In Vitro and In Vivo Anti-Candida spp. Activity of Plant-Derived Products

Candidiasis therapy, especially for candidiasis caused by Candida non-albicans species, is limited by the relatively reduced number of antifungal drugs and the emergence of antifungal tolerance. This study evaluates the anticandidal activity of 41 plant-derived products against Candida species, in both planktonic and biofilm cells. This study also evaluates the toxicity and the therapeutic action of the most active compounds by using the Caenorhabditis elegans–Candida model. The planktonic cells were cultured with various concentrations of the tested agents. The Cupressus sempervirens, Citrus limon, and Litsea cubeba essential oils as well as gallic acid were the most active anticandidal compounds. Candida cell re-growth after treatment with these agents for 48 h demonstrated that the L. cubeba essential oil and gallic acid displayed fungistatic activity, whereas the C. limon and C. sempervirens essential oils exhibited fungicidal activity. The C. sempervirens essential oil was not toxic and increased the survival of C. elegans worms infected with C. glabrata or C. orthopsilosis. All the plant-derived products assayed at 250 µg/mL affected C. krusei biofilms. The tested plant-derived products proved to be potential therapeutic agents against Candida, especially Candida non-albicans species, and should be considered when developing new anticandidal agents.


Introduction
Infections caused by yeasts belonging to the genus Candida affect especially immunocompromised individuals, children, elderly, individuals hospitalized in Intensive Care Units (ICU), and users of invasive devices [1]. Vulvovaginitis caused by Candida and Candida-associated stomatitis also represent important infections in the field of Public Health [2,3]. The Citrus limon EO provided MIC values of 250 µg/L against C. tropicalis and C. glabrata (Table 1). The Litsea cubeba EO yielded MIC values of 62.5, 250, and 250 µg/mL against C. krusei, C. glabrata, and C. orthopsilosis, respectively (Table 1). Finally, the Citrus reticulata EO afforded MIC values of 250 µg/mL against both C. krusei and C. orthopsilosis (Table 1). The first three products were tested against Candida in the following assays.
According to the adopted criteria for antimicrobial activity, none of the five plant extracts tested here (Table 2) were active against the evaluated Candida strains, their MIC values were higher than 2000 µg/mL. Among the seven plant-derived compounds tested herein (Table 3), gallic acid showed the greatest activity; its MIC values were 125, 31.25, 250, and 250 µg/mL against C. krusei, C. glabrata, C. parapsilosis, and C. orthopsilosis, respectively, so this acid was selected for further studies. Against the C. parapsilosis ATCC 22019 and the C. krusei ATCC 6258 reference strains, amphotericin B (AMB) gave MIC values of 0.25 and 1.00 µg/mL, respectively. The EOs were extracted from Cupressus sempervirens and Citrus limon leaves and Litsea cubeba fruits in 0.65%, 1.5%, and 1.0% yield, respectively. GC-MS and GC-FID analyses helped to identify 13, 9, and 11 chemical constituents in the EOs extracted from Cupressus sempervirens (total 99.1%) and Citrus limon (total 98.1%) leaves and Litsea cubeba fruits (total of 96.2%), respectively. Sabinene, terpinen-4-ol, citral, limonene, neral, and geraniol were the major compounds in these EOs. Table 4 lists all the identified compounds, retention indexes, and relative area percentages (% RA). The Litsea cubeba EO was tested against C. krusei, and gallic acid was assayed against