An Overview on Conventional and Non-Conventional Therapeutic Approaches for the Treatment of Candidiasis and Underlying Resistance Mechanisms in Clinical Strains

Fungal infections and, in particular, those caused by species of the Candida genus, are growing at an alarming rate and have high associated rates of mortality and morbidity. These infections, generally referred as candidiasis, range from common superficial rushes caused by an overgrowth of the yeasts in mucosal surfaces to life-threatening disseminated mycoses. The success of currently used antifungal drugs to treat candidiasis is being endangered by the continuous emergence of resistant strains, specially among non-albicans Candida species. In this review article, the mechanisms of action of currently used antifungals, with emphasis on the mechanisms of resistance reported in clinical isolates, are reviewed. Novel approaches being taken to successfully inhibit growth of pathogenic Candida species, in particular those based on the exploration of natural or synthetic chemicals or on the activity of live probiotics, are also reviewed. It is expected that these novel approaches, either used alone or in combination with traditional antifungals, may contribute to foster the identification of novel anti-Candida therapies.


Relevance of Candidiasis within the Spectrum of Fungal Infections
In recent years, the number of fungal infections has risen significantly, being today estimated to affect, yearly, around 150 million people and cause 1.5 million deaths [1,2]. These infections range from superficial rushes in the mucosas, in the skin or in the nails, to systemic infections, in which the fungal cells disseminate in the bloodstream and may end up colonizing any major internal organ [1]. Candida species are among the more relevant etiological agents causative of superficial and invasive fungal infections. Vulvovaginal candidiasis (the common name attributed to infections caused by Candida spp. in the vaginal tract) is estimated to affect 70% to 75% of women worldwide, 5% to 8% of these in a recurrent manner [3]. The incidence of invasive candidiasis annually is estimated to be 700,000 infections, with associated mortality rates close to 50%, especially in countries where no adequate antifungal therapy is available [1,2]. Different from other relevant fungal pathogens, such as those belonging to the Aspergillus or Cryptococcus genera, Candida spp. are part of the human commensal microbiota colonizing the skin, the genitourinary or the gastrointestinal tracts [4]. Under certain conditions, such as reduced activity of the host immune system, prolonged use of antibiotics or chemotherapy, these commensal populations can overgrow, triggering more serious (in some cases life-threatening) infections [2]. C. albicans is the more relevant species in causing superficial and invasive candidiasis, but a growing incidence of non-albicans Candida species (usually known as NACS) is reported [5]. C. glabrata, C. tropicalis, C. parapsilosis and C. krusei are among the more relevant NACS, accounting, together with C. albicans, for more than 80% of all described cases of candidiasis [6,7]. The crude mortality rate associated to infections caused by NACS has been reported to exceed those attributed to C. albicans (ca. 37%) reaching in the highest cases~50% for C. glabrata and~59% for C. krusei [8]. This epidemiological shift from C. albicans to NACS is believed to result from a selective pressure caused by the massive utilization of azoles in prophylactic and active treatments that resulted in the selection of species innately more tolerant to these drugs. The use of better diagnosis methods to identify isolates in the clinical setting is another relevant factor as in the past the identification of the Candida isolates may have not been as accurate as it is today [9].

Available Antifungals against Candida spp. and Their Modes of Action
The development of antifungal drugs is limited by the similarity between fungal and human cells, making it therefore difficult to identify molecules that specifically target the microbial cell while not damaging the host. The classes of antifungals available include azoles, polyenes and echinocandins. These target the biosynthesis of ergosterol or the cell wall, two cellular traits absent in mammalian cells ( Figure 1). 5-fluorocytosine, a fluoropyrimidine, is also used to treat candidiasis but in this case the mechanism is more general as it targets DNA synthesis ( Figure 1). A small description of the mechanisms of action of these molecules and the underlying resistance mechanisms is provided in the following subsections.

