Dynamics associated with the efflux pumps in biofilms and persister cells of Cryptococcus neoformans and Cryptococcus gattii

ABSTRACT

Cryptococcosis is a systemic mycosis affecting immunocompromised individuals (mainly C. neoformans) and immunocompetent hosts (mainly C. gattii). Effective treatment remains challenging due to antifungal resistance, driven by efflux pumps, target enzyme mutations, biofilm formation, and persister cells. This study evaluated efflux pump gene expression and its role in biofilm development and resistance in persister cells of both species, as well as the effects of combined treatments—fluconazole + FK506 and fluconazole + octyl protocatechuate—on biofilms. Materials and Methods included microscopy, biofilm characterization during formation, treatment assays, induction of persister cells, and relative quantification of efflux pump genes AFR1, AFR2, AFR3, and MDR1. Biofilms of both species were robust, with increased extracellular matrix linked to efflux pump overexpression. Combined treatments reduced biofilm growth and extracellular matrix production while downregulating AFR1, AFR2, AFR3, and MDR1, highlighting a link between efflux pumps and matrix export. Persister cells displayed low metabolic activity under amphotericin B treatment, followed by slow regrowth and reactivation of efflux pumps. These findings demonstrate multifaceted roles of efflux pumps in antifungal resistance and biofilm development, confirm the presence of persister cells in C. neoformans and C. gattii, and suggest novel therapeutic strategies for cryptococcosis.

IMPORTANCE

Cryptococcosis is a fungal infection with mortality exceeding 75%, predominantly affecting immunocompromised individuals—including transplant recipients and people living with HIV—but also capable of infecting healthy people. Treatments are lengthy, toxic, and increasingly ineffective due to antifungal resistance. Two major contributors to therapeutic failure are biofilms—protective fungal communities encased in a self-produced matrix—and persister cells, a dormant subpopulation that survives lethal drug concentrations and reactivates infection once treatment stops. This study reveals that efflux pumps—proteins that expel drugs from fungal cells—are not only resistance mechanisms but also active contributors to biofilm construction. Critically, persister cells reactivate their efflux pumps upon recovery, providing a biological explanation for infection recurrence. Two drug combinations—fluconazole + FK506 and fluconazole + octyl protocatechuate—synergistically disrupted biofilms and reduced persister cell viability while potentially lowering drug doses and toxicity, pointing toward safer, more effective strategies for cryptococcosis.

INTRODUCTION

Cryptococcus neoformans and Cryptococcus gattii constitute the primary species complexes of the genus Cryptococcus and are recognized as pathogenic fungi capable of causing cryptococcosis, a disease that can disseminate systemically within the host. Cryptococcosis is among the four most prevalent fungal infections worldwide, with an estimated 194,000 annual cases and a mortality rate exceeding 75% according to global epidemiological data. Multicenter studies have highlighted its growing incidence and classified it as one of the fungal infections of the greatest clinical relevance (1, 2). Species of Cryptococcus are geographically widespread, with C. neoformans being more prevalent in urban environments and C. gattii in wooded regions in addition to their capacity to adapt to diverse regions and hosts (immunocompromised—mainly C. neoformans; immunocompetent—mainly C. gattii) (3).

Cryptococcosis is managed with an induction therapy regimen consisting of a combination of amphotericin B (AMB) and 5-fluorocytosine (5-FC) for 4 weeks, followed by an 8-week maintenance phase and a consolidation phase lasting 6–12 months with fluconazole (FLC) (1). Despite its efficacy, antifungal treatment presents significant limitations. AMB is associated with nephrotoxicity, while 5-FC can induce hepatotoxicity. Moreover, clinical isolates resistant to FLC have been reported. The rising incidence of invasive mycoses and the emergence of antifungal resistance are matters of concern. Antifungal resistance compromises treatment efficacy, particularly due to the prolonged use of these agents, a particularly critical issue in immunocompromised patients (4).

Resistance arises through several mechanisms: ERG11 mutations, the activation of alternative metabolic pathways, the overexpression of efflux pumps (primarily ABC and MFS transporters), biofilm formation, and the emergence of persister cells (5–9). Biofilms are structured communities encased in a self-produced extracellular polymeric matrix (EPM), conferring resistance up to 1,000-fold higher than planktonic cells, with elevated efflux pump activity (7, 8). Persister cells are a dormant biofilm subpopulation capable of surviving lethal antifungal concentrations and reinitiating infection upon stress removal (9).

Among the therapeutic alternatives under investigation, octyl protocatechuate, a phenolic compound, acts synergistically with azoles and exhibits antifungal and antibiofilm properties (10–12). FK506 (tacrolimus) has shown potential to reverse resistance when combined with conventional antifungals by reducing efflux pump-mediated drug extrusion (13). Given the urgent need for improved strategies, this study aimed to evaluate efflux pump gene expression and its role in biofilm development and persister cell resistance in Cryptococcus neoformans and Cryptococcus gattii and to assess the efficacy of fluconazole + FK506 and fluconazole + octyl protocatechuate combination treatments (1, 2, 4, 8).

MATERIALS AND METHODS

Microorganism culture

Cryptococcus neoformans (ATCC 90112) and Cryptococcus gattii (ATCC 56990) were obtained from the Clinical Mycology Laboratory, School of Pharmaceutical Sciences, UNESP Araraquara, maintained on Sabouraud Dextrose Agar at 37°C, and subcultured every 24 h prior to experiments.

