Antifungal Potential of Marine Organisms of the Yucatan Peninsula (Mexico) against Medically Important Candida spp.

Invasive fungal infections represent a global health threat. They are associated with high mortality and morbidity rates, partly due to the ineffectiveness of the available antifungal agents. The rampant increase in infections recalcitrant to the current antifungals has worsened this scenario and made the discovery of new and more effective antifungals a pressing health issue. In this study, 65 extracts from marine organisms of the Yucatan Peninsula, Mexico, were screened for antifungal activity against Candida albicans and Candida glabrata, two of the most prevalent fungal species that cause nosocomial invasive fungal infections worldwide. A total of 51 sponges, 13 ascidians and 1 gorgonian were collected from the coral reef and mangrove forest in the Yucatan Peninsula (Mexico) and extracted with organic solvents. Nine crude extracts showed potent antifungal activity, of which four extracts from the sponge species Aiolochroia crassa, Amphimedon compressa, Monanchora arbuscula and Agelas citrina had promising activity against Candida spp. Bioassay-guided fractionation of the M. arbuscula extract revealed the remarkable fungicidal activity of some fractions. Analysis of the chemical composition of one of the most active fractions by UHPLC-HRMS and NMR indicated the presence of mirabilin B and penaresidin B, and their contribution to the observed antifungal activity is discussed. Overall, this work highlights marine organisms of the Yucatan Peninsula as important reservoirs of natural products with promising fungicidal activity, which may greatly advance the treatment of invasive fungal infections, especially those afflicting immunosuppressed patients.


Introduction
Fungi are important components of most ecosystems on Earth [1]. They are also part of the human microbiota and, under particular circumstances, can cause healththreatening invasive infections, in which the fungus reaches the bloodstream or any major internal organs [2]. The yeast Candida spp. asymptomatically colonizes the skin, mucosal surfaces and the gastrointestinal tract of most healthy individuals [3]. However, some aggressive drug therapies or immunosuppressive infections may promote their transition from commensals to pathogens [4]. As a result, invasive fungal infections caused by commensal Candida spp. are the most prevalent severe fungal infections among hospitalized patients [4,5]. The mortality rates of patients with invasive candidiasis are unacceptably The large size and biological diversity of the oceans make them promising natural sources of bioactive molecules [13]. In the marine environment, organisms subjected to a plethora of stimuli produce molecules with unique structural, chemical, and biological characteristics. These marine natural products (MNPs) are a rich source of potential alternative drugs [14], many of which have already entered clinical phase trials [13].
Mexico has been considered as one of the three areas in the world with the greatest terrestrial and marine biodiversity. In particular, the Gulf of Mexico and the Caribbean Sea, which meet in the Yucatan channel, constitute two outstanding marine ecosystems [15]. The single geographical location of the Yucatan channel promotes the abundance of The large size and biological diversity of the oceans make them promising natural sources of bioactive molecules [13]. In the marine environment, organisms subjected to a plethora of stimuli produce molecules with unique structural, chemical, and biological characteristics. These marine natural products (MNPs) are a rich source of potential alternative drugs [14], many of which have already entered clinical phase trials [13].
Mexico has been considered as one of the three areas in the world with the greatest terrestrial and marine biodiversity. In particular, the Gulf of Mexico and the Caribbean Sea, which meet in the Yucatan channel, constitute two outstanding marine ecosystems [15]. The single geographical location of the Yucatan channel promotes the abundance of highly diverse and unique marine species, which represent a potential source of bioactive compounds [15][16][17][18]. The antifungal potential of the marine organisms of the Yucatan Peninsula (YP) has not been intensively investigated [15]. The macroalgae are the only group of marine organisms in the YP that have been searched for antifungal activity. Indeed, as far as we know, there are only two reports on this topic. The first, by Morales et al. [19], evaluated the antifungal activity present in the extracts of marine macroalgae against Trichophyton mentagrophytes. In the second, the antifungal activity of marine macroalgae extracts on Pseudocercospora fijiensis, Colletotrichum gloeosporioides and Fusarium oxysporum was tested [20].
In this work, we report the promising antifungal activity of 65 extracts from several marine invertebrate species of the YP against yeast species that cause life-threatening infections.

