Fungal Biofilms as a Valuable Target for the Discovery of Natural Products That Cope with the Resistance of Medically Important Fungi—Latest Findings

The development of new antifungal agents that target biofilms is an urgent need. Natural products, mainly from the plant kingdom, represent an invaluable source of these entities. The present review provides an update (2017–May 2021) on the available information on essential oils, propolis, extracts from plants, algae, lichens and microorganisms, compounds from different natural sources and nanosystems containing natural products with the capacity to in vitro or in vivo modulate fungal biofilms. The search yielded 42 articles; seven involved essential oils, two Brazilian propolis, six plant extracts and one of each, extracts from lichens and algae/cyanobacteria. Twenty articles deal with the antibiofilm effect of pure natural compounds, with 10 of them including studies of the mechanism of action and five dealing with natural compounds included in nanosystems. Thirty-seven manuscripts evaluated Candida spp. biofilms and two tested Fusarium and Cryptococcus spp. Only one manuscript involved Aspergillus fumigatus. From the data presented here, it is clear that the search of natural products with activity against fungal biofilms has been a highly active area of research in recent years. However, it also reveals the necessity of deepening the studies by (i) evaluating the effect of natural products on biofilms formed by the newly emerged and worrisome health-care associated fungi, C. auris, as well as on other non-albicans Candida spp., Cryptococcus sp. and filamentous fungi; (ii) elucidating the mechanisms of action of the most active natural products; (iii) increasing the in vivo testing.


Introduction
In recent decades, fungi has emerged as a major cause of life-threatening invasive human infections, in particular among immunocompromised patients [1][2][3], particularly those with human immunodeficiency virus (HIV), cancer patients receiving chemotherapy, transplant recipients, extremely aged persons and subjects in intensive care units [4,5]. Fungal infections lead to mortalities estimated in 1.5 million per year, having a great impact on global human health [5]. The main purpose of this review is to provide an updated analysis of the natural products with a capacity of inhibiting fungal biofilms, published from 2017 to May 2021.
We collected papers on essential oils (EOs), propolis, extracts from plants, algae, lichens and microorganisms, metabolites obtained from these sources and nanosystems. graphed by Stoodley et al. [49], Ramage et al. [50], Mayer et al. [51] and other authors [43,[52][53][54]. In stage 1, planktonic cells adhere to a biotic or abiotic surface, followed by a yeast-to-hyphal transition (Figure 1a,b). In the second stage, EPM is produced, resulting in a firmly adhered irreversible attachment. In stage 3 an early biofilm architecture is developed, while in stage 4 the biofilm reaches maturation in a three-dimensional structure ( Figure 1c). Finally, in stage 5, single planktonic cells are dispersed from the mature biofilm. This dispersion is an important step in the fungal biofilm development cycle, which can induce either bloodstream or invasive fungal infections [55] leading to an increased pathogenicity [56] and high risk of mortality [57,58]. This consortium of cells offers the optimum conditions for the fungi to obtain nutrients, dispose waste products and protect from the environment [58]. Once the biofilm is formed, the sessile cells communicate with each other through quorum sensing (QS) molecules that induce the fungal population to cooperate in diverse defense behaviors such as virulence and biofilm formation [59][60][61]. The sesquiterpene farnesol and the aromatic alcohol tyrosol are the two reported natural occurring QS molecules [62,63]. The first inhibits filamentation and biofilm formation in C. albicans [64] while tyrosol stimulates the germ tube formation and, thus, filamentation [63].
(b) (a) (c) Figure 1. (a) The five-stage process involved in the development of biofilm: 1. adherence of yeasts to a surface followed by yeast-to-hyphal transition; 2. the exopolymeric matrix (EPM) is produced resulting in a firmly adhered "irreversible" attachment; 3. early biofilm architecture is developed; 4. the biofilm reaches maturation in a three-dimensional structure and 5. single planktonic cells are dispersed from the mature biofilm. Reproduced from Stoodley et al., 2002 [49]. Image  Figure 1a). Bar = 10 µm. Yeasts, hyphae, and pseudohyphae can be observed. Most EPM was lost during the SEM procedures. Reproduced from Ramage et al. [50] with permission.

Natural Products as an Important Source of Antifungal Drugs
Most antifungal agents in clinical use are natural or natural-derived compounds, mainly isolated from microbial strains [104]. For example, the polyenes nystatin and amphotericin B were isolated from Streptomyces spp. [105]; the spiro-diketone griseofulvin, was obtained from Penicillum griseofulvum [106]; the echinocandins caspofungin and anidulafungin are semisynthetic derivatives from the natural pneumocandin B, which was isolated from the fungus Glarea lozoyensis [107]. In turn, micafungin derived from the lipopeptide FR901379 was isolated from the fungus Coleophoma empetri [107], a plant pathogen associated with postharvest fruit rot in cranberries [108]. Under this scenario, it is clear that nature has provided a robust platform for finding novel scaffolds for the discovery of antifungal drugs.
Natural antifungal products were typically discovered using the traditional antifungal assays, which measure the inhibition of growth of yeasts and filamentous fungi in their planktonic state, thus free-floating in the culture medium [109]. Recent works have reported extracts either from plants [110][111][112], algae [113,114], endophytic fungi [115] or marine fungi [116], as well as secondary metabolites such as phenols [111,117], flavonoids [118], naphtoquinones [119] and terpenoids [120], which showed growth inhibitory properties of fungi in their planktonic state. Both recent and previous reviews reported several pure natural products that have shown antifungal activity assessed with this type of techniques. However, none of these compounds have become leads for the development of new antifungal drugs [121].
The aim of this present review is to provide an update (2017-May 2021) on the natural products (extracts and secondary metabolites) that possess the ability to in vitro or in vivo modulate fungal biofilms not only constituted by C. albicans and some non-albicans Candida yeasts, but also by fungi from other genera. When available, the mechanism of action of these natural products has been included. To this purpose, most data published in the literature from 2017 were collected with the objective of drawing conclusions that can be useful for future research. Previously, some reviews on this subject were published [122][123][124][125][126] by Nazzaro et al. [123] and Singla and Dubey [124], with very few references each regarding the effect against fungal biofilms after 2017.