Polyenes
The better studied and more largely used polyene is amphotericin B which was the first antifungal developed for the treatment of disseminated candidiasis [10]. Nystatin is also used against Candida although only to treat oral infections [11]. The high lipophilicity of polyenes renders them able to penetrate the phospholipid bilayer of the plasma membrane where they bind ergosterol and promote the formation of pores ( Figure 2) [12]. Necessarily, this perturbs the action of the plasma membrane as a selective barrier and a matrix for proteins. Despite its potent effect against Candida, the usefulness of amphotericin B is limited by its nephrotoxicity [13]. Although safer formulations to vehicle this drug have been developed (mostly based on the use of liposomes), its high cost remains an impediment and it is mostly used as a second-line therapy [13]. All Candida species show susceptibility to polyenes but in the case of C. glabrata and C. krusei the use of maximal doses is recommended (Table 1) [13].

Azoles
Azoles comprise the largest family of antifungals used against Candida. The first azoles used in clinical practice were clotrimazole and miconazole that were approved for use in 1969, followed by ketaconazole in 1981 [14]. These three drugs are all imidazoles since they harbour an imidazole ring in their structure (as shown in Figure 1). The usefulness of clotrimazole and miconazole as antifungals was limited by their inhibitory effect on the human hepatic CYP enzymes [15]. As a response to that, in the early 90s the triazoles fluconazole and itraconazole were introduced in the market, showing improved pharmacokinetic profile, a broader spectrum of antifungal activity and a lower inhibitory effect against the human CYP450 system [14]. In the early 2000s voriconazole emerged, being advantageous by showing higher activity against the more azole-resilient NCAS species, compared to fluconazole or itraconazole [14]. Today imidazoles are mostly used for the treatment of superficial candidiasis, while triazoles are preferred for the treatment of invasive candidiasis [14,16]. Regardless of the family they belong to, azoles act by inhibiting the activity of the lanosterol-14α-demethylase enzyme (encoded by the ERG11 gene) that is involved in ergosterol biosynthesis. As a result of this inhibition, azole-exposed fungal cells accumulate toxic sterols in the plasma membrane dramatically affecting its permeability ( Figure 2) [12]. C. glabrata and C. krusei show less susceptibility to azoles than the remaining Candida spp. and higher doses are recommended to treat infections caused by these species (Table 1).

Fluoropyrimidines
The fluoropyrimidine more commonly used in the treatment of candidiasis is 5-flucytosine (5-FC), which enters fungal cells through cytosine transporter(s) being afterwards metabolized via the pyrimidine salvage pathway to 5-fluorouracil (5-FU), considered the active form of 5-FC ( Figure 2). 5-FU incorporates in RNA, causing premature chain termination, and inhibits the activity of thymidylate synthase, an enzyme essential for DNA synthesis ( Figure 2) [12,17]. With the exception of C. krusei, the remaining Candida spp. are susceptible to 5-FC (Table 1). Although the enzymes that drive conversion of 5-FC into 5-FU are not present in mammalian cells [12], bacteria living in the human gut were shown to efficiently convert 5-FC into 5-FU [17] thereby explaining the toxic effects reported in patients under 5-FC therapy. Due to its toxic effects, 5-FC is given to patients in low concentration and in combination with other antifungals [13].

Echinocandins
Echinocandins are the only new class of antifungals discovered in recent years [1]. These are commercially available in three forms: caspofungin, anidulafungin and micafungin. Two more recent molecules, rezafungin and biafungin, have been recently described but its use in the clinical setting is not yet established as their efficacy is still under assessment in clinical trials. Compared to the already available echinocandins, rezafungin and biafungin show higher activity, lower toxicity and fewer drug interactions [18][19][20]. Echinocandins act by inhibiting the catalytic subunits of β-(1,3)-d-glucan synthase, essential for cell wall synthesis. Consequently, no elongation of (1,3)-β-d-glucans is observed in fungal cells exposed to echinocandins, rendering them highly susceptible to lysis ( Figure 2) [12]. Echinocandins show efficacy against all Candida species, although C. parapsilosis has been found to be intrinsically less susceptible [13]. Due to their safety profile and fungicidal activity, echinocandins are frequently used as the primary treatment of invasive candidiasis [13]. Table 1. General susceptibility patterns of Candida species to antifungal drugs used in the treatment of candidiasis (adapted from [21,22]

