL of 0.2% w/v NaCl to induce haemolysis of any erythrocytes. After isotonisation with the same volume of 1.6% w/v NaCl, the total and differential cell numbers were measured using a Sysmex XT-2000iV analyser (Kobe, Japan).

HDM Model

Forty-eight hours after HDM challenge, animals were anaesthetised with urethane (1.75 mg/kg; i.p.) and received 0.5 mL Evans blue dye i.v. (0.5% in saline). Bronchoalveolar lavage was performed 1 h later by injecting 5 mL saline into the airways using a tracheal cannula. The fluid was withdrawn and re-injected three times and collected for assessment of Evans blue extravasation, as an index of vascular permeability, inflammatory cell counts and cytokines. Blood samples were collected by cardiac puncture just before the bronchoalveolar lavage with syringes containing ACD (Citrate Dextrose Solution). Plasma was separated by centrifugation (1500 rpm, 5 min), supernatants collected and kept at − 80 °C until cytokine analysis.

Cytokine and Protein Measurement

Mouse BALF cytokine and mediator concentrations were measured using a multiplex assay using a Magpix system (Luminex). Additional proteins were measured using protein-specific ELISAs: IL-13 was measured as per manufacturer’s instructions (Krishgen Biosystems).

In guinea pigs, IL-4, IL-6, and IL-13 concentrations were determined in BALF and plasma by conventional ELISA (AssayGenie GPFI00031; Reed Biotech RE3186GP and Krishgen Biosystems KLY0128, respectively).

Lung Immunohistochemistry

Left lung lobes from 4 guinea pigs per group were removed and placed in a sterile container containing 10% neutral buffered formalin for 48 h before transfer to 70% ethanol for processing. Following fixation and processing in paraffin wax, sections (5 μm) were transversely cut for staining. Sections were stained with: (1) Haematoxylin and eosin (Sigma-Aldrich, GHS232 and 318906, respectively) to examine general morphology of the peribronchoalveolar infiltrated area (µm²); (2) Periodic acid Schiff staining (Schiff-PAS) (Sigma-Aldrich, 395B1-KT) to measure the percentage of marked areas showing goblet cell metaplasia; (3) Masson’s trichrome staining (Abcam, AB150686) to measure the percentage of collagen fibre deposition within a predefined marked area of tissue. Images were captured using a Leica DM2000 LED microscope with × 20 or × 40 magnification and analysed using ImageJ software (Fiji) (https://imagej.nih.gov/ij) or Image-Pro Plus v4.1. At least six randomly selected images per sample were analysed.

Cutaneous Allergen Response Test

A skin test was performed with allergen to confirm sensitisation [20] of guinea pigs as described elsewhere. Under urethane anaesthesia (Sigma, 2 mg/kg i.p.), each animal was injected with 0.5 mL of Evans blue dye (i.v. 0.5% in sterile 0.9% saline), followed immediately by injections of 100 µL of saline, HDM, or HDM diluted at 1:5, 1:50 or 1:500 in saline (HDM stock solution 2 mg/mL i.d.), at 5 sites/animal. After 1 h, 1 mm punch biopsies were collected, weighed, and incubated in formamide (4 mL/g tissue) for 24 h. An untreated skin sample was used as a control to determine dry/wet weight ratio (dried 72 h, 37 °C). The absorbance of Evans blue extracted from the skin sites was determined by spectrophotometry at 620 nm and the concentration was calculated using a standard curve (1000–1.5 µg/mL). Clean formamide was used as a negative control. The corrected dry/wet weight ratio was applied, and the volume of extravasated Evans blue expressed as µg of Evans blue (EB) per g of dry tissue.

Statistical Analysis

All raw data collected were recorded on Microsoft Excel® spreadsheets. In vivo data are reported as total and differential number of cells per mL of BALF or cytokine concentration (pg/mL) of BALF, with provision of individual animal data and group means ± S.E.M. In vitro data are reported as the means of technical replicates ± S.E.M. Data were analysed and results represented using GraphPad Prism 10.1. Outliers were identified and removed using the Robust Regression and Outlier Removal (ROUT) method in GraphPad Prism 10.1. Inter-group deviations for cell counts and cytokines were statistically analysed by a one-way ANOVA. In the case of significant differences in mean values among the different treatments, comparisons versus the vehicle control were carried out using a Dunnett’s test. In case the equal variance test failed, a Kruskal–Wallis ANOVA on ranks followed by a Dunn’s test was utilised.