Reported Antibiofilm Activities of EOs, Propolis and Extracts from Plants, Algae and Cyanobacteria
Different studies involving EOs with antibiofilm activity are recorded in Table 1. Peixoto et al. [127] reported that Laurus nobilis L. (Lauraceae) EO at 2× minimum inhibitory concentration (MIC, 1000 µg/mL) inhibited the initial adhesion of C. albicans, while at 2× and 4× MIC, it inhibited biofilm formation and also reduced the eradication of mature biofilms with no significant difference when compared to the positive control, nystatin. In another study, Manoharan et al. [128]  (coriander) were demonstrated to inhibit more than 90% of biofilm formation when tested at 0.01%. Among them, cascarilla bark and helichrysum oil and their main components, α-longipinene and linalool, significantly reduced the yeast-to-hyphal transition, adherence and biofilm formation and greatly inhibited invasive hyphal growth in the nematode C. elegans. Serra et al. [129] tested different commercial EOs against two C. albicans strains. The results showed that only Pelargonium graveolens L'Hér. (Geraniaceae) (geranium) and Melisa officinalis L. (Lamiaceae) (melissa) EOs, eradicated mature biofilms with MBEC 80 of 22.3 and 17.9 µg/mL, respectively for geranium and of 13.3 µg/mL on both strains, for melissa. Out of the 12 EOs tested, only Pelargonium graveolens and Melissa officinalis eradicated mature biofilms [129] 2018 Pogostemon heyneanus, Cinnamomum tamala and Cinnamomum camphora C. albicans ATCC 90028; C. glabrata MTCC 6507 and C. tropicalis MTCC 310 The three EOs produced a reduction in the hyphal formation with P. heyneanus EO showing the maximum inhibition P. heyneanus and C. tamala disrupted mature biofilms Candida biofilms EPM was reduced. A large reduction of sugars was observed [130] 2019 Foeniculm vulgare EO (fennel oil) 10 isolates of C. albicans The MBEC 50 of fennel oil for 7/10 tested strains was 2-to 6-fold the MIC [131] 2020 Cymbopogon citratus, Cuminum cyminum, Citrus limon and Cinnamomum verum C. tropicalis isolates T26, U7 and V89 C. citratus EO reduced biofilm formation of all C. tropicalis tested strains. C. limon and C. cyminum EOs showed minor effects [132] 2021 As reported by Banu et al. [130], EOs from Cinnamomum tamala (Buch.-Ham.) T. Nees and Eberm. (Lauraceae) (Indian cassia), Pogostemon heyneanus Benth (Lamiaceae) (Indian patchouli) and Cinnamomum camphora (L.) J.Presl (Lauraceae) (camphor) inhibited about 54%-65% of biofilms formed by C. albicans, C. glabrata and C. tropicalis. In addition, the three EOs reduced the Candida spp. preformed biofilms, with an inhibition range of 55%-67% at their MBICs (0.5%-5% v/v); with C. tamala being the most active plant sp., P. heyneanus EO showed the maximum inhibition of yeast-to-hyphal transition. On the other hand, the EOs from P. heyneanus and C. tamala disrupted Candida spp. mature biofilms and reduced the thick aggregation of the yeast cells. This result was confirmed by the observation of a decrease of sugars present in the EPM layer.
The capacity of Foeniculum vulgare Mill. (Apiaceae) EO (fennel oil) to eradicate 10 strains of C. albicans biofilms was studied by Bassyouni et al. [131]. The MBEC 50 of fennel oil for eradicating the 18-h-old biofilm was 2-to 16-fold of MIC, in 7/10 tested strains. Sahal et al. [132] investigated the antifungal and biofilm inhibitory effects of EOs by using C. tropicalis biofilms coated on different biomaterials. Treatments with 2%-8% of C. citratus EO coated on silicone rubber resulted in a 45%-76% reduction in biofilm formation of all the strains. Likewise, Cuminum cyminum L. (Apiaceae), Citrus limon (L.) Osbeck (Rutaceae) and Cinnamomum verum J.Presl (Lauraceae) EOs, were also effective in inhibiting the C. tropicalis biofilms in polystyrene plates at sub-MIC values. Therefore, the mentioned extracts, in particular C. citratus EO, could be used as an antibiofilm agent on silicone rubber prostheses and medical devices. C. tropicalis biofilms pose a serious risk for skin infections and may cause a shorter lifespan of the prosthesis. In an additional study, Choonharuangdej et al. [133] tested the efficacy of C. verum and C. citratus EOs for eradicating C. albicans biofilms established on heat-polymerized polymethyl methacrylate (PMMA) material and determined whether they were able to retard the formation of fungal biofilms and/or eradicate them. Results showed that cinnamom oil at 0.8 µL/mL (8× MIC) and lemongrass oil at 6.4 µL/mL (16× MIC), both coated on PMMA, inhibited the formation of C. albicans biofilms by 70.0% after 24 h of treatment. In contrast, at 8× MIC (0.8 and 3.2 µL/mL, respectively), both EOs eradicated 99% of the pre-established C. albicans biofilm in 1 h. Table 2 shows the relevant studies on antifungal propolis with antibiofilm activity. Galletti et al. [134] evaluated the activity of green propolis collected in Paraná (Brazil) against Fusarium spp. biofilms that frequently cause disseminated infections in immunocompromised patients, with a high rate of mortality [135]. In the mentioned work, the authors used clinical isolates of F. oxysporum, F. solani and F. subglutinans and a standardized F. solani strain. The used propolis proved to be efficient to reduce both