C. glabrata and C. krusei. The tested agents were fungistatic at all the concentrations ( Figure 1A-C). The Citrus limon EO at 0.5 × MIC (125 µg/mL) or 1 × MIC (250 µg/mL) exerted fungicidal action against C. tropicalis ( Figure 1D) after 4 h. The Citrus limon EO at 2 × MIC (500 µg/mL) had a fungicidal effect on C. tropicalis ( Figure 1D) and C. glabrata ( Figure 1E) after 2 h. A fungicidal effect emerged after exposure of C. orthopsilosis ( Figure 1F) to the C. sempervirens EO at 0.5 × MIC (15.6 µg/mL), 1 × MIC (31.25 µg/mL), and 2 × MIC (62.5 µg/mL) for 8, 6, and 4 h, respectively. This same EO at 0.5 × MIC exhibited fungistatic activity against C. glabrata after 8 h. This EO, at 1 × MIC or 2 × MIC, displayed a fungicidal effect against C. glabrata after 12 h ( Figure 1G). Table 5 depicts the minimal biofilm-inhibiting concentration (MBIC) and the minimal biofilm-eradicating concentration (MBEC) obtained with the Litsea cubeba, Citrus limon, and Cupressus sempervirens EOs and gallic acid. The Cupressus sempervirens EO gave the best antibiofilm activity; the MBIC and MBEC values ranged between 62.5 and 1000 µg/mL against all the Candida species (Table 5). The lowest MBIC and MBEC values were achieved against Candida krusei at 62.5 and 250 µg/mL, respectively (Table 5).  The antifungal activity findings obtained in the in vitro assays were confirmed by using an in vivo infection model, namely the Caenorhabditis elegans-Candida infection assay, which is regarded as an infection model to study Candida-associated infections. Initially, the toxicity of the selected plant-derived products was evaluated by testing approximately 15-20 late-L4 larvae in each well of a 96-well microplate, exposed to concentrations of the selected products of 0.5 × MIC, 1 × MIC, and 2 × MIC at 25 • C for 24 h. The Litsea cubeba and Cupressus sempervirens EOs at concentrations between 31.25 and 125 µg/mL as well as gallic acid at concentrations between 15.62 and 250 µg/mL were not toxic (p > 0.05) against C. elegans as compared to untreated larvae (data not shown). In turn, the Citrus limon EO at 0.5 × MIC (125 µg/mL) was not toxic (p > 0.05), but this same EO at 1 × MIC (250 µg/mL) and 2 × MIC (500 µg/mL) was significantly toxic (p < 0.05 and p < 0.0001, respectively) (data not shown).
The C. glabrata-infected larvae were treated with the Cupressus sempervirens ( Figure 2A) and Citrus limon ( Figure 2B) EOs and gallic acid ( Figure 2C) at 25 • C for four days. Only in the presence of Cupressus sempervirens was a higher frequency of viable larvae maintained ( Figure 2A). In contrast, the larvae infected with C. krusei, and treated with the Litsea cubeba EO ( Figure 2D) at concentrations between 31.25 and 125 µg/mL or gallic acid ( Figure 2E) at concentrations between 62.5 and 250 µg/mL, were not cured of candidiasis.
The Citrus limon EO at concentrations between 125 and 500 µg/mL was used to treat larvae infected with C. tropicalis ( Figure 2F). After 24 h, a small percentage of dead larvae (5-10%) was detected. However, after 48 h, 35% and 55% of the larvae died at EO concentrations of 125 µg/mL and 250-500 µg/mL, respectively. At the end of four days, only 40% and 10-15% of the larvae survived, respectively, at the same concentrations.
Lastly, worms infected with C. orthopsilosis were treated with the Cupressus sempervirens EO ( Figure 2G). Exposure to this EO (15.62 to 62.5 µg/mL) increased the survival of C. elegans worms infected with C. orthopsilosis as compared to the treated control. At four days postinfection, 80-85% of the infected and treated larvae survived.

