The crude extract of E. uniflora leaves was obtained at 10% (w/v) with acetone: water (7:3, v/v) by turbo extraction in four rounds of 5 min. The extract was filtered and concentrated by rotary evaporation at 40°C (RV10 Basic, IKA®). The residue was frozen (−80°C, 3 days) and then lyophilized (Model L101, Liotop®) to yield a Crude Extract (CE). Approximately 20 g of CE were reconstituted in water (200 mL) and this solution was fractionated six times with 20 mL hexane. The resulting aqueous fraction was partitioned twelve more times with 20 mL ethyl acetate. This ethyl acetate fraction (EAF) was concentrated, frozen and lyophilized. The EAF was standardized as 17.82% m/m (expressed in rutin) of total flavonoids by UV/VIS and 6.56 % m/m (expressed as myricitrin) by HPLC (Ramos et al., 2017).

In order to evaluate C. albicans resistance to phagocytosis by polimorphonuclears neutrophils, (PMNs), 48 clinical isolates obtained from renal transplant recipients with oral candidiasis (Project approved under 152/07 protocol) and both references strains, C. albicans ATCC90028 and C. albicans SC5314 were used. These strains were kept in the culture collection of the Medical and Molecular Mycology Laboratory, Department of Clinical and Toxicological Analyses, Federal University of Rio Grande do Norte.

For all the other experiments, we selected C. albicans strain 111R. We previously observed this is a highly filamentous strain after induction of morphogenesis in the presence of serum and other morphogenesis inducing media (Silva-Rocha et al., 2015).

The isolates were stored at −80°C in YPD liquid medium (dextrose 20 g/L, peptone 20 g/L, yeast extract 10 g/L) containing 20% (v/v) glycerol. The cryotubes of 2 mL of capacity (Cralplast; Cotia Sao Paulo, Brazil) were thawed on ice and 100 μL of cells suspension of each strain were added to 5 mL of YPD liquid medium and incubated in a shaker afterwards (Tecnal, TE-420, São Paulo, Brazil), at 35°C for 48 h for the reactivation and verification of viability. Subsequently, 100 μL of each cell suspension was inoculated on the surface of Sabouraud Dextrose Agar (SDA; Oxoid, UK) containing 300 μg/mL of chloramphenicol (Park–Davis), using a Drigalsky loop. The plates were incubated at 37°C for 48 h. Yeast colonies were plated on CHROMagar® Candida (CHROMagar Microbiology, Paris, France) to check for purity and screening for different color colonies. Species identification was based on the characteristics of the cells observed microscopically after cultivation on corn meal agar containing Tween 80, as well as classical methodology (Yarrow, 1998).

For all the virulence factors evaluated in vitro, all C. albicans strains (phagocytosis) and 111R only (further assays) were initially grown in NGY medium (Difco Neopeptone 1 g/L, Dextrose 4 g/L; Difco yeast extract 1 g/L) for 18–24 h in a rotatory shaker (Tecnal, TE-420, Sao Paulo, Brazil) at 30°C, 200 rpm. This culture medium produces an inoculums size of about 2 × 10⁸ cells/mL. Cultures were spectrophotometrically measured at a wavelength of 600 _nm ranging from 0.8 to 1.2 (Biochrom Libra S32). Subsequently, C. albicans cells were diluted to obtain the specific inoculum needed for each attribute of virulence evaluated in vitro (Chaves et al., 2007). For the experiments performed after growth in the presence of the EAF of E. uniflora, 1,000 μg/mL of the natural product was added to the NGY broth. The experiments were performed in a manner that both control (in the absence of the EAF fraction) and test (in the presence of the EAF) assays had the same number of C. albicans viable cells after CFU counting, after 48 h of incubation in SDA.

PMNs freshly isolated from blood samples of the same healthy volunteer on the day of the experiment (Fradin et al., 2005) were suspended in Eagle's minimal essential medium (Gibco) + 20 mM HEPES, pH 7.2, and standardized to 8 × 10⁵ PMN/mL. C. albicans strain 111R cells grown overnight in NGY broth in the presence and absence of the EAF of E. uniflora were washed and resuspended to contain 5 × 10⁶ yeasts/mL in HEPES-buffered Eagle's minimal essential medium containing one-tenth volume of fresh human plasma for opsonization purposes. Equal volumes (100 μl) of PMNs and yeast suspensions were mixed and incubated at 37°C for 3 h with rotation at 50 rpm (Tecnal, TE-420, Sao Paulo, Brazil). Phagocytosis of C. albicans was determined by counting the number of yeasts cells inside or attached to 100 PMNs with optical microscopy (CX21, Olympus, Japan; Chaves et al., 2012). The assay was performed in triplicate.

