Pharmacodynamic and immunological interactions of amphotericin B formulations and voriconazole with human neutrophils against mature Scedosporium apiospermum and Fusarium spp. biofilms

Abstract

Background

Mould infections caused by Scedosporium apiospermum and Fusarium solani species complex (FSSC) biofilms are rising among immunocompromised and immunocompetent patients. Little is known about the immunomodulatory effects of antifungal agents against these moulds. We examined the effects of deoxycholate and liposomal amphotericin B (DAmB, LAmB) and voriconazole on antifungal activities and immune responses of neutrophils (PMNs) against mature biofilms compared with their planktonic counterparts.

Methods

Antifungal activity of human PMNs exposed to mature biofilms and planktonic cells for 24 h was determined at effector-to-target ratios of 2:1 and 5:1, alone or combined with DAmB, LAmB and voriconazole, assessed as fungal damage by XTT assay. Cytokine production was evaluated by multiplex ELISA, following PMN stimulation with biofilms in the presence/absence of each drug.

Results

All drugs showed additive or synergistic effects with PMNs against S. apiospermum at 0.03–32 mg/L. They showed antagonism primarily against FSSC at 0.06–64 mg/L. Increased IL-8 was produced by PMNs exposed to S. apiospermum biofilms plus DAmB or voriconazole compared with PMNs exposed to biofilms alone (P < 0.01). During combined exposure, IL-1β was increased, an effect only counteracted by increased levels of IL-10 caused by DAmB (P < 0.01). LAmB and voriconazole caused similar IL-10 levels with those released by biofilm-exposed PMNs.

Conclusions

The synergistic, additive or antagonistic effects of DAmB, LAmB or voriconazole on biofilm-exposed PMNs are organism-specific, with FSSC exhibiting greater resilience than S. apiospermum to antifungals. Biofilms of both moulds caused dampened immune responses. The drug-mediated immunomodulating effect on PMNs, evidenced by IL-1β, enhanced host protective functions.

Introduction

Invasive infections due to non-Aspergillus hyphomycetes, such as Scedosporium apiospermum and Fusarium solani species complex (FSSC) are an important cause of morbidity and mortality, especially among immunocompromised hosts. They are able to cause a broad spectrum of infections such as vision-threatening keratitis, as well as superficial, invasive or disseminated infections.^1–3 Treating these infections remains a real challenge as many parameters are taken into consideration for clinical efficacy, such as degree of resistance against antifungals, innate immune response, immunosuppression, implanted medical devices or biofilm formation.^1,4

Biofilms are composed of fungal cells embedded in a self-secreted gelatinous matrix composed of extracellular polymeric substances. The presence of the extracellular matrix has been shown to impair antifungal penetration, making fungal cells highly resistant to many antifungal agents.^5–8 Several studies on biofilm formation have demonstrated for both S. apiospermum and FSSC that amphotericin B formulations are more effective than voriconazole against both organisms.^9–12

The refractoriness of biofilm-related infections to commonly used antifungal therapies highlights the importance of innate host defences against them. The innate immune response varies, depending on species and fungal growth form.¹³ Human neutrophils are essential in the first line of defence against invasive fungal infections. Upon infection, recruited human neutrophils exhibit a variety of mechanisms, antifungal oxidative and non-oxidative ones, and also contribute to the signalling network inducing the synthesis of pro-inflammatory cytokines in order to recruit phagocytic cells and promote fungal death or growth delay.¹⁴ Clinical immunomodulation approaches hold promise for the treatment of fungal infections in patients with an impaired immune system. However, most immunomodulating agents are currently under evaluation in clinical trials.¹⁵ Antifungal agents have been shown to modulate immune responses by releasing pro- and anti-inflammatory cytokines, affecting production of reactive oxygen species or enzymatic pathways or indirectly affecting phagocytic activities by alteration of antifungal morphology.¹⁶ Although resistance profiles of S. apiospermum and FSSC biofilms to antifungals have been reported in several studies,^{9–12,17–19} a knowledge gap exists on the influence of antifungals to modulate the fungicidal activity and cytokine production of innate immune cells against biofilms. Knowledge on the interactions among pathogen, antifungal agent and innate immune cells is vital in the fight against drug-resistant infections.