Incidence of Antifungal Resistance and Underlying Mechanisms
In recent years, the number of resistant strains among Candida increased prominently, especially among NACS [23][24][25] (Table 2). Among the different antifungal classes, the highest percentage of resistance is observed for azoles, as detailed in Table 2. It is believed that this growing emergence of resistance to azoles is linked to the massive use of fluconazole in prophylaxis of patients considered at risk of suffering an infection caused by Candida [26][27][28][29]. The use of agricultural fungicides structurally similar to clinical azoles exerted another layer of pressure for the selection of more azole-tolerant Candida strains [30,31]. Although resistance to echinocandins and amphotericin B is very low, a slight, but detectable, increase in the emergence of resistant strains has been observed for C. glabrata and C. krusei [32] (Table 2). Table 2. Percentage of isolates among the five more prevalent Candida spp. exhibiting resistance to azoles, echinocandins, flucytosine or amphotericin B (amphoB), as reported by surveillance epidemiological studies [22,[32][33][34][35][36][37][38][39]. ** The high range of percentages found for C. krusei results from this species showing highly divergent susceptibilities to different imidazoles or triazoles (e.g., most strains are largely resistant to fluconazole but susceptible to voriconazole).

Molecular Mechanisms Underlying Resistance to Antifungals in Clinical Strains
In this section the main mechanisms behind resistance of clinical Candida strains to the different classes of antifungals will be described. In general, these mechanisms of resistance can be summarized as involving the evolution of adaptive responses aiming to counteract the deleterious effects of the antifungal (e.g., reducing drug efficacy by changing the target) or to reduce the internal concentration of the drug (e.g., through the overexpressing drug-efflux pumps). The mechanisms already characterized as underlying resistance to to azoles in clinical isolates were gathered in Table 3, while those conferring tolerance to echinocandins, polyenes or 5-FU in clinical strains are detailed in Table 4.