Results

Glasmacinal Inhibits Cell Migration and Cytokine Release in a Dose-Dependent Manner in OVA Challenged Animals

Differential cell count analysis from BALF vehicle-treated mice after OVA challenge demonstrated significant increases in all cell subtypes (Fig. 1). Glasmacinal-treated mice exhibited a dose-dependent decrease in total, eosinophil, macrophage, and neutrophil cell counts, which was significant following 1 and 2 mg/kg. The positive control, budesonide, reduced all cell subtypes.

Fig. 1.

Glasmacinal inhibits infiltration in a dose-dependent manner. Bar graphs show effects on cells in BALF of OVA sensitised (days 1 and 7) and challenged (days 15, 16 and 17) animals treated once daily from days 4–17 with glasmacinal (0.4–2.0 mg/kg, p.o.), or vehicle (10 mL/kg, p.o.). Positive control animals were dosed with budesonide (10 mg/kg, p.o.) on days 15, 16 and 17. BALF was collected 24 h after the final OVA challenge. Each column represents the mean and each bar represents S.E.M. of n = 8–10. Vehicle + OVA is compared to each treatment groups using one-way ANOVA followed by Dunnett’s test. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05 versus vehicle

BALF was analysed for Th1 and Th2 cytokines (Fig. 2). Glasmacinal-treated mice had a dose-dependent reduction in cytokines which was significant for IL-13, IL-6, IL-1β, IL-5 and IL-17 following 1 and 2 mg/kg, with 2 mg/kg also significantly reducing IL-10. The positive control, budesonide, reduced all cytokines.

Fig. 2.

Murine OVA model and Th1 and Th2 cytokines. Bar graphs show the IL-13 (A), IL-5 (B), IL-10 (C), IL-6 (D), IL-1β (E) and IL-17 (F) concentrations (pg/mL) in BALF supernatant of OVA sensitised (days 1 and 7) and challenged (days 15, 16 and 17) animals treated once daily from days 4 to 17 with glasmacinal (0.4–2.0 mg/kg, p.o.), or vehicle (10 mL/kg, p.o.). Positive control animals were dosed with budesonide (10 mg/kg, p.o.) on days 15, 16 and 17. BALF was collected 24 h after the final OVA challenge. Each column represents the mean and each bar represents S.E.M. of n = 8–10. Vehicle + OVA is compared to each treatment groups using one-way ANOVA followed by Dunnett’s test. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05 versus vehicle

Pro-inflammatory cytokines and other mediators were also measured in BALF (Fig. 3). One and 2 mg/kg glasmacinal significantly reduced MCP-1 (or CCL2), matrix metalloprotease-12 (MMP-12) and thymic stromal lymphopoietin (TSLP), and 2 mg/kg significantly reduced TNF-α. Budesonide, reduced all these inflammatory mediators.

Fig. 3.

Murine OVA model and Th1 cytokines and other mediators. Bar graphs showing the TNF-α (A), MCP-1 (B), MMP-12 (C), TSLP (D) concentrations (pg/mL) in BALF supernatant of OVA sensitised (days 1 and 7) and challenged (days 15,16 and 17) animals treated once daily from day 4 to 17 with glasmacinal (0.4–2.0 mg/kg, p.o.), or vehicle (10 mL/kg, p.o.). Positive control animals were dosed with budesonide (10 mg/kg, p.o.) on days 15, 16 and 17. BALF was collected 24 h after the final OVA challenge. Each column represents the mean and each bar represents S.E.M. of n = 8–10. Vehicle + OVA is compared to each treatment groups using ANOVA followed by Dunnett’s test. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05 versus vehicle

Glasmacinal Inhibits Cell Migration and Reduces Vascular Permeability in HDM Challenged Animals

In vehicle-treated HDM-exposed guinea pigs, eosinophils and macrophages infiltrated into BALF. Two week-pretreatment with either AZM or glasmacinal (2 mg/kg) significantly reduced total cell (Fig. 4A) and eosinophil cell numbers (Fig. 4B).

Fig. 4.

Glasmacinal (Glas.) ameliorates inflammatory cell infiltration into the lung. Bar graphs show the effect on total cells (A) and differential cell populations (B) in BALF 48 h after HDM challenge in guinea pigs. Animals were treated once daily from days 3 to 17 with glasmacinal (2.0 mg/kg, p.o.), AZM, (2.0 mg/kg, p.o.) or vehicle. Each column represents the mean and S.E.M. of n = 5–7 animals. Vehicle + HDM-challenged animals were compared to each treatment group using ANOVA followed by Tukey’s post hoc test. ***p < 0.001; **p < 0.01; *p < 0.05

Qualitative analysis of the lung tissue sections stained with conventional H&E showed an accumulation of leukocytes around the airways and visible hypertrophy of the epithelial layer of HDM-exposed animals in comparison to sham animals (Fig. 5A, 1–4). Treatment with AZM or glasmacinal markedly reduced the accumulation of cells around the airways, suggesting that the inhibition of BALF cell counts was probably due to less inflammatory cells migrating into the lung tissue (Fig. 5A, 5).