Morphogenesis was induced according to Chaves et al. (2007). C. albicans cells were grown overnight in NGY broth (30°C–200 rpm) in the presence and absence of the EAF of E. uniflora. C. albicans cells were diluted in saline solution and standardized to 1 × 10⁶ cells/ml.

In order to induce morphogenesis, C. albicans cells suspension was inoculated to 150 ml of 20% (v/v) Fetal Bovine Serum (FBS; Sigma-Aldrich; Missouri, Saint Louis, USA) in YPD broth in 200 ml of capacity flasks. The strain was incubated under mechanical rotation at 37°C, 200 rpm, for a period of 3 h for further protein extraction.

A concentration of 1 × 10⁶ cells/ml of C. albicans 111R cells was used. After morphogenesis induction, C. albicans cells were transferred to 15 mL conic tubes (Cralplast; Cotia Sao Paulo, Brazil), centrifuged for 20 min, 13,000 rpm, 4°C (280R, FANEN) and washed three times with ice cold ultrapure water. Total protein extraction was performed with C. albicans cell pellets. Therefore, 5 ml of Tris-HCl 10 mM (pH 8.8) buffer containing a protease inhibitor solution cocktail (Protease Inhibitor Cocktail for use with mammalian cell and tissue extracts, DMSO solution, Sigma-Aldrich; Missouri, Saint Louis, USA) was added to the yeast cells (995 μL Buffer, 5 μL of protease inhibitor) subsequently macerated with liquid nitrogen. The samples were centrifuged at 13,000 rpm, 4°C and the supernatant was collected. The total protein concentration of the supernatant was quantified by Bradford method (Bradford, 1976) using Bovine Serum Albumin (BSA; Sigma-Aldrich; Missouri, Saint Louis, USA) as a standard and the absorbance readings were performed at 595 wavelength by spectrophotometer.

Protein profiles were analyzed with Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis, SDS-PAGE, under reducing conditions using the discontinuous buffer system (Laemmli, 1970; Studier, 1973).

Total soluble proteins were separated and analyzed by two-dimensional electrophoresis (Görg et al., 2000). Firstly the proteins were submitted to Ettan IPGphor 3 (GE Healthcare) according to the following parameters: First gradient—500 V, second gradient—1,000 V, third gradient—8,000 V, fourth gradient—8,000 V, 20°C, 50 mA. The second dimension was carried out for the separation of proteins according to molecular mass, and held on 12.5% (w/v) polyacrylamide gel according to Laemmli (1970). Gels from tests and control experiments extracted proteins were stained with Coomassie Brilliant Blue G-250, according to Neuhoff et al. (1988). The gels were scanned (Labscan, HuntersLab). The analysis of data was performed with the ImageMaster 2D Platinum 6.0 GE software.

The spots resulting from gel electrophoresis were then removed and cut into segments of ~1 mm³. The fragments were subsequently destained with 400 μL of decolorizing solution containing acetonitrile (ACN) 50% (v/v) in ammonium bicarbonate 25 mM, stirred with a vortex for 10 min between each washing until the complete removal of the dye. Gel excised fragments were then completely dehydrated with 200 μL of 100% ACN and than dried in a concentrator (Vacfuge Plus™, Eppendorf). The fragments were subsequently rehydrated with 40 μL of Dithiothreitol (20 mM DTT in 50 mM ammonium bicarbonate) for 40 min at 56°C. After rehydration was carried out, alkylation with 40 μL of iodoacetamide solution (55 mM iodoacetamide in 50 mM ammonium bicarbonate) was performed. A second dehydration was performed with 200 μL of ACN (100%). Subsequently, 15 μL of trypsin solution in acetic acid (100 ng/μL) was added to the dried gel fragments which were then rehydrated in ice. A 40 μL volume of 25 mM ammonium bicarbonate was added and the fragments were incubated at 37°C for 14h.The reaction was stopped with 15 μL of blocking solution (ACN 50% (v/v) and formic acid 5% (v/v). Three elutions were carried (repated twice each time) and supernatants were collected after all elutions. The first elution was performed with 40 μL of methanol 60% (v/v) and formic acid 1% (v/v), and then the samples sonicated for 15 min at 40°C. For the second elution 40 μL of ACN 50% (v/v) and formic acid 1% (v/v) was added and then samples were sonicated for 15 min at 40°C. A volume of 40 μL of ACN 100% was added to dehydrate gel fragments, subsequently sonicated for 15 min. The eluted peptides were then subjected to concentration in a Concentrator (Vacfuge Plus™, Eppendorf) during 30 min. After concentration, samples were resuspended in 10 μL of trifluoroacetic acid (TFA) 0.1% (v/v) in 100% ACN and subjected to purification before mass spectrometry analysis.

