Pseudomonas aeruginosa increases the susceptibility of Candida albicans to amphotericin B in dual-species biofilms

Abstract Background Biofilms are the leading cause of nosocomial infections and are hard to eradicate due to their inherent antimicrobial resistance. Candida albicans is the leading cause of nosocomial fungal infections and is frequently co-isolated with the bacterium Pseudomonas aeruginosa from biofilms in the cystic fibrosis lung and severe burn wounds. The presence of C. albicans in multispecies biofilms is associated with enhanced antibacterial resistance, which is largely mediated through fungal extracellular carbohydrates sequestering the antibiotics. However, significantly less is known regarding the impact of polymicrobial biofilms on antifungal resistance. Results Here we show that, in dual-species biofilms, P. aeruginosa enhances the susceptibility of C. albicans to amphotericin B, an effect that was biofilm specific. Transcriptional analysis combined with gene ontology enrichment analysis identified several C. albicans processes associated with oxidative stress to be differentially regulated in dual-species biofilms, suggesting that P. aeruginosa exerts oxidative stress on C. albicans, likely through the secretion of phenazines. However, the mitochondrial superoxide dismutase SOD2 was significantly down-regulated in the presence of P. aeruginosa. Monospecies biofilms of the sod2Δ mutant were more susceptible to amphotericin B, and the susceptibility of these biofilms was further enhanced by exogenous phenazines. Conclusions We propose that in dual-species biofilms, P. aeruginosa simultaneously induces mitochondrial oxidative stress, while down-regulating key detoxification enzymes, which prevents C. albicans mounting an appropriate oxidative stress response to amphotericin B, leading to fungal cell death. This work highlights the importance of understanding the impact of polymicrobial interactions on antimicrobial susceptibility.


Introduction
Candida albicans is a commensal and opportunistic fungal pathogen of humans that is frequently co-isolated from infection sites with the Gram-negative bacterium Pseudomonas aeruginosa. 1 Extensive research has shown that these two microbes undergo a complex range of interactions, the outcome of which appears to be dependent on the surrounding environment.For example, P. aeruginosa has been shown in vitro to bind and kill C. albicans hyphae, 2 whereas in vivo, P. aeruginosa and C. albicans have been shown to have an agonistic relationship, resulting in enhanced pathogenesis. 3P. aeruginosa secretes a broad range of virulence factors including phenazines, siderophores, haemolysins and phosphatases, which are regulated by one or more quorumsensing systems.Several of these secreted products have been shown to directly affect C. albicans, with the phenazines resulting in the induction of reactive oxygen species (ROS)-induced stress, and inhibition of fungal growth and morphogenesis, 4,5 whereas siderophores impose nutrient restriction.On the other hand, C. albicans secretes the quorum-sensing molecule farnesol, which inhibits the Pseudomonas quinolone signal pathway, 6 and imposes ROS-induced stress on other microbes. 7,8. albicans and P. aeruginosa are most commonly co-isolated from the lungs of cystic fibrosis patients, 9 and from wounds of burn victims, 10 where the two microbes tend to grow together in biofilms.Biofilms are communities of cells encased in an extracellular matrix (ECM), which provides protection from the host's immune responses, and increases the resistance of microbes to antimicrobial agents.Cells in a monospecies biofilm can exhibit MIC values 100-1000-fold greater than their planktonic counterparts.However, in mixed-species biofilms, the microbes interact through secreted signalling molecules, direct cell-cell interactions and through nutrient competition, 11 all of which have the potential to affect antimicrobial resistance.Furthermore, each species contributes specific extracellular polysaccharides to the ECM, and fungi produce hyphae that act as molecular scaffolds that together alter the structure and composition of the biofilm.These interactions affect the susceptibility of the microbes to antimicrobial therapy.3][14][15] Although the precise molecular mechanisms behind this enhanced antibacterial resistance are not fully understood, it is thought that the contribution of fungal extracellular polysaccharides to the ECM sequesters the antibiotic molecules and limits their diffusion through the biofilm.Furthermore, microbe-microbe interactions and nutrient competition may alter the transcriptional profile of bacterial cells, priming the cells to be more tolerant to the antibiotic.However, how these microbe-microbe interactions affect the response to antifungal treatment is unknown.
Currently there are only a handful of classes of antifungal drugs on the market, the most widely used of which are the azoles (fluconazole, miconazole, etc.), which target ergosterol biosynthesis; the echinocandins (caspofungin and anidulafungin), which target the synthesis of β-glucan; and the polyenes (amphotericin B and nystatin), which target ergosterol.The widespread use of antifungals in both hospital and agricultural settings has resulted in the rapid rise of antifungal resistance, with the WHO identifying MDR Candida infections as a major cause for concern.With limited new antifungal drugs in the pipeline, there is an urgent need to preserve and improve the efficacy of the available antifungals.Here, we demonstrate that in a dualspecies biofilm P. aeruginosa enhanced the susceptibility of C. albicans to the antifungal amphotericin B but had minimal effect on the susceptibility of the fungus to fluconazole or caspofungin.The enhanced susceptibility of C. albicans to amphotericin B was mediated by the P. aeruginosa-dependent repression of the fungal superoxide dismutase, SOD2, resulting in C. albicans being exposed to enhanced reactive oxygen stress.