Azoles
Resistance to azoles in Candida has been largely associated to modifications or overexpression of the drug target Erg11, modifications in the ergosterol pathway or overexpression of genes encoding drug-efflux pumps (Table 3). Numerous single nucleotide polymorphisms (SNPs) were reported to occur in the azole-target enzyme Erg11 encoded by C. albicans, C. krusei or C. tropicalis, it being thought that these mutations reduce the inhibitory effect of the azole over the enzyme [40][41][42][43][44]. Overexpression of ERG11 has also been described as a mechanism driving resistance in C. albicans, C. parapsilosis and C. tropicalis isolates [41,42,45,46]. The higher transcription of ERG11 in these azole-resistant isolates has been shown to result from these strains upregulating or encoding hyperactive forms of the Upc2 transcription factor, a strong positive regulator of ERG11 gene [47][48][49] (Figure 3). Differently, the CgERG11 allele encoded by C. glabrata azole-resistant isolates is, in the vast majority of the cases, identical to the one encoded by susceptible strains [50][51][52][53]. No link between the overexpression of CgERG11 and increased resistance to azoles could also be established in C. glabrata [51,53,54] suggesting that this species has evolved responses to handle azole stress distinct from those verified in C. albicans or C. parapsilosis.
The induction of the activity of drug-efflux pumps has been observed in several azole-resistant isolates belonging to the different Candida species [42,46,49,[55][56][57][58][59][60][61][62][63][64][65][66][67][68][69][70]. The more studied drug efflux pumps linked to azole resistance are those belonging to the ATP-binding cassette (ABC) superfamily which include in C. albicans CaMdr1, CaCdr1 and CaCdr2 [55,56,71]; in C. glabrata, CgCdr1, CgCdr2 and CgPdh1 [59][60][61]; in C. krusei, CkAbc1 and CkAbc2 [69]; in C. parapsilosis CpCdr1 and in C. tropicalis CtCdr1 [42,46]. More recently, multi drug resistance (MDR) transporters belonging to the Major Facilitator Superfamily (MFS) have also been implicated in tolerance of different Candida species to azoles including CaMdr1 in C. albicans, C. parapsilosis and C. tropicalis [42,46,49] and CgTpo1_1, CgTpo3 and CgQdr2 in C. glabrata [72]. Although the influence of these transporters in mediating resistance in clinical isolates has not been studied at the same extent as those of the ABC superfamily, promising results had been obtained in a recent study showing a positive correlation between the expression of the C. glabrata CgAqr1, CgTpo1_1, CgTpo3 and CgQdr2 MFS-MDR transporters and resistance to clotrimazole [70]. In this study it was also shown that the deletion of CgTPO3 abolishes resistance to clotrimazole in one of the identified resistant clinical isolates [70]. The model that is generally accepted to explain the positive effect of the ABC and MFS transporters in drug resistance is their role in directly mediating the extrusion of the drugs, however, from the biochemical point of view this model is difficult to accept considering the wide structural divergence of the hypothesized substrates, [73]. Indeed, more recent studies performed in the eukaryotic model yeast S. cerevisiae show that ABC and MFS-MDR transporters have physiological substrates whose transport may affect the partition of the drugs between the intra-and the extracellular environment, [73]. Specifically, some MDR transporters have been shown to influence the lipid composition of the plasma membrane, by promoting the transport of phospholipids and/or ergosterol, which thereby may affect the diffusion rate of the drugs across the membrane, [74]. It was recently shown that deletion of the poorly characterized C. albicans ABC transporter CaRoa1 results in increased membrane rigidity and, consequently, in a reduced intracellular concentration of azoles [75]. Further studies in Candida spp. are required to clarify whether the observed positive effect of ABC and MFS-MDR transporters in reducing internal concentration of azoles is exerted directly or indirectly, via the transport of another physiological substrate and the relevance of these mechanisms in driving resistance in clinical isolates.
In all cases described so far, the higher activity of MDR pumps is linked to their higher expression in the azole-resistant isolates [72]. In C. albicans and in C. glabrata the transcriptional regulation of these drug-efflux pumps is under a tight control of the pleiotropic drug resistance network (or PDR) that in C. glabrata is dependent of the CgPdr1 regulator [76] while in C. albicans is controlled by CaTac1 [77] ( Table 3). Further studies, exploring gene-by-gene or genome-wide approaches, have implicated other regulators in the transcriptional regulation of drug-efflux pumps or ergosterol metabolism under azole stress including CaMrr1 and CaCap1 in C. albicans [78,79]; CgStb5 in C. glabrata [80], Upc2 in C. albicans, C. tropicalis, C. parapsilosis and C. glabrata [48,49,57,81] and CpTac1 and CpMmr1 in C. parapsilosis [41,82]. The knowledge gathered on the regulatory associations between known transcription factors involved in azole resistance and MDR pumps is briefly summarized in Figure 3. In the case of the less-studied species C. tropicalis and C. krusei, the regulators of the identified drug-efflux pump-encoding genes are not yet identified. Nonetheless, similarity searches revealed that these species encode proteins showing similarity to CaTac1 (CTRG_05307 in C. tropicalis) and to CaMrr1 (CTRG_02269 in C. tropicalis and JL09_3889 in C. krusei) [83]. . Schematic representation of regulatory associations between regulators involved in azole resistance and genes encoding multidrug resistance efflux pumps demonstrated to be involved in azole resistance in Candida spp. The information concerning the regulatory associations between transcription factor and target genes was retrieved from the PathoYeastract database [84]. Although ABC-MDR and MFS-MDR transporters involved in azole resistance in C. krusei and C. tropicalis had been identified, until so far the regulators of these genes remain to be characterized.
The overexpression of drug-efflux pump-encoding genes results, in many cases, from the occurrence of activating mutations in the coding sequence of the corresponding regulators [41,45,78,82,[85][86][87][88][89]. This type of mechanism has been documented for CgPdr1 in C. glabrata; for CaTac1, CaMrr1 and CaUpc2 in C. albicans and for CpMrr1 and CpTac1 in C. parapsilosis [41,45,78,82,85,88]. An important feature of these "hyper-active" alleles is that they become active even when azoles are absent [78,85,88]. Interestingly, it was recently shown that C. albicans strains harboring CaTac1 gain-of-function alleles exhibit a decreased fitness in vivo, specially when challenged with stresses other than azoles [90]. It thus seems that specialization of the cells to improve azole stress at the expense of CaTac1 hyper-activation results in reduced capacity to handle unrelated stresses. In the same line, the expression of CgPdr1 gain-of-function alleles were also hypothesized to be linked with a reduced tolerance of C. glabrata to organic acids [51,68,89,91].