Screening of a Library of Marine Extracts from the Yucatan Peninsula for Antifungal Activity
A library of 65 extracts (51 sponges, 13 ascidians and 1 gorgonian), collected from the coral reef and mangrove forest in the Yucatan Peninsula in Mexico, was screened for antifungal activity against C. albicans and C. glabrata. The crude extracts were resuspended in the smallest possible volume of DMSO, yielding the stock concentrations listed in Table 1. To ensure total solubilization, all the extracts were sonicated prior to use. Next, 5 µL of each extract (Table 1) was added to the wells of a 96-well plate, containing a cellular suspension of C. glabrata or C. albicans. Growth was recorded after 48 h at 30 • C (C. albicans) or 37 • C (C. glabrata), by measuring OD 600 . Growth ratios were determined in comparison to the control cells and those below 0.5 were considered to be active extracts. For the concentrations tested (Table 1), 9 of the 65 extracts, obtained from 8 sponge species, showed antifungal activity against C. glabrata ( Figure 2). The active extracts were from Aiolochroia crassa (collected from two different locations: Mahahual in the Quintana Roo state (MA18-4) and Alacranes Reef in the Yucatan state (E50)), Amphimedon compressa, Monanchora arbuscula, Leucetta floridana, Agelas sceptrum, A. citrina, A. dilatata, and Haliclona (Rhizoniera) curacaoensis. Four of these extracts (A. crassa (MA18-4), A. compressa (E29), M. arbuscula (E35) and A. citrina (CZE56)) also had antifungal activity against C. albicans ( Figure 3). To the best of our knowledge, there are no previous studies on the antifungal activity of the following three sponges: H. (Rhizoniera) curacaoensis, A. crassa, and A. dilatata.
The antifungal activity displayed by the A. citrina extract may be due to the presence of agelasidines (Table 10, Figure 10). It has been reported that the alkaloid (-)-agelasidine C shows strong antifungal activity on C. albicans [21]. Moreover, (-)-agelasidine C and agelasidines E and F, isolated from A. citrina, also showed activity against C. albicans [21,22].
As for A. compressa, methanol extracts obtained from this species have already demonstrated antifungal activity against C. albicans [23]. Accordingly, 8,8 -dienecyclostellettamine (Table 1) isolated from this species is active against C. glabrata and C. albicans [24].
L. floridana-derived extracts are known to display antifungal activity on C. albicans [26]. In this study, we found that L. floridana crude extracts are very active against C. glabrata ( Figure 2).
Batzelladine L, batzelladine D, norbatzelladine L, and ptilomycalin A were isolated from M. arbuscula with high antifungal activity against clinically important fungi. Batzelladin L is active against the filamentous fungus Aspergillus flavus [27], batzelladines D and norbatzelladine L are active against Saccharomyces cerevisiae, a yeast that is phylogeneti-cally close to C. glabrata [28], and the alkaloid ptilomycalin A is active against the yeast Cryptococcus neoformans [29].
We determined the minimum inhibitory concentration (MIC) of the four most active extracts against both Candida spp. ( Table 2). The extracts were serial diluted in DMSO and 5 µL was added to the wells of a 96-well plate, containing cellular suspensions of C. glabrata or C. albicans. Growth was recorded after 24 and 48 h at 30 • C (C. albicans) or 37 • C (C. glabrata), by measuring OD 600 . The MIC was defined as the drug concentration where the relative OD 600 fell at least 50% below the control (DMSO alone). M. arbuscula (E35) was the most active extract against both Candida spp. with an MIC of 3.91 µg/mL ( Table 2). The extract of A. compressa was more active against C. glabrata than against C. albicans and the opposite was observed for the extract of A. citrina. The A. crassa extract was the least active against both species. The MIC for A. compressa, M. arbuscula, and A. citrina did not change between 24 and 48 h.
The potent activity of the crude extract from M. arbuscula (E35) on C. albicans and C. glabrata led us to further explore this extract.