Discussion
In Brazil, plant-derived products have gained importance because of the publication of Resolution 971 (3 May 2006) [13] and Act 5813 (22 June 2006) [14], which regulate the National Policy on Integrative and Complementary Practices and the National Policy on Medicinal and Phytotherapeutic Plants, respectively. These regulations introduced the use of medicinal plants and phytotherapeutic drugs into the Unified Health System (SUS) and aimed to ensure safe access of the Brazilian population to these medications as well as their rational application, promoting the sustainable use of the national biodiversity.
In this scenario, this study evaluated the antifungal potential of plant-derived products (essential oils, Brazilian native plant extracts, and plant constituents) that have been employed as antimicrobials in folk medicine. Among the EOs assessed herein, the Cupressus sempervirens, Citrus limon, and Litsea cubeba EOs are noteworthy. According to literature data, the EO extracted from Cupressus sempervirens leaves exhibits similar chemical composition to the composition identified here, albeit with different percentages. For instance, Selim et al. [15] reported that α-pinene (48.6%), δ-3-carene (22.1%), limonene (4.6%), and α-terpinolene (4.5%) are the main constituents of Cupressus sempervirens studied in Saudi Arabia, whereas Ibrahim et al. [16] described α-pinene (21.15%), terpinen-4-ol (6.98%), allo-ocimene (24.00%), and α-cedrol (23.68%) as the main components of Egyptian Cupressus sempervirens. As for Cupressus sempervirens growing in Brazil, we determined some of these substances at lower concentrations as well as the compounds sabinene (20.3%) and citral (20.0%), which were reported at higher concentrations in the EO extracted from Cupressus sempervirens leaves for the first time ( Table 4).
The Cupressus sempervirens anticandidal activity against C. albicans has been demonstrated, but it has not been defined as fungistatic or fungicidal [17,18]. Here, we showed the greater susceptibility of C. glabrata and C. orthopsilosis to the Cupressus sempervirens EO (Table 1). These Candida species have been cited as causing a significant increase in Candida infections in the last few years [19]. Additionally, C. glabrata is the NAC species that has been the most commonly isolated from the environment and by health practitioners in a Brazilian Tertiary Hospital [20], and C. orthopsilosis has been identified as the prevalent organism among yeasts isolated from the hydraulic system of a hemodialysis facility [21].
Exposure to the Cupressus sempervirens EO completely inhibited C. orthopsilosis ( Figure 1F) and C. glabrata ( Figure 1G) cells. The fungicidal action of this EO could be partly attributed to its major constituents such as sabinene, citral, and terpinen-4-ol, which have been reported to display antimicrobial effects [22][23][24]. Moreover, at 2 x MIC, this EO provided the same effect as 4 µg/mL AMB against C. orthopsilosis ( Figure 1F), which is the best antifungal drug concentration with fungicidal action that has been described in in vivo studies [25].
The C. elegans larvae infected with C. glabrata (Figure 2A) or C. orthopsilosis ( Figure 2G) and treated with the Cupressus sempervirens EO at concentrations between 15.62 and 62.5 µg/mL had significantly (p < 0.05) higher survival as compared to the infected larvae control four days postinfection. This suggested that this EO might be a valuable antifungal agent against Candida infections. Besides that, this EO showed antibiofilm activity (MBIC or MBEC) against all the tested Candida non-albicans species (Table 5), which pointed out that it could be an adjuvant in the treatment of biofilm-associated Candida-non-albicans infections.
The Citrus limon EO exhibited anticandidal activity (MIC) against all the assayed strains (Table 1). At concentrations between 125 and 500 µg/mL and between 250 and 500 µg/mL, this EO displayed a fungicidal effect against C. tropicalis ATCC 13803 ( Figure 1D) and C. glabrata ATCC 2001 ( Figure 1E), respectively. The antifungal mechanism of the Citrus limon EO is associated with its main component, limonene [26], which was detected at a similar concentration to the concentration reported by Campelo et al. [27]. Limonene damages the C. albicans cell wall/membrane, thereby modifying cellular adhesion and plasticity, pH, and ionic content [26]. In addition, such damage causes oxidative stress and consequent DNA damage, resulting in cell cycle modulation and apoptosis as demonstrated by Thakre et al. [28]. The high concentration of limonene in the EO could contribute to its nematocidal effect [29].
The Litsea cubeba EO is known to possess diverse biological properties, among which the antimicrobial action is worthy of note [30]. Here, this EO afforded the best activity against C. krusei ATCC 6258, with MIC and MBIC values of 62.5 and 250 µg/mL, respectively. Its anti-Candida effect can be justified by the presence of chemical constituents such as limonene, citral, neral, terpinen-4-ol, and geraniol (Table 4), which have previously been described to present anti-Candida activity [26,31]. Moreover, in agreement with a previous study [32], we confirmed the Litsea cubeba EO nematocidal activity.
Gallic acid had great inhibitory action against C. glabrata and C. krusei (Table 1), but better antibiofilm activity against C. krusei (Table 5). Previous studies have shown antifungal activity for gallic acid against C. albicans and filamentous fungi [33][34][35][36]; however, its activity against Candida biofilms has been poorly investigated. This compound can inhibit biofilm formation [34], as confirmed by the C. krusei antibiofilm result (MBIC) recorded here. Gallic acid toxicity to C. elegans larvae has been reported at concentrations starting from 120 µg/mL [36].
The fungistatic and fungicidal actions of the EOs make them promising alternatives to treat superficial candidiasis; that is, to treat oral candidiasis and denture stomatitis by topical administration, since they can be included in mouth rinses or toothpastes [37]. Interestingly, we detected that all the assayed plant-derived products at 250 µg/mL had an effect on C. krusei biofilms ( Table 5), suggesting that these products might be valuable antifungal agents in the therapy against C. krusei biofilm-associated infections. This organism is an important pathogenic Candida species that is frequently refractory to conventional antimicrobial agents and has been isolated from patients with oral candidiasis [38]. Further studies using an in vivo biofilm-associated animal model (e.g., a rat model of acute dermal toxicity) are necessary to confirm that the EOs might be useful to treat candidal biofilm-associated infections, especially topical infections.

