Fluconazole Alters the Polysaccharide Capsule of Cryptococcus gattii and Leads to Distinct Behaviors in Murine Cryptococcosis

Cryptococcus gattii is an emergent human pathogen. Fluconazole is commonly used for treatment of cryptococcosis, but the emergence of less susceptible strains to this azole is a global problem and also the data regarding fluconazole-resistant cryptococcosis are scarce. We evaluate the influence of fluconazole on murine cryptococcosis and whether this azole alters the polysaccharide (PS) from cryptococcal cells. L27/01 strain of C. gattii was cultivated in high fluconazole concentrations and developed decreased drug susceptibility. This phenotype was named L27/01F, that was less virulent than L27/01 in mice. The physical, structural and electrophoretic properties of the PS capsule of L27/01F were altered by fluconazole. L27/01F presented lower antiphagocytic properties and reduced survival inside macrophages. The L27/01F did not affect the central nervous system, while the effect in brain caused by L27/01 strain began after only 12 hours. Mice infected with L27/01F presented lower production of the pro-inflammatory cytokines, with increased cellular recruitment in the lungs and severe pulmonary disease. The behavioral alterations were affected by L27/01, but no effects were detected after infection with L27/01F. Our results suggest that stress to fluconazole alters the capsule of C. gattii and influences the clinical manifestations of cryptococcosis.


Introduction
Deaths due to cryptococcosis, particularly the variant caused by Cryptococcus gattii, an emergent fungal pathogen that affects immunocompetent individuals, are increasing [1][2][3]. This fungus is acquired from the environment through inhalation of desiccated yeast cells or spores. The lungs are the most strongly affected site, followed by the central nervous system (CNS), where meningoencephalitis is often detected [2,4].
The therapy for cryptococcosis caused by C. gattii is very important for reduction of mortality [3,5] and fluconazole is recommended for the treatment of mild to moderate pulmonary symptoms of C. gattii-induced cryptococcosis. Amphotericin B (alone or in combination with 5-flucytosine) followed by maintenance with fluconazole is recommended for severe cases [4] but long-term maintenance therapy with fluconazole often leads to fungal resistance [6]. Although the antifungal combination represents an important alternative to the conventional therapy [7], efficacy may also be impaired by polyenes and azoles used in combination [8]. These difficulties in treating cryptococcosis may result in treatment failure or relapse.
Fungal resistance can be confirmed in vitro: Resistant strains are able to grow in the presence of high concentrations of fluconazole [9,10]. Special attention should be paid to the growing resistance to fluconazole. The ARTEMIS global antifungal surveillance study revealed increases in reduced susceptibility to fluconazole in C. gattii [6], but it is still unclear whether and how this phenomenon affects the disease caused by this agent.
The aim of this study was to evaluate the influence of reduced susceptibility to fluconazole on the disease caused by C. gattii in C57BL/6 mice and whether this azole alters polysaccharide from cryptococcal cells. Our results suggest that the development of reduced susceptibility to fluconazole alters the polysaccharide capsule and the clinical manifestations of cryptococcosis, influencing whether the disease is restricted to the lungs or is disseminated to the CNS.

Ethics statement
The protocol of animal studies was approved by the Comitê de É tica em Experimentação Animal (CETEA) from Universidade Federal de Minas Gerais (Protocol 170/2011) and animal experiments were performed in strict accordance with the Brazilian Federal Law 11,794 establishing procedures for the scientific use of animals. All mice were housed in clean bedding (five mice per cage) with food and water ad libitum in a controlled environment with a 12 h light/dark cycle at 23uC. All mice were monitored twice daily. For intratracheal inoculation, mice were anesthetized by intraperitoneal injection of ketamine hydrochloride (50 mg mL 21 ) and xylazine (0.02 mg mL 21 ) in sterile saline. All efforts were made to minimize suffering. Any mice that appeared moribund (e.g. intense piloerection, convulsions, lack of locomotor activity) were euthanized immediately. Mice were euthanized under anesthesia (i.p. injection of ketamine hydrochloride (50 mg mL 21 ) and xylazine (0.02 mg mL 21 ) in sterile saline) by cervical dislocation by experienced animal handlers.