the total biomass (assessed with CV dye) and the number of viable cells (quantified with XTT) for all evaluated isolates. In addition, the CW fluorescence assay showed that biofilm structure was lost, leaving only isolated damaged cells. In a recent paper, Martorano Fernandes et al. [136] evaluated the inhibitory effects of Brazilian red propolis on C. albicans and a co-culture of C. albicans-C. glabrata biofilms. Metabolic activity determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, and cell viability assessed with CFU counts and surface roughness (optical profilometry) were evaluated. Results showed that red propolis had high inhibitory mono sp-biofilm effects but low activity against co-cultured two spp. biofilm formation, compared to chlorhexidine. The surface roughness (Sa parameter) within the mono-sp. and the co-culture biofilms statistically differed among groups. Table 3 shows the different studies on extracts from plants, lichens, algae and cyanobacteria with antibiofilm activity.   Hydroethanol extracts of leaves, pulps, seeds and barks of several Eugenia spp. (Myrtaceae) such as E. leitonii Legrand, E. brasiliensis Lam., E. myrcianthes Nied. and E. involucrata DC. were tested for their capacity to eradicate mature C. albicans biofilms [137]. Among them, extracts from E. leitonii seeds and E. brasiliensis seeds and leaves reduced C. albicans biofilm viability by 54%, 54% and 55% at 156.2, 156.2 and 312.5 µg/mL (10× MIC), respectively, with better activity than that of nystatin which showed a 42% reduction. At these concentrations, all extracts caused damage to biofilm architecture and integrity, which could be observed by SEM. Alizadeh et al. [138] determined by CV assay that the ethanol extract of Malva sylvestris L. (Malvaceae) root at 0.78 and 1.56 mg/mL (MIC and 2× MIC, respectively) reduced C. albicans biofilm growth. By light microscopy, the authors observed that the extract was able to decrease biofilm thickness and cellular density. Silva et al. [139] determined that the hydroethanol extract of Anadenanthera colubrina Vell. Brenan (Fabaceae) barks, from the Caatinga biome (Brazil), showed the capacity to eradicate mature biofilms formed by four C. albicans strains, and one of each C. parapsilopsis and C. krusei, causing, in the last two, a 100% decrease of biofilms at 500× MIC. The Algerian Clematis flammula L. (Ranunculaceae) ethanol leaves (CFL) and Fraxinus angustifolia Vahl. (Oleaceae) leaves (FAL) and bark (FAB) extracts at 500 µg/mL showed 36.8%, 62.4% and 54.8% inhibition of C. albicans biofilm formation, respectively, which was probably related to the disruption of the cell surface hydrophobicity (CSH) and to the inhibition of germ tube and hyphae formation [140]. After four h incubation, 66.32% of hyphal form was seen in the control group, while 3.96%, 2.11% and 1.65% of hyphae were formed in the presence of CFL, FAL and FAB, respectively [140]. As reported by Dwivedi et al. [141], the Hibiscus sabdariffa L. (Malvaceae) flower DMSO extract inhibited the yeast-to-hypha transition with hyphae showing morphological changes, and also adherence of C. albicans cells (80% at 6.25 mg/mL). It also inhibited biofilm formation as well as disrupted the pre-formed C. albicans biofilm by 50% when tested at 3.12 mg/mL. The hexane extract of purple leaves from Orthosiphon aristatus (Blume) Miq. (Lamiaceae) was tested on different biofilm stages. [142]. Treatments of C. albicans with the extract at 2 mg/mL, showed a 69.2% decrease in cell viability (assessed with MTT) at the adhesion stage, while fluconazole at 6 µg/mL caused a 54.7% reduction. It is important to highlight that a 50% reduction in cell density compared to the negative control was observed in the presence of 1.3 mg/mL of the extract, as determined by the CV assay. At the development stage, a 57.1% and 57.3% inhibition on C. albicans growth, in the presence of the extract and fluconazole respectively, was observed. A low inhibition of approximately 20% was observed on mature biofilms after treatment with both the extract and fluconazole [142]. The ability to inhibit the formation (named in this paper 'anti-maturation') or to eradicate preformed 24 h-old C. albicans biofilms (named 'antibiofilm') were evaluated for 38 lichen acetone extracts, by XTT assay [143]. Among them, eleven extracts showed antibiofilm activity, with seven displaying both anti-maturation and antibiofilm properties. Of them, extracts from Evernia prunastri (L.) Ach (Parmeliaceae) and Ramalina fastigiata (Pers.) Ach. (Ramalinaceae) were the most promising ones, with half inhibitory concentration (IC 50 ) values <4 µg/mL for anti-maturation. E. prunastri, Cladonia uncialis (L.) Weber ex F.H.Wigg (Cladoniaceae), R. fastigiata and Xanthoparmelia conspersa (Ehrh. ex Ach.) (Parmeliaceae) extracts showed IC 50 values <10 µg/mL for antibiofilm eradication. Cepas et al. [144] tested 675 hexane, ethyl acetate and methanol extracts of 225 microalgae and cyanobacteria against C. albicans and C. parapsilopsis biofilms, which were inhibited by 308 extracts. Among the 11 phylum, the lowest activity was reported for Euglenophyta, with MBIC 50 and MBIC 90 90 values of 64 and 512 µg/mL, respectively, against C. parapsilopsis.