In order to gain insight into the immunopathogenesis of infections due to S. apiospermum and FSSC, we investigated the immunomodulatory potential of deoxycholate amphotericin B (DAmB), liposomal amphotericin B (LAmB) and voriconazole on the oxidative antifungal activities of human polymorphonuclear leucocytes (PMNs) against biofilms of three clinical isolates of each organism and their planktonic counterparts. Additionally, we studied whether the drugs under evaluation can modify immune responses of PMNs exposed to S. apiospermum and FSSC biofilms, to the benefit of the host in its fight against the fungus.

Materials and methods

Clinical isolates, growth conditions and biofilm formation

Three isolates of S. apiospermum and three isolates of FSSC were recovered from lung tissue, bronchial secretions, trauma sites and corneas of adult patients with proven invasive scedosporiosis or fusariosis. Stocks were maintained on Sabouraud dextrose agar (Mast Group Ltd, Liverpool, UK) slants at −30°C. Clinical isolates were revived on potato dextrose agar plates at 37°C for 4 to 6 days and the sporangiospores were subsequently harvested in PBS solution (0.02 M phosphate, 0.15 M NaCl; pH 7.2) containing 0.05% Tween 80. The concentration of each organism (per mL of solution) was determined using a haemocytometer. MLST identified the three FSSC isolates as Fusarium petroliphilum, Fusarium metavorans and Fusarium solani. Both molecular identification and biofilm formation were determined as previously described.⁹ Biofilms of each strain were formed on 96-well flat-bottomed polystyrene plates using 10⁵ cfu/mL in RPMI-1640 (AppliChem GmbH, Darmstadt, Germany) at 37°C for 48 h. Prior to antifungal or PMN incubations, formed biofilms were washed once with sterile distilled water to remove planktonic cells.

Antifungal agents and susceptibility assessment

The antifungal agents used were DAmB (Sigma–Aldrich), LAmB (Gilead Sciences, Inc.) and voriconazole (Sigma–Aldrich). DAmB and voriconazole were dissolved in DMSO (Sigma–Aldrich) to a stock concentration of 5 and 18 mg/mL, respectively, and stored at −30°C. LAmB was reconstituted in sterile water to 5 mg/mL (powder containing 50 mg of amphotericin B dissolved in 10 mL of sterile water) and was stored at 4°C. Each drug was further diluted to 1024 mg/L in RPMI-1640 medium and used to produce double-fold serial dilutions ranging from 0.007 to 256 mg/L. For comparative purposes, the susceptibility profile of planktonic cells and biofilms to each antifungal agent was determined by the XTT reduction assay, as previously described.⁹

The MIC₅₀ of both growth forms was determined as the minimum antifungal drug concentration that caused ≥50% fungal damage compared with untreated controls. The average planktonic MIC₅₀ of DAmB, LAmB and voriconazole against S. apiospermum and FSSC was 0.25/0.5/0.125 mg/L and 0.125/0.25/1 mg/L, respectively. The corresponding average biofilm MIC₅₀ of DAmB, LAmB and voriconazole against S. apiospermum and FSSC was 1/2/32 mg/L and 0.5/1/>256 mg/L, respectively.⁹

Isolation of human neutrophils

PMNs were isolated from healthy volunteers by dextran sedimentation of RBCs and separated from the remaining leucocytes by centrifugation over Ficoll (Lymphocyte Separation Medium, Gibco BRL/Life Technologies Ltd, Paisley, Scotland) at 400× g for 20 min. Following osmotic lysis of contaminating RBCs, PMNs were washed and resuspended in Hanks’ balanced salt solution (HBSS) without Ca²⁺ and Mg²⁺ ions (Gibco BRL) and counted on a haemocytometer by trypan blue exclusion staining (Sigma Chemical Co., St. Louis, USA). The cell viability was >95% for all blood donor experiments.¹³ PMNs were adjusted to 10⁶ cells/mL in RPMI-1640 supplemented with 10% pooled human serum.