Biofilm formation

Biofilm formation followed Martinez and Casadevall (14) with modifications by Gualque et al. (15). Cultures in Sabouraud Dextrose Broth (150 rpm, 24 h) were harvested, washed twice with 0.85% saline, and resuspended at 1 × 10⁷ cells/mL. Suspensions (100 μL) were added to polystyrene 96-well plates (TPP) and incubated at 37°C for 2 h for initial adhesion. After washing with PBS and replacing with RPMI-1640, biofilms were incubated for 12 h (early-stage), 24 h (intermediate-stage), or 48 h (mature). Negative controls contained medium only; positive controls included medium with yeast inoculum. All assays were performed in triplicate.

Multiparametric multiphoton microscopy

Suspensions were added to 24-well plates with glass coverslips, incubated at 37°C, stained with 100 μg/mL Calcofluor White and 100 μg/mL Fluorescein Diacetate for 60 min in the dark, and visualized on a Carl Zeiss LSM 780/LSM 7MP (FMRP-USP; 40 × oil immersion; FIJI 2.14.0).

Biofilm extracellular matrix extraction

Biofilms were resuspended in sterile water, vortexed, sonicated for 20 min, and centrifuged at 1,500 × g for 20 min; the supernatant was used for downstream analyses (16).

Biofilm characterization

SEM was performed after glutaraldehyde fixation (2.5%, 4°C, 1 h), graded ethanol dehydration, and gold sputter-coating on a Jeol JSM-6610LV (UNESP Araraquara). Biomass was quantified with 0.1% crystal violet (570 nm) (17), extracellular matrix with 1% safranin (492 nm) (18), metabolic activity by XTT assay (490 nm) (14), extracellular polysaccharides by phenol-sulfuric acid method with glucose standard (485 nm) (19), proteins by Bradford with BSA standard (595 nm) (20, 21), ergosterol by KOH-methanol saponification and hexane extraction (282 nm) (22), eDNA by phenol:chloroform:isoamyl alcohol extraction after RNase treatment (260 nm) (23), and eRNA using RNeasy Plant Mini Kit (Qiagen) with NanoDrop (24).

RT-qPCR gene expression

Target genes AFR1, AFR2, AFR3, and MDR1 were quantified by RT-qPCR. Total RNA was extracted with RNeasy Plant Mini Kit (Qiagen), treated with DNase I (Sigma-Aldrich), and integrity verified on an Agilent 2100 Bioanalyzer. cDNA was synthesized using Universal cDNA Synthesis Kit II (Exiqon). qPCR was performed with Maxima SYBR Green/ROX Master Mix (2×; Thermo Scientific) on an Applied Biosystems 7500 cycler. Primers are listed in Table 1. Data were analyzed using the 2^⁻ΔΔCt method with β-tubulin as the endogenous reference; gene expression controls for biofilms: untreated planktonic samples; for treated conditions: respective untreated controls. All experiments were performed in three independent biological replicates. Data were expressed as mean ± SD.

TABLE 1.

List of primers used for RT-qPCR analysis

	C. neoformans
	Fw	Rv
βTUB (CNAG_04948)	ACCCTGTGCTCGATGTTCTG	CCAAGACCGGAACCAGTACC
AFR1 (CNAG_00730)	GAGTTCACCAGCTCGGAAAG	CACCTTCGATCACACCAATG
AFR2 (CNAG_00869)	GGTTCCGACTACATGGCTGT	GAGTTCACCAGCTCGGAAAG
AFR3 (CNAG_06909)	GCCCATAGCACTAGGAGTCG	ATGTCATCCTGCGAATAGCC
MDR1 (CNAG_00796)	CTCTTGATCACATCGCGAAA	ACCGACAATCTTGCTCTGCT

Treatment assays

The octyl protocatechuic acid derivative was provided (Dr. Luis Otávio Regasini) through the Biota (FAPESP) and Sisbiota (CNPq) projects (UNESP Araraquara). Fluconazole (FLC), FK506, and octyl 3,4-dihydroxybenzoate (OCT) were tested alone and in combination (FLC + FK506; FLC + OCT) at 0.421–1,728 μg/mL in biofilm assays (2, 12, 24, 48 h) and at 0.125–512 μg/mL for planktonic cells per CLSI M27-A3 (25). Minimum inhibitory concentration (MIC) was defined as the concentration causing 50% reduction in metabolic activity. Synergy was assessed by fractional inhibitory concentration index (FICI): synergistic if FICI <0.5, indifferent if 0.5–4.0, and antagonistic if >4.0 (25).

Recovery, induction, and treatment of persister cells

Mature 48 h biofilms were treated with 100 μg/mL AMB at 37°C for 48 h (after, XTT assay). After washing, cells were resuspended in PBS, serially diluted, plated on Sabouraud Dextrose agar, and incubated at 37°C for up to 10 days for CFU counting. Cells surviving AMB were considered persister cells (26, 27). MIC and MFC assays used AMB at 100–800 μg/mL and FLC, FK506, and OCT at 2–4,096 μg/mL by resazurin assay (25, 28). All in triplicate.

Statistical analysis

Data were analyzed using unpaired t-test with normality assessment and two-way ANOVA with Tukey’s post hoc test (GraphPad Prism 8.0). Statistical significance was defined as P < 0.05.