Reported Antibiofilm Activities of Pure Natural Compounds
Pure natural compounds with antibiofilm activity are summarized in Table 4 and detailed below. Liu et al. [145] demonstrated that the formyl-phloroglucinol meroditerpenoid, eucarobustol E (EE), isolated from Eucalyptus robusta Sm. (Myrtaceae), suppressed 73% of C. albicans biofilm formation at 32 µg/mL, destroyed nearly all mature biofilms (92%) at 128 µg/mL, blocked the yeast-to-hyphal transition and reduced the CSH at 16 µg/mL. EE downregulated the expression of genes involved in hyphal growth (EFG1, CPH1, TEC1, EED1, UME6, and HGC1), cell surface proteins (ALS3, HWP1, and SAP5) and in ergosterol biosynthesis (ERG6, ERG13, ERG252, ERG11, ERG10, and ERG7). This activity resulted in the reduction of ergosterol, which alters cell membrane functions, leading to cell death. In turn, EE upregulated the farnesol-encoding gene DPP3, which negatively regulated biofilm formation. According to the authors, EE differs from clinical antifungal agents in their antibiofilm mechanisms and so it is certainly worth considering for further development as an antifungal drug. Shi et al. [146] reported that berberine (BBR) inhibited biofilm formation of 13 strains of C. tropicalis and one strain of each C. albicans, C. parapsilosis and C. glabrata with MBICs ranging from 64 to 256 µg/mL. The mRNA expression of ERG11 and of the efflux proteins CDR1 and MDR1 were 1.43-2.10-fold upregulated by BBR at 16 µg/mL. Behbehani et al. [147] found that the lignan magnolol (2-(2-hydroxy-5-prop-2-enylphenyl)-4-prop-2-enylphenol), isolated from Magnolia officinalis Rehder and E.H. Wilson (Magnoliaceae), showed strong antibiofilm formation activity against C. albicans, C. dubliniensis and C. glabrata, with 69.5%, 46.7% and 35.6% at 32 µg/mL, respectively, as determined by MTT assay. Six EO's components such as thymol, carvacrol, cinnamaldehyde, citral, menthol and eugenol were tested by Kumari et al. [148]. These compounds proved to be effective on biofilm formation and on preformed C. neoformans and C. laurentii biofilms in the following order: thymol > carvacrol > citral > eugenol = cinnamaldehyde > menthol with MBIC 80 ranging from 32 to 128 µg/mL and MBEC 80 ranging from 64 to 256 µg/mL, determined by XTT assay. SEM and CLSM showed the absence of EPM, reduction in cellular density and alteration of the surface morphology of biofilm cells. Cryptococcosis is a systemic infection [149] very difficult to treat due to the ability of these fungi to form biofilms resistant to standard antifungal treatments. Kumari et al. [150] deepened the study of the C. neofomans antibiofilm activity of thymol (16 µg/mL), carvacrol (32 µg/mL) and citral (64 µg/mL) using field emission scanning electron microscopy (FE-SEM), atomic force microscopy (ATM) and Fourier transform infrared spectroscopy. The three terpenes appear to act through the interaction with ergosterol or the inhibition of its biosynthesis, and the disruption of the biofilm cell surface, with pore formation and efflux of the K+/intracellular content. Morphological changes and qualitative/quantitative alterations in the EPM and in cellular components of C. neoformans biofilm cells were also observed. The terpenes-treated cells showed 35%−45% reduction in total carbohydrates, with variation in the type of glycosyl residues. Li et al., showed that the eudesmane sesquiterpene ent-isoalantolactone (ent-iLL) showed inhibition of the yeast-to-hyphal conversion of a mutant of C. albicans in assays performed in liquid and solid media at 8 µg/mL and 4 µg/mL, respectively [151]. In addition, ent-iLL at 16 µg/mL reduced the presence of ergosterol in the membrane, through inhibiting the activity of Erg11 and Erg6 [151]. The polyphenol curcumin (Cur) was evaluated for their antibiofilm properties against C. albicans by Alalwan et al. [152]. The MBIC 80, was 200 µg/mL, as determined by XTT assay. Furthermore, Cur at 50 µg/mL was able to decrease C. albicans adhesion to a PMMA denture base material, an effect that could be enhanced by pre-treatment of the yeasts with the polyphenol. Regarding its molecular effects, Cur down-regulated the adhesin ALS3, with minimal effect on its related ALS1. On the contrary, the clustered aggregative and flocculation genes AAF1, EAP1, and ALS5 transcripts were up-regulated. Three gingerols (6-, 8-and 10-gingerols) and three shogaols (6-, 8-, and 10-shogaols) isolated from Zingiber officinale Roscoe (Zingiberaceae) showed antibiofilm and anti-virulence activities against a fluconazole-resistant C. albicans strain. Results showed that only 6-shogaol at 10 µg/mL and 6-and 8-gingerols at 50 µg/mL, significantly reduced the C. albicans biofilm formation, suggesting that the increase in the length of the side chain decreased the activity. CLSM showed that biofilms treated with 6-gingerol and 6-shogaol were reduced in density and in thickness. In addition, both compounds inhibited hyphal growth in embedded colonies and free-living planktonic cells, and prevented cell aggregation, which was confirmed by SEM. Both entities significantly altered the expressions of some hypha-specific (HWP1 and ECE1), biofilmrelated (HWP1 and RTA3) and multidrug transporter (CDR1 and CDR2) related genes [153]. Yan et al. [154] reported that the 1,4-naphtoquinone derivative shikonin (SK), at 32 µg/mL, inhibited almost totally C. albicans biofilm formation and eradicated the preformed mature biofilms, as evidenced with XTT assay and confirmed by CLSM. SK inhibited hyphae formation, showing a complete inhibition at 0.5 µg/mL in Lee s medium and reduced CSH by 70.3% at 2 µg/mL. Several hypha and adhesion specific genes such as of ECE1, HWP1, EFG1, CPH1, RAS1, ALS1, ALS3 and CSH1 were downregulated while TUP1, NRG1 and BCR1 were upregulated by SK. In addition, SK induced the production of farnesol at 8 µg/mL, which enhanced its antibiofilm activity. Saibabu et al. [155] showed that different C. albicans strains treated with 62.5 µg/mL of vanillin showed few or no adherence to buccal epithelial cells. By MTT assay, it was observed the adherence of Candida cells to polystyrene surface was reduced by 52%, the biofilm formation was decreased by 49%, and the mature biofilm eradication decreased by 52%. At 125 µg/mL, vanillin protected C. elegans against C. albicans infection and enhanced its survival [155]. Kischkel et al. [156] evaluated farnesol on preformed F. solani complex biofilms particularly formed by Fusarium keratoplasticum, which is the most prevalent fungi related to biofilm formation in hospital water systems and in internal pipelines. Farnesol showed activity against F. keratoplasticum preformed biofilms, and was effective also during its formation at different times (at adhesion and at 24, 48 and 72 h) with a complete inhibition at 700 µM, as assessed by counting the CFU number and by CV and XTT assays.