Evaluation of antifungal activity of PMNs against planktonic cells and biofilms alone or in combination with antifungal agents

Planktonic cells (2 × 10⁵ cfu/mL) and 48 h biofilms (10⁵ cfu/mL) of S. apiospermum and FSSC were coincubated with DAmB, LAmB or voriconazole and PMNs, at an effector-to-target ratio of 2:1 or 5:1 at 37°C in an atmosphere of 5% CO₂ for 24 h. Three clinically relevant drug concentrations were selected for DAmB and LAmB: one was the MIC₅₀ and the other two were two- and three-fold dilutions above and below the MIC₅₀ value, respectively. We chose to use two drug concentrations for voriconazole because MIC₅₀ of FSSC biofilms was >64 mg/L. Specifically, planktonic cells of S. apiospermum were treated with 0.03/0.25/1 mg/L DAmB, 0.06/0.5/2 mg/L LAmB and 0.015/0.125 mg/L voriconazole. Biofilms of S. apiospermum were exposed to 0.125/1/16 mg/L DAmB, 0.25/1/16 mg/L LAmB and 0.5/32 mg/L voriconazole. Similarly, for FSSC the drug concentrations used for planktonic cells were 0.03/0.125/0.5 mg/L DAmB, 0.06/0.25/1 mg/L LAmB and 0.06/1/4 mg/L voriconazole, whereas the corresponding drug concentrations against biofilms were 0.06/0.5/4 mg/L DAmB, 0.25/2/32 mg/L LAmB and 8/64 mg/L voriconazole. Control wells contained drug-free organisms. Antifungal activity was assessed by the XTT reduction assay. Each experimental condition was tested in octuplicate. Eight independent experiments were performed and two healthy blood donors/experiment were used.

Evaluation of cytokine production against biofilms of S. apiospermum and FSSC

For cytokine evaluation, one strain from each fungal organism was randomly chosen. Cytokines were measured in the culture supernatants of PMNs exposed to 48 h biofilms of S. apiospermum or FSSC in the presence or absence of each drug at 37°C and 5% CO₂ for 3 h. The effector-to-target ratio used was 5:1 and the concentrations of DAmB, LAmB and voriconazole were 1/2/32 mg/L and 0.5/2/64 mg/L against the biofilms of each organism, respectively. Cytokines released in culture supernatants were quantitated using the 16-plex human cytokine assay of multiplex ELISA designed to detect IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12p70, IL-13, IL-15, IL-17, IL-23, IFN-γ, TNF-α and TNF-β (Quansys Biosciences, West Logan, UT, USA). Chemiluminescence intensities were captured by microplate Q-view imager and log-based regression data analysis with a log-to-log curve fitting model was performed using Quansys Q-view software. Two replicates were used per experimental condition and two independent experiments were performed, including two different donors per experiment.

Statistical analysis

The average values of the replicate wells from each condition were used in the data analysis to determine the mean ± standard error (SE) for all the experiments. The percent antifungal damage of the PMNs + drug treatment was compared with PMNs or drug alone by ANOVA (n = 8, P < 0.05). Synergism was defined as an antifungal effect caused by the combination treatment that was significantly greater than the effect produced by the PMNs alone plus the effect of the drug alone. Additivity was defined as significantly greater damage caused by the combination than by either PMNs or drug alone. Antagonism was defined as the effect of the combination that was significantly less than the effect produced by either PMNs or antibiotic alone. The statistics program InStat (GraphPad, Inc., San Diego, CA, USA) was used. Parametric analysis of variance (ANOVA) with Dunnett’s multiple comparisons test was used for statistical comparisons of cytokine concentrations between PMNs alone, PMNs treated with antifungal agents alone, PMNs treated with planktonic cells or biofilms and PMNs treated with a combination of planktonic cells or biofilms and drugs. A P value of <0.01 was considered statistically significant.