Cryptococcus gattii and antifungal drug susceptibility testing
We tested 12 strains of C. gattii (from the culture collection of the Laboratório de Micologia/Universidade Federal de Minas Gerais, Brazil). Initially, the MIC for fluconazole (Sigma-Aldrich, St. Louis, Missouri, USA) was determined in drug-supplemented solid culture medium in Sabouraud's dextrose agar (SDA). The MICs for fluconazole and amphotericin B (Sigma-Aldrich) were also determined by the microdilution method [8,[11][12][13]. Drug susceptibility testing was performed in three independent experiments in duplicate.

Reduced susceptibility to fluconazole induction
The synopsis of the methodology is presented in Figure 1. After determining the MIC of fluconazole on SDA, an average of five colonies from the highest MIC assay was selected for culture on SDA plates supplemented with fluconazole in a stepwise manner. The strain able to grow at the highest fluconazole concentration was chosen for the remaining tests. L27/01 strain of C. gattii was cultivated in high fluconazole concentrations and developed decreased drug susceptibility. After induction of reduced azole susceptibility, this phenotype was named L27/01 F . To verify the maintenance of reduced fluconazole susceptibility, the L27/01 F phenotype was cultured in SDA without drug 170 times, and the MIC was determined [11] every 5 subcultures ( Figure 1A-1H).

Intratracheal infection
C57BL/6 male mice/group (6 to 8 weeks old) were used in all experiments. Mice (n = 8/group) were anesthetized i.p. injection of ketamine hydrochloride (50 mg mL 21 ) and xylazine (0.02 mg mL 21 ) in sterile saline, then inoculated i.t. with 30 mL of 1610 6 CFU/animal of L27/01 or L27/01 F strains or PBS only (control). Animals were monitored twice daily for survival. To verify in vivo resistance, fluconazole (10 mg/kg/day) was administered daily i.p., from 5 to 11 days post infection (d.p.i.) ( Figure 1J) and lungs were removed for the determination of colony forming units (CFU) as described below.
Polysaccharides (PS) released from L27/01 or L27/01 F cells were isolated from culture supernatants by filtration and extraction of capsular components with DMSO. Ultrafiltration using an Amicon (Millipore, Danvers, MA) ultrafiltration cell (cutoff = 100 kDa) was performed as described previously [20]. The final PS was quantified and surface-associated capsular PS was isolated from yeast cells [21]. Further, yeast cells (10 6 ) were incubated for 1 h with mAb 18B7 (10 mg/ml). Then, cells were incubated with FITC-labeled goat anti-mouse IgG (Fc specific) antibody (Sigma) for 30 min at 37uC [22]. After determining the MIC of fluconazole on SDA, an average of five colonies obtained from the highest fluconazole concentration were selected for culture on SDA plates supplemented with this drug. L27/01 (A) was the strain able to grow at the highest fluconazole concentration (B) and was chosen for culture on SDA plates supplemented with this drug. This strain was cultured in solid medium with increasing concentrations (B-C) of fluconazole until growth ceased at 100 mg/mL (D). To verify the maintenance of resistance to fluconazole and cross-resistance between this drug and amphotericin B, the selected strain was cultured in SDA without drug every 48 h 170 times (E), and the MIC test was performed by microdilution every 5 subcultures (F). Colonies grown at 95 mg/mL were maintained in this concentration, and the strain was named ''L27/01 F '' (G). L27/01 strain grew in the absence of drug (H). The genetic similarity between L27/01 and L27/01 F strains was evaluated by randomly amplified polymorphic DNA (RAPD)-PCR, PFGE and ITS Sequencing. CAP59, CAP64, ARF-1, ERG11, UXS-1 levels by real-time PCR were evaluated. Lipid evaluation was performed to compare the ergosterol content of L27/01 and L27/ 01 F cell membranes. Also, Urease and Laccase activities of L27/01 and L27/01 F strains were determined (I). Evaluation of murine cryptococcosis after inoculation with L27/01 or L27/01 F strain: survival curve (J). Investigation of whether fluconazole affects the polysaccharide (PS) capsule (K). The phagocytosis assay was performed to assess the influence of PS capsules from L27/01 and L27/01 F strains on phagocytosis and intracellular proliferation rate (IPR) in murine peritoneal macrophages from C57BL/6 mice (L). Cryptococcal cell dissemination, immune response and behavioral The effective diameter and polydispersity of PS preparations were measured by QELS in a 90Plus/BI-MAS multi-angle particle sizing analyzer (Brookhaven Instruments Corp., Holtsville, NY) as described by Frases et al. [23]. Zeta potential (), particle mobility and shift frequency of polysaccharide samples were calculated in a Zeta potential analyzer (ZetaPlus, Brookhaven Instruments Corp., Holtsville, NY) [24].