A recent report from Wang et al. [157] described the isolation of 14 new terpenoids from the liverwort Heteroscyphus coalitus (Hook.) Schiffner (Geocalycaceae), including eight ent-clerodane diterpenoids, four labdane diterpenoids, heteroscyphsic acids A-I, heteroscyphins A-E, a harziane type diterpenoid and one guaiane sesquiterpene together with a known ent-junceic acid. At 4-32 µg/mL most of these compounds inhibit hyphal and biofilm formation of an efflux pump deficient strain of C. albicans, but not of the wild type of strain. The most effective molecule was heteroscyphin D, which suppressed the ability of C. albicans to adhere to A549 cells and to form biofilms with a complete inhibiton at 8 µg/mL, determined by XTT assay. In addition, this compound was able to modulate the transcription of related genes in this fungus, as described in Table 4. Das et al. [158] showed that artemisinin (Ar), the sesquiterpene lactone isolated from Artemisia annua L. (Asteraceae), and scopoletin (Sc), the 7-hydroxy-6-methoxy coumarin present in several spp., were tested on mature biofilms of different albicans and non-albicans Candida strains. The results demonstrated that Ar was more active than Sc in disrupting the preformed EPM-covered biofilm structure and in killing the sessile cell population at their respective MBEC 10 . In the same year, Lemos et al. [159] demonstrated that Sc, at its MIC (50 µg/mL), was able to reduce the preformed biofilms of a resistant C. tropicalis strain in 68.2%, and to inhibit the biofilm formation. Kipanga et al. [160] showed that the drimane sesquiterpene dialdehydes warburganal and polygodial, obtained from Warburgia ugandensis Sprague (Canellaceae), inhibited C. albicans biofilm formation with MBIC 50 of 4.5 and 10.8 µg/mL, respectively, and with MBIC 50 of 49.1 and 50.6 µg/mL, respectively, against C. glabrata. Regarding biofilm eradication, warburganal and polygodial showed MBEC 50 of 16.4 and 16.0 µg/ml, respectively, against C. albicans but did not inhibit C. glabrata biofilm eradication. The higher potency of warburganal over polygodial for inhibiting biofilm formation and eradication could be attributed to the hydroxyl group present at position C-9, that differentiates both sesquiterpenes. The triterpenoid saponins gypenosides, isolated from Gynostemma pentaphyllum (Thunb.) Makino (Cucurbitaceae), showed MBIC 80 > 128 µg/mL against two fluconazole-resistant strains of C. albicans, as determined by XTT. No significant reduction in the density and in the length of the hyphae were observed [161].
Zhao et al. [162] showed that turbinmicin, a highly functionalized polycyclic compound isolated from the marine microbiome, completely disrupted extracellular vesicle (EV) delivery, during biofilm growth at 16 µg/mL, and this impaired the subsequent assembly of the biofilm EPM. C. albicans biofilm EVs have a pivotal role in EPM production and biofilm drug resistance [163]. This property was observed by a combination of flow cytometry, image confirmation, and fluorescence sensitivity on C. albicans, C. tropicalis, C. glabrata, C. auris and A. fumigatus. By SEM, it was determined that turbinmicin at 2.5 µg/mL eliminated EPM, thus rendering the drug-resistant biofilm communities susceptible to the antifungal effects of TBM itself, as well as to clinical antifungal agents. Zainal et al. [164] demonstrated that allicin, the organosulfur compound obtained from Allium sativum L. (Amaryllidaceae), was able to eradicate 50% of C. albicans biofilm formed on self-polymerized acrylic resin when administered at a sub-MIC concentration of 4 µg/mL. Feldman et al. [165] reported that cannabidiol (CBD) obtained from Cannabis sativa L. (Cannabaceae) exerted an inhibitory effect on biofilm formation with a MBIC 90 of 100 µg/mL. At 25 µg/mL, the metabolically active cells (assessed by MTT) in 24, 48 and 72 h-biofilms decreased by 48%, 64% and 87%, respectively. At 1.56 and 3.12 µg/mL, mature biofilms decreased by 28% and 44%, respectively. Furthermore, CBD reduced the thickness of fungal biofilm and EPM production accompanied by a downregulation of genes involved in EPM synthesis. At 25 µg/mL, CBD inhibited 90% of the ATP synthesis and enhanced mitochondrial membrane potential and ROS levels. At 1 4 MBIC 90 , CBD produced upregulation of yeast-associated genes and downregulation of hyphae-specific genes. As reported by Ivanov et al. [166], camphor inhibited more than 50% of the biofilm biomass in C. albicans strains at 62-250 µg/mL, while in C. tropicalis, the inhibition was achieved at 175 µg/mL. On the other hand, eucalyptol showed the same effect in the tested C. albicans strains at higher concentrations (>3000 µg/mL) and in C. tropicalis, C. parapsilosis and C. krusei at >1000 µg/mL [166]. Camphor, applied at 125 µg/mL, reduced ROS generation in one strain of C. albicans, although eucalyptol failed to exert this effect. The compounds did not interfere with ERG11 expression. Neosartorya fifischeri antifungal protein 2 (NFAP2), a novel member of small cysteine-rich and cationic antifungal proteins from filamentous ascomycetes (crAFPs), showed low activity against C. auris biofilms [167]. When NFAP2 was tested in combination with clinical antifungals, an enhancement of the activity was observed. The results of combinations were not included in this review.       6-S at 10 µ g/mL was more effective in suppressing hyphal formation than 6 g at 50 µ g/mL 6-S at 10, 50, and 100 µ g/mL inhibited 85, 94, and 94% biofilm formation, respectively. 6 g and 8 g at 50 µ g/mL inhibited by 88 and 80%, respectively the biofilm formation 6 g and 6-S significantly altered the expressions of some hypha-specific (HWP1 and ECE1), biofilm-related (HWP1 and RTA3) and multidrug transporter (CDR1 and CDR2) related genes. 80% of C. elegans infected with C. albicans survived in the presence of both compounds at 50 μg/mL [153] 6-shogaol (6-S) Fluconazole-resistant C. albicans DAY185 6-S at 10 µg/mL was more effective in suppressing hyphal formation than 6 g at 50 µg/mL 6-S at 10, 50, and 100 µg/mL inhibited 85, 94, and 94% biofilm formation, respectively. 6 g and 8 g at 50 µg/mL inhibited by 88 and 80%, respectively the biofilm formation 6 g and 6-S significantly altered the expressions of some hypha-specific (HWP1 and ECE1), biofilm-related (HWP1 and RTA3) and multidrug transporter (CDR1 and CDR2) related genes. 