Purification was performed using Reverse Phase-Pippete Zip-Tip (Millipore) following the manufacturer's instructions. Peptide samples were resuspended in 0.1% (w/v) TFA solution. The ZipTips were equilibrated with 10 μL of a solution containing ACN and 0.1% TFA and the solution discarded afterwards. This procedure was performed three times. A 10 μL volume of water and 0.1% TFA solution was gently homogenized and further discharged for three times. To bind the samples inside the pipette tips to columns, 10 μL of sample peptides were aspirated for 7 to 10 times to increase fixation to columns. After binding of the peptides to columns, two washing steps were performed. Firstly, 10 μL of a 5% (v/v) methanol and 0.1% TFA solution was added and the process repeated three times. The elution process was ended after gentle homogenization of peptides the addition of 5 μL of 0.1% TFA solution in 50% ACN and subsequently the samples were transferred to clean 2 mL microcentrifuge tubes (Salvato and Carvalho, 2010).

After trypsin digestion and purification, the peptides were then analyzed by mass spectrometry. The peptides were identified by mass spectrometry by MALDI-TOF/TOF equipment (TOF/TOF™ 5800 System, Sciex), the fragmentation spectra (MS/MS) containing the molecular weights of the peptides were evaluated using the MASCOT database (MatrixScience). Morphogenesis assays and proteomic analysis were perdormed in triplicate.

The evaluation of the effect of the EAF of E. uniflora on filamentation of C. albicans was performed in vivo according to Solis and Filler (2012). Twenty-four adult mice (Mus musculus) of 12 weeks old weighing 18–25 g were used. The animals were obtained from the Animal House of the Center of Health Sciences, Federal University of Rio Grande do Norte, kindly provided by Dr. Antonia Cláudia Jácome da Câmara. Water and food were provided ad libitum throughout the study period. The day before animal's experimentation, C. albicans 111R was grown overnigth in NGY broth in the absence (control experiment) and presence (test experiment) of 1,000 μg/mL of the EAF of E. uniflora at 30°C, 200 rpm for 18–24 h. The animals were weighed a day before inoculation to calculate the dose of the immunosuppressant (prednisolone, EMS; Hortolandia, Sao Paulo, Brazil) subcutaneously administered to each individual (100 mg/kg in 200 μL). Prednisolone was administered a day previously to infection and 3 days after infection. An inoculum of 1 × 10⁶ cells of C. albicans 111R strain was prepared in PBS buffer. The animals were intraperitoneally anesthetized with a mixture of ketamine (Francotar®, Virbac;110 mg/kg) and xylazine (Calmiun®, Agener União) (10 mg/kg). After anesthetizing the animals, a sterile swab was inserted in microcentrifuge tubes containg C. albicans inoculum, all the cells suspension (100 μL) was dropped and rubbed on the oral cavity of the animals during 60 s. CFU counts were performed before inoculation to verify if cells were viable.

The animals were divided in three different groups: Group I: Animals that were infected with C. albicans strain 111R grown in NGY broth the absence of E. uniflora EAF. Group II: Animals that were infected with C. albicans strain 111R grown in NGY broth in the presence of E. uniflora EAF and Group III: Animals that were infected with C. albicans strain 111R grown in NGY broth in the absence of E. uniflora EAF, where the fraction was administered to the animals tongue after infection.