Strains and media
C. albicans strains were maintained on YPD (yeast peptone dextrose) agar and grown in YPD broth, whereas P. aeruginosa strains were grown and maintained on LB medium.All biofilm assays were performed in Mueller-Hinton broth (MHB).The strains used in this study are listed in Table S1 (available as Supplementary data at JAC Online).

Biofilm assay
Biofilm assays were adapted from previously described methods. 15In brief, C. albicans was grown overnight in YPD and P. aeruginosa was grown overnight in LB, and cultures were washed in PBS.C. albicans strains were resuspended at 1 × 10 6 cells/mL and P. aeruginosa strains were resuspended to an OD 600 of 0.2 in MHB.Each well of 96-well plates contained 100 μL (1 ×10 5 cells) C. albicans and 10 μL (2.5 ×10 6 cells) of P. aeruginosa (final ratio fungus:bacterium = 1:25).Plates were incubated at 37°C for 2 h to allow cells to adhere, at which point the medium was replaced with fresh sterile medium, and plates incubated statically at 37°C for 24 h.Cells not part of the biofilm were removed and medium replaced with fresh MHB containing the appropriate concentration of antifungal agent, and plates incubated for 2 h or 18 h.Medium was replaced with 100 μL PBS containing 50 μg/mL DNase I and plates were incubated at 37°C for 1 h to degrade the ECM.Biofilms were detached from the plates by scraping, serially diluted, and plated onto YPD agar supplemented with 100 μg/mL tetracycline to determine C. albicans cfus, or cetrimide agar to quantify P. aeruginosa cfus.
To assess the effects of phenazines or H 2 O 2 on antifungal resistance, compounds were added to single-species biofilms at the indicated concentrations, after the adherence step, and biofilms formed as outlined above.
To assess the effects of dead P. aeruginosa on antifungal resistance, overnight bacterial cultures were washed in PBS and cells inactivated by either heat-killing at 100°C in PBS for 1 h, or fixed in 4% paraformaldehyde (PFA) for 1 h at room temperature.Inactivated cells were washed with PBS and resuspended in MHB to an OD 600 of 0.8 to mimic the P. aeruginosa cfus recovered from mature biofilms.Biofilms were then established and quantified as described above.
To assess the impact of biofilm supernatants on the susceptibility of C. albicans to amphotericin B, supernatants from multiple 24 h biofilms were pooled, filter sterilized, snap frozen and stored at −20°C for a maximum of 6 days.Single-species biofilms were then established in MHB containing 50% biofilm supernatant.

RNA sequencing
Biofilms were formed as described previously. 16C. albicans (SC5314) was grown overnight in YPD at 37°, 200 rpm and P. aeruginosa (PAO1) was grown overnight at 37°C, 200 rpm in LB, respectively.Cultures were washed with PBS; C. albicans was resuspended to 1 ×10 6 cells/mL and P. aeruginosa was resuspended to an OD 600 of 0.2 in MHB.Biofilms were grown in six-well plates, containing 3 mL C. albicans (3 ×10 6 cells) and 300 μL (75 × 10 6 cells) of P. aeruginosa (final ratio 1:25).Plates were incubated at 37°C for 2 h to allow cells to adhere, after which nonadhered cells were removed and replaced with 3 mL of fresh MHB.Plates were incubated at 37°C for 24 h, then medium was replaced with fresh MHB, and biofilms incubated for a further 4 h.Medium was replaced with 2 mL of PBS containing 50 μg/mL DNase I and incubated at 37°C for 1 h to degrade extracellular DNA.Biofilms were detached from the plate by scraping, triplicate biofilms were then pooled, and 50 μL was serially diluted and plated on cetrimide agar or YPD supplemented with 100 μg/mL tetracycline to check for contamination.Remaining biofilm cells were centrifuged at 3500 rpm at 4°C for 5 min, and pellets snap frozen in liquid nitrogen.Four biological replicates were shipped to GeneWiz ® , UK, for RNA extraction, sequencing and basic bioinformatic analysis.

RNA extraction and sequencing
RNA was extracted from the biofilms by GeneWiz ® using the Qiagen RNeasy Plus mini kit.Library preparation was done in the following stages: (i) ribosomal RNA depletion; (ii) RNA fragmentation and random priming; (iii) first-and second-strand cDNA synthesis; (iv) end repair, 5′ phosphorylation and dA-tailing; and (v) adapter ligation, PCR enrichment and sequencing.Paired-end sequencing was performed using Illumina HiSeq 4000 (2 × 150 bp configuration, single index, per lane).