Bioassay-Guided Fractionation of the M. arbuscula Extract
The M. arbuscula crude extract was partitioned using the modified Kupchan procedure to obtain the following five fractions: E35-WF, E35-BF, E35-HF, E35-WMF, and E35-DF ( Table 3). The three most active fractions against both C. glabrata and C. albicans were E35-DF, E35-BF, and E35-WMF (Table 3, Figures 4 and 5), with E35-DF being the most active fraction. The fraction E35-DF, which required the slightest amount to produce a significant impact on yeast growth (Table 3), was then subjected to solid phase extraction (SPE) using an RP-18 cartridge. The procedure generated seven sub-fractions, R1-R7 (Table 4), whose fungistatic and fungicidal activity were evaluated on C. glabrata and C. albicans.   The fraction E35-DF, which required the slightest amount to produce a sign impact on yeast growth (Table 3), was then subjected to solid phase extraction (SPE) an RP-18 cartridge. The procedure generated seven sub-fractions, R1-R7 (Table 4), fungistatic and fungicidal activity were evaluated on C. glabrata and C. albicans.
We found that all the sub-fractions were active against C. glabrata and C. albican the sub-fractions R2, R3, R4 and R5 exceling in terms of antifungal efficacy at the co trations assayed ( Figure 6 and Table 5).    The fraction E35-DF, which required the slightest amount to produce a sign impact on yeast growth (Table 3), was then subjected to solid phase extraction (SPE) an RP-18 cartridge. The procedure generated seven sub-fractions, R1-R7 (Table 4), w fungistatic and fungicidal activity were evaluated on C. glabrata and C. albicans.
We found that all the sub-fractions were active against C. glabrata and C. albicans the sub-fractions R2, R3, R4 and R5 exceling in terms of antifungal efficacy at the co trations assayed ( Figure 6 and Table 5).   We found that all the sub-fractions were active against C. glabrata and C. albicans, with the sub-fractions R2, R3, R4 and R5 exceling in terms of antifungal efficacy at the concentrations assayed ( Figure 6 and Table 5).  Table 4. Inoculum: growth prior to sub-fraction addition; untreated control cells.
The MICs of the most active sub-fractions (R2 to R5) were determined next ( Table 5). With the exception of sub-fraction R4, all the other sub-fractions had lower MICs than the original E35-FD fraction (Table 5), confirming the success of the fractionation step. In addition to C. albicans and C. glabrata, other Candida spp. are emerging as important pathogens. Among them are C. krusei, C. tropicalis, and C. parapsilosis, which together with the former are responsible for more than 90 percent of all yeast infections [30]. Therefore, the most active fractions R2-R5 were also tested against C. krusei, C. tropicalis, and C. parapsilosis (Table 6). All of the sub-fractions showed activity against C. krusei, C. tropicalis, and C. parapsilosis, with sub-fraction R4 being the most active one.
By determining the minimum fungicidal concentration (MFC), we also found that sub-fractions R4 and R5 generally had the highest fungicidal activity (lower MFCs) against all species at 24 and 48 h (at concentrations that were two to eight times higher than the MIC). The exception was C. parapsilosis, for which the sub-fractions R2 and R3 had the strongest activity at 48 h (Table 7).  Table 4. Inoculum: growth prior to sub-fraction addition; untreated control cells. The MICs of the most active sub-fractions (R2 to R5) were determined next (Table 5). With the exception of sub-fraction R4, all the other sub-fractions had lower MICs than the original E35-FD fraction (Table 5), confirming the success of the fractionation step.
In addition to C. albicans and C. glabrata, other Candida spp. are emerging as important pathogens. Among them are C. krusei, C. tropicalis, and C. parapsilosis, which together with the former are responsible for more than 90 percent of all yeast infections [30]. Therefore, the most active fractions R2-R5 were also tested against C. krusei, C. tropicalis, and C. parapsilosis (Table 6). All of the sub-fractions showed activity against C. krusei, C. tropicalis, and C. parapsilosis, with sub-fraction R4 being the most active one.
By determining the minimum fungicidal concentration (MFC), we also found that sub-fractions R4 and R5 generally had the highest fungicidal activity (lower MFCs) against all species at 24 and 48 h (at concentrations that were two to eight times higher than the MIC). The exception was C. parapsilosis, for which the sub-fractions R2 and R3 had the strongest activity at 48 h (Table 7).   To confirm the presence of mirabilin B (1) and penaresidin B (2) (Figure 8) in the R4 sub-fraction, the carbon chemical shift signals of the 13 C NMR spectrum of this sub-fraction (see Supplementary Materials, Figures S4 and S6) were compared to those reported for these compounds in the literature, using a Pearson's chi-squared goodness of fit test ( ) with Yates continuity correction ( Table 9). The 13 C NMR spectrum of the R4 sub-fraction indicated the presence of the main carbon chemical shifts of compounds 1 and 2. A total of 19 [M + H] + ion adducts that corresponded to 19 UHPLC signals were detected, including 3 for R2, 2 for R3, 7 for R4, and 7 for R5 (Table 8).    To confirm the presence of mirabilin B (1) and penaresidin B (2) (Figure 8) in the R4 sub-fraction, the carbon chemical shift signals of the 13 C NMR spectrum of this sub-fraction (see Supplementary Materials, Figures S4 and S6) were compared to those reported for these compounds in the literature, using a Pearson's chi-squared goodness of fit test (χ 2 ) with Yates continuity correction ( Table 9). The 13 C NMR spectrum of the R4 sub-fraction indicated the presence of the main carbon chemical shifts of compounds 1 and 2. The chi-squared goodness of fit test revealed that the experimental values for compound (1) did not differ significantly, at a 99% confidence level, from the reported values (χ 2 Y = 0.0460; p-value = 0.8302). The same was observed for compound (2) (χ 2 Y = 0.0405; p-value = 0.8405). Thus, we conclude that the experimental data do not differ significantly from those expected for both compounds.
Mirabilin B was identified in M. arbuscula (previously known as Monanchora unguifera) and demonstrated activity against Cryptococcus neoformans [34], but there are no data on its activity against Candida spp. Penaresidin B, isolated from the marine sponge Penares sp., has no antifungal activity against C. neoformans, Aspergillus niger, or C. albicans [35]. Although at this stage, we cannot rule out the contribution of the other compounds (Table 8), it may well be that the antifungal activity observed in R4 results from mirabilin B or from its synergetic interaction with those unidentified NPs.