RNA extraction, reverse transcription and real-time PCR
Total RNA from L27/01 and L27/01 F strains was extracted with RNeasy kit (Qiagen, Germany), and RT-PCR was performed using the enzyme MMLV (Promega, Madison, WI, USA). cDNA was used to determine CAP59, CAP64, MDR1, ARF-1, ERG11, UXS-1 levels by real-time PCR using 200 nM of specific primers and SYBR Green Master Mix (Applied Biosystem, USA, 10 mL) ( Figure 1I). The primers used (available in Table S1) were designed according to the DNA sequences available from the C. gattii database from the Broad Institute (http://www. broadinstitute.org/annotation/genome/ cryptococcus_neoformans_b/MultiHome.html). Analyses were performed using the delta-delta Ct method and normalized using the C. gattii actin gene.

Phagocytosis assay, Intracellular Proliferation Rate, measurement of ROS production by macrophages and enzymatic activity of the antioxidant system
The phagocytosis assay was performed as described previously [1] to assess the influence of PS capsules from L27/01 and L27/ 01 F strains on phagocytosis, on the intracellular proliferation rate (IPR) in murine peritoneal macrophages from C57BL/6 mice, and on the ability to neutralize ROS during the phagocytosis assay. Briefly, murine macrophages (2610 5 cells/mL) were obtained 3 days post thioglycollate (3%) injection and maintained in RPMI 1640 supplemented with 10% fetal bovine serum. Macrophages were immediately infected with L27/01 or L27/01 F yeast cells (6610 4 cells/mL) opsonized with 10% murine serum, with or without PS (10 mg/L), and incubated for 3 or 27 h at 37uC under 5% CO 2 for phagocytic index, intracellular proliferation rate (IPR) determination, ROS quantification, superoxide dismutase (SOD) and peroxidase (PER) measurement. The phagocytic index was calculated as the percentage of cells with internalized C. gattii after 27 h, while the IPR was calculated as the quotient of the intracellular yeast cell numbers at 27 h (time point which was the maximum intracellular yeast number) and 3 h [1] ( Figure 1L). 29,79-dichlorofluorescein diacetate (DCFH-DA; Invitrogen, Life Technologies, Carlsbad, CA, USA) was used for ROS quantification and the fluorescence was measured with a fluorometer (Synergy 2 SL Luminescence Microplate Reader; Biotek) [13]. A cell-free extract from C. gattii-infected macrophages or fungus or PS capsule alone (control groups) was obtained for SOD and PER activities [13].

Behavioral analysis
The SHIRPA protocol for behavioral and functional assessment of neurological diseases was used ( Figure 1M). The tasks are grouped into five functional categories: neuropsychiatric state, motor behavior, autonomic function, muscle tone and strength, and reflex and sensory function (available in Table S2) [27,28]. The mice were examined daily and the score for each functional categories was calculated as the total of the evaluated parameters according to Lackner et al. [27] and Pedroso et al. [28] using the EpiData 3.1 software.