Bioinformatic analysis
Sequence quality of each sample was evaluated by determination of the number of reads, the yield (Mbases), the mean quality score and the percentage of reads over 30 bases in length.FastQC software was used to determine per base sequence quality and per sequence GC content.Sequence reads were trimmed to remove adapter sequences and nucleotides with poor quality, using Trimmomatic v.0.36.The trimmed reads were mapped to the C. albicans reference genome, available on Dual-species biofilms enhance antifungal susceptibility ENSEMBL, using the STAR aligner v.2.5.2b.For dual-species biofilms, reads were first mapped to the P. aeruginosa PAO1 genome available on ENSEMBL, and then unmapped reads were aligned to the C. albicans genome.Unique gene hit counts were calculated using featureCounts from the Subread package v.1.5.2.Only unique reads that fell within exonic regions were counted (the maximum hits per read was set to 10 by default).One biological replicate of the C. albicans monospecies biofilms (CA 0M-6) had a significant number of reads that did not align to the C. albicans genome or the P. aeruginosa genome and was therefore excluded from downstream analysis.Differential gene expression analysis was performed using DESeq2.Principal component analysis (PCA) was performed to reveal the similarities within and between groups, with PCA plots included in the output (Figure S1).The DESeq2 output file (C.albicans_expression_ CA_vs_PACA.xlsx) and the raw and normalized reads and transcript per million (TPM) values for all C. albicans (C_albicans_counts.xlsx)and P. aeruginosa (P_aeruginosa_count_data.xlsx) genes are available at the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/),under accession number GSE167137.
To reveal the global genomic responses, defined differentially expressed genes were compared with genes under oxidative stress, heatshock stress and hyperosmotic stress with different timings, 0, 10, 30 and 60 min, 17 respectively, using Pearson correlation coefficient in custom R scripts (cor.test).

Enrichment analysis
9][20] Gene ontology (GO) analysis was performed using the Candida Genome Database (CGD) Gene Ontology Slim Mapper. 21,22This tool maps the annotations from each gene list to GO Slim terms, which are broad, high-level GO terms that are specific to the selected Candida species-in this case C. albicans. 22KEGG pathway enrichment analysis was done using KOBAS 3.0 software, 23 as described previously. 24

Quantification of ergosterol
Biofilms were formed as described above.After the 24 h incubation, biofilms were washed with PBS and fixed with 4% PFA for 1 h at room temperature.Biofilms were disrupted and then stained with filipin (0.05 mg/mL) for 2 h at room temperature.Stained biofilm cells were washed with PBS, mounted onto glass microscope slides and imaged on a Zeiss AxioObserver at ×63 magnification, acquiring 30-33 images per sample.To quantify the amount of ergosterol, the median fluorescent intensity of the filipin staining was quantified using FIJI software.Up to five cells per image were outlined, as well as five spherical areas containing no cells (background fluorescence readings).The mean corrected total cell fluorescence (CTCF) was calculated for each image, in which CTCF = integrated density − (area of selected cell × mean fluorescence of background readings).Data were analysed using one-way analysis of variance (ANOVA) and the Holm-Šidák multiple comparisons test.

Measurement of ROS
To quantify the amount of ROS in the single-and dual-species biofilms, biofilms were grown in triplicate for 24 h, washed with PBS and then stained with 10 μM 2′,7′-dichlorodihydrofluorescein diacetate (H 2 DCFDA) at 37°C for 45 min, before a final wash in PBS.Images were acquired using a Zeiss LSM880/Elyra/Axio Observer.Z1 Confocal Microscope (Carl Zeiss Inc.) with a ×10 objective.A 488 nm wavelength laser was used to excite samples, with consistent laser power and gain maintained throughout, and fluorescent signal between 500 and 550 nm wavelengths captured.ROS staining was quantified through mean fluorescence intensity of three different 3 722.5 μm 2 regions per technical biofilm replicate, per condition, across three independent biological experiments.Biofilms grown in the presence of 5 mM H 2 O 2 served as a positive control for ROS production, and background fluorescence was checked against unstained control biofilms.Data were analysed using one-way ANOVA and the Holm-Šidák multiple comparisons test.