The animals from group III were intraperitoneally anesthetized with a mixture of ketamine (110 mg/kg) and xylazine (10 mg/kg) 2 days after the infection was established, and were subsequently treated by admiministring 50 μL of the E. uniflora EAF (1,000 μg/mL) onto the dorsum of the tongue for 3 times in intervals of 10 min.

The animals were sacrificed 5 days after infection with excessive dose of ketamine (110 mg/kg) and xylazine (10 mg/kg). A sterile swab was rubbed on the oral cavity of animals and introduced into a microcentrifuge tube containing 1 ml of 0.9% (w/v) saline. The samples were diluted 1:10 and seeded on the surface of 90 × 120 mm Petri dishes containing Sabouraud dextrose agar with 300 μg/mL chloramphenicol and incubated at 37°C for 48 h to perform CFU counting.

The animals tongues were excised with the aid of a sterile scalpel and immediately fixed in formaldehyde 10% (v/v) in PBS for further histopathological analysis of the tissue. After fixation steps, the samples were dehydrated with graded ethanol and diaphanized xylene. The samples were embedded in paraffin and sliced in eight mm of thickness, for the manufacture of the slides and then stained with hematoxylin-eosin for subsequent evaluation step of the cellular infiltrate and PAS (Periodic Acid Schiff) to assess the fungal load in a light microscope. In order to evaluate tissue damage resulting from the infection, the formation of tissue ulcers, erosion process, inflammatory response, as well as the type of response and presence and fungal burden at the infection site were evaluated.

Cell viability was studied in A549 cell lines at a concentration of 2.5 × 10² cells per well. E. uniflora ethyl acetate fraction was tested in concentrations of 8,000, 4,000, 2,000, 1,000, 500, 250, 125, 62.5, 37, and 18 μg/mL. Cytotoxicity was assessed by the MTT assay (bromide (3- [4,5-dimethyl-thiazol-2-yl] -2,5-diphenyltetrazolium bromide) (Sigma-Aldrich; Missouri, Saint Louis, USA), used at a concentration of 5 mg/ml according with Mosmann (1983) with some modifications.

The starting solution with the EAF fraction of E. uniflora was subsequently serially diluted in the specific medium used for cell line in the previously prepared microplates containing the mammalian cells. The plates were incubated at 37°C, 5% CO₂ under light for 24 h. The solution containing the fraction was removed and 10 μL of MTT solution was added to the wells, following another incubation step at 37°C for a period of 4 h, as previously described (Denizot and Lang, 1986; Gerlier and Thomasset, 1986; Ferrari et al., 1990). After the incubation time, the MTT solution was removed and added to 100 μL of isopropanol for diluting the precipitate. Viable cells changed their color from yellow to violet. Untreated cells constituted the positive control (viable cells), and cells treated with hydrogen peroxide (Sigma-Aldrich, St. Louis, MO, USA) constituted the negative control (dead cells). The spectrophotometric reading of the plates was performed in the microplate ELISA reader (BioRad model 3550) with a wavelength of 570 nm. The percentage of viable cells was calculated as follows:

In a previous study performed by our group, we realized that the extract of E. uniflora was able to impair proper hypha formation in both solid and liquid media (YPD+20% serum, Spider and N-acetyl-D-glucosamine; Silva-Rocha et al., 2015). In addition, the typical oval morphological form of C. albicans blastoconidia was altered; showing invaginations of the cell wall and an increase in the size of citoplasmic vacuoles. We hypothesized that the extract was somehow acting on the cell wall, and then we investigated how C. albicans cells grown in the presence of E. uniflora would interact with phagocytic cells. Therefore, we performed phagocytosis assays with PMNs.

PMNs were co-incubated with C. albicans cells that were previously grown overnight in the presence and absence of the EAF of E. uniflora. In the absence of the ethyl acetate fraction, C. albicans strains had a broad variability of number of yeasts phagocytized by 100 PMNs, ranging from 11 ± 3.2 (Strain 111R) to 176 ± 11.6 and 176 ± 6.1 (Strains 44 and 10Li, respectively) yeasts/100 PMNs. The mean value of the number of C. albicans phagocytized by 100 PMNs for all the strains was 120.36 ± 36.71 yeasts/100 PMNs (Table 1).

