LC-MS Analysis and Antifungal Activity of Turnera subulata Sm.

Fungi of the Candida genus are responsible for invasive candidiasis, which affects people all over the world and has high mortality rates. This is due to their virulence factors, which give them great resistance and pathogenicity. In addition, the emergence of multidrug-resistant strains makes it difficult to treat these infections. In this way, natural products have emerged as an alternative to standard drugs, where plants known for their medicinal properties such as Turnera subulata become attractive to research. The present work aimed to analyze the ethanol extract of Turnera subulata leaves against standard strains of Candida albicans, Candida krusei and Candida tropicalis using broth microdilution techniques. The identification of the compounds in T. subulata leaves by LC-MS revealed the presence of a wide variety of substances such as carboxylic acids and terpenes, with flavonoids and fatty acids being more evident. The antifungal assays showed that the extract was not able to inhibit the growth of the tested strains at concentrations with a clinical relevance. However, at higher concentrations, it was able to inhibit the fungal dimorphism of C. albicans and C. tropicalis. It is possible that the T. subulata extract has potential as an inhibitor of fungal virulence factors without affecting the cell viability. Further research should be carried out in order to assess its inhibitory potential for other fungal virulence factors.


Introduction
Fungal infections caused by Candida, Cryptococcus and Aspergillus species kill thousands of people annually. This happens because these infections are difficult to treat and are often neglected. Although there are antifungal drugs that are widely used in medicine with relative effectiveness, mortality rates remain high as these microorganisms are capable of developing a resistance to this class of drugs [1,2].
The development of new drugs and diagnostic tools is important to avoid these problems and requires extensive knowledge about the biology of fungal pathogens, especially

Identification of Chemical Composition
The identification of the compounds in T. subulata leaves revealed the presence of a wide variety of substances such as carboxylic acids and terpenes, with flavonoids and fatty acids being more evident, as represented in the chromatogram ( Figure 1). Table 1 presents the chromatographic and mass spectral data such as the molecular ionic mass, retention time and fragmentation pattern for the compound identification.
Of the 18 peaks observed in the chromatograms, 12 were identified; among these, compounds 1, 2 and 3 exhibited [M-H]-at 191, 133 and 130 m/z, respectively. These were identified as quinic acid, malic acid and leucine by an authentic comparison of the pattern [22]. Compounds 6 and 7 showed a deprotonated [M-H]-ion at 593 m/z and 577 m/z, respectively. Compound 6 was identified as rhamnosyl isoorientin due to the presence of a fragment at 473 m/z corresponding with a loss of the C-hexose moiety. A fragment at 429 corresponded with a loss of the O-rhamnose moiety and a water molecule as a fragment ion at 327 m/z corresponded with the additional loss of the C-hexose moiety [23]. Compound 7, presenting fragments at 413 and 293 m/z that corresponded with C-glycosyl flavones and O glycosylated in the sugar portion, was identified as rhamnosyl vitexin [24]. Compound 9 presented a mass [M-H]-at 305 m/z; it was identified as a hexose derivative formed by the dehydration of disaccharides [25]. Compound 10, a deprotonated molecule with a mass [M-H]-at 187 m/z, showed a prominent fragmented ion in m/z due to the loss of the water portion and was identified as azelaic acid [26].
Compounds 11, 12 and 14 were identified as apigenin derivatives. Compound 11 produced a molecular anion in the apigenin fragment at 269 m/z, pointing to the presence of apigenin 7-O-neohesperidoside (rhoifolin). Compound 12 was identified as apigenin 7-O-rutinoside (isorhoifolin) due to the fragment ions at 431 (M-rhamnose) m/z and 269 (M-rutinose) m/z [27]. molecule with a mass [M-H]-at 187 m/z, showed a prominent fragmented ion in m/z due to the loss of the water portion and was identified as azelaic acid [26].
Compounds 11, 12 and 14 were identified as apigenin derivatives. Compound 11 produced a molecular anion in the apigenin fragment at 269 m/z, pointing to the presence of apigenin 7-O-neohesperidoside (rhoifolin). Compound 12 was identified as apigenin 7-O-rutinoside (isorhoifolin) due to the fragment ions at 431 (M-rhamnose) m/z and 269 (M-rutinose) m/z [27].    Table 2 shows the results of the intrinsic activity performed by the broth microdilution technique. The inhibition potential of the products tested against Candida compared with the standard drug fluconazole was identified. The results showed that the products were able to inhibit 50% of the microorganism population (IC 50 ) only at high concentrations and only for C. albicans and C. krusei; fluconazole alone was effective against the three strains tested, where C. albicans was inhibited by 16.70 µg/mL, C. tropicalis by 9.30 µg/mL and C. krusei by 133.32 µg/mL.  When cultivated in the growth medium added to the ethanolic extract of the leaves of Turnera subulata, C. krusei did not show any changes in its dimorphic potential at any of the concentrations tested. The presence of hyphae and pseudohyphae was recorded, as seen in Figure 4. For C. albicans and C. tropicalis, there was inhibition only at the highest HSA-8192 µ g/mL concentration. The images presented in Figure 3 show the growth control and the control of the effect of fluconazole on the fungal dimorphism. In the micromorphology reading, it could be observed that the properly depleted medium stressed the Candida strains, driving the morphological transition and causing the emission of hyphae and pseudohyphae.