Results

Antifungal activity of PMNs alone or in combination with DAmB, LAmB or voriconazole against S. apiospermum and FSSC planktonic and biofilm growth forms

Both organisms exhibited similar susceptibility profiles to PMN exposure as the PMN-mediated damage was greater against planktonic cells than biofilms, whereas the maximum damage effected by PMNs was less than 60% and 20%, respectively. Specifically, after 24 h of treatment, PMN-mediated damage against S. apiospermum planktonic cells reached 55% ± 5% compared with 12% ± 2% against biofilms, whereas the corresponding damage against FSSC was 45% ± 10% compared with 4.5% ± 2% against biofilms (Figures 1a and b and 2 a and b).

Figure 1.

Antifungal activity of PMNs against S. apiospermum planktonic cells (a) and biofilms (b) alone or combined with DAmB at concentrations of 0.03, 0.25, 1 mg/L (a) and 0.125, 1, 16 mg/L (b), LAmB 0.06, 0.5, 2 mg/L (a) and 0.25, 2, 16 mg/L (b) or voriconazole (VRC) 0.015, 0.125 mg/L (a) and 0.5, 32 mg/L (b). Additive (*), synergistic (✦) or antagonistic (×) effects are indicated (n = 8; P < 0.01). The percent antifungal damage of the PMNs + drug treatment was compared with PMNs or drug alone.

Figure 2.

Antifungal activity of PMNs against FSSC planktonic cells (a) and biofilms (b) alone or combined with DAmB at concentrations of 0.03, 0.125, 0.5 mg/L (a) and 0.06, 0.5, 4 mg/L (b), LAmB 0.06, 0.25, 1 mg/L (a) and 0.25, 2, 32 mg/L (b) or voriconazole (VRC) 0.06, 1, 4 mg/L (a) and 8, 64 mg/L (b). Additive (*) or antagonistic (×) effects are indicated (n = 8; P < 0.01). The percent antifungal damage of the PMN s+ drug treatment was compared with PMNs or drug alone.

However, when voriconazole (0.5 and 32 mg/L), DAmB (0.125 mg/L) and LAmB (0.25 mg/L) were added to PMNs exposed to biofilms of S. apiospermum, additive and synergistic effects were observed compared with the damage induced by PMNs alone on biofilms (voriconazole: 19.5% ± 5.5% and 69.5% ± 2.6%, DAmB: 31% ± 5.1%, LAmB: 15% ± 3.5%, 56% ± 8.4% versus 12% ± 2.1%; Figure 1b). Of note, 1 and 16 mg/L DAmB and 2 and 16 mg/L LAmB led to antagonistic effects against S. apiospermum biofilms, suggesting that higher drug concentrations may produce toxicity issues when amphotericin formulations are used in the clinic (Figure 1b). For planktonic cells, additive effects were observed with all drug combinations tested compared with PMNs alone, enhancing the activity of PMNs from 3% to 30%, depending on the drug concentration used (Figure 1a).

In contrast, except for the additive effect observed with the combination treatment of PMNs + LAmB at 2 mg/L (56% ± 11.3% versus 4.5% ± 2%), the remaining drug combinations against FSSC biofilms produced antagonistic effects compared with the damage induced by the comparator component of PMNs against biofilms (Figure 2b). Similar antagonistic effects were also observed against planktonic cells of FSSC, except for the condition of PMNs + DAmB used at 0.125 mg/L and 0.5 mg/L (54% ± 14% and 84% ± 9%, P < 0.01; Figure 2). It is worth noting that while each drug caused an overall increased damage either against biofilms (induced damage: 53% to 90%) or their planktonic counterparts (induced damage: 37% to 93%), the combination of PMNs + drug produced primarily a negative effect, suggesting that inhibitory factors in the tripartite interaction of drug–FSSC–PMNs could lead to fungal growth.