RESULTS

Biofilm characterization of Cryptococcus neoformans (ATCC 90112) and Cryptococcus gattii (ATCC 56990)

Scanning electron microscopy of biofilms

At the 2-h stage, microscopy (Fig. 1 and 2) revealed sparse yeasts with little evidence of aggregation. The yeasts were mostly in a planktonic state, with minimal production of microcolonies. The morphology was comparable between C. neoformans and C. gattii, suggesting that both organisms follow a similar initial pattern of pre-adhesion to the surface. By 12 h, in both species, an increase in cell density was observed, with more visible clusters of cells forming a more organized structure, indicative of microcolony formation and extracellular matrix production. After 24 and 48 h, biofilms of both strains appeared more firmly attached, with a more prominent extracellular matrix. Adhesion filaments were also observed in the 24- and 48-h biofilms of both species.

Fig 1.

Scanning electron microscopy (SEM) of the biofilm of C. neoformans ATCC 90112. (A1) 2 h—Pre-adhesion at 500×. (A2) 2 h at 1,000×. (A3) 2 h at 10,000×. (B1) 12 h—Initial adhesion at 1,000×. (B2) 12 h at 2,500×. (B3) 12 h at 10,000×. (C1) 24 h—Intermediate adhesion at 1,000×. (C2) 24 h at 2,500×. (C3) 24 h at 10,000×. (D1) 48 h—Maturation at 1,000×. (D2) 48 h at 2,500×. (D3) 48 h at 10,000×.

Fig 2.

SEM of C. gattii ATCC 56990 biofilm. (A1) 2 h—Pre-adhesion at 1,000×. (A2) 2 h at 2,500×. (A3) 2 h at 5,000×. (B1) 12 h—Initial adhesion at 1,000×. (B2) 12 h at 2,500×. (B3) 12 h at 10,000×. (C1) 24 h—Intermediate adhesion at 1,000×. (C2) 24 h at 2,500×. (C3) 24 h at 10,000×. (D1) 48 h—Maturation at 1,000×. (D2) 48 h at 2,500×. (D3) 48 h at 10,000×.

Kinetics of biofilm formation without treatment

During the first 2 h, C. neoformans and C. gattii cells showed initial adhesion with low biomass, reduced absorbance, minimal extracellular content, and low metabolic activity. By 12 h, biomass and absorbance increased, microcolonies emerged, and extracellular matrix production began, accompanied by higher metabolic activity. Between 24 and 48 h, biofilms reached maturity, with maximal biomass, structured clusters, high absorbance, and peak metabolic activity. From 24 h onward, parameters stabilized, indicating completed maturation. Both species behaved similarly although C. gattii showed slightly higher values.

Quantification of target gene expression in biofilms (pre-treatment)

Regarding biofilm analyses, substantially similar patterns were observed between C. neoformans and C. gattii. The expression of AFR1 and MDR1 increased up to 12 h, reaching a peak, and then progressively declined until 48 h (Fig. 3). The AFR1 gene exhibited the highest expression among all genes, followed by MDR1. The expression of AFR2 and AFR3 was lower under all conditions but followed a similar pattern, with an initial increase followed by a decrease or stabilization after 24 h. These results suggest that AFR1 and MDR1 play a significant role during the early stages of biofilm development in C. neoformans, facilitating initial adaptation during formation. The subsequent decrease in expression after the peak may indicate adaptive regulation as the biofilm matures and stabilizes extracellular matrix levels and metabolic activity.

Fig 3.

(A) Efflux pump gene expression in the biofilm form of C. neoformans ATCC 90112. AFR1 (green); AFR2 (dark blue); AFR3 (purple); MDR1 (light blue). (B) Efflux pump gene expression in C. gattii ATCC 56990 biofilm. Gene expression control: planktonic samples.

Treatment of planktonic and biofilm conditions with selected compounds

The MIC of fluconazole indicates limited efficacy when used alone, particularly against C. gattii. The MICs of FK506 and the octyl derivative for C. neoformans and C. gattii suggest moderate efficacy. The results indicate that both FK506 and the octyl derivative potentiated the action of fluconazole, reducing the amount required to inhibit fungal growth. C. gattii exhibited slightly lower susceptibility to both individual compounds and their combinations compared to C. neoformans (Table 2). Both combinations (fluconazole + FK506 and fluconazole + octyl derivative) demonstrated significant synergy (FICI < 0.5) against C. neoformans and C. gattii, indicating that these combinations are more effective than the compounds used individually (Table 3).

TABLE 2.

Minimum inhibitory concentration (MIC) for inhibiting fungal growth in the planktonic form

Minimum inhibitory concentration (MIC)—planktonic form
Strain	Fluconazole	FK506	Octyl	Fluconazole + FK506	Fluconazole + octyl
ATCC Cn 90112	2.0 µg/mL	1.0 µg/mL	1.0 µg/mL	0.250 µg/mL	0.125 µg/mL
ATCC Cg 56990	4.0 µg/mL	2 µg/mL	2.0 µg/mL	0.500 µg/mL	0.250 µg/mL

TABLE 3.

Fractional inhibitory concentration index (FICI)—planktonic form

Fractional inhibitory concentration index (FICI)—planktonic form
Strain	Fluconazole + FK506	Fluconazole + octyl
ATCC Cn 90112	0.375	0.187
ATCC Cg 56990	0.375	0.187

Regarding biofilms, MICs increased with incubation time in both species, reflecting enhanced resistance over time. C. gattii generally required higher concentrations than C. neoformans (Tables 4 and 5). FICI values were below 0.5 in all tested conditions, confirming synergy; the synergistic effect was consistently stronger with the octyl derivative.

TABLE 4.