Statistical analysis
Results are shown as means 6 SEM. Statistical analysis of all data were performed using GraphPad Prism version 5.0 with p, 0.05 considered significant. Survival curve was plotted by Kaplan-Meier analysis and results were analyzed using the log rank test. Weight variation was analyzed with Wilcoxon Rank-Sum Test. The results of MDR1, CAP59 and CAP64 genes expression, capsule size, ergosterol, enzymes and CFU/g organ or BALF were analyzed by analysis of variance (ANOVA) and the nonparametric Friedman test, Student's t test. Also, the results of phagocytosis assay, intracellular proliferation rate, measurement of ROS production by macrophages and enzymatic activity of the antioxidant system, myeloperoxidase activity and quantification of cytokines and chemokines were analyzed by analysis of variance (ANOVA) and the nonparametric Friedman test, Student's t test. Statistical analyses were carried out using 90Plus/BI-MAS software for effective diameter and polydispersity (Brookhaven Instruments Corp., Holtsville, NY). Zeta Plus software was used for Zeta potential, (Brookhaven Instruments Corp., Holtsville, NY). SHIRPA data were analyzed with Wilcoxon Rank-Sum Test (28).

L27/01 and L27/01 F show differences in susceptibility to fluconazole
Initially, 12 strains of C. gattii were used in susceptibility tests with fluconazole (by microdilution and supplemented solid medium) and with amphotericin B (by the microdilution method) (data not shown). During the tests, we chose an average of five colonies of the strain that was able to grow at the highest concentration of fluconazole for successive subcultures every 48 hours with increasing concentrations of fluconazole: 32; 64; 128 and 256 mg/mL. As the growth was interrupted at 128 mg/mL, we reduced the drug range concentration in solid medium to: 64; 70; 75; 80; 85; 90; 95; 100; 105 and 110 mg/mL. Thus, growth was observed up to a concentration of 95 mg/mL since the growth was interrupted at 100 mg/mL. L27/01 strain was chosen due to growth at the highest fluconazole concentration. The minimum inhibitory concentrations (MICs) for fluconazole against the L27/ 01 strain varied from 8.0 to 16.0 mg/mL and for amphotericin B it was 0.06 mg/mL. This strain was chosen to be tested in the other experiments and was cultured in solid medium with increasing concentrations of fluconazole until growth ceased at 100 mg/mL. Colonies grown at 95 mg/mL were maintained in this concentration ( Figure 1C-G), and the phenotype was named L27/01 F ( Figure 1G). This procedure showed significantly reduced susceptibility to fluconazole which persisted until the 170 th subculture in drug-free medium as confirmed by the microdilution (MIC = 128 mg/mL). For amphotericin B, the MIC increased from 0.06 to 1 mg/mL, demonstrating cross-resistance with fluconazole. Phenotypic examination revealed that L27/01 strain produced mucous colonies, while the stress caused by fluconazole (L27/01 F strain) led to dry colonies (data not shown). RAPD (data not shown) and PFGE (Figure 2A) revealed identical banding patterns between L27/01 and L27/01 F , which were clustered on the same clade with 100% similarity. We performed the sensibility test of the two strains to different temperatures and we verified no diferences between them to high temperatures (data not shown), corroborating the results of PFGE. The DNA sequence of the ITS1 region PCR products (555 bp) of L27/01 and L27/01 F strains were analyzed from the BLAST search and corresponded to C. gattii isolate deposited in GenBank (accession nu. JN939462.1) with 100% identity and 100% query coverage. Also, the alignment of these sequences displayed 100% similarity between L27/01 and F strain (data not shown). Furthermore, L27/01 and L27/01 F strains were subjected to RT-PCR to test for expression of genes related to efflux pumps (AFR1 and MDR1) or ergosterol synthesis (ERG11). The L27/01 F strain overexpressed the MDR1 gene ( Figure 2B) (p,0.05), but expression of ARF1 and ERG11 genes was similar in L27/01 and L27/01 F strains (data not shown). Ergosterol content was 50% lower (p,0.05) in the L27/01 F strain compared to L27/01 ( Figure 2C).
Reduced susceptibility to fluconazole in strain L27/01 F in vitro was associated with in vivo antifungal resistance Interestingly, animals infected with L27/01 and treated with 10 mg/kg/day of fluconazole showed a significant (p,0.05) reduction of fungal burden in the lungs, while animals infected with L27/01 F presented no reduction in CFU compared with untreated infected mice ( Figure 2G).
Reduced virulence of C. gattii was associated with alterations in the polysaccharide capsule: reduced susceptibility to fluconazole impairs GXM synthesis and secretion, but does not block polysaccharide assembly.
Since we found distinct profiles of disease caused by the L27/01 and L27/01 F strains, we further investigated whether the reduction of susceptibility to fluconazole affects the main virulence factor of the genus Cryptococcus: the polysaccharide capsule. L27/ 01 F strain showed reduced ability to form a polysaccharide capsule compared to the L27/01 strain ( Figure 3A-C). Genes related to capsule (CAP59 and CAP64) were quantified by RT-PCR; CAP59 and CAP64 expression were significantly (p,0.05) higher in the L27/01 strain (40% and 30%, respectively) ( Figure 3I). However, the expression of a gene related to polysaccharide assembly (UXS1) was similar between strains ( Figure 3I). No significant differences in extracellular polysaccharide levels were apparent (2.3 mg/mL for the L27/01 F strain and 2.5 mg/mL for L27/01). DMSO extracts of whole cells revealed differences between fractions (5.3 mg/ml for L27/01 and 2.5 mg/ml for L27/01 F strain), which is consistent with results from India ink ( Figure 3A-B) and scanning electron microscopy ( Figure 3D-E).