Fig. 5.

Histological analysis of guinea pig lung tissue. Representative images of all groups: (A) H&E showing total cells; (B) PAS showing mucus accumulation; (C) Masson's trichome showing peribronchial collagen deposition. Images were taken with a 20 × objective lens using a bright field microscope (scale bars depict 50 µm). Animals were treated once daily from days 3 to 17 with glasmacinal (2 mg/kg, p.o.), AZM (2 mg/kg, p.o.), or vehicle. Arrows indicate positively marked areas. (5) Quantification of each column represents the mean and S.E.M. of n = 4 animals. Vehicle treated, + HDM challenged animals were compared to each treatment group using two-ways ANOVA followed by Tukey’s post hoc test. ****p < 0.0001; ***p < 0.002; *p < 0.05

Similar qualitative evaluation of PAS-stained lung sections revealed that glasmacinal treatment, as with AZM, was associated with reduced mucus production (Fig. 5B, 1–5). Assessment of remodelling by Masson’s trichome staining further showed decreased collagen deposition in all treatment groups relative to the HDM control group (Fig. 5C, 1–5).

Glasmacinal Affects the Levels of both Systemic and Local Cytokines in HDM Model

Cytokines were measured in BALF and plasma (Fig. 6). Th2 cytokine IL-4 in plasma was significantly reduced following treatment with glasmacinal and AZM compared to vehicle plus HDM (Fig. 6B). No treatment effect was demonstrated on BALF IL-13 (Fig. 6E). BALF IL-6 was significantly reduced by glasmacinal compared to vehicle plus HDM (Fig. 6C).

Fig. 6.

Glasmacinal reduced cytokines in BALF and plasma after HDM challenge. Bar graphs show the effect of glasmacinal on cytokines in BALF (A, C, E) and plasma (B, D, F) of guinea pigs HDM-sensitised (days 1, 2, 3 and 7, 8 and 9, s.c.) and challenged (day 15, i.n.). Each column represents the mean and S.E.M. of n = 5–7 animals. In vehicle treated, + HDM challenged animals, IL-4 in BALF and IL-6 and IL-13 in plasma were unchanged, obviating a treatment effect on these endpoints. Vehicle-treated, + HDM-challenged animals were compared to each treatment group using 2-way ANOVA followed by Tukey’s multiple comparison test. ****p < 0.0001; ***p < 0.002; **p < 0.01

Glasmacinal does not Affect Sensitisation to Allergen

Skin testing with HDM resulted in dose-dependent acute oedema, as assessed by the intensity of Evans blue dye extravasation, demonstrating successful sensitisation to HDM. AZM or glasmacinal pre-treatment did not impact sensitisation (Fig. 7).

Fig. 7.

Effects of glasmacinal (glas.) and AZM on the HDM-induced skin response. Bar graph shows the effect of glasmacinal on the skin test response of HDM sensitised (days 1, 2, 3 and 7, 8 and 9, s.c.) and challenged (day 15, i.n.) guinea pigs. Animals were treated once daily from day 3 to 17 with vehicle, glasmacinal (2.0 mg/kg, p.o.) or AZM (2.0 mg/kg, p.o.). Each column represents the mean and S.E.M. of n = 5–7 animals

Discussion

In the studies described here, we demonstrate that the novel, non-antibiotic macrolide, glasmacinal, reduces allergen-induced infiltration of eosinophils into the lung in two species. This extends previous findings that glasmacinal inhibits neutrophilic inflammation in the lung in murine models [4] and COPD patients [6].

Eosinophils are important in the pathogenesis of asthma [21] and a sub-group of patients with COPD [22]. Eosinophilic inflammation is treated by ICS and monoclonal antibodies that inhibit cytokines such as IL-5 and IL-4/IL-13. However, there is concern about the long-term safety of ICS (e.g. pneumonia risk in COPD [9, 11]) and biologics require systemic administration and are expensive. Chronic use of antibiotic macrolides to prevent exacerbations of respiratory diseases is off-label and used primarily in patients having predominantly neutrophilic inflammation, although chronic use of AZM has also shown benefit in patients with eosinophilic asthma [19] and COPD [18]. Our data in two established models of eosinophilic inflammation indicate that glasmacinal may be beneficial in patients with eosinophilic lung inflammation and support the inclusion of such patients in future clinical trials of glasmacinal. Ultimately, glasmacinal may become an option for patients who are intolerant of, or who have incomplete responses to, current treatments targeting eosinophilic inflammation and provide an alternative to the current practice of using chronic AZM off-label.