Minimum inhibitory concentration (MIC) for inhibiting fungal growth at 2, 12, 24, and 48 h of biofilm development

Minimum inhibitory concentration (MIC)—biofilm (2 h)
Strain	Fluconazole	FK506	Octyl	Fluconazole + FK506	Fluconazole + octyl
ATCC Cn 90112	6.75 µg/mL	3.37 µg/mL	3.37 µg/mL	0.842 µg/mL	0.421 µg/mL
ATCC Cg 56990	13.5 µg/mL	6.75 µg/mL	6.75 µg/mL	1.68 µg/mL	0.842 µg/mL

TABLE 5.

Fractional inhibitory concentration index (FICI)—biofilm (all time points)

Fractional inhibitory concentration index (FICI)—biofilm (2 h)
Strain	Fluconazole + FK506	Fluconazole + octyl
ATCC Cn 90112	0.374	0.373
ATCC Cg 56990	0.187	0.187

Multiphoton microscopy of biofilms

Regarding C. neoformans biofilms (Fig. 4A1 through D2/1), the control exhibited a high density of viable cells, with intense green fluorescence and few signs of dead cells, indicating a thick and healthy biofilm. Treatment with fluconazole alone at 3.37 μg/mL caused a slight reduction in cell viability, reflected by decreased green intensity and lower density in the three-dimensional structure. When fluconazole was combined with FK506 at the same concentration, further reduction of viable cells was observed, as evidenced by reduced green fluorescence and an even lower cell density in the volumetric profile. Finally, the combination of fluconazole with the octyl derivative at the same concentration was the most effective among the tested conditions, resulting in a less dense biofilm structure and increased blue fluorescence, indicating a higher number of dead cells and a significant decrease in viability. For C. gattii biofilms (Fig. 4A1 through D2/2), a similar response pattern was observed.

Fig 4.

Multiphoton microscopy of C. neoformans ATCC 90112 (A1–D2/1) and C. gattii ATCC 56990 (A1–D2/2) biofilms treated with the MIC of the fluconazole + octyl derivative combination. Surface plots shown in A2–D2/1 and A2–D2/2. Calcofluor White (blue) stains the fungal cell wall; Fluorescein Diacetate (green) stains viable cells. Control: mature biofilm without treatment.

Characterization of biofilms (post-treatment)

After determining the MICs, selected compounds were applied at the lowest MIC for each biofilm stage to compare efficacy (Fig. 5A through H, Table 4). Untreated biofilms of both species grew robustly over time. Fluconazole alone showed limited inhibition; FK506 was more effective; the octyl derivative was most effective as a monotherapy. Combined fluconazole + FK506 caused substantial reductions; fluconazole + octyl was consistently the most effective combination, producing the most pronounced reductions in all parameters.

Fig 5.

(A and B) Biomass levels after treatment in C. neoformans and C. gattii biofilms, respectively. Statistical comparisons indicated (* = significance). (C and D) Extracellular matrix levels. (E and F) Metabolic activity. (G and H) Extracellular polysaccharide levels. Control: untreated biofilm sample.

Quantification of target gene expression in biofilms (post-treatment)

The same concentrations were used for all tested compounds, corresponding to suboptimal doses (2 × lower than MIC) of the most effective treatments observed (FLC + OCT). In both species (Fig. 6 and 7), treatment with fluconazole alone significantly increased the expression of efflux pump genes. The combination of fluconazole with octyl derivative was particularly effective in reducing the expression of these pumps, indicating that octyl may potentiate the antifungal effect of fluconazole. The fluconazole + FK506 combination also reduced gene expression, but to a lesser extent than fluconazole + octyl. All treatments, except fluconazole alone, induced the downregulation of the efflux pump genes evaluated.

Fig 6.

AFR1, AFR2, AFR3, and MDR1 efflux pump gene expression in the biofilm form of C. neoformans ATCC 90112 following treatment. (A) AFR1. (B) AFR2. (C) AFR3. (D) MDR1. Gene expression control: respective untreated biofilm samples. All comparisons were statistically significant (* = significance).

Fig 7.

AFR1, AFR2, AFR3, and MDR1 efflux pump gene expression in the biofilm form of C. gattii ATCC 56990 following treatment. (A) AFR1. (B) AFR2. (C) AFR3. (D) MDR1. Gene expression control: respective untreated biofilm samples. All comparisons were statistically significant (* = significance).

Recovery and induction of persister cells

In graph A (Fig. 8), it can be observed that the metabolic activity of C. neoformans and C. gattii biofilms was significantly reduced compared to the growth control. The absorbance of C. neoformans and C. gattii biofilms was almost null, whereas the control showed an absorbance of approximately 3.0. Regarding graph B, it is noted that even after the biofilms were subjected to selective pressure, with extensive damage and drastic reduction in metabolic activity due to exposure to very high concentrations of AMB, growth still occurred, albeit more slowly and gradually, of the persister cells on Sabouraud Dextrose agar. Therefore, results indicate that amphotericin B treatment was highly effective in reducing the metabolic activity and cell viability of C. neoformans and C. gattii biofilms, exerting significant selective pressure and enabling the isolation of persister cells from the rest of the biofilm populations.

Fig 8.

(A) Metabolic activity of C. neoformans ATCC 90112 and C. gattii ATCC 56990 biofilms treated with 100 µg/mL amphotericin B for 48 h. Growth control–Cn 90112: P < 0.0008***. Growth control–Cg 56990: P < 0.0006***. (B) Colony-forming units after biofilm treatment and resuspension on Sabouraud Dextrose agar. Medium control–Cn 90112: P < 0.0008***. Medium control–Cg 56990: P < 0.00009****. Cn 90112–Cg 56990: P < 0.0486*.