Evaluation of cytokine production from PMNs exposed to biofilms of S. apiospermum and FSSC in the presence of DAmB, LAmB or voriconazole

Multiplex analysis of cytokines in the culture supernatants of each experimental condition detected IL-8, IL-1β and IL-10. For both S. apiospermum and FSSC, the amounts of IL-8 and IL-10 were significantly increased in PMNs exposed to DAmB, LAmB or voriconazole compared with that produced by PMNs exposed to biofilms or the combination treatment (P < 0.01; Figures 3a and c and 4a and c). Furthermore, for S. apiospermum, whereas DAmB and voriconazole caused PMNs to respond with significantly increased IL-8 levels against biofilms compared with those produced when PMNs were stimulated with biofilms alone (P < 0.01; Figure 3a), such an effect was not observed in the case of FSSC biofilms, as similar IL-8 amounts were released by PMNs in response to either of the above treatments (not significant; Figure 4a). The three drugs caused significantly increased IL-1β levels to be released by PMNs in the tripartite interaction compared with those produced by either biofilm- or drug-treated PMNs (P < 0.01; Figures 3b and 4b). The inflammatory response was counteracted only by DAmB which, in the combination treatment, caused the release of significantly elevated IL-10 levels by PMNs compared with those produced by biofilm-treated PMNs (P < 0.01, Figures 3c and 4c).

Figure 3.

Profiles of IL-8 (a), IL-1β (b) and IL-10 (c) cytokines released after treatment of human PMNs (white column) with 1 mg/L DAmB, 2 mg/L LAmB and 32 mg/L voriconazole (VRC) (grey columns), S. apiospermum biofilms alone (black column) or with the combination of biofilms plus each antifungal agent at the same concentrations (striped columns). Statistically significant differences are indicated by asterisks (n = 4; P < 0.01).

Figure 4.

Profiles of IL-8 (a), IL-1β (b) and IL-10 (c) cytokines released after treatment of human PMNs (white column) with 0.5 mg/L DAmB, 2 mg/L LAmB or 64 mg/L voriconazole (VRC) (grey columns), FSSC biofilms alone (black column) or with the combination of biofilms plus each antifungal agent at the same concentrations for 3 h (striped columns). Statistically significant differences are indicated by asterisks (n = 4; P < 0.01).

Discussion

In this study we found that the combined treatment of PMNs with DAmB or LAmB show synergistic or additive activities at lower therapeutic concentrations compared with voriconazole against S. apiospermum biofilms and their planktonic counterparts. In contrast, combined treatment of PMNs with AmB or voriconazole against FSSC show primarily antagonistic activities.

Amphotericin B formulations have been previously shown to variably enhance antifungal activity of PMNs against early formed hyphal elements of Aspergillus fumigatus and F. solani. LAmB augments the conidiocidal activity of both murine and human PMNs, suggesting a possible immunomodulatory role on PMN antifungal functions by augmenting conidial susceptibility to PMNs (see Simitsopoulou et al.¹⁶ and references therein). Earlier studies conducted in our laboratory investigating the combined effects of amphotericin B formulations with PMNs against S. apiospermum and Scedosporium prolificans hyphae demonstrated that the combination treatment produced significantly enhanced hyphal damage compared with PMNs or drugs alone.¹³ Voriconazole, posaconazole or itraconazole exhibited synergistic or additive effects with PMNs against S. apiospermum and S. prolificans hyphal elements.²⁰ Conventional amphotericin B and lipid formulations of amphotericin B displayed enhanced antihyphal activity of human phagocytes against A. fumigatus or F. solani, although the latter fungal organism showed a tendency to be more resistant to oxidative antifungal mechanisms and presented decreased susceptibility to antifungal peptides or to other non-oxidative antifungal mechanisms.^13,21