General Experimental Procedures
The separation was performed using a Waters XBridge column C18, 2.1 × 150 mm, 3.  13 C NMR spectra were recorded on a Bruker Avance 500 spectrometer at 125 MHz, using CDCl 3 . The chromatographic analysis was performed on an UltiMate 3000 UHPLC (Thermo Fisher Scientific, Waltham, MA, USA).

Statistical Analyses
A Pearson's chi-squared goodness of fit test (χ 2 ) was applied to determine whether our data (experimental chemical shift values of 13 C NMR for compound (1)) were significantly different from those expected (reported chemical shift values of 13 C NMR for compound 1). The same procedure was carried out for compound (2). It is worth mentioning that both χ 2 tests were applied with the Yates continuity correction to reduce the approximation error, and thus prevent overestimation of the statistical significance for small data [36]. Herein, the chi-squared statistic was as follows: where O i would represent the observed values, and E i would be the expected values.

Animal Collection and Identification
Samples of animals were collected by snorkeling and scuba diving in different coastal zones of the Yucatan Peninsula, Mexico, during the following three different periods: September-December 2016, January-March 2017, and September 2018. The selected species were collected from the following two different regions: Mexican Caribbean (Cozumel Island, Rio Indio, Mahahual, and Bermejo, Quintana Roo) and Campeche Bank (Alacranes Reef and Progreso, Yucatan) ( Figure 9). The sponges were identified at the ICMyL-UNAM (Mexico), while the ascidians were identified at the University of Vigo (Spain) and Autonomous University of Yucatan (Mexico). Taxonomic information, collection sites, and previous reports on the antifungal activity of the species/genus of the 65 marine organisms are shown in Table 10. The structures of the compounds with antifungal activity previously isolated from marine species, whose extracts were tested in this work, are depicted in Figure 10. Table 10. Taxonomic information and previously reported antifungal activity of the marine species studied in this work.