Multiphoton microscopy of persister cells

In both species (Fig. 9), persister cells were visualized with a greenish coloration transitioning between stronger and weaker/less intense shades. All other cells were dead and stained with Calcofluor White. Treatment with amphotericin B resulted in the survival of only 1%–2% of the cells present in mature biofilms according to the visualization shields.

Fig 9.

Multiphoton microscopy of C. neoformans ATCC 90112 and C. gattii ATCC 56990 biofilms. (A and B) Mature biofilm cells treated with 100 µg/mL amphotericin B for 48 h and stained with Calcofluor White (blue; fungal cell wall) and Diacetate Fluorescein (green; viable cells). Yellow targets: persister cells.

Determination of MIC and MFC of persister cells against selected compounds

The results indicate that the antifungal combinations fluconazole + FK506 and fluconazole + octyl are significantly more effective at inhibiting the growth of persister cells compared to the use of each antifungal alone (Tables 6 and 7). The fluconazole + octyl combination was the most effective, reducing the MIC to 64 μg/mL for C. neoformans and 128 μg/mL for C. gattii, although both species maintained a resistant profile. Fractional inhibitory concentration index (FICI) values below 1 indicate synergy between the antifungals. Here, the FICI values for both antifungal combinations were well below 1, demonstrating synergistic interactions. The fluconazole + octyl combination exhibited the strongest synergy (FICI = 0.156 for C. neoformans and 0.187 for C. gattii), confirming the MIC results (Table 8). The results also showed that the minimum fungicidal concentration (MFC) of amphotericin B was the same for both strains, suggesting that persister cells of C. neoformans and C. gattii have similar sensitivity to amphotericin B. However, they exhibited a high resistance profile, exceeding four times the induction dose (100 μg/mL) and more than one thousand times the fungicidal dose for planktonic cells (0.250 μg/mL for C. neoformans and 0.500 μg/mL for C. gattii).

TABLE 6.

Minimum inhibitory concentration (MIC) to inhibit the growth of persister cells

Minimum inhibitory concentration (MIC)—persister cells
Strain	Fluconazole	FK506	Octyl	Fluconazole + FK506	Fluconazole + octyl
ATCC Cn 90112	2,048 µg/mL	512 µg/mL	512 µg/mL	128 µg/mL	64 µg/mL
ATCC Cg 56990	2,048 µg/mL	1,024 µg/mL	1,024 µg/mL	256 µg/mL	128 µg/mL

TABLE 7.

Minimum fungicidal concentration (MFC) of persister cells

Minimum fungicidal concentration (MFC)—planktonic cells
Strain	Amphotericin B
ATCC Cn 90112	0.250 µg/mL
ATCC Cg 56990	0.500 µg/mL

TABLE 8.

Fractional inhibitory concentration index (FICI)—persister cells

Fractional inhibitory concentration index (FICI)—persister cells
Strain	Fluconazole + FK506	Fluconazole + octyl
ATCC Cn 90112	0.312	0.156
ATCC Cg 56990	0.375	0.187

Quantification of target gene expression in persister cells

After the recovery period (10 days), persister cells of C. neoformans and C. gattii exhibited minimal but relevant overexpression of efflux pump genes (Fig. 10A). AFR1 was the most highly expressed gene in both species, while AFR2 showed the lowest expression. C. gattii maintained higher expression levels than C. neoformans. High-dose AMB (100 μg/mL) caused a drastic reduction, effectively abolishing the expression of all evaluated efflux pump genes (Fig. 10B). Persister cells treated with the MIC of fluconazole from mature biofilms exhibited the overexpression of all four efflux pump genes in both species (Fig. 10C), underscoring the insufficiency of this antifungal and confirming that C. gattii showed greater relative induction of AFR1 compared to C. neoformans.

Fig 10.

(A) Efflux pump expression in persister cells of C. neoformans ATCC 90112 and C. gattii ATCC 56990 post-recovery. Gene expression controls: planktonic samples. (B) Efflux pump expression in persister cells treated with 100 µg/mL amphotericin B after recovery. Gene expression controls: respective untreated persister samples. (C) Efflux pump expression in persister cells following post-recovery treatment with the MIC of fluconazole from mature biofilms (Cn = 432 µg/mL; Cg = 864 µg/mL). AFR1 Cn–Cg: P < 0.0015**; AFR2: P < 0.020*; AFR3: P < 0.0015**; MDR1: P < 0.0029**. Gene expression controls: respective untreated persister samples.

DISCUSSION

Fungal biofilms represent the main form of microbial growth that significantly contributes to treatment resistance and the persistence of infections. The ability to form biofilms provides fungi with adaptive advantages, such as adherence to favorable growth sites, protection against desiccation, predation, and resistance to biocides. The capacity of Cryptococcus to establish biofilms on both abiotic surfaces (e.g., medical devices) and biotic substrates has direct implications for the pathogenesis of infections caused by these microorganisms (29). Yeast cells may exploit this adaptive condition to gain access to the bloodstream and subsequently invade internal organs. This scenario is particularly concerning, as disseminated fungal infections are associated with high mortality rates (2, 30). The present study demonstrated that both C. neoformans and C. gattii strains exhibited a similar growth pattern and adhered consistently to the abiotic surface tested (polystyrene), forming dense and well-structured biofilms (Fig. 1, 2 and 11A through C), encompassing all stages in a coordinated manner (pre-adhesion, initial adhesion/microcolony formation, intermediate adhesion, and maturation). These findings reinforce the observations of Martinez and Casadevall (29).