80% of C. elegans infected with C. albicans survived in the presence of both compounds at 50 µg/mL [153]  The three compounds inhibit biofilm formation and eradicate mature biofilms by the following mechanisms: (i) ergosterol biosynthesis inhibition and selectively interaction via ergosterol binding, (ii) disruption of the biofilm cell surface with reduction in cell height, (iv) alterations in the fatty acid profile which attenuate the cell membrane fluidity with enhanced permeability, resulting in pore formation and efflux of the K+/intracellular content, (v) mitochondrial depolarization caused higher levels of ROS. Then, the oxidative stress caused a significant decline in the amount of EPM and capsule sugars (mannose, xylose, and glucuronic acid), leading to a reduced capsule size and an overall negative charge on the cell surface The filamentation in Lee's media was completely inhibited by 0.5 µ g/mL of SK SK at 4 µ g/mL inhibited biofilm formation by 65.4%, while the biofilm growth was almost totally inhibited when exposed to 32 µ g/mL of SK SK at 32 µ g/mL destroyed mature biofilms by 92 The three compounds inhibit biofilm formation and eradicate mature biofilms by the following mechanisms: (i) ergosterol biosynthesis inhibition and selectively interaction via ergosterol binding, (ii) disruption of the biofilm cell surface with reduction in cell height, (iv) alterations in the fatty acid profile which attenuate the cell membrane fluidity with enhanced permeability, resulting in pore formation and efflux of the K+/intracellular content, (v) mitochondrial depolarization caused higher levels of ROS. Then, the oxidative stress caused a significant decline in the amount of EPM and capsule sugars (mannose, xylose, and glucuronic acid), leading to a reduced capsule size and an overall negative charge on the cell surface The filamentation in Lee's media was completely inhibited by 0.5 µ g/mL of SK SK at 4 µ g/mL inhibited biofilm formation by 65.4%, while the biofilm growth was almost totally inhibited when exposed to 32 µ g/mL of SK SK at 32 µ g/mL destroyed mature biofilms by 92 The three compounds inhibit biofilm formation and eradicate mature biofilms by the following mechanisms: (i) ergosterol biosynthesis inhibition and selectively interaction via ergosterol binding, (ii) disruption of the biofilm cell surface with reduction in cell height, alterations in the fatty acid profile which attenuate the cell membrane fluidity with enhanced permeability, resulting in pore formation and efflux of the K+/intracellular content, (v) mitochondrial depolarization caused higher levels of ROS. Then, the oxidative stress caused a significant decline in the amount of EPM and capsule sugars (mannose, xylose, and glucuronic acid), leading to a reduced capsule size and an overall negative charge on the cell surface [150] Antibiotics 2021, 10 The three compounds inhibit biofilm formation and eradicate mature biofilms by the following mechanisms: (i) ergosterol biosynthesis inhibition and selectively interaction via ergosterol binding, (ii) disruption of the biofilm cell surface with reduction in cell height, (iv) alterations in the fatty acid profile which attenuate the cell membrane fluidity with enhanced permeability, resulting in pore formation and efflux of the K+/intracellular content, (v) mitochondrial depolarization caused higher levels of ROS. Then, the oxidative stress caused a significant decline in the amount of EPM and capsule sugars (mannose, xylose, and glucuronic acid), leading to a reduced capsule size and an overall negative charge on the cell surface The filamentation in Lee's media was completely inhibited by 0.5 µ g/mL of SK SK at 4 µ g/mL inhibited biofilm formation by 65.4%, while the biofilm growth was almost totally inhibited when exposed to 32 µ g/mL of SK SK at 32 µ g/mL destroyed mature biofilms by 92 The three compounds inhibit biofilm formation and eradicate mature biofilms by the following mechanisms: (i) ergosterol biosynthesis inhibition and selectively interaction via ergosterol binding, (ii) disruption of the biofilm cell surface with reduction in cell height, (iv) alterations in the fatty acid profile which attenuate the cell membrane fluidity with enhanced permeability, resulting in pore formation and efflux of the K+/intracellular content, (v) mitochondrial depolarization caused higher levels of ROS. Then, the oxidative stress caused a significant decline in the amount of EPM and capsule sugars (mannose, xylose, and glucuronic acid), leading to a reduced capsule size and an overall negative charge on the cell surface The filamentation in Lee's media was completely inhibited by 0.5 µ g/mL of SK SK at 4 µ g/mL inhibited biofilm formation by 65.4%, while the biofilm growth was almost totally inhibited when exposed to 32 µ g/mL of SK SK at 32 µ g/mL destroyed mature biofilms by 92 The three compounds inhibit biofilm formation and eradicate mature biofilms by the following mechanisms: (i) ergosterol biosynthesis inhibition and selectively interaction via ergosterol binding, (ii) disruption of the biofilm cell surface with reduction in cell height, (iv) alterations in the fatty acid profile which attenuate the cell membrane fluidity with enhanced permeability, resulting in pore formation and efflux of the K+/intracellular content, (v) mitochondrial depolarization caused higher levels of ROS. Then, the oxidative stress caused a significant decline in the amount of EPM and capsule sugars (mannose, xylose, and glucuronic acid), leading to a reduced capsule size and an overall negative charge on the cell surface The three compounds inhibit biofilm formation and eradicate mature biofilms by the following mechanisms: (i) ergosterol biosynthesis inhibition and selectively interaction via ergosterol binding, (ii) disruption of the biofilm cell surface with reduction in cell height, (iv) alterations in the fatty acid profile which attenuate the cell membrane fluidity with enhanced permeability, resulting in pore formation and efflux of the K+/intracellular content, (v) mitochondrial depolarization caused higher levels of ROS. Then, the oxidative stress caused a significant decline in the amount of EPM and capsule sugars (mannose, xylose, and glucuronic acid), leading to a reduced capsule size and an overall negative charge on the cell surface   Although the main objetive of this review is to provide published data on the in vitro or in vivo antibiofilm activity of natural products, mention should be made to the method reviewed by Jha et al., 2020 [17]. These authors performed a multiple target-based structural bioinformatic study to recognize molecules that have the capacity of targeting proteins with a vital role in wall synthesis (FKS2 and FKS1), ergosterol synthesis (ERG11, ERG1, ERG24 and ERG3), and drug transport, such as Flu1 and the kinases CST20, HST7 and CEK1 involved in CHP1 pathway. The Cph1 transcription factor is involved in biofilm and pseudohyphae formation [168]. According to the results obtained, 2-O-prenylcoumaric, acid, 4-coumaryl acetate, coniferaldehyde, coniferyl alcohol, cycloserinehybrid and 2coumaric acid targeted all proteins, while 110 molecules bound to at least four of them [17]. Among the evaluated compounds, vanillin showed interactions with ERG11, ERG1, ERG24, FKS2, CST20, HST7 and CEK1 while BBR targeted all of these proteins except Flu1 [17]. Thymol interacted to all proteins, except ERG24 and ERG1, and carvacrol bound to ERG24, ERG1 HST7, CST20, CEK1 and Flu1. Likewise, eugenol targeted ERG24, ERG1, FKS1, FKS2, FUR1 and Flu1 and cinnamaldehyde interacted with all of the proteins except FKS1, FKS2, FUR1 and Flu1. These data would explain, at least in part, the antibiofilm activity of the mentioned small molecules, which has been reported above and also in Table 4, while opening the possibility to test experimentally the effects on biofilms of the rest of the ligands.