Flucytosine, Echinocandines and Polyenes
Acquired resistance to polyenes in clinical isolates is rare and the few studies correlate that phenotype with a reduction of ergosterol content in the plasma membrane of the resistant isolates [50,[92][93][94][95][96]. These events are generally associated with the occurrence of SNPs that inactivate genes of the ergosterol biosynthetic pathway and thereby alter the sterol content of the membrane, this being described in C. albicans, C. glabrata and C. tropicalis (detailed in Table 4) [50,[92][93][94][95][96]. No significant link between the activity of drug-efflux pumps and resistance to polyenes has been identified in resistant isolates belonging to the different Candida species. Although amphotericin B-resistant isolates had been identified in C. krusei and C. parapsilosis [97,98], the underlying mechanisms remain to be disclosed ( Table 2). The echinocandin-resistance phenotype exhibited by the small number of identified resistant Candida isolates was attributed to mutations in the β-1,3-glucanase-encoding genes FKS1 and FKS2 genes [99][100][101][102][103]. These mutations are thought to reduce the sensitivity of the proteins to the drug [100] (Table 4) [99][100][101][102][103]. The naturally high tolerance to echinocandins of C. parapsilosis as well as of the closely related species C. orthopsilosis and C. metapsilosis, was also suggested to result from these species encoding a CpFKS1 allele less sensitive to echinocandins [104]. Up to now, increased activity of drug-efflux pumps has not been identified as a relevant mechanism by which clinical isolates acquire resistance to echinocandins. Concerning flucytosine, resistance in clinical isolates has been linked to modifications on the coding sequence of the Fcy cytosine permease or in the uracil phosphotransferase Fur1 (Table 4 and Figure 1) [105][106][107][108][109][110]. Resistance of some C. tropicalis isolates was linked to the emergence of mutations in CtURA3 gene, encoding the enzyme involved in the metabolization of UMP, the natural substrate of thymidylate synthase ( Figure 1) [108,109]. It is thought that this mutation increases the synthesis of UMP compensating for the loss of this metabolite that will occur with formation of 5-FdUMP ( Figure 1) [105]. Overexpression of CkABC1 and CkABC2 Overexpression of CpMDR1 and CpCDR1 Increased activity of CpUpc2 Overexpression of CtCDR1 and CtMDR1