Fluconazole affects physical and structural properties of the L27/01 F strain capsule
Considering that the physical properties of the cryptococcal capsule are related to binding by monoclonal antibodies, we examined the fluorescence pattern of the 18B7 monoclonal antibody (IgG1) bound to the L27/01 F and L27/01 strains ( Figure 3G-H). The results showed differences in epitope presentation in capsular-PS for both strains with lower fluorescence intensity for L27/01 F strain, revealing differences in opsonine patterns that may influence phagocytosis.
Hydrodynamic properties of L27/01 F and L27/01 PS were calculated from dynamic light scattering measurements. The L27/ 01 strain showed a longer PS capsule size distribution with a maximum size of 7000 nm; L27/01 F cells presented a polydispersed size distribution shorter than L27/01 ( Figure 3J-K). For secreted PS, L27/01 F and L27/01 cells showed no significant differences in effective diameter ( Figure 3L-M). Conversely, an important finding is that capsular PS differed in Zeta potential between the parental and L27/01 F strains ( Figure 3F). However, secreted PS did not show mobility differences between the two strains ( Figure 3F).

L27/01 F strain is more internalized by macrophages
Considering the differences in PS capsule between the two strains and the well characterized anti-phagocytic properties of GXM in previous studies, the phagocytosis assay was performed in vivo by counting internalized yeasts in BALF cells and in vitro with murine peritoneal macrophages infected with C. gattii. The influence of the polysaccharide of L27/01 and L27/01 F strains on the phagocytic index ( Figure 4A) and intracellular proliferation rate (IPR) ( Figure 4B) were determined. The results demonstrated similar profiles in vitro and in vivo; the L27/01 F strain was more susceptible to internalization than the L27/01 strain (p,0.05). The L27/01 strain demonstrated a greater ability to proliferate inside macrophages and the proliferation was reduced when the L27/01 F polysaccharide capsule was added to the assay (p,0.05).
In addition, we evaluated whether fluconazole altered the antioxidant properties of the polysaccharide capsule. Reactive oxygen species (ROS) were determined in the supernatant from the phagocytosis assay with the L27/01 or L27/01 F polysaccharide (PS) capsule ( Figure 4C). The results demonstrated that ROS ( Figure 4C) production was significantly higher (p,0.05) when the macrophages were infected with the L27/01 F PS as well as when it was added to the culture of the L27/01 strain. Alternately, the L27/01 PS capsule added to the L27/01 F strain improved neutralization of ROS. These results show that stress to fluconazole compromises resistance to ROS.
For further investigation of how PS may neutralize ROS inside macrophages, we determined the activities of SOD and PER in yeast cells after 3 h of the phagocytosis assay. The results show that SOD activity was significantly higher for the L27/01 strain (p, 0.05) and for the L27/01 F strain plus L27/01 PS (p,0.05) ( Figure 4D). When the total PER activity ( Figure 4E) was measured, the L27/01 strain and the L27/01 PS capsule showed higher (p,0.05) PER activity than the L27/01 F strain and its polysaccharide capsule. Since different profiles were found in the phagocytosis assay and since the fungal translocation to the