Reported Antibiofilm Activities of Nanosystems Containing Natural Products
The published literature related to the antibiofilm properties of nanosystems are shown in Table 5. Quatrin et al. [169] prepared nanoemulsions (NE) containing 5% E. globulus EO and 2% sorbitan monooleate (Span 80) with the aqueous phase containing 2% Tween 80. EO-NE were tested for their capacity to inhibit biofilm formation of three Candida spp. in high density polyethylene substrates. EO-NE at 22.5 mg/mL reduced C. albicans, C. tropicalis and C. glabrata biofilm formation by 84.5%, 61.3% and 84%, respectively. The biofilms were quantified by CV staining and corroborated by ATM and the fluorescence CW technique.
Due to M. alternifolia EO (tea tree oil (TTO)) showed antibiofilm activity against C. albicans spp. in previous studies but possesses poor solubility and high volatility [170], Souza et al. [171] prepared TTO-nanoparticles and tested them against C. albicans, C. glabrata, C. parapsilopsis, C. membranafaciens and C. tropicalis biofilms. The results obtained showed that TTO and TTO-nanoparticles decreased by 67% and 72% respectively the C. albicans and C. glabrata biofilm biomass visualized with CV and confirmed by CW and ATM. The antibiofilm activity was more evident against C. glabrata. TTO-NP decreased EPM and proteins in biofilms and, therefore, TTO-NP can penetrate more easily through the EPM, releasing the TTO into the biofilm and resulting in a better antimicrobial activity. TTO-NPs can disturb the fungal membrane by blocking the respiratory chain through the inhibition of the enzyme succinate dehydrogenase of the internal fungal cell mitochondrial membrane. CU-SL at sub-inhibitory concentrations of 9.37 µg/mL significantly suppressed fungal adhesion CU-SL at 9.37 µg/mL inhibited biofilm development and maturation Four major transcriptional genes that promote biofilm formation, ROB1, EFG 1, TEC1, BRG 1 and NDT80 were down-regulated, as well as the adhesin genes ALS1, SAP8 and EAP1, the hyphal regulatory genes SAP4, HWP1 RAS1 and HYR1 and ERG11 [174] ATCC: American type cultuire collection (Manassas, VA, USA); EO: essential oil; NE: nanoemulsions; NP: nanoparticles. Other culture collection acronyms can be found in the respective reference. Considering that free Cur at 200 µg/mL inhibited almost completely the C. albicans biofilm formation, Ma et al. [172] prepared Cur-chitosan nanoparticles (CSNP). However, CSNP-Cur exhibited a slightly less inhibitory effect than free Cur. On the contrary, 400 µg/mL of CSNP-Cur eradicated the preformed C. albicans biofilms by 91.38%, being more effective than free Cur. SEM and CLSM showed that CSNP-Cur reduced polymicrobial biofilm thickness as well as killed microbial cells embedded in biofilms on silicone surfaces. In a recent paper, Gumus et al. [173] prepared nanoparticles of juglone (JU) encapsulated in poly D,L-lactic-co-glycolic acid (PLGA). As determined by plate count technique, these PLGA-JU nanoparticles completely inhibited biofilm formation and pre-established biofilms, at doses equivalent to 1.25 and 0.625 mg/mL of JU, respectively, whereas fluconazole did not cause a significant inhibition, even at the highest dose applied (10 mg/mL). PLGA-JU were more active than free JU and fluconazole. The inhibition of biofilm formation seems to be due to PLGA-JU reducing the cell adhesion, and the number of viable cells and altered membrane structures at both mentioned doses.
Rajasekar et al. [174] synthesized a nanocomplex formed by the surfactant sophorolipid (SL) derived from non pathogenic yeasts such as Starmerella/Candida bombicola and Cur, and tested it against C. albicans biofilms.
Sub-inhibitory concentrations of Cur-SL (9.37 µg/mL) significantly decreased fungal adhesion to glass coverslips, as well as biofilm development, maturation, and filamentation. Concomitantly, a significant downregulation of a select group of biofilm adhesins and hyphal regulatory genes was observed.

Discussion
In this review, the natural products that demonstrated activity against fungal biofilms from 2017 to May 2021 were collected, with the aim of offering an overview of the progress made in this area to curb difficult-to-eradicate fungal infections. The type of natural products that showed better antibiofilm activities, the fungal species of the target biofilms, the type of assays used and the mechanisms of action were analyzed in order to detect the advances performed in this period that can be the basis of future works.