When C. albicans cells were grown overnight in the presence of E. uniflora extract, a significant reduction in the number of cells phagocytized by the PMNs was observed (Figure 1B). The number of C. albicans phagocytized by 100 PMNs ranged from 10 ± 1 to 10 ± 5.5 yeasts/100 PMNs (Strain 18A and 111R) to 91 ± 9.1 yeasts/100 PMNs (Strain 46; Table 1). A comparison among the mean value of phagocytosis of the control group (absence of the EAF) with the test group shows a significant reduction in the number of cells phagocytized by PMNs (120.36 ± 36.71 vs. 44.68 ± 19.84 yeasts/100 PMNs; Table 1). Phagocytosis of C. albicans cells by PMNs and hypha formation may be observed when cells were cultivated in the absence of the EAF of E. uniflora (Figure 1A).

Surprisingly, when C. albicans cells were grown overnight in the presence of the E. uniflora EAF, a significant reduction in the number of cells phagocytized by the PMNs was observed (Figure 1B). The number of C. albicans phagocytized by 100 PMNs ranged from 10 ± 1 to 10 ± 5.5 yeasts/100 PMNs (Strain 18A and 111R) to 91 ± 9.1 yeasts/100 PMNs (Strain 46; Table 1). A comparison among the mean value of phagocytosis of the control group (absence of the fraction) with the test group shows a significant reduction in the number of cells phagocytized by PMNs (120.36 ± 36.71 vs. 44.68 ± 19.84 yeasts/100 PMNs; Table 1).

It is important to emphasize again that C. albicans 111R is a highly filamentous strain found in our previous publication (Silva-Rocha et al., 2015). Due to its highly filamentous phenotype, which can be easily observed even when this strain is not incubated in hypha inducing conditions; this strain was not promptly phagocytized. Therefore, we could not observe a significant reduction in the number of phagocytized cells when this strain was previously grown in the presence of the EAF, because this strain was already poorly phagocytized (P > 0.05). However, because this was the most filamentous strain found in our study and we would like to prove that this natural product was able to impair morphogenesis, we chose it to perform our proteomic assays.

Because we observed that C. albicans 111 R was the most filamentous strain, but also showed reduced ability to form filaments in the presence of inducing media after growth in the presence of the EAF, we selected this strain to analyze which proteins are differentially expressed in the of the referred fraction. The induction of filamentation was performed in YPD+20% serum at 37°C, for 3 h and proteins were analyzed by bidimensional electrophoresis. A total of 39 proteins, conserved in the three different 2D-gels (triplicate assays), were selected and then identified by MALDI-TOF mass spectrometry. Proteins identification and their biological functions are described in Table 2.

The proteins identified were classified within seven distinct categories based on their biological function, as follows: Energy (Pam16, Atp2, Rcf1, Cagl0g03245g, Cox4, Ndk1, and Cawg_02398), Protein Metabolism (Gtp1, Cyp9, Rpl3, Pca2, Mrpl4, Rpn6, Dbp10, Tef1, and Mrh4), Glucose Metabolism (Eno1, Fba1, and Med7), Transport (Yet2, Sec1, Acb2, and Mth1), Nucleic Acid Metabolism (Tpp1, Rex3, and Lsm3), Cell Structure (Gim4, Atg26, and End3) and Stress Response (Ssb1). Three proteins (Mcpb, Ynr1, and Spg4) were not included in any of the categories, while two of them (Irc10 and Spac24c9.04) had unknown functions (Table 3).

From the total 39 proteins identified and classified, 23.1% were included in the “Protein Metabolism” category, generally associated with protein synthesis and structural formation of ribosomes. A total of 17.9% of identified proteins were included in the “Energy” category, and mitochondrial biogenesis and ATP synthesis were the main functions observed. The three proteins classified in “Glucose metabolism” group were related to glycolytic pathways. All proteins categorized in the “Transport” group were correlated with the maintenance of cell functions by promoting the transport of glucose and protein in the cytoplasm of C. albicans cells.

Proteins involved in the “Nucleic Acid Metabolism” were associated with biological functions of DNA repairing, RNA processing and protein folding. Membrane synthesis and endocytosis are functions related to the “Cell structure” group. Only a single protein (Ssb1) was included in the “Stress Response” group and its biological function is related to protein biosynthesis and stress response.