Essential Oils
Citrus reticulata (peel), Citrus reticulata Blanco (peel and fresh and dry leaves), Citrus reticulata var. Briefly, distilled water (500 mL) was added to the plant material (100 g) and transferred to a Clevenger-type apparatus. The essential oil (EO) was collected, and the remaining water was eliminated with anhydrous sodium sulfate, which was followed by filtration. This method was used in triplicate, and the obtained EOs were kept under refrigeration (4 • C). The mean quantities of the EOs (w/w) were obtained on the basis of the plant material weight and data from three experiments.

Plant Extracts
Anacardium occidentale L. and Anacardium othonianum cashew nuts were obtained from a local market in Franca (Oct/2013) and from Montes Claros de Goias (Mar/2017). Voucher specimens (SPFR 16040 and HJ3793) were deposited in the Biology Department Herbarium in the Plant Systematics Laboratory of the Faculty of Philosophy, Sciences and Letters of Ribeirão Preto, University of São Paulo, Brazil (Herbarium SPFR) and in the Herbarium Jataiense Germano Guarim Neto of the Goiás Federal University, Brazil (Herbarium HJ). The air-dried powdered A. occidentale and A. othonianum nuts were extracted with ethanol.
Vochysia divergens was collected in the Pantanal area, in the State of Mato Grosso (16 • 35 22" S and 56 • 47 83" W) in January 2017. A voucher specimen UFMT 39559 was deposited in the Herbarium of the Mato Grosso Federal University (UFMT), Brazil (Herbarium UFMT). The V. divergens stem barks were powdered and exhaustively extracted by maceration at room temperature; ethanol was employed. After filtration, the solvent was removed under reduced pressure to yield the ethanolic extract.
The Curcuma longa L. extracts were provided by Dr. Marco Túlio Menezes Carvalho, from the State University of Minas Gerais, MG, Brazil. The dried rhizomes were obtained from a local market (September 2017) in Passos, MG (20 • 43 13 ' S and 46 • 36 36 ' W), and the powdered material was stored in the dark. The extracts were obtained by maceration of the powdered curcuma rhizomes, at room temperature; boiling water or ethanol/water (50:50, v/v) was employed.