P. aeruginosa increases the susceptibility of C. albicans to amphotericin B
Previously we have shown that dual-species biofilms of C. albicans and P. aeruginosa have increased tolerance to antibiotics, as a result of the fungal ECM sequestering the drugs. 15Therefore, we hypothesized that P. aeruginosa may also impact the antifungal susceptibility of C. albicans.To test this hypothesis, 24 h mono-or dual-species biofilms were treated with antifungal drugs for either 18 h (fluconazole and caspofungin) or 2 h (amphotericin B) and then C. albicans culturability was quantified by plating.The susceptibility of C. albicans in dual-species biofilms to fluconazole and caspofungin was not affected compared with the monospecies control (Figure 1a, b).However, treatment of the dual-species biofilms with the polyene amphotericin B resulted in a significant reduction in C. albicans cfus, with negligible fungal cells detected when the dual-species biofilms were treated with 2 μg/mL amphotericin B (Figure 1c).On the other hand, treatment of monospecies biofilms with amphotericin B still resulted in significant fungal growth, even when treated with 8 μg/mL amphotericin B. Amphotericin B had no impact on P. aeruginosa at any of the concentrations tested, whereas at high concentrations fluconazole reduced the culturability of P. aeruginosa (Figure S1a, b).Therefore, these data suggest that the presence of P. aeruginosa in the biofilm results in a significant increase in the susceptibility of C. albicans to the polyene class of antifungals.
To determine the role of P. aeruginosa in this interaction, we grew C. albicans biofilms in the presence of heat-killed or fixed P. aeruginosa cells and added live P. aeruginosa to mature C. albicans monospecies biofilms just prior to the addition of amphotericin B. C. albicans susceptibility to the polyene was only increased in the presence of live bacteria, which had been directly incorporated into the C. albicans biofilm (Figure 2a, b).Furthermore, establishing biofilms in the presence of supernatants from mature biofilms did not affect the susceptibility of C. albicans to amphotericin B (Figure 2c), suggesting that bacterial viability and the architecture of the biofilm are important for the observed phenotype.In agreement with this, the susceptibility of C. albicans to amphotericin B was not altered when C. albicans was co-cultured with P. aeruginosa under planktonic conditions (Figure 2d).
To confirm that the impact of P. aeruginosa on C. albicans amphotericin B susceptibility is a general trait of P. aeruginosa, we screened several P. aeruginosa clinical isolates for their ability to suppress the resistance of C. albicans to the polyene.The viability of C. albicans was significantly reduced in all dual-species biofilms treated with amphotericin B, although not as significantly as in the presence of PAO1 (Figure 2e).However, increasing the incubation time of the biofilms with the antifungal from 2 to 18 h killed all fungal cells (Figure 2f), despite P. aeruginosa cfu counts remaining consistent (Figure S1d).Therefore, P. aeruginosa Alam et al.  enhances the susceptibility of C. albicans to amphotericin B in dual-species biofilms.

P. aeruginosa increases the susceptibility of clinical C. albicans isolates and Candida dubliniensis to amphotericin B
To confirm that the impact of P. aeruginosa on the susceptibility of C. albicans to amphotericin B is a general response of C. albicans, we tested a series of clinical isolates.All C. albicans isolates tested exhibited reduced culturability after treatment with amphotericin B in dual-species biofilms (Figure 3a).Next, to assess whether a similar effect was observed with C. albicans strains with increased resistance to amphotericin B, we tested the erg6 mutant strain DSY4, which has a higher MIC for amphotericin B than SC5314 due to altered sterol composition of the cell membrane, but still displays good growth rates in rich media.In the presence of P. aeruginosa, DSY4 was more susceptible to amphotericin B, as observed for SC5314 (Figure 3b).Therefore, strains with increased MIC values for amphotericin B, resulting from defects in ergosterol biosynthesis, are still impacted by the presence of P. aeruginosa.
To determine whether P. aeruginosa impacts the susceptibility of other Candida species to amphotericin B, we grew P. aeruginosa in dual-species biofilms with non-albicans Candida species for 24 h, and then treated the biofilms with 2 μg/mL amphotericin B. C. dubliniensis behaved similarly to C. albicans, being more susceptible to amphotericin B in the presence of P. aeruginosa.However, C. tropicalis, C. parapsilosis, C. krusei and C. glabrata still displayed significant culturability after treatment with amphotericin B in both mono-and dual-species biofilms (Figure 3c).The culturability of C. tropicalis was significantly reduced in the presence of P. aeruginosa, suggesting that the bacterium has some impact on C. tropicalis.On the other hand, the culturability of the other Candida species in dual-species biofilms was comparable to that of monospecies biofilms, indicating that P. aeruginosa has negligible impact on these Candida species.In addition, the amount of P. aeruginosa was not affected by the presence of the different Candida species, or the antifungal (Figure S1d).

C. albicans mutants with reduced ergosterol content are more resistant to amphotericin B
One of the main modes of action of amphotericin B is to bind ergosterol in the fungal membrane, forming pores resulting in fungal cell lysis.Genes involved in the synthesis of ergosterol are regulated by the Upc2 transcription factor. 25Therefore, the upc2Δ mutant synthesizes significantly less ergosterol than WT cells. 25To determine whether the increased susceptibility of C. albicans to amphotericin B in the presence of P. aeruginosa is dependent on ergosterol, we quantified the susceptibility of the upc2Δ mutant to amphotericin B in both mono-and dual-species biofilms.Deletion of UPC2 attenuated the ability of P. aeruginosa to impact the susceptibility of C. albicans to the polyene, with significant fungal culturability observed in dual-species biofilms treated with amphotericin B (Figure 4a).Therefore, we hypothesized that P. aeruginosa increases the susceptibility of C. albicans to the polyene by increasing the amount of ergosterol in the Alam et al.Alam et al.

Downregulated Upregulated biofilm formation biological process carbohydrate metabolic process cell adhesion cell budding cell cycle cell development cell wall organization cellular homeostasis cellular protein modification process cellular respiration conjugation cytokinesis cytoskeleton organization DNA metabolic process filamentous growth generation of precursor metabolites and energy growth of unicellular organism as attached cells hyphal growth interspecies interaction between organisms lipid metabolic process nucleus organization organelle organization other pathogenesis protein catabolic process protein folding pseudohyphal growth regulation of biological process response to chemical response to drug response to stress ribosome biogenesis
fungal cell membrane.However, the levels of ergosterol were similar between mono-and dual-species biofilms (Figure 4b) suggesting that although ergosterol levels are important for the mode of action of amphotericin B, P. aeruginosa does not directly increase ergosterol incorporation into the fungal cell membrane.