CNS was previously associated with the ability to survive in the macrophage intracellular environment, we verified the kinetics of the infection caused by the two strains. After, 0.5, 1, 7, or 15 days postinoculation with yeast cells, the fungal burden in lungs and brain were significantly higher in mice infected with the L27/01 strain ( Figure 5A and B). The ability of the L27/01 F strain recovered from the organs to grow at higher concentrations of fluconazole was maintained, as verified in medium supplemented with this drug (data not shown). The L27/01 F strain did not disseminate to the brain until 30 days post-inoculation, while the L27/01 strain colonized the CNS within 12 hours ( Figure 5B). Analysis of bronchoalveolar lavage fluid (BALF) ( Figure 5C) revealed that the fungal burden of L27/01 strain increased within 7 days postinoculation (p,0.05), while in the L27/01 F strain the burden remained lower in the extracellular space. These results were confirmed in three independent experiments in triplicate.
In the histopathological analysis, yeasts varying from 5 to 30 mm in size were observed free within the cytoplasm in the bronchial, bronchiolar and alveolar lumen. Infiltrate resulting from mild to intense inflammation was composed of neutrophils and macrophages, sometimes forming granulomas ( Figure 5E and F). Thickening of the alveolar walls was also observed. The intensity of lesions in the lung differed between strains 15 days postinfection, being mild in mice inoculated with the L27/01 strain and moderate to intense in mice inoculated with the L27/01 F strain. In brain tissue, cryptococcomas, characterized by a focal mass containing multiple grouped yeast cells associated with the destruction of the neuropil, were only observed in the hippocampus of mice infected with the L27/01 strain; in other areas, no alteration was detected ( Figure 5G). Although cryptococcomas were not associated with inflammatory infiltration, degeneration and neuronal death were observed in the hippocampus (Figure 5G).

L27/01 F strain increases the recruitment of host cells to the lungs, but decreases production of IFN-c, TNF-a, IL-10 and CXCL1/KC
Mononuclear leukocytes and neutrophils were more recruited to the lungs by the L27/01 F strain than the L27/01 strain (p,0.05) ( Figure 6A-B). The MPO activity in the lungs ( Figure 6C) confirmed higher neutrophil influx with the F strain (p,0.05).

L27/01 leads to significant behavioral alterations associated with neurocryptococcosis
Mice infected with L27/01 or L27/01 F strains were assessed using the SHIRPA protocol. The results demonstrated significant changes (p,0.05) in the behaviors of the two infected groups. In the L27/01 group, muscle tone and strength ( Figure 7A) were altered 9 day post-inoculation (p,0.05), while this clinical function was altered in the L27/01 F group only 22 days post-inoculation. Motor behavior ( Figure 7B) was significantly altered after 16 days for the L27/01 strain (p,0.05). Neuropsychiatric state ( Figure 7C) was significantly altered after 14 days for the L27/01 strain (p, 0.05). Autonomous ( Figure 7D), reflex, and sensory function ( Figure 7E) were altered 19 days post-infection with L27/01 cells. Motor behavior, neuropsychiatric state, autonomous function, reflex and sensory functions were not altered in mice infected with the L27/01 F strain.