Species (Code Used in This
Antifungal Activity Previously Reported Reference The sponges were identified at the ICMyL-UNAM (Mexico), while the ascidians were identified at the University of Vigo (Spain) and Autonomous University of Yucatan (Mexico). Taxonomic information, collection sites, and previous reports on the antifungal activity of the species/genus of the 65 marine organisms are shown in Table 10. The structures of the compounds with antifungal activity previously isolated from marine species, whose extracts were tested in this work, are depicted in Figure 10. Table 10. Taxonomic information and previously reported antifungal activity of the marine species studied in this work.

Species (Code Used in
This Study)
[25]   Figure 10. Selected structures of compounds with reported antifungal activity isolated from marine organisms.

Preparation of the Organic Extracts
Tissue slices of each species were exhaustively extracted three times in a lapse of 24 h each, with a 500 mL mixture of dichloromethane-methanol (1:1), at 25 °C. The solvent was filtered and then removed under vacuum at 40 °C with a rotatory evaporator. The extracts were stored at −20 °C in tightly sealed glass vials.

Preparation of the Organic Extracts
Tissue slices of each species were exhaustively extracted three times in a lapse of 24 h each, with a 500 mL mixture of dichloromethane-methanol (1:1), at 25 • C. The solvent was filtered and then removed under vacuum at 40 • C with a rotatory evaporator. The extracts were stored at −20 • C in tightly sealed glass vials.
3.5. Antifungal Assays 3.5.1. Screening of the Marine Extracts C. glabrata (ATCC2001) and C. albicans (SC5314) were maintained in yeast peptone dextrose (YPD) agar plates and grown at 37 • C or 30 • C, respectively. Crude extracts were dissolved in DMSO, and 5 µL was added to the wells of a 96-well plate, containing 95 µL of RPMI-1640 medium at pH 7. Extract concentrations ranged from 125 to 12.5 µg/mL ( Table 1). Cellular suspensions of C. glabrata or C. albicans (3 × 10 3 CFU/mL) were prepared from fresh cultures grown overnight on YPD agar plates, and 100 µL was added to each well. Growth in RPMI-1640 medium was recorded after 48 h, by measuring OD 600 . The growth condition without an extract/fraction but with DMSO (control condition, 2.5% DMSO) was used as the normalization condition, after background (RPMI-1640 medium) subtraction. Growth ratios below 0.5 were considered for further analyses.

Antifungal Susceptibility Testing
The minimal inhibitory concentration (MIC) of C. glabrata and C. albicans was determined by conducting broth microdilution assays in accordance to the CLSI (Clinical Laboratory and Standards Institute) standard method (M27-A3) [74], with few modifications. Growth in RPMI-1640 medium was recorded after 24 and 48 h at 30 • C (C. albicans) or 37 • C (C. glabrata), by measuring OD 600 . The growth condition without an extract/(sub-)fraction, but with DMSO (final concentration 2.5%), was used as the normalization condition, after background (RPMI-1640 medium) subtraction. The MIC was set as the lowest extract/(sub-)fraction concentration at which there was a ≥50% decrease in growth compared to the control (cells grown in the presence of 2.5% of DMSO). At least, three independent assays were performed for each crude extract/fraction. Fluconazole (ACROS Organics) was used as a reference antifungal. The range of tested concentrations for each extract/(sub-)fraction is listed in Table S1. The minimal fungicidal concentration (MFC) was assessed by spotting 5 µL of the above cultures onto YPD agar plates. Growth was recorded after 24 or 48 h at 30 • C (C. albicans) or 37 • C (C. glabrata). The MFC corresponds to the concentration of the fraction that decreases the number of cells compared to the initial inoculum.