C. albicans undergoes a significant transcriptional response to the presence of P. aeruginosa
To elucidate how P. aeruginosa increases the susceptibility of C. albicans to amphotericin B, global transcriptional analysis was performed on C. albicans in both mono-and dual-species biofilms.The presence of P. aeruginosa resulted in a significant transcriptional response in C. albicans, with 431 genes being significantly up-regulated and 739 genes being significantly down-regulated (Padj < 0.05, and log 2 -fold cut-off ≥1 or ≤ −1) (Figure S2, File S1).In agreement with our ergosterol staining data, KEGG pathway enrichment analysis identified ergosterol biosynthesis to be among the biological processes that are significantly down-regulated in the dual-species biofilms, with 15 of the genes involved in ergosterol biosynthesis (ERG1, ERG3-6, ERG9-13, ERG24, ERG25, ERG251, HMG1 and MVD1) being significantly down-regulated (Figure 5a).Therefore, the increased susceptibility of C. albicans to amphotericin B in the presence of P.
aeruginosa is not due to increased ergosterol levels.From the significantly down-regulated genes, 24 were associated with fungal cell wall organization, with genes required for chitin and glucan synthesis (CHT2, CHT3, ENG1, EXG2 and KRE1), filamentous growth (ALS2-4, HWP1, HWP2, HYR1, SFL2) and O-mannan biosynthesis (PMT1, PMT2, PMT4 and DMP1) being down-regulated in dual-species biofilms.P. aeruginosa binds to C. albicans hyphae via O-mannans resulting in fungal cell death. 2,26Therefore, the down-regulation of these processes might prevent bacterial attachment and ensure fungal survival.Other processes that were significantly down-regulated included genes involved in ribosome biogenesis, translation and carbon metabolism (Figure 5a), suggesting that C. albicans is under stress in the dualspecies biofilms.In agreement with this, genes involved in iron acquisition (FRP1, FTR2, FHT1) and amino acid metabolism were significantly up-regulated (Figure 5b, c), suggesting that there is nutrient competition in dual-species biofilms.Response to stress was also one of the GO terms enriched in our analysis (Figure 5c).
To further understand the type of stress the bacterium imposes on C. albicans, we compared our transcriptional response with published transcriptional responses to heat shock, osmotic stress and oxidative stress. 17Among the conditions tested, genes differentially regulated in the dual-species biofilms showed a significant positive correlation with oxidative stress (Table S2), Dual-species biofilms enhance antifungal susceptibility suggesting that P. aeruginosa might impose oxidative stress on C. albicans.In support of this, CAP1, a key transcription factor involved in the oxidative stress response, was significantly upregulated in dual-species biofilms.KEGG and GO term analysis identified peroxisome function (CTN3, ECI1, FAA21, FOX2, MDH1-3, OPT3, PEX5, PEX6, POT1, POX1-3, PXA1, PXA2, PXP2, C2_00390C_A and CR_02570C_A), phagophore assembly (APG7, ATG1, ATG9, AUT7, C1_00860W_A, C1_03330C_A and C1_12150C_A) and oxidoreductase activity (51 genes including GPX2, YHB1, YHB4, YHB5 and SOD6) to be significantly upregulated in C. albicans in response to P. aeruginosa (Figure 5b,  c).In addition to functioning in fatty acid oxidation, the fungal peroxisome is involved in detoxifying hydrogen peroxide, providing further evidence that P. aeruginosa may be imposing reactive oxygen stress on C. albicans.Up-regulation of genes associated with the phagophore suggests that autophagy has been induced.Taken together, these results suggest that C. albicans is under significant stress, likely oxidative stress, in the dual-species biofilms.