Discussion
Cryptococcus gattii isolates are frequently reported as less susceptible to azoles than other Cryptococcus species [29][30][31]. Based on the Epidemiological Cutoff Values (ECVs) defined by Espinel-Ingroff et al. [30], L27/01 F strain showed the ability to develop secondary resistance to fluconazole because the higher MICs (increased from 8-16 mg/mL to 128 mg/mL) persisted until the 170 th repeated culture in drug free medium. RAPD and PFGE analysis revealed identical banding patterns between L27/01 and L27/01 F . According to Torres et al. [32], aneuploid strains can show increased sensitivity to high temperatures, but we verified no diferences between them to high temperatures (data not shown) in sensibility test to temperature. Although cultivating Cryptococcus in fluconazole supplemented media may also generate fluconazoleheteroresistant clones [33][34][35], L27/01 F may not be a heteroresistance phenotype, since the higher MIC values persisted after 170 subcultures in drug free medium. Heteroresistance is also a phenomenon present in Candida albicans [36,37].
Interestingly, the reduced susceptibility to fluconazole was confirmed in the in vivo tests, as this drug was unable to reduce the CFU in the lungs of animals infected with L27/01 F . Additionally, environmental isolates of fluconazole-resistant C. gattii have also been reported, demonstrating the possibility of primary resistance to fluconazole [38]. The data regarding the influence of fluconazole in C. gattii virulence are still scarce, but some evidence has been reported for C. neoformans [33,39]. Special attention should be directed to prophylaxis with fluconazole, which increases the possibility of cross-resistance to other drugs [40], as we demonstrated for amphotericin B in the L27/01 F strain. The reduced ergosterol levels in the L27/01 F strain probably caused the cross-resistance to amphotericin B, confirming previous data [8]. On the other hand, overexpression of the MDR1 gene was the probable mechanism of significant decreased susceptibility to fluconazole in the L27/01 F strain, increasing efflux to the extracellular environment. Different from our data, Gast et al. [41] found overexpression of ERG11 gene as responsible for reduced susceptibility to fluconazole in C. gattii strains. Overall, we think that the mechanisms of acquired activities of L27/01 and L27/01 F strains: one enzymatic unit was calculated per 10 5 and 10 8 cells of C. gattii for urease and laccase activity, respectively, after 24 hours (D). Eight mice per group were inoculated by the intratracheal route with 10 6 cells of L27/01 or L27/01 F . Animals administered with PBS represent the non-infected (control) group. Animals were monitored daily. Survival curve (E). Weight variation (Score) (F). Each 1 (one) gram received or lost corresponds to 10 points added or subtracted, respectively. *The vertical dotted line indicates a significant difference from the appointed day. CFU/g of lungs (G) from animals infected with 1610 6 cells of L27/01 or L27/01 F strains treated i.p. with fluconazole at 10 mg/kg/day (n = 6). NI: not infected group. NT: non treated group. Black bars refer to L27/01 (&) and grey bars refer to L27/01 F (m). Data represent the mean of three independent experiments in triplicate. *P,0.05 was considered to be significant. doi:10.1371/journal.pone.0112669.g002 resistance may be strain-dependent and MDR1 was prominent in our study, with no role developed by ERG11, since the ergosterol levels were significantly reduced in L27/01 F .
The less virulent phenotype of L27/01 F was associated with lower activities of urease and laccase [18,19], probably compromising its inability to disseminate to the CNS together with poor capsule formation. The lower expression of CAP59 and CAP64 genes confirmed these results. Although the exact role of the encoded Cap proteins remains unknown, it has been suggested that CAP59 plays a role in the trafficking of GXM PS molecules [42], whereas proteins encoded by CAP64 are involved in the acetylation of GXM [43]. The CAP64 gene is reportedly essential for virulence [44,45]. The data allowing for the description and characterization of less susceptible C. gattii are still scarce, but some data on C. neoformans suggests that fluconazole and other drugs may promote capsule shedding, leading to poorly encapsulated or acapsular cells [46], but without a correspondence in murine model since most of these studies focus in vitro analyses. Our findings may have clinical implications because a large capsule may facilitate fluconazole penetration into the yeasts due to its more hydrophilic profile [47]. Altogether, a relationship between the ergosterol biosynthetic