Antifungal Resistance Driven by Large-Scale Genomic Alterations
A recent genomic analysis has unveiled an important role for the inactivation of the CgMSH2 gene as a driver of resistance to azoles, echinocandins and amphotericin B in C. glabrata while colonizing the host [129]. The CgMSH2 gene encodes a protein involved in DNA repair and its inactivation (promoted by frameshift mutations in the coding sequence) leads to increased genetic diversity in the C. glabrata population. As such, isolates harbouring inactive CgMSH2 alleles rapidly acquired resistance to azoles, echinocandins or amphotericin B resulting from the rapid acquisition of beneficial mutations in CgPdr1, in CgFks1 or CgFks2 or in CgErg6 [129]. After this pioneering work, several epidemiological studies have focused their attention on the prevalence of resistant strains harbouring inactivated CgMSH2 alleles, the percentages observed ranging between 5% and 17% [129][130][131][132]. Around 50% of the susceptible isolates examined in these studies were also found to harbour inactivated CgMSH2 alleles [129][130][131][132], suggesting that this mechanism does not per se assure antifungal resistance. Deletion of the C. albicans CaMSH2 gene was also found to result in drug resistance [133], however, up to now this mechanism has not been described to underlie the resistance phenotype in clinical isolates. The genomic plasticity exhibited by C. albicans and C. glabrata has also been found to contribute to increased drug resistance in these species. In specific, in azole-resistant clinical C. glabrata isolates it has been described the duplication of chromosomes that include the CgCDR1, CgPDH1 or CgERG11 genes, as well as the formation of mini-chromosomes harbouring several copies of genes encoding CgCDR1 or CgPDH1 [124,134]. The diploid nature of C. albicans has also been found to underlie the appearance of hyperactive alleles of genes involved in azole-resistance (e.g., CaERG11 and CaTAC1) [120,[135][136][137]. More recently, mis-translation of serine tRNAs in leucine at CUG codons, a well known specific trait of C. albicans, has also been linked with an accelerated resistance rate to fluconazole [137,138] while loss of heterozygosity was reported to underlie resistance to flucytosine in C. tropicalis [139].

Novel Approaches for the Development of Anti-Candida Agents
The persistent increase in the emergence of strains resistant to currently used antifungals has been paving the way for the development of new approaches that can be used to prevent growth of Candida spp. and that can be further considered as interesting alternatives as new anti-Candida therapies. The main results obtained in these different approaches are described in the following sections, together with a discussion on what are the current challenges or limitations in knowledge that still persist.

Phytotherapeutics
Systematic testing of compounds from natural sources including substances/extracts produced by animals, plants or microorganisms, have resulted in the identification of many molecules that inhibit growth of Candida cells. When produced by plants these substances are named phytotherapeutics, these being attractive since they are naturally perceived by consumers as less toxic and safer than common pharmaceuticals [140]. Currently, there is an increasing number of phytotherapeutics being identified as efficient against Candida species including extracts isolated from garlic (Allium sativum L., Tulbaghia alliace or Tulbaghia violacea), coconut (Cocos nucifera) or virgin coconut oil, mint (Mentha piperita L.) or sage (Salvia officinalis L.) [140,141]. A frequent limitation of this type of approaches is the difficulty in isolating the molecules responsible for the observed inhibitory effect over Candida since frequently these natural extracts are complex and used without further processing.

Redesign of "Old Antifungals"
The "redesign" of common antifungals is also an approach that has been explored to obtain molecules with inhibitory potential against Candida. The most paradigmatic examples are the new formulations of amphotericin B, which include lipid-associated and liposomal formulations showing higher fungal targeting and reduced toxicity against the host [142]. Within the same line azole-like molecules have also been obtained showing increased antifungal potency against all Candida species, compared to the efficacy exhibited by fluconazole [143]. These modified azoles, named ATTAF-1 and ATTAF-2, share general structural features with triazole alcohols, however, their mode of action appears to differ from the one of fluconazole which is an important trait to sensitize resistant strains [143]. Further on, Shrestha et al. (2017) developed a series of 27 alkylated variants of fluconazole, some of which presented a low hemolytic activity, low cytotoxicity and strong inhibitory potential against several Candida species [144]. Although the range of minimal inhibitory concentrations obtained was fairly wide, these compounds proved efficient against both C. albicans and non-C. albicans species and were observed to target the ergosterol biosynthetic pathway by inhibiting the sterol 14a-demethylase enzyme instead of targeting the ERG11 gene [144].