In the OVA model, associated with a strong adaptive Th2 inflammation, glasmacinal suppressed eosinophil activation via reduction of Th2 cytokines. The concomitant reductions in OVA-induced neutrophil, and to a less extent macrophage infiltration, in Th1 cytokines, as well as TSLP, MCP-1 and MMP-12 also suggest that glasmacinal acts broadly across the inflammatory cascade, impacting early epithelial-derived markers through to macrophage mediators. This confirms previous findings demonstrating that glasmacinal inhibits neutrophilic inflammatory responses, as well as promoting the bronchial epithelial barrier [4, 23].

In the HDM model, a more complex natural allergen model, glasmacinal reduced both localised inflammatory and systemic signalling. However, the lack of effect on BALF IL-13, a main effector in the HDM model, may indicate that the timing of the endpoint assessments, 48 h after HDM challenge, was not optimal. Macrolides are known to exert biphasic, time-dependent actions [24], making it difficult to pinpoint the optimal timepoints for measuring delayed, potentially anti-inflammatory effects, especially between models and species. Importantly, glasmacinal did not affect HDM-induced sensitisation in the skin, which suggests it was not affecting the process of allergen presentation for Th2 activation and antibody generation. However, HDM extract contains a complex mixture of allergens, which limits the assessment of effects on the generation of specific antibody responses. Glasmacinal treatment decreased mucus generation in lung histological sections, which would be an additional desirable effect in patients with mucus hypersecretion. The specificity of this action of glasmacinal—whether in addition to or a result of leukocyte effects—remains to be investigated. Additionally, at the histological level, both glasmacinal and AZM appear to preserve the overall tissue structure following HDM challenge, supporting previous studies on the epithelial protective effects of the two treatments [23, 25, 26].

Translation from animal models to clinical benefit is uncertain but is mitigated by using two species and positive comparator controls. While glasmacinal reduced the levels of selected pro-inflammatory cytokines, this was not consistent across both models, and the exact mechanism by which glasmacinal and AZM act as anti-inflammatory agents remain unknown. Both models employ systemic rather than airway sensitisation, which may produce distinct immune responses and potentially different treatment responsiveness. Future studies using airway sensitisation models may help clarify the translational relevance of these findings. In addition, treatment was initiated during sensitisation, so the observed effects of the OVA model may reflect modulation of sensitisation and/or induction rather than established exacerbation responses, but the lack of effect on skin sensitisation argues against this. Future studies initiating treatment after sensitisation are needed to distinguish these mechanisms. Despite these limitations, the data support the inclusion of patients with eosinophilic inflammation in future clinical trials with glasmacinal.

In conclusion, our results suggest that glasmacinal has potential as a treatment for respiratory diseases with lung eosinophilia. Glasmacinal could facilitate restriction of AZM use to short-term treatment of respiratory infections, therefore limiting antimicrobial resistance development. The reported results, together with observations that glasmacinal is effective against airway neutrophil inflammation, suggest glasmacinal has broad anti-inflammatory effects of relevance to the treatment of COPD and asthma.

Author Contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Y.R.V. (Figs. 4, 5, 6, 7) and Pharmidex (Figs. 1, 2, 3). The first draft of the manuscript was written by J.A.K., C.P.P and M.J.P. All authors, J.A.K, Y.R.V, V.N., C.P.P. and M.J.P., contributed to writing, reviewing, and editing versions of the manuscript. All authors read and approved the final manuscript.

Funding

This study was supported by the European Innovation Council Accelerator Grant 947081.

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

Financial interests: Authors J.A.K., M.J.P., and V.N. are employees of EpiEndo Pharmaceuticals. Authors J.A.K., M.J.P., V.N., and C.P.P. are either shareholders or have options in EpiEndo Pharmaceuticals. C.P.P. also declares he has received consultancy and speaker fees from Recipharm and Eurodrug; he is also an NED and has equity in Ananda Pharma. Author Y.R.F. declares they have no financial interests. M.J.P. teaches at and receives travel expenses from Goethe University Frankfurt, is a co-founder of and holds equity in Phialogics AG and is a shareholder in Cyclone Therapeutics Inc. and Oxford Biomedica plc.

Ethics Approval

These studies were performed in line with the principles of the Declaration of Helsinki. The study involving guinea pigs was reviewed and approved by the local Animal Welfare and Ethical Review Board (AWERB) of King’s College London and experiments were carried out under UK Home Office Project Licence PP2935715 in strict accordance with the Animals Act 1986 and the ARRIVE guidelines. Studies in mice were carried out at Pharmidex in the UK covered by a UK Home Office Project Licence P8AE03703 and followed pre-prepared protocols. The study involving guinea pigs was also approved.