Driven by our earlier studies demonstrating that amphotericin B formulations and voriconazole exhibit enhancing effects of variable degrees on human phagocytes against early hyphal elements of Scedosporium and FSSC, we proceeded to investigate whether such antifungal responses remain significant against mature biofilms of these organisms. To our knowledge, our study is the first to investigate PMN activity against mature biofilms and planktonic cells of Scedosporium and FSSC. Although Scedosporium and Fusarium are morphologically very similar to many other moulds with regular branched hyaline hyphal septation, based on our results the two organisms seem to have different antifungal strategies to counteract the antifungal activities of drugs and immune cells.

In this respect, our findings show that S. apiospermum amplify the demonstrated enhanced activity of PMNs in the presence of amphotericin B formulations and voriconazole, mainly at therapeutic dosages. However, the antagonism observed between PMNs and antifungal agents against FSSC biofilms, but also against their planktonic counterparts, shows that Fusarium may have developed key strategies that successfully promote fungal survival at the expense of the human host.

Additionally, both S. apiospermum and FSSC biofilms cause dampened immune responses exhibiting similar cytokine profiles. Following treatment of PMNs with biofilms and DAmB, LAmB or voriconazole, the drugs were shown to have an inflammation-increasing effect, which is counteracted by significantly increased levels of the anti-inflammatory cytokine IL-10, an effect however demonstrated only by DAmB, as LAmB and voriconazole cause comparable IL-10 levels with those produced when PMNs are exposed to biofilms alone. Interestingly, the drug-mediated immunomodulating effect observed with drug-treated PMNs does not seem to work to the benefit of the host shifting immune responses to an immune-enhancing effect. Differences in the biofilm structure and extracellular matrix components that make up the biofilm surface may dampen signal responses to fungal invasion resulting in weak stimulation of PMNs and reduced cytokine production.

Despite the constantly increasing clinical interest in Scedosporium and FSSC infections, little has been accomplished in recent years to improve the understanding of host defences against them and the role of cytokines in clinical practice. Two study groups investigated the cytokine levels in tears and aqueous humour of patients with fungal keratitis, with the majority of them induced by F. solani. Both showed predominant infiltration of PMNs in infected cornea and production of Il-1β, Il-6, Il-8, Il-10 and Inf-γ.^22,23 Use of the above cytokines, as adjunctive therapy for treatment of invasive fungal infections, sounds appealing but has not been systemically studied.

The observed antibiofilm activities of human neutrophils were probably due to oxidative and non-oxidative killing mechanisms with minimal involvement of immune cell responses. Moreover, the enhancing or antagonistic effects exhibited by antifungal agents acting upon PMNs exposed to biofilms were organism-specific with FSSC exhibiting greater resilience than S. apiospermum to host and antifungals. There is still much to be deciphered about the potential interplay between antifungal agents, fungal biofilms and phagocytes. So far, limited clinical data exist to assess the benefits of immunomodulatory therapy. Synergistic effects of antifungal agents with cytokines may offer new perspectives in invasive fungal infections.

Funding

This research was cofounded by Greece and the European Union [European Social Fund (ESF)] through the Operational Program Human Resources Development, Education and Lifelong Learning 2014–2020 in the context of the project ‘Pharmacodynamic and immunomodulatory activities of antifungal agents against clinically important hyphomycetes’ (MIS: 5047921).

Transparency declarations

E.R. has received research grant support from Pfizer and Gilead, has served as consultant to Astellas, Gilead, Pfizer and Merck, and has been in the speakers’ bureaus of Merck, Aventis, Astellas and Pfizer. The remaining authors have no relevant disclosures.