Candida Species
Reference strains of six Candida species, including C. albicans SC 5314, C. glabrata ATCC 2001, C. parapsilosis ATCC 22019, C. krusei ATCC 6258, C. tropicalis ATCC 13803, and C. orthopsilosis ATCC 96141 were used in this study. The strains were maintained at −70 • C in sterile distilled water plus 50% glycerol and subcultured in Sabouraud dextrose agar (SDA, Difco, Detroit, MI) and CHROMagar Candida medium (Becton Dickinson and Company, Sparks, MD) at 37 • C for 24 h to ensure purity and viability.

Minimum Inhibitory Concentration Determination
The in vitro antifungal susceptibility assays of all the natural products were performed by the broth microdilution method according to the adapted protocol M27-S4 from the Clinical and Laboratory Standards Institute [39]. Sterile microtiter plates (Corning Inc., NY, USA) were used. The inoculum size was 2.5 × 10 3 cells/mL. The final concentration of amphotericin B (AMB) and the tested products ranged from 0.03 to 16 µg/mL and from 3.90 to 2.000 µg/mL, respectively. AMB and all the natural products were solubilized in DMSO (2%) and diluted in Roswell Park Memorial Institute (RPMI 1640, Sigma-Aldrich, St. Louis, MO, USA) medium added with 0.2% glucose. The C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 strains and AMB were included as quality control [39]. The minimum inhibitory concentration (MIC) was determined with the fluorometric indicator resazurin at 0.01% (w/v) [40]. MIC was defined as the lowest antifungal/product concentration that maintained a blue hue. The AMB breakpoint adopted herein was ≤1 µg/mL, whereas AMB >1 µg/mL was considered as resistant [41]. The wells where micro-organism growth occurred displayed the pink color. Plant-derived products were considered active when MIC was <100 µg/mL, moderately active when MIC ranged from 100 up to 500 µg/mL, and weakly active when MIC was >500 and <1000 µg/mL. Above 1000 µg/mL, the products were considered inactive [42]. All the tests were conducted in triplicate.

Identification of the Chemical Composition of the EOs
Gas chromatography-flame ionization detection and gas chromatography-mass spectrometry analyses were accomplished with Shimadzu QP2010 Plus and GCMS2010 Plus (Shimadzu Corporation, Kyoto, Japan) systems. The GC-MS and GC-FID conditions and the identification of the chemical constituents of the EOs were carried out in agreement with the methodology proposed by Santos et al. [43].

Time-Kill Curves
All the compounds and natural products were evaluated at 0.5 × MIC, 1 × MIC, and 2 × MIC for each more susceptible Candida strain, at predetermined incubation time points (0, 2,4,6,8,12,24, and 48 h). AMB was used at 4 µg/mL [25]. The Candida strains were subcultured on Sabouraud Dextrose Agar (SDA) plates at 35 • C for 24 h, suspended in 5 mL of RPMI 1640, and adjusted to a 0.5 McFarland turbidity standard (1 to 5 × 10 6 cells/mL) with a nephelometer (ATB 1550, BioMérieux, France). Next, the micro-organism suspension was diluted in RPMI 1640, to give a new suspension containing 2.5-5.0 × 10 4 cells/mL, and the plant product was added at an appropriate concentration. The latter suspension was incubated at 35 • C for 48 h, and 100 µL aliquots were removed at each time point. Tenfold serial dilutions were performed, and 10 µL aliquots were plated on SDA plates and incubated at 35 • C for 24 h. The mean colony-forming unit (CFU) value was converted to the respective log CFU/mL value. The data were plotted as log CFU/mL against time point. A fungicidal activity was considered to exist when micro-organism growth decreased by at least 3 log 10 CFU/mL as compared to the initial inoculum, to result in a reduction of 99.99% of CFU/mL. In turn, a fungistatic activity was considered to exist when micro-organism growth was less than 99.9% or <3 log 10 in CFU/mL as compared to the initial inoculum [22,44]. The experiments were conducted in triplicate.