P. aeruginosa increases the susceptibility of C. albicans to amphotericin B through the phenazine-mediated induction of ROS stress
Given that our transcriptomics data suggested that C. albicans might be under oxidative stress in the dual-species biofilms, we quantified the amount of ROS in single-and dual-species biofilms using the ROS-sensitive dye H 2 DCFDA in combination with confocal microscopy.Single-species biofilms of either C. albicans or P. aeruginosa exhibited minimal staining with H 2 DCFDA, suggesting that the levels of ROS in these biofilms was below detectable levels.As a positive control, single-species biofilms were grown in the presence of H 2 O 2 , a known ROS inducer.In the C. albicans single-species biofilms in response to 5 mM H 2 O 2 , some of the hyphal cells showed significant staining with H 2 DCFDA, suggesting that the intracellular ROS levels in these cells were higher than in the non-treated single-species biofilm cells.However, there was a significant increase in the number and fluorescence intensity of fungal cells stained with H 2 DCFDA in the untreated dualspecies biofilms, confirming that the intracellular ROS levels in C. albicans were significantly increased in dual-species biofilms (Figure 6).Therefore, in the dual-species biofilms, P. aeruginosa exerts oxidative stress on C. albicans through the generation of intracellular ROS.
In addition to binding ergosterol and inducing cell lysis, amphotericin B has been shown to exert antifungal activity through the generation of ROS stress. 27At the same time, P. aeruginosa is also known to induce ROS stress in C. albicans. 5Therefore, we hypothesized that ROS induction by both P. aeruginosa and amphotericin B at the same time may result in enhanced antifungal activity.In support of this hypothesis, treatment of C. albicans monospecies biofilms with 10 mM H 2 O 2 resulted in enhanced susceptibility of the fungus to amphotericin B, comparable to the effects observed in the presence of P. aeruginosa (Figure 7a).Similar effects were observed when C. albicans monospecies biofilms were established in the presence of menadione, an inducer of mitochondrial ROS.However, C. albicans biofilms grown in the presence of hypoxanthine and xanthine oxidase, which induce cellular ROS, did not increase the susceptibility of C. albicans to amphotericin B (Figure 7a).Therefore, it is likely  Alam et al.
that P. aeruginosa increases C. albicans susceptibility to the polyene through the induction of mitochondrial ROS.P. aeruginosa has been shown to induce ROS stress in C. albicans through the secretion of phenazines. 5In agreement with phenazine-mediated ROS stress in C. albicans, deletion of genes involved in phenazine biosynthesis attenuated the ability of P. aeruginosa to enhance the susceptibility of C. albicans to amphotericin B (Figure 7b, Figure S1e).Addition of exogenous phenazines to C. albicans monospecies biofilms confirmed that phenazine methosulphate (PMS), and to a lesser extent, phenazine-1-carboxylic acid (PCA) mediated ROS stress and therefore enhanced C. albicans susceptibility to amphotericin B (Figure 7c).

P. aeruginosa increases the susceptibility of C. albicans to amphotericin B by increasing mitochondrial ROS, while suppressing SOD2 expression
Superoxide radicals are removed through the actions of superoxide dismutase (Sod) enzymes.The C. albicans genome contains six SOD genes (SOD1-6), which vary in their subcellular localization.In response to ROS stress, many organisms up-regulate these enzymes to detoxify the superoxide radicals.However, analysis of our transcriptomic data confirmed that, in the presence of P. aeruginosa, SOD1-5 were down-regulated in C. albicans, whereas SOD6 was up-regulated (Table 1).Sod2 is a mitochondrial Sod enzyme, essential for the detoxification of mitochondrial ROS.Previous transcriptional studies confirm that, in response to oxidative stress, SOD2 is significantly up-regulated to detoxify ROS, and is usually up-regulated in response to amphotericin B treatment. 28,29Given that both phenazines and amphotericin B target mitochondrial ROS production, and that menadione mimics the effects of P. aeruginosa on C. albicans susceptibility to amphotericin B, P. aeruginosa-induced repression of SOD2 could enhance ROS toxicity.In agreement with this, deletion of SOD2 resulted in significantly increased C. albicans susceptibility to amphotericin B in the absence of P. aeruginosa, whereas deletion of SOD1 and SOD3-6 had minimal effect on amphotericin B susceptibility (Figure 8a).Furthermore, addition of exogenous PMS to sod2Δ monospecies biofilms further enhanced the susceptibility of C. albicans to amphotericin B (Figure 8b).To determine whether P. aeruginosa-dependent suppression of SOD2 was the driving force behind the increased susceptibility of C. albicans to amphotericin B, we expressed SOD2 from a constitutive promoter and reassessed C. albicans viability upon amphotericin B treatment.Constitutive expression of SOD2 resulted in increased resistance of C. albicans to amphotericin B in the presence of P. aeruginosa (Figure 8c), confirming that P. aeruginosa-mediated repression of SOD2 expression contributes to the enhanced susceptibility of C. albicans to amphotericin B in dual-species biofilms.