Bioassay-Guided Fractionation of the M. arbuscula Crude Extract
Sliced bodies of M. arbuscula (wet weight, 29.8 g; dry weight, 15.3 g) were exhaustively extracted, as previously described, to obtain 1.70 g of a crude residue. Liquid-liquid fractionation of 1.65 g of crude extract with H 2 O/CH 2 Cl 2 (1:1 v/v) produced an aqueous and organic phase. The aqueous phase was extracted with n-butanol (200 mL) to yield 217.0 mg of the final aqueous fraction (WF) and 756.0 mg of the n-butanol fraction (BF), after removal of the solvents under reduced pressure. The organic phase was concentrated under reduced pressure and was further partitioned between 10% aqueous CH 3 OH (400 mL) and hexane (2 × 400 mL) to produce, after removing the solvent under reduced pressure, 672.2 mg of the hexane fraction (HF). The H 2 O content (% v/v) of the methanolic fraction was adjusted to 50% aqueous CH 3 OH, and the mixture was extracted with CH 2 Cl 2 (100 mL) to afford, after removing the solvent under reduced pressure, 106.3 mg of the CH 2 Cl 2 fraction (DF) and 755.8 mg of the remaining aqueous methanolic fraction (WMF). The dichloromethane fraction (DF) was subjected to solid phase extraction (SPE) with RP-18 (Merck KGaA), using a stepped gradient from H 2 O to CH 3 OH and then CH 2 Cl 2 (H 2 O (100%), H 2 O/CH 3 OH (2:1, 1:1, and 1:2), CH 3 OH (100%), CH 3 OH/CH 2 Cl 2 (1:1), and CH 2 Cl 2 (100%), yielding seven fractions (R1-R7). The fractions were concentrated under reduced pressure, producing the following weights: R1: 3.8 mg, R2: 1.4 mg, R3: 22.3 mg, R4: 17.7 mg, R5: 39.8 mg, R6: 17.5 mg and R7: 3.8 mg. Fractions R2-R5 were subjected to UHPLC/HRMS analysis and the mobile phase consisted of the following compounds: (A) H 2 O with 0.1% formic acid (v/v); (B) CH 3 CN with 0.1% formic acid (v/v) at a flow rate of 400 µL/min. A combination of gradient and isocratic elution was used, starting with 99% A and 1% B, changing to 1% of A and 99% of B in 13 min, followed by 2 min of isocratic at 99% of B, 1 min gradient from 99% to 1% of B and finally, 4 min of isocratic at 99% of A.

De-Replication
De-replication of the sub-fractions was performed by ultra high-performance liquid chromatography/high-resolution mass spectroscopy (UHPLC/HRMS) on Q Exactive Focus (v/v). The mass spectrometer operated in the positive ESI mode. The exact mass of the components was compared against the Antimarin ® database and for the components with no matches in the database, the predicted molecular formula and exact mass were searched in the database platform SciFinder ® . If a plausible match was found, considering the exact mass/molecular formula, the molecule was considered as a putative component of the fraction. Finally, 13 C NMR spectra were recorded on a Bruker Avance 500 spectrometer at 125 MHz, respectively, using CD 3 OD for confirming the presence of the main chemical shifts of the compounds found.

Conclusions
This work shed light on the great antifungal potential of marine natural products produced by invertebrates of the Yucatan Peninsula. Three of the nine sponge species whose extracts were active against C. albicans and C. glabrata (H. (Rhizoniera) curacaoensis, A. crassa and A. dilatata) have never been associated with antifungal activity, and therefore may represent a new source of antifungal compounds.
The fact that most of these extracts were more effective against C. glabrata is particularly interesting, as this yeast is more tolerant to the current antifungals than C. albicans.
M. arbuscula stood out as the most active species against both C. glabrata and C. albicans. This observation is in line with several reports that highlight the antifungal activity of MNPs isolated from this organism, such as batzelladine L, batzelladine D, norbatzelladine L, and ptiolomycalin A. However, by combining a bioguided fractionation with a de-replication methodology, we found that the activity of M. arbuscula crude extract cannot be ascribed to these compounds. Interestingly, in one of the most active sub-fractions, we found several compounds, of which we identified two-mirabilin B and penaresidin B. Mirabilin B stands out as a promising drug candidate because the pure compound is active against another yeast species-C. neoformans-and its synthesis has already been reported. In the future, it would be interesting to further explore the antifungal and antibiofilm properties of mirabilin B on Candida spp. This would be particularly important given the fungicidal activity of the sub-fraction where mirabilin B was found, which makes the future isolation and identification of the molecules responsible for that activity a possible new strategy to combat life-threatening fungal infections that affect immunocompromised individuals.