Evaluation of the Effects on Biofilm
Two groups of tests were completed to evaluate the activity of the tested plant products against Candida biofilms: i) inhibition of biofilm formation was assessed to determine the minimal biofilm-inhibiting concentration (MBIC) and ii) effect on the preformed biofilm was analyzed to determine the minimal biofilm-eradicating concentration (MBEC). Three EOs, namely the Litsea cubeba, Citrus limon, and Cupressus sempervirens EOs, and one plant-derived compound (gallic acid) were selected after MIC determination. A sterile 96-well flat-bottom microtiter plate (Corning) was used. Inhibition/eradication of biofilm formation by the Candida strains was assayed according to a previously described methodology [45]. The EOs and gallic acid (concentration range from 1.95 to 2000 µg/mL) were dissolved in DMSO and two-fold diluted in RPMI 1640 medium at 35 • C for 48 h. Biofilm viability was measured by the tetrazolium salt (sodium 3 -[1-(phenylaminocarbonyl)-3,4-tetrazolium]-bis(4-methoxy-6-nitro) hydrated benzene sulfonic acid, XTT) reduction assay, by adding 100 µL of XTT-menadione to each well. Wells containing culture medium/biofilms/XTT/menadione (positive control) and wells containing culture medium/XTT/menadione (negative control) were included. The optical density (OD) was read on a microtiter plate reader (Asys -Eugendorf, Salzburg, Austria) at a wavelength of 492 nm [45]. The effective concentration of the EO/chemical compound capable of reducing ≥90% OD as compared to the control free of a chemical substance (100% of survivors) was considered as the MBIC or MBEC. Each experimental condition was tested in triplicate, and the arithmetic mean of the results was used to present the results.

In Vivo Toxicity to Caenorhabditis elegans
Acute toxicity was measured after exposure of Caenorhabditis elegans to the tested plant product for 24 h. The AU37 [glp-4(bn2) I; sek-1(km4) X] mutant strain was kindly provided by the São Paulo State University, Dr. Júlio de Mesquita Filho, Instituto de Ciência e Tecnologia, Departamento de Biociências e Diagnóstico Bucal, São José dos Campos, SP, Brazil. Nematodes grew on nematode growth medium (NGM) agar plates, seeded with Escherichia coli OP50 and incubated at 16 • C. They were synchronized by treatment with sodium hypochlorite, transferred, and incubated in NGM without E. coli OP50. The worms were then washed with NaCl 50 mM. Around 20 worms were added to the wells of 96-well microplates containing culture broth (60% 50mM NaCl, 40% BHI (brain heart infusion broth), cholesterol (10 µg/mL), kanamycin (90 µg/mL), and ampicillin (200 µg/mL)). Selected plant-derived products were added to each well at 0.5 × MIC, 1 × MIC, or 2 × MIC; AMB was added at 1.0 µg/mL. The 96-well microplates were maintained at 25 • C, and individual worm survival was assessed after 24 h, Nematodes were considered dead when they were rod-shaped and did not respond to touching [46,47]. Two independent experiments were carried out for each treatment.

Infection Assay of Caenorhabditis elegans-Candida Species
C. elegans AU37 fed with E. coli OP50 were maintained at 16 • C and synchronized as described above. Candida species were grown on BHI-agar, and worms in stage L4 were added to the plates containing each Candida species. Next, the plates with Candida and worms were incubated at 25 • C for 2 h. Then, the worms were washed with 50 mM NaCl, and C. elegans suspension was adjusted to contain 15-20 L4 larvae per well in a 96-well microplate [46]. In each well, the plant product was added at 0.5 × MIC, 1 × MIC, or 2 × MIC concentrations. The plates were incubated at 25 • C for four days. Worms not treated with a plant-derived product and worms infected with Candida species served as controls. Worm survival was expressed as a percentage of worm viability at day zero. Three independent experiments with at least three replicates were performed.

Statistical Analysis
Statistical analyses were done with the Program GraphPad Prism 5.0 (GraphPad Software Inc., San Diego, CA, USA). The survival curve of C. elegans was plotted by using the Kaplan-Meier method, and the survival differences were analyzed by log-rank (Mantel-Cox). A value of p < 0.05 was considered statistically significant.