Discussion
Antimicrobial resistance is a global threat, with MDR strains frequently isolated in the clinic.It is now estimated that, by 2050, 10 million people per year will die as a result of antimicrobialresistant infections. 30Although not as frequently mentioned as antibiotic resistance, antifungal resistance places a huge burden on patients and healthcare providers.This is exemplified by the isolation of MDR strains of Candida auris, with several strains being resistant to the three major classes of antifungals. 31As a result, the WHO and the CDC have declared antimicrobialresistant Candida infections a global health threat.Therefore, there is an urgent need for the development of new classes of antifungal drugs, or for increased efficacy of existing treatment options.Here, we have shown that P. aeruginosa enhances the susceptibility of C. albicans to amphotericin B through phenazinedependent induction of mitochondrial ROS stress in conjunction with the suppression of ROS-detoxifying enzymes.
Owing to the nephrotoxicity of amphotericin B, it is usually used as a last-resort treatment.However, liposomal formulations have greatly reduced the toxicity of the drug.Amphotericin B targets ergosterol in the fungal membrane, impairing membrane integrity, and has been shown to induce mitochondrial ROS stress. 27In agreement with this, the C. albicans sod2Δ mutant was more susceptible to amphotericin B than the deletion of genes encoding the other superoxide dismutases.P. aeruginosa has been shown to induce ROS stress in C. albicans through the secretion of phenazine compounds. 32Only PMS, a synthetic analogue of 5-methyl phenazine-1-carboxylic acid betaine (5MPCA), was able to increase the susceptibility of C. albicans to amphotericin B. P. aeruginosa secretes 5MPCA during dual-species growth with C. albicans, 33 but not during monoculture, 34 which may explain why culture supernatants from P. aeruginosa monocultures did not affect the susceptibility of C. albicans to amphotericin B. However, supplementation of C. albicans monospecies biofilms with PMS did not fully recapitulate the presence of the bacterium, suggesting that additional factors are required.5MPCA can be modified by amino acids including arginine to form red  The parental control strain and the sod2Δ mutant were grown in single-species biofilms for 24 h and then treated with 2 μg/mL amphotericin B for 2 h.(c) The parental control strain and the SOD2 overexpression (CAI4-SOD2) strain were grown in single-or dual-species biofilms for 24 h and then treated with 2 μg/mL amphotericin B for 2 h.Data are the log 10 -(mean) ± the SEM from three biological replicates.Dotted line represents the minimal level of detection of the assay.Data were analysed using twoway ANOVA and the Holm-Šidák multiple comparisons test.
Alam et al.
pigmented proteins. 32Our transcriptional analysis identified amino acid biosynthesis, particularly arginine biosynthesis, to be significantly up-regulated in dual-species biofilms.Therefore, increases in amino acid biosynthesis might increase the bioactivity of the phenazine in dual-species biofilms.Supplementation of sod2Δ biofilms with PMS has a greater impact on the susceptibility of C. albicans to the polyene.SOD2 is usually up-regulated in the presence of amphotericin B and other inducers of mitochondrial ROS stress. 28,29However, transcriptional analysis confirmed that, in the presence of P. aeruginosa, SOD2 and several other detoxifying enzymes were down-regulated.Therefore, it would appear that P. aeruginosa may simultaneously induce ROS stress while down-regulating SOD2 expression, which likely results in the stress surpassing the capacity of the detoxification system, thus leading to cell death.P. aeruginosa is highly genetically variable. 35Although treatment of dual-species biofilms containing PAO1 with amphotericin B resulted in minimal fungal survival after 2 h, biofilms containing PA14 and P. aeruginosa clinical isolates required a longer incubation with the drug to completely eradicate the fungus (although a significant reduction in fungal culturability was observed after 2 h).Although all P. aeruginosa species produce 5MPCA, the production of the phenazine intermediate is variable between isolates. 33Therefore, it is likely that, in this assay, the species produce 5MPCA at lower levels compared with the PAO1 strain, which may explain why extended incubation times are required to fully eradicate the fungus.
Other classes of antifungals also induce ROS production in C. albicans as part of the mode of action, 36 suggesting that P. aeruginosa should also increase the susceptibility of the fungus to other antifungals.Fluconazole targets ergosterol biosynthesis, reducing ergosterol levels in the fungal membrane, and resulting in the accumulation of toxic sterols. 37,38As a result, fluconazole is fungistatic against C. albicans.However, in vivo P. aeruginosa has been shown to make fluconazole fungicidal toward C. albicans. 39et, in our biofilm assay, fluconazole had minimal effect on the culturability of C. albicans, whether in the absence or presence of the bacterium.However, glucans in the ECM sequester fluconazole, preventing it from penetrating the biofilm. 40Therefore, it is likely that, as fluconazole is added to mature biofilms, the glucan component of the ECM sequesters the drug, preventing it from reaching the fungal cells.
Although C. albicans is still the main species causing candidiasis, the incidence of infections involving non-albicans Candida species is increasing. 41Therefore, it is important for antifungal therapies to be effective across all Candida species.P. aeruginosa enhanced the susceptibility of C. dubliniensis to amphotericin B, similarly to C. albicans, which is not surprising, as C. dubliniensis is the most closely related species to C. albicans.Amphotericin B had a negligible effect on the viability of this isolate of C. krusei both in the presence or absence of P. aeruginosa, suggesting that this isolate is resistant to the polyene.Estimates suggest that up to 15% of C. krusei isolates are resistant to amphotericin B owing to reduced ergosterol content in the cell membrane. 42Despite the other Candida species being susceptible to the polyene, their susceptibility was not significantly enhanced in the presence of P. aeruginosa.The interactions between P. aeruginosa and nonalbicans Candida species have not been as extensively studied as C. albicans.Therefore, it is possible that either 5MPCA does not induce ROS stress, that the bacterium is not able to downregulate the ROS detoxifying enzymes in these species, or that these species can tolerate greater levels of ROS than C. albicans.
4][45] Therefore, in vivo, much higher concentrations of drug are required to eradicate the infection, but cannot always be achieved.Our data confirm that, in a mixed-species biofilm, C. albicans is more susceptible to amphotericin B than in monospecies biofilms.Given that in vivo biofilms are normally composed of multiple species, especially in traumatic wounds and in the cystic fibrosis lung where C. albicans and P. aeruginosa are frequently co-isolated, this suggests that C. albicans might be more susceptible to amphotericin B than the other classes of antifungal drugs.However, confirmation of our simple in vitro biofilm results in more clinically relevant biofilm models and in vivo infection settings is required, to confirm whether this enhanced susceptibility to amphotericin B is maintained.Given that we, and others, have shown that C. albicans promotes drug tolerance in multiple bacterial species, 12,14,15 treatment of polymicrobial biofilms with an initial course of antifungal, prior to the initiation of antibacterial treatment, may help to overcome some of the antimicrobial resistance observed.However, this requires further investigation.