Antifungal Tests
When cultivated in the growth medium added to the ethanolic extract of the leaves of Turnera subulata, C. krusei did not show any changes in its dimorphic potential at any of the concentrations tested. The presence of hyphae and pseudohyphae was recorded, as seen in Figure 4. For C. albicans and C. tropicalis, there was inhibition only at the highest HSA-8192 µg/mL concentration.

Discussion
Fungal infections affect people all over the world. In most of these infections, the isolated fungi are of the Candida genus, with C. albicans being the most common. However, many non-albicans Candida species are pathogenic. This pathogenicity is due to virulence factors such as the ability to develop biofilms, which gives them a great resistance. In addition, these fungi can develop a drug resistance [30,31]. Thus, natural products represent an alternative in the treatment of these infections, which can promote a reduction in the fungal virulence or even promote the action of drugs in combined therapies [32].
In a study carried out by Santos et al. [16] using an ethanol extract of T. subulata leaves, there was no clinically relevant antifungal activity against strains of C. albicans ATCC 40227, C. krusei ATCC 40147 and C. tropicalis ATCC 13803 when the MIC of the product was ≥ 1024 μg/mL. In addition, in a combined activity with drugs, the extract did not show any changes in the MIC when associated with amphotericin B and nystatin, but showed a potentiating effect of antifungal activity against C. tropicalis when associated with metronidazole.
In a study carried out by Morais [33], the crude extract as well as the hexane fractions and ethyl acetate from T. subulata also did not show a significant antifungal activity