Fig 11.

(A) Biomass levels during biofilm formation kinetics of C. neoformans ATCC 90112 and C. gattii ATCC 56990. (B) Extracellular matrix levels (P < 0.0098**). (C) Mitochondrial metabolic activity. (D) Extracellular polysaccharide quantification (P < 0.0199*). (E) Extracellular protein quantification (P < 0.0193*). (F) Extracellular ergosterol quantification (P < 0.0259*). (G) Extracellular DNA quantification (P < 0.0437*). (H) Extracellular RNA quantification (P < 0.0337*). Controls: growth medium without fungal inoculum.

The biofilm ECM is a complex protective network of proteins, polysaccharides, lipids, and nucleic acids (31). Cryptococcal polysaccharides act as a scaffold for structural support, protection, and adhesion (14, 32). Proteins participate in adhesion, stress defense, metabolic adjustments, and resistance (20). Ergosterol is essential for membrane integrity, affecting membrane ATPase activity; defects in its biosynthesis increase sensitivity to azoles and oxidative stress (33, 34). eDNA strengthens biofilm antifungal resistance in C. albicans and A. fumigatus, while eRNA confers mechanical stability (35–37). The ECM distinguishes biofilms from planktonic cells and facilitates resistance through efflux pump induction and protective structuring (38, 39).

According to the treatment results obtained in the present study, biofilms of C. neoformans and C. gattii were significantly less susceptible than planktonic cells to all compounds tested (Tables 2 and 4). These findings are consistent with other studies reporting that the biofilm phenotype confers resistance to antifungal therapy (8, 39). Biofilm formation is the most common mode of microbial growth in nature, representing approximately 65% of all clinical infections. Due to the resistance levels associated with this phenotype, biofilms pose a major obstacle in clinical practice (40). Planktonic cells generally rely on irreversible genetic modifications to sustain a resistant phenotype, whereas biofilms can persist through their physical presence and high cell density, which provides an intrinsically resistant phenotype, independent of defined genetic alterations. Furthermore, biofilms can be up to 1,000 times more resistant to antifungal agents than planktonic cells, although the mechanisms underlying this strong antagonism to antifungal therapy are not yet fully understood (38, 41). As observed in the susceptibility assays of early-stage biofilms (2 h and 12 h) in this study, although resistance is stronger and more robust in later phases of development, resistance to antifungal compounds can be detected just a few hours after surface adhesion even though MIC values are lower than those observed in mature biofilms (14, 42). Moreover, some studies suggest that the expression of multidrug resistance efflux pumps is higher in young biofilms and has been associated with increased resistance to azoles, although this mechanism does not appear to be exclusive or predominant in mature biofilm resistance (27, 38, 43, 44). An important detail, for example, is that the transcriptional profile of efflux pumps observed in C. gattii ATCC 24065 biofilms in studies such as that of Tavares et al. (44) was similar to that described for the reference strain R265 under infection conditions, suggesting that the pattern may be conserved. Our results (Fig. 11) can support this premise using ATCC’s, showing that efflux pump genes (AFR1, AFR2, AFR3, and MDR1—particularly AFR1 and MDR1) display higher transcript expression in young biofilms (between 12 h and 24 h) compared to mature biofilms (48 h), even though mature biofilms (48 h) exhibited equal or higher MIC values than young biofilms.

In the present study, treatment results revealed that fluconazole, when used alone, was the compound with the highest MIC values, both for the planktonic form and especially for the biofilm form (where MIC values indicated resistance in both strains and species tested—Table 4). Previous studies have shown that resistance to azoles (such as fluconazole, itraconazole, and voriconazole) in C. neoformans can be associated with alterations in the target enzyme lanosterol 14α-demethylase, encoded by the ERG11 gene (45–47). Resistance can also be related to transport changes caused by efflux pumps, such as Afr1 (typically composed of two transmembrane domains and two cytoplasmic domains) (7, 48, 49). In Cryptococcus, chromosome 1 harbors both the genes involved in ergosterol biosynthesis and the major efflux pump genes for azoles. Studies have shown that duplication of this chromosome can lead to heteroresistance to azoles and that both gene sets play a crucial role in this duplication, consequently driving phenotypic manifestations such as the overexpression of efflux pumps (47, 48, 50). ABC and MFS efflux pump genes are overexpressed in C. gattii biofilms according to mapped transcriptional profiles (44). Other studies also demonstrated that Cryptococcus biofilms can exhibit marked resistance to fluconazole, reporting MIC values well above 32 or 64 µg/mL (14, 51). In this context, although studies such as those by Stone et al. (47), Basso et al. (48), and Chang et al. (49) were conducted with clinical isolates, our results revealed a similar scenario, where ABC transporter genes (AFR1, AFR2, AFR3) and the MFS transporter gene (MDR1) were overexpressed (Fig. 3) in fluconazole-resistant natural biofilms in C. neoformans and, more prominently, in C. gattii (Table 4), reinforcing the translational relevance of the efflux pump expression patterns observed in the present study with ATCC reference strains. Although AFR1 and MDR1 are the genes most directly associated with fluconazole resistance and AFR2 acts synergistically when co-overexpressed with AFR1 and MDR1 (49), recent evidence indicates that AFR3 also contributes to resistance against this azole, even if less specifically. Inactivation of this transporter can result in increased susceptibility, reduced GXM, and decreased ergosterol content (52, 53). Comparing Fig. 6, 7, 11F and 12D, we can suggest that AFR3 may be important for the maintenance of ergosterol content.