New Compounds Obtained by Chemical Synthesis
The synthesis of entirely new compounds obtained by chemical synthesis, either or not involving metallic elements, has also been largely explored to obtain compounds with anti-Candida potential. Table 5 provides a systematic overview on a large cohort these "new chemicals". Many of those new chemicals have silver in their structure, which is interesting since silver has been used since the times of ancient Greece as an antimicrobial. Examples of the Ag-containing compounds synthesized include those containing camphorimines, tetraazoles, albendazoles or phenantrolines as ligands (Table 5) [145][146][147][148]. Complexes with other metal centers like copper, cobalt, nickel or iron; or even with metals not usually used as bioagents, such as tin, chromium, cadmium or lead [149][150][151][152][153][154][155][156][157], have also been synthesized and shown to display moderate anti C. albicans activity (Table 5). Polinuclear complexes (based on Cu, Cd or Ni), particularly those harbouring ferrocenyl derived ligands, were also reported to have high activity against C. albicans (Table 5) [158]. More recently the use of a Ru(III) perylene complex has also been reported to be interesting as anti-Candida agent through photodynamic inactivation [159]. Although some of these complexes revealed a marked potential to constitute novel anti-Candida agents, their mechanism of action remains elusive in most cases, being also necessary to investigate their ability to inhibit growth of strains that are resistant to currently used antifungals. Another aspect of relevance is the fact that in the majority of the studies performed the compounds were not tested against NACS or against clinical strains that are, in general, more difficult to inhibit than laboratory strains.

Nanoparticles
Considering the recent interest in the use of nanoscale materials as antimicrobial agents, due to their high surface area to volume ratio that gives them unique chemical and physical properties [160], a number of studies have focused on the development and exploration of silver nanoparticles (AgNPs) as anti-Candida agents [161][162][163]. In these studies, silver nanoparticles are synthesized using organic or inorganic reductive agents (e.g., silver nitrate or citrate) [161,162] which promote the formation of metallic silver (Ag 0 ), followed by agglomeration into oligomeric clusters that eventually result in the formation of metallic colloidal silver particles [164]. The exact mechanism by which AgNPs exert toxicity against Candida spp. remains a bit elusive, although evidence has been obtained suggesting that they may perturb the cellular envelope causing a disruption of the plasma membrane potential and consequent damage and leakage of cell constituents [163]. Concerning this matter, an interesting result was obtained with camphorimine-based complexes, being demonstrated that C. albicans, but not C. parapsilosis, C. tropicalis or C. glabrata, were able to mediate the conversion of Ag(I) into AgNPs [145].