The pattern of C. albicans differentially expressed proteins in the presence of the EAF of E. uniflora is described in Table 4. Eight proteins were overexpressed in the presence of the EAF (Pam16, Med7, Atp2, Rcf1, Eno1, Ynr1, Spg4, and Ssb1), while 10 proteins showed reduced expression, as follows: Gtp1, Nuf2, Cyp9, Acb2, Mug97, Cox4, Yet2, Mth1, Ndk1, and Gim4.

Several biological processes were affected when the EAF of E. uniflora was added to C. albicans 111R NGY broth and morphogenesis was further induced. Proteins involved in generation of energy by ATP synthesis (Atp2), mitochondrial biogenesis (Pam16 and Rcf1) and glycolytic pathway (Med7 and Eno1) were overexpressed in the presence of the fraction. Heat Shock Protein Ssb1, involved in C. albicans stress response was also overexpressed in the presence of the EAF (Table 4).

Proteins involved in the structural organization of C. albicans were down regulated during morphogenesis after growth in the presence of the EAF. Cytoplasmic transport of proteins (Acb2) and cell division process (Nuf2 and Mug97) also showed decreased protein expression in the presence of the EAF of E. uniflora. Protein synthesis (Gtp1 and Cyp9) and folding (Gim4) were also negatively induced in the presence of the fraction (Table 4).

The effect of the EAF of E. uniflora in C. albicans hypha formation and pathogenicity was evaluated in vivo under three different conditions: Group I: C. albicans strain 111R grown in NGY broth the absence of the E. uniflora EAF. Group II C. albicans strain 111R grown in NGY broth the presence of E. uniflora EAF and Group III: C. albicans strain 111R grown in NGY broth in the absence of E. uniflora EAF, where the natural product was administered to the animal's tongue after infection.

In the macroscopic analysis of the tongue dorsum of mice infected with C. albicans grown in the absence of the E. uniflora EAF, the presence of lesions with pseudomembranous white plaques corresponding to establishment of oral candidiasis was observed (Groups I and III; Figure 2A) while the group of mice infected in the oral cavity with 111R strain grown overnight in the presence of the EAF of E. uniflora was macroscopically less affected with the infection and pseudomembranous white plaque in the dorsum of tongue was not observed (Group II; Figure 2B). Two days after the infection, the animals from group III were treated by administering the fraction on the surface of the tongue, reducing the progression of the formation of white pseudomembranous confluent plaques when compared with group I, which did not received treatment.

A sterile swab was rubbed in the oral cavity of animals and introduced into a tube containing 1 ml of 0.9% saline to perform direct examination and culture on SDA to analyze the fungal burden by CFU/ml counts. In the microscopical analysis of the sample collected from the mice belonging to Group I, we observed the presence of cellular clumps of long hyaline and septate hyphae (Figure 3A). The direct examination of samples obtained from mice belonging to Groups II and III showed visually shortest pseudo-hyphae and true-hyphae (Figure 3B). This difference was more pronounced for Group II samples.

Oral cavity samples were seeded on the surface of SDA, incubated at 37°C to perform CFU counting. The number of CFU/mL is described in the Figure 4. A reduction in the number of CFU from 2.36 Log₁₀ CFU/ml to 1.85 Log₁₀ (Groups I and II, respectively) and from 2.36 Log₁₀ CFU/ml to 1.92 Log₁₀ (Groups I and III, respectively) was observed. The histopathological analysis of the mice tongue infected with C. albicans was performed. The mucosa damage and fungal presence were evaluated with HE and PAS staining. The presence and absence of ulceration, erosion, inflammatory response and type of inflammatory cells were analyzed (Figures 5A–H). The animals belonging to Group I showed both more extensive tissue damage and fungal presence, specifically filamentous forms (long pseudo-hyphae and true hyphae), while the animals belonging to Groups II and III had less tissue damage and fungal presence (Figures 5A–H).

To ensure the safety and evaluate the cytotoxic effect of the EAFof E. uniflora in human cells, a MTT assay was performed using A549 cells, in concentrations ranging from 8,000 to 18 μg/mL. The extract was not toxic against the cells tested, which showed high viability over than 80% of cells even when the fraction concentration was 8-folds higher than the concentration used to perform all the experiments of the present study (Figure 6).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.