Discussion
Fungal infections affect people all over the world. In most of these infections, the isolated fungi are of the Candida genus, with C. albicans being the most common. However, many non-albicans Candida species are pathogenic. This pathogenicity is due to virulence factors such as the ability to develop biofilms, which gives them a great resistance. In addition, these fungi can develop a drug resistance [30,31]. Thus, natural products represent an alternative in the treatment of these infections, which can promote a reduction in the fungal virulence or even promote the action of drugs in combined therapies [32].
In a study carried out by Santos et al. [16] using an ethanol extract of T. subulata leaves, there was no clinically relevant antifungal activity against strains of C. albicans ATCC 40227, C. krusei ATCC 40147 and C. tropicalis ATCC 13803 when the MIC of the product was ≥1024 µg/mL. In addition, in a combined activity with drugs, the extract did not show any changes in the MIC when associated with amphotericin B and nystatin, but showed a potentiating effect of antifungal activity against C. tropicalis when associated with metronidazole.
In a study carried out by Morais [33], the crude extract as well as the hexane fractions and ethyl acetate from T. subulata also did not show a significant antifungal activity against Similarly, the present work showed that EELTS did not show an antifungal activity against the three tested Candida strains, although flavonoids were the main component of the extract. It is known that flavonoids exhibit diverse biological activities, including antifungal activities [34]. These activities are extensively reported in the literature and, according to Jin et al. [35], flavonoids can act on the cell wall as well as biofilm formation and fungal dimorphism. This may elucidate the results obtained, where EELTS was able to inhibit the development of hyphae at the highest concentrations against C. albicans and C. tropicalis.
Fatty acids present in EELTS may also have contributed to the inhibition of fungal dimorphism in C. albicans and C. tropicalis. Studies show that fatty acids can reduce the virulence of these fungi such as biofilm formation, hyphae growth and cell aggregation [36]. The plasticity of the fungal cells of the genus Candida has been frequently associated with an increase in the virulence and, for this reason, there is great interest in researching the compounds capable of inhibiting these factors. It is more favorable to reduce the virulence of fungi without interfering with their cell viability, thus being able to prevent the development of resistance [3,5].
Interestingly, although the fluconazole activity was effective against C. krusei, it was still more resistant than C. albicans and C. tropicalis. According to Sampaio et al. [37], this difference in the antifungal activity of fluconazole between Candida species may be due to subtle changes between them, which give them resistance. In agreement, Arendrup and Patterson [38] reported that this resistance in Candida does not occur in the same way among their species. For C. albicans, the prolonged use of antifungals followed by recurrent infections such as chronic mucocutaneous candidiasis increases the chances of an acquired resistance. For several non-albicans Candida species such as C. krusei, there is less susceptibility to several classes of antifungals.

Plant Collection
Fresh leaves of Turnera subulata were collected from the Araripe National Forest (FLONA; Araripe Apodi) in a locality known as Barreiro Grande (07 • 21 S and 039 • 28 W), located in the municipality of Crato in the south of the State of Ceará (Brazil, Crato).

Preparation of Ethanol Extract
A total of 500 g of fresh leaves of Turnera subulata was crushed and then subjected to an exhaustive removal in 95% ethanol for 72 h. The extraction solution was subjected to solvent distillation on a rotary evaporator under a reduced pressure at an average temperature of 50 • C [39]. After distilling the solvent to dryness, the ethanolic extract of the fresh leaves of Turnera umifolia (EELTS) was obtained, with a percentage yield of 1.1%.

LC-MS Conditions
The analyses were performed using an Acquity UPLC (Waters, Milford, MA, USA) system coupled to a Xevo Quadrupole and Time-of-Flight mass system (QTOF, Water, Milford, MA, USA). A Waters Acquity BEH C18 column was used for the separation condition (150 mm × 2.1 mm; 1.7 µm) and set at 40 • C. An injection volume of a 5 µL aliquot of ethanolic extract was subjected to an exploratory gradient. The mobile phase was composed of deionized water (A) and acetonitrile (B) and both contained formic acid (0.1% v/v). The extracts were subjected to an exploratory gradient as follows: 2-95% B (15.

Growth Media
A Sabouraud Dextrose Agar (SDA) medium purchased from HIMEDIA ® was prepared according to the manufacturer's instructions. Sabouraud Dextrose Broth (CSD), purchased from KASVI ® and doubly concentrated, was used in the assays to evaluate the antifungal activity. For the analysis of fungal dimorphism, a potato dextrose agar (PDA) medium, purchased from Difco ® , was used. The growth media were solubilized with distilled water and sterilized in an autoclave at 121 • C for 15 min.

Inoculum Preparation
The strains were initially kept in test tubes containing SDA under refrigeration (8 • C). For the minimum inhibitory concentration (MIC) and minimum fungicide concentration (MFC) tests, the fungi were initially cultivated in Petri dishes containing SDA and incubated at 37 • C for 24 h (overnight). After this, the suspensions of microorganisms were prepared in tubes containing 4 mL of a sterile solution (0.9% NaCl). These suspensions were then shaken in a vortex mixer and the turbidity was compared and adjusted according to the 0.5 McFarland scale, which corresponded with an inoculum of approximately 10 5 colony-forming units (mL-CFU/mL) [40].