Use of Probiotics and Antimicrobial Peptides
For a long time, it has been known that the use of probiotics can be beneficial for the treatment of mucosal candidiasis, specially, for vaginal candidiasis. In this sense, a few products are currently available in the market mostly based on the use of lactobacilli, considering the well-known track record of these species as probiotics [165]. A few examples of these products are described in Table 6.
Probiotics are defined as "live microorganisms that, when administered in adequate amounts, provide a health benefit on the host" [166]. In the vaginal tract, the microbiota is largely dominated by lactobacilli, with L. gasseri, L. jensenii and L. crispatus being among those most abundant [167,168]. A decreased abundance in these microbial species appears to correlate with increased activity of pathogens, including of C. albicans and C. glabrata [169,170]. These results support the long-standing use of probiotics in the treatment of vaginal candidiasis. The mechanisms by which these lactobacilli species inhibit growth of pathogens, and of Candida in particular, remains to be elucidated, as well as the genes that mediate this interaction. Nonetheless, the evidence gathered so far (obtained using lactobacilli species differing from those that are indigenous of the vaginal tract) suggest that production of lactic-acid-concomitant acidification of the vaginal pH is on the basis of the protective effect of lactobacilii against vaginal pathogens [171]. Although this can be hypothesized for bacteria that are generally sensitive to low pH, this is not the case of yeasts that grow very well under acidic pHs. Indeed, a recent study performed with the supernatant of vaginal lactobacilli species (L. crispatus, L. gasseri and L. vaginalis) showed no significant correlation between the amount of lactic acid present and the inhibition of Candida [172] and concentrations of lactic acid similar to those found to be present in the vaginal tract (even under conditions of eubiosis) were also found not to significantly affect growth of C. albicans or C. glabrata [173,174]. It thus remains to be established what is the contribution that lactic acid production may have in the inhibition of Candida and of other vaginal pathogens. Other mechanisms by which vaginal lactobacilli are hypothesized to control the activity of Candida species in the vaginal tract is by competing for adhesion sites in the epithelial cells, by secreting biosurfactants that may decrease fungal binding to host surfaces and by secreting to the environment hydrogen peroxide and bacteriocins [171]. Interestingly, it was recently shown that invasive candidiasis from the gut can be restrained by commensal bacteria [175] which opens the door to the development of probiotics not only for the treatment of vaginal candidiasis but also for patients that may be at a high risk of developing systemic candidiasis caused by commensal Candida populations found in the gut such as those subjected to massive invasive surgeries.

Conclusions
Although extensive knowledge about the molecular mechanisms by which Candida spp. surpass the deleterious effects of antifungals has been collected, the translation of that knowledge to the understanding of which of these mechanisms play a role in the stressful environment of the host is still limited. In this review, we aimed at providing that picture, focusing what is actually described to mediate resistance in clinical isolates. The modification of the drug target and the overexpression of genes playing a detrimental role in antifungal tolerance determined by the adjustment of regulatory circuits (through modification of pivotal regulators in drug resistance such as CaTac1 or CgPdr1) and/or the occurrence of chromosomal rearrangements, comprise the vast majority of what is known to mediate antifungal tolerance in resistant isolates. However, there is still a road to pursue in this since the resistance of several resistant isolates cannot be explained by these mechanisms strongly suggesting that other antifungal-resistance genes remain to be identified. It is possible that the difficulty in mimicking in the laboratory the stressful environment of the host complicates the identification of these genes and, in this field, it is expected that extensive genomic analyses of resistant isolates may help to shed light on this. The full clarification of this panoply of resistance genes and mechanisms is essential not only to improve the success of treatments and improve the outcomes of candidiasis, but also to develop more efficient diagnosis tools that could rapidly provide clinicians a fast response on how to fine-tune treatments. It also seems clear that the development of non-conventional therapies, focused on biological targets other than those that are targeted by already used antifungals, is essential considering the persistent increase in the emergence of strains resistant to azoles and, less significantly, to echinocandins. Although much has been done in this field and promising results had been obtained, especially in the identification of new chemicals showing a robust anti-Candida effect, it remains to be established in many cases if indeed these compounds are able to sensitize antifungal-resistant isolates, and what their spectrum of activity against NACS is. In almost all cases it is also lacking the characterization of the toxicological effects of these drugs/compounds/probiotics in mammalian cells as well as their pharmacokinetic profile. Further investigation in this field is therefore essential to assure that alternative antifungals will be provide to the community in the mid-term.
Author Contributions: This review was conceptualized by N.P.M. that also coordinated the writing of the overall manuscript together with all the remaining authors. S.B.S. and R.S. collaborated in the collection of information and writing of the Sections 1-3, while N.A.P., M.F.N.N.C. and M.J.P. compiled information and collaborated in the writing of Section 4. All authors have read and agreed to the published version of the manuscript.
Funding: Funding received from FCT through its support to iBB-Institute for Bioengineering and Biosciences and to Centro de Química Estrutural (contracts UID/BIO/