Fig 12.

(A and B) Extracellular protein levels after treatment in C. neoformans and C. gattii biofilms. (C and D) Extracellular ergosterol levels. (E and F) Extracellular DNA levels. (G and H) Extracellular RNA levels. All comparisons were statistically significant (* = significance). Control: untreated biofilm sample.

Although FK506 displayed a lower MIC than fluconazole alone, combined treatments proved more promising. Fluconazole + FK506 drastically reduced MIC values in both planktonic cells and biofilms (Tables 2 and 4), with FICI indices demonstrating significant synergy (Tables 3 and 5). Biofilm characterization showed that biomass, metabolic activity, and ECM components (polysaccharides, proteins, ergosterol, eDNA, and eRNA) were significantly affected (Fig. 5 and 12), as were efflux pump genes AFR1, AFR2, AFR3, and MDR1 (Fig. 6 and 7). Calcineurin inhibitors suppress the fungal stress response, reducing MFS efflux pump expression and increasing intracellular FLC accumulation; our MDR1 expression results corroborate these findings (54, 55, 56, 57). Fluconazole + FK506 synergism in C. neoformans is driven by the inhibition of multidrug resistance pumps, with Afr1, Afr2, Afr3, and Mdr1 transmembrane domains showing affinity for FK506 substrates (58, 59). This synergism was clinically validated: Rossato et al. (60) demonstrated synergistic effects in 70% of C. neoformans isolates (60), and Reichert-Lima et al. (26) identified relevant synergistic interactions among 80 Brazilian clinical isolates (26).

Similarly, although the protocatechuic acid derivative octyl (octyl protocatechuate) exhibited a lower and more effective MIC than fluconazole, the results of the combined treatments in both planktonic cells and biofilms proved to be more promising. The combination of fluconazole + octyl significantly reduced MIC values when compared to the compounds used individually (Tables 2 and 4). The FICI indices demonstrated a significant effect among the combinations tested (Tables 3 and 5). Biofilm characterization after treatment clearly revealed the phenotypes observed. In this characterization, all compounds were applied at the same concentration, corresponding to the lowest MIC value determined for each phase of biofilm formation (see Table 4) in order to allow direct comparison of antifungal and antibiofilm efficacy. The expression of all target genes (AFR1, AFR2, AFR3, and MDR1—Fig. 6 and 7) was significantly reduced. Moreover, both biomass and metabolic activity, as well as the extracellular matrix and its main components (polysaccharides, proteins, ergosterol, eDNA, and eRNA), were affected by the synergism between fluconazole + octyl. As demonstrated by Soares (12), Medina-Alarcón (61), and collaborators, octyl protocatechuate acts synergistically with antifungals such as fluconazole, enhancing damage to the cell membrane and potentiating antifungal activity against Trichophyton and P. brasiliensis and P. lutzii, respectively, while showing very low toxicity in pulmonary fibroblast lineage MRC5 (98% viability). Another study reported that protocatechuic acids can inhibit biofilm formation and enhance antifungal efficacy against Candida albicans (62). Treatment with nonyl protocatechuate in T. rubrum and T. mentagrophytes caused severe biofilm damage: cell wall collapse and leakage of cytoplasmic content. The antimicrobial activity of phenolic acid derivatives increases with the number of alkyl side chains (12, 63). The hydrophilic portion of phenolic compounds can bind to the hydrophilic region of the cell membrane, allowing penetration and disruption of lipid bilayers (64). These results demonstrate that this compound primarily targets the cell membrane by an amphiphilic nature, disrupting the plasma membrane and altering its fluidity and integrity, mainly through interference with ergosterol biosynthesis and downregulation of the ERG11 gene (65). Similarly, another phenolic compound, gallic acid, also exhibited antifungal activity when combined with fluconazole, reducing MIC values. This acid was able to affect the expression of key efflux pump genes in Cryptococcus gattii, namely, AFR1 and MDR1 (66).

Phenotypically, treatments with FK506 and octyl—alone and combined with fluconazole—highlighted a crucial role of efflux pumps in biofilms. FK506 reduces and inhibits these pumps (58, 67); similar effects occur with protocatechuic acid derivatives (66). These inhibitory effects were observed across all biofilm formation, also reducing the components of the ECM components—consistent with the role of ABC and MFS transporters in exporting quorum sensing mediators and polymeric substances (33, 68, 69). Their inhibition by FK506 and octyl enhanced FLC efficacy through increased intracellular ergosterol availability, disrupting membrane homeostasis and reducing biofilm density (Fig. 5 and 12, Table 4).

Persister cells are consistently identified in a “dormant” state, which is directly associated with reduced metabolic activity. As shown in (Fig. 10A), the identified cells exhibited low metabolic activity, possibly corroborating the dormancy characteristics of persister cells observed in biofilms of other fungi (70). The main glucose-utilization pathways, such as glycolysis, and pentose phosphate pathway, are negatively regulated in persister cells (70, 71). Studies indicate that the downregulation of glycolytic proteins in persister cells results in impaired energy storage and may also increase the rigidity of the plasma membrane, thereby hindering the action of amphotericin B. Glucose deprivation induces a dormant state, as observed in these cells. It is important to note that while similar effects occur in glucose-deprived planktonic cultures, no persister cells are observed under those conditions. In the absence of glucose, fungal cells are forced to utilize alternative carbon sources (oxaloacetate, phosphoenolpyruvate, and fructose-1,6-bisphosphate), which yield trehalose and glycogen as products. The proteomic signature of persister cells from Candida albicans and Saccharomyces cerevisiae suggests that these cells can survive using such alternative carbon sources (9, 70).