Figure 1 .Figure 2 .
Figure 1.P. aeruginosa enhances the susceptibility of C. albicans to amphotericin B in dual-species biofilms.Twenty-four-hour preformed singleor dual-species biofilms were treated with increasing concentrations of (a) fluconazole, (b) caspofungin or (c) amphotericin B for 18 h (a and b) or 2 h (c).Data are the log 10 (mean) ± the SEM from three biological replicates.Dotted line represents the minimal level of detection of the assay.Data were analysed using two-way ANOVA and the Holm-Šidák multiple comparisons test (*P < 0.05, **P < 0.01, ns, not significant).

Figure 3 .Figure 4 .
Figure 3. P. aeruginosa also increases the susceptibility of C. albicans clinical isolates and C. dubliniensis to amphotericin B. (a) Clinical isolates of C. albicans were grown in single-or dual-species biofilms with PAO1 for 24 h and then biofilms were treated with 2 μg/mL amphotericin B for 2 h.Data are the log 10 (mean) ± the SEM from three biological replicates.(b) The DSY4 strain, which has increased resistance to amphotericin B, was grown in single or dual species biofilms with PAO1 for 24 h and then biofilms were treated with amphotericin B for 2 h.Data are the log 10 (mean) ± the SEM.(c) Medically important Candida species were grown in single-or dual-species biofilms with PAO1 for 24 h and then biofilms were treated with 2 μg/mL amphotericin B for 2 h.Data are the log 10 (mean) ± the SEM from three biological replicates.Dotted line represents the minimal level of detection of the assay.Data were analysed using two-way ANOVA and the Holm-Šidák multiple comparisons test.

Figure 5 .
Figure 5. C. albicans undergoes a significant transcriptional response in dual-species biofilms.(a) KEGG enrichment analysis of pathways significantly down-regulated in dual-species biofilms.Ergosterol and O-mannan biosynthesis are highlighted.(b) KEGG enrichment analysis of pathways significantly up-regulated in dual-species biofilms.ROS detoxification and amino acid metabolism are highlighted.(c) GO slim term analysis of biological processes differentially regulated between single-and dual-species biofilms.Genes of interest in purple are significantly down-regulated, whereas genes in blue are significantly up-regulated (see text for details).This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 6 .
Figure 6. C. albicans is under ROS stress in dual-species biofilms.(a) Single-and dual-species biofilms were grown in MHB or MHB containing 5 mM H 2 O 2 for 24 h.Biofilms were stained with H 2 DCFDA and imaged.(b) Biofilms were excited at 488 nm and emission captured between 500 and 550 nm.Mean fluorescence intensity was taken at three 850 × 850 μm regions per biofilm, with 9 biofilms assessed per condition (27 regions total), with unstained control biofilms to account for background fluorescence.Scale bar represents 100 μm.Data were analysed using one-way ANOVA and the Holm-Šidák multiple comparisons test.This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 7 .
Figure 7. Increased susceptibility of C. albicans to amphotericin B is partially mediated by P. aeruginosa phenazines.(a) SC5314 single-species biofilms were grown in the presence of H 2 O 2 , menadione, and hypoxanthine and xanthine oxidase for 24 h, and then treated with 2 μg/mL amphotericin B for 2 h.(b) SC5314 was grown in single-or dual-species biofilms with P. aeruginosa strains defective in iron and phenazine biosynthesis.(c|) SC5314 single-species biofilms were grown in the presence of 40 μM purified phenazine compounds for 24 h, and then biofilms were treated with 2 μg/mL amphotericin B for 2 h.Data are the log 10 (mean) ± the SEM from three biological replicates.Dotted line represents the minimal level of detection of the assay.Data were analysed using two-way ANOVA and the Holm-Šidák multiple comparisons test.PCA, phenazine-1-carboxlic acid; PMS, phenazine methosulphate; PYO, pyocyanin; 1-HP, 1-hydroxyphenazine.

Figure 8 .
Figure 8. Overexpression of mitochondrial SOD2 prevents hypersusceptibility of biofilms to amphotericin.(a) C. albicans SOD mutants were grown in single-species biofilms and then treated with 2 μg/mL amphotericin B for 2 h.(b)The parental control strain and the sod2Δ mutant were grown in single-species biofilms for 24 h and then treated with 2 μg/mL amphotericin B for 2 h.(c) The parental control strain and the SOD2 overexpression (CAI4-SOD2) strain were grown in single-or dual-species biofilms for 24 h and then treated with 2 μg/mL amphotericin B for 2 h.Data are the log 10 -(mean) ± the SEM from three biological replicates.Dotted line represents the minimal level of detection of the assay.Data were analysed using twoway ANOVA and the Holm-Šidák multiple comparisons test.

Table 1 .
Expression of superoxide dismutase genes