Regarding the type of natural products that showed antibiofilm activities, 17 of the 42 manuscripts involved natural mixtures such as EOs (seven manuscripts) propolis (two manuscripts) or extracts prepared with solvents (total eight manuscripts), six of them from plants, one from lichens and one from microalgae or cyanobacteria. Two papers dealt with nanosystems including EOs. The antibiofilm activities of the seven EOs do not add a great novelty to the already known antibiofilm activity of these products. However, the nanosystems prepared with EOs open a great avenue for further research. In addition, the paper authored by Quatrin et al. [169], which tested a nanosystem with E. gobulus EO (EO-NE), not only used C. albicans as in most publications on antibiofilm EOs, but also used C. tropicalis, C. glabrata and C. auris although the EO-NE was devoid of activity in the latter sp. [169]. Considering that C. glabrata is a fungus intrinsically resistant to most antifungal drugs [175], the activity found against this fungus is encouraging for further research. With respect to extracts obtained with solvents from different spp., no preference for a botanical family was observed, and in most cases the antibiofilm activities should be deepened since the mechanism to achieve this effect was not analyzed.
Twenty three of the 42 retrieved articles tested pure natural compounds, twenty of them alone and three included in nanosystems. Regarding the main type of pure compounds that showed a capacity for inhibiting biofilms on its own, nine of them were terpenes with four monoterpenes derivatives, four sesquiterpenes and one diterpene, which is in accordance with previous reports [124]. Two manuscripts studied the coumarin Sc, and one evaluated dammarane-type glycosides (gypenosides). A phloroglucinol, an alkaloid, a neolignan and some diphenols were investigated along with four compounds belonging to other types of chemical families. Cur was included in two nanosystems. Most of the compounds (21 compounds) were derived from plants, with one compound obtained from fungi (farnesol) and one obtained from a sea squirt microbiome (turbinmicin).
With respect to the fungal spp. tested in the 42 papers included in this review, 38 of these involved Candida spp. biofilms, two comprised Fusarium spp., two involved Cryptococcus spp. and one, A. fumigatus plus Candida spp. Among Candida biofilms, 23 studies included only C. albicans and 12 used different non-albicans Candida spp., but also involved C. albicans. Two papers used only C. tropicalis biofilms. Unfortunately, only two papers used C. auris among the non-albicans Candida spp. The newly described yeast C. auris has emerged as a resistant fungal pathogen responsible for hospital outbreaks, especially in intensive care units [176], leading to high mortality rates [177]. C. auris is at present a worrisome healthcare problem and new antifungals that overcome its resistance are highly needed [177]. The protein NFAP2 and the highly functionalized polycyclic turbinmicin completely disrupted EV delivery during C. auris biofilm growth [164]. Turbinmicin also showed activity against A. fumigatus biofilms.
Regarding anti-C. neoformans and C. laurentii biofilms, only thymol, carvacrol and citral proved to be effective for inhibiting the formation and eradication of the mentioned biofilms; however, importantly, their mechanisms of action were studied [150,152]. These are significant findings, since this fungus can colonize the central nervous system, highlighting its significance as a critical pathogenic yeast.
Propolis from Brazil and farnesol were the sole natural samples which showed a capacity for inhibiting Fusarium spp. This mold causes mildly superficial to fatally disseminated fungal infections, being the second most common mold causing opportunistic invasive infections after Aspergillus [178]. Keratitis is still the most common infection produced by Fusarium spp. [75,179]. The 2005-2006 outbreak of keratitis due to contact lens infections was attributed partly to the ability of F. keratitis to form biofilms [180,181].
Regarding the type of tests performed, most papers used only in vitro assays. Of them, twelve articles studied the inhibition of cell-to-hyphal transition or cell adhesion, 31 reported inhibition of biofilm formation, and 25 described eradication capacity.
It is worth taking into account that for the antibiofilm compounds to be useful in clinics, it is crucial to determine whether a drug is able to penetrate and eradicate the pre-formed biofilms [102]. The fact that the reported works have not assessed the biofilm eradication capacity but only inhibitory effects on biofilm formation might contribute to a poor correlation between biofilm susceptibility and clinical outcomes.
Only 15 papers deepened the study by giving evidence of the mechanisms of action. Of them, 11 were performed with pure compounds, three with nanosystems, and only one with an extract. It is important to highlight that sometimes the mechanisms of antibiofilm action were demonstrated to be multifactorial [150,182].
Only five of the 42 analyzed papers used in vivo assays, of which four were performed with C. elegans and only one, used rats. This is worrying since animal models of Candida biofilm infections are relevant for identifying novel antifungals useful in clinics [105].

Conclusions and Perspectives
From 2017 to May 2021, there has been an active research on natural products with antifungal biofilm activity. It is clear that great progress has been made and that the newly discovered natural antibiofilm agents could provide novel agents for biofilm-associated infections. Of them, the protein NFAP2 and the highly functionalized polycyclic turbinmicin have demonstrated interesting antibiofilm properties and deserve further research.
However, several types of studies must be deepened. For example, fungal biofilms different from those of Candida spp.
should be prepared and used as targets. Among them, it is necessary to investigate the behavior of natural samples on filamentous fungal biofilms, which are scarcely studied in comparison to Candida spp. such as those of the genera Aspergillus, Fusarium and Trichophyton [183,184].
To be useful for future development, all papers should analyze biofilm eradication capacities in addition to the study of the inhibition of biofilm formation and other properties. Besides, in vivo assays should be included in all papers dealing with activity against biofilms.
It is worth taking into account that several clinical trials involving natural antibiofilm agents are in progress, and some of them exhibited promising perspectives, as recently reviewed by Lu et al. [185].

Materials and Methods
The search for suitable papers was performed in electronic databases by using the following keywords: "biofilm", "fungal infections", "sessile cells", "secondary metabolites", "natural products", "repurposing", "antifungal drugs", and "antibiofilm". Additional papers matching the search criteria were included after surveying the references from the selected articles.
The information gathered was divided into three groups: (I) natural extracts including EOs, propolis and extracts from plants, lichens, algae and cyanobacteria, (II) pure natural compounds, and (III) nanosystems. This last section was sub-divided by EOs included in nanosystems and pure natural compounds included in nanosystems.