Biofilms treated with 100 μg/mL AMB for 48 h confirmed selective pressure, leaving only persister cells assessed by CFU and multiphoton microscopy (Fig. 8B and 9) (27). Slow but consistent colony growth was observed for up to 10 days. MIC tests (Table 6) and MFC tests (Table 7) demonstrate extreme tolerance to fluconazole (MIC 2,048 μg/mL) and AMB (MFC 400 μg/mL—>1,000-fold above planktonic fungicidal doses). Tolerance implies survival after transient exposure to fungicidal concentrations; persisters represent a primary case of this phenomenon (72). A biphasic killing pattern is observed: most cells are killed, but a small proportion survives and can reproduce the pattern upon re-exposure (38, 73, 74). The development of AMB resistance is rare, confirming that surviving colonies represent genuine persister cells (75, 76).

Our results showed that the main efflux pumps of persister cells from C. neoformans and C. gattii are virtually not expressed during amphotericin B treatment (Fig. 10B). This suggests that ABC efflux pumps may not be crucial for the survival of these cells during exposure to high doses of amphotericin B. One indication supporting this hypothesis is that energy availability (ATP) is stored as a reserve in these cells during stress, allowing them to resume proliferation under favorable conditions later (9, 27). This premise aligns with our data, which showed the reactivation of these pumps once favorable conditions were restored (Fig. 10A).

Treatment with FK506 and the octyl derivative alone, as well as their combination with fluconazole, reduced the MICs of persister cells. This demonstrates that, in conjunction with the gene expression analysis after recovery (Fig. 10A) and during subsequent exposure to fluconazole (Fig. 10C), these cells reactivated the expression of ATP-dependent efflux pumps (ABC). Literature and biofilm data indicate that these transporters, in addition to MFS transporters, are targets of FK506 and phenolic compounds. Finally, the efficacy of the fluconazole + FK506 and fluconazole + octyl combinations was notable (Table 8), significantly reducing these MICs and suggesting that such combinations could represent promising alternatives in the complex context of cryptococcosis therapy.

From a specific clinical standpoint, the results of this study may lead to significant implications for the management of recurrent and persistent cryptococcal infections (77, 78). Persister cells, by surviving high concentrations of antifungals in a dormant state and subsequently reactivating the expression of the efflux pump, represent a plausible biological basis for treatment failure and infection recurrence—particularly in immunocompromised patients undergoing prolonged maintenance antifungal therapy with fluconazole. The ability of persistent biofilm-derived cells to resist amphotericin B concentrations far exceeding the standard fungicidal dose for planktonic forms underscores why conventional therapeutic regimens may prove insufficient in the context of device-associated or tissue-disseminated cryptococcal infections. In this scenario, the synergistic combinations explored here—fluconazole + FK506 and fluconazole + octyl derivative—offer a double advantage: greater antifungal efficacy and potentially a reduction in the total drug burden (46, 47, 78–81). By reducing the effective inhibitory concentrations required, these combinations could mitigate the risks of nephrotoxicity and hepatotoxicity associated with high doses of amphotericin B or prolonged azole regimens, respectively (54, 55, 61, 82). These observations encourage further preclinical and clinical investigations of combination strategies targeting both efflux-mediated resistance and biofilm persistence as complementary approaches to reduce cryptococcosis treatment failure (8, 9, 67).

Conclusions

C. neoformans and C. gattii biofilms exhibit structural complexity and heterogeneity, with a matrix rich in biomolecules essential for their development. Efflux pumps play a crucial role in biofilm formation and maintenance, aiding in the transport of extracellular molecules and contributing to antifungal resistance. Treatments with fluconazole combined with FK506, and particularly with the octyl derivative, have demonstrated greater efficacy in disrupting biofilm structure and in the formation and expression of the studied efflux pump genes. The synergistic combinations identified in this study offer a promising pathway to reduce total antifungal exposure—and, therefore, the nephrotoxic and hepatotoxic burden associated with conventional regimens—while maintaining or improving therapeutic efficacy. Persister cells of these species can survive high concentrations of antifungals, alternating between dormant and active states. During exposure to amphotericin B, efflux pumps were not essential for their survival, although their expression was subsequently reactivated. Combinations of fluconazole + FK506 and fluconazole + octyl (especially) were more effective against these cells. Clinically, the persistence of cells embedded in biofilms, capable of reactivating efflux pump expression after exposure to antifungals, may lead to a possible mechanistic explanation for recurrent cryptococcal infections (particularly in immunocompromised patients) on prolonged maintenance therapy. These findings elucidate resistance mechanisms in biofilms, point to the presence of persister cells, and suggest potential new therapeutic strategies for cryptococcosis. It is important to mention, however, that this study was conducted with ATCC reference strains (C. neoformans ATCC 90112 and C. gattii ATCC 56990). Clinical isolates may exhibit distinct phenotypic and genotypic profiles due to prior exposure to antifungal agents, host immune pressures, and environmental stressors, which can lead to divergent biofilm formation and persister cell capacities, as well as distinct efflux pump expression patterns. Future investigations comparing ATCC reference strains with clinical isolates—including those recovered from patients with recurrent or refractory cryptococcosis—will be crucial to validate and expand the biological and clinical relevance of these findings.

DATA AVAILABILITY

The data supporting the findings of this study are available from the corresponding author upon reasonable request. No novel sequences or data sets were deposited in public repositories.