Drugs and Reagents
Dimethylsulfoxide (DMSO-Dynamic) was used to dilute the extract. The antifungal fluconazole at a dose of 150 mg (PRATI-Donaduzzi) was diluted in distilled water and used as a reference drug for the antifungal tests. In the preparation of the matrix solution of the extract, 0.15 g was weighed and then solubilized in 1 mL of DMSO. The extract and fluconazole were diluted again in sterile distilled water in order to obtain the desired concentration for the tests (16,384 µg/mL). The assay was performed using DMSO with a final concentration lower than 10% (the pilot assay performed in the lab indicated that DMSO concentrations lower than 10% did not affect the final results).

Intrinsic Activity of the Antifungal Effect of EELTS and Fluconazole
An antifungal test with EELTS and fluconazole was performed using the broth microdilution technique and 96-well polystyrene plates. A total of 100 µL of a double-concentrated SDB medium was added to each well, plus the fungal suspension (10%). Subsequently, 100 µL of the natural product at a concentration of 16,384 µg/mL was deposited in the first well, from which the serial microdilution was carried out until the penultimate well; the concentrations ranged from 8192 to 8 µg/mL. The last well was reserved for the growth control. Controls for the sterility of the medium and the dilution of the natural product and fluconazole were also performed [41].
The plates were incubated at 37 • C for 24 h. After this period, they were taken to be read in an ELISA spectrophotometer device (Termoplate ® ) with a wavelength of 630 nm [42]. The results provided the minimum inhibitory concentration (MIC) of the tested products as well as the IC 50 . The tests were performed in quadruplicate.

Determination of the Minimum Fungicide Concentration
With the aid of a sterile rod, the aliquots were transferred from each well of the MIC test plate, where the concentrations varied from 8192 to 8 µg/mL, to Petri dishes containing SDA. The plates were incubated at 37 • C for 24 h. After this period, the plates were checked for the growth of Candida colonies. The MFC was defined as the lowest concentration of the natural product capable of inhibiting the growth of fungal colonies [43] In the observation of the morphological alterations of the Candida strains against the EELTS extract, the technique of microculture for yeasts was used, using the depleted PDA medium in humid chambers. Intrinsic activity concentrations were considered, with concentrations of 8192 µg/mL being evaluated as HSA (a higher concentration assessed) as well as 2048 µg/mL (HSA/4) and 512 µg/mL (HSA/16). A total of 3 mL of the medium associated with the product tested was poured onto glass slides in a humid chamber. After the solidification of the medium, the yeast was seeded with the aid of a calibrated loop of 1 µL; two parallel streaks were then drawn, which were covered with sterile coverslips. The plates were incubated at 37 • C. After 24 h, the slides were observed under an optical microscope with a 40 × objective. A yeast growth control was performed as well as a fluconazole control for comparative purposes. The microcultures were photographed with an attached camera with a 5 x zoom. The tests were performed according to Sidrim and Rocha [44] and Mendes [45], with a few modifications.

Statistical Analysis
Quadruplicates were performed for each test and a two-way ANOVA analysis of variance with Fisher's test was applied to each sample. The IC 50 values were computed by a linear regression for the interpolation into standard curves relating to the percentage (%) of the growth values and the product concentration in µg/mL using GraphPad Prism software, version 5.0.

Conclusions
Although EELTS did not affect the cell viability of the three tested Candida strains, it was possible to observe, at high concentrations, an inhibitory effect on dimorphism in C. albicans and C. tropicalis. These results suggested that the product could act on these fungi directly on their virulence factors without affecting the cell viability. Further studies are needed to check its activity against other virulence factors such as biofilm formation. In addition, it is possible that this product could modulate the action of drugs, enhancing their activity.