pathway and the virulence of cryptococcal cells is plausible [48]. L27/01 cells were able to modulate the recruitment of host cells to the lung. The virulence of the L27/01 strain was higher in this murine model owing to intracellular replication and lack of susceptibility to macrophages, most likely due to the properties of its PS capsule [47,49]. According to Johnston and May [50], one speculative possibility is that phagocytosis of Cryptococcus by macrophages can fail due to capsular PS shedding from the fungal cell, causing the internalization of only PS fractions and releasing of cryptococcal cell. The similarity between strains in the expression of the UXS1 gene, which is related with the incorporation of PS fibers to the capsule [51] showed that L27/ 01 F is able to increase the PS content of the cell, corroborating the results found for intracellular proliferation, ROS and antioxidant enzymes: in these assays, addition of L27/01 PS to the L27/01 F strain provided protection. Fluconazole also reduced the negative charge and most likely the electrostatic repulsion between the L27/01 F strain and macrophages, probably increasing phagocytosis [46].
It is still unclear how the PS capsule neutralizes oxidative bursts and promotes fungal survival inside macrophages, although these properties have been mentioned previously [47]. In our study, the higher susceptibility of the L27/01 F strain to ROS was associated with impairment of the antioxidant properties of the PS, which showed lower SOD and PER activities during phagocytosis. These data provide new insights into the influence of antifungal drugs on the virulence factors of C. gattii, demonstrating a connection between fluconazole resistance and the antioxidant system.
The murine model developed in this study clearly mimics the natural route of infection and revealed the ability of the L27/01 strain to disseminate to the CNS. This was confirmed with histopathological analyses, which showed typical cryptococcomas, as well as behavioral alterations. Altogether, this model permits evaluation of the ability of different C. gattii strains to affect the CNS, and also the host response to the fungus. The higher fungal   burden in the lungs of mice infected with the L27/01 strain induced pro-inflammatory cytokines and, in turn, increased IL-10 concentrations, but this response did not prevent death. These data corroborate a previous study of murine cerebral infection with C. gattii [12]. In mice infected with the L27/01 F strain, higher cellular recruitment led to intense inflammatory reactions and alveolar thickening. TNF-a and IFN-c increase the activity of phagocytes against cryptococcal cells; furthermore, IL-10 inhibits macrophage activation. Chemokines have a clear role in the recruitment of neutrophils to the lung, and mortality may result from an increased response of cytokine production [25,26]. Overall, our data suggest that the immunomodulatory effects of L27/01 F may be linked to antigenic variation of surface proteins or carbohydrates [52] caused by fluconazole. In addition, the smaller size (and size distribution) of PS fibers compared to the parental strain may contribute to immunomodulation. However, we cannot exclude that molecular mechanisms such as epigenetic and chromosomal disomy [33,35] may play a role in the phenotypic change of cryptococcal cells. Aneuploidy or isochromosome formation are also important in drug-resistant C. albicans [53,54].
The altered behaviors of mice in the SHIRPA [27,28] assay were similar to those described in humans with cryptococcosis [2]. The clinical domains not altered by the L27/01 F strain confirmed no dissemination to the CNS. To our knowledge, this is the first study to demonstrate behavioral changes in a murine model of cryptococcosis. The hippocampal neurodegeneration observed in mice infected with the L27/01 strain may account for the distinct behavioral profile [28,55].
In conclusion, fluconazole causes changes in the PS capsule as an adaptation to the cellular stress caused by this antifungal agent. This phenotype persists, even upon removal of stress after many generations, and is correlated with in vivo susceptibility, revealing distinct profiles in recruitment of host cells and clinical manifestations of cryptococcosis, which has clinical implications. In addition, our results demonstrate a murine model suitable for pathogenicity studies and assessment of neurocryptococcosis and confirm that intracellular proliferation plays a key role in infection. We suggest that further studies with antifungal-resistant C. gattii strains are needed in order to better define the pathogenesis of infection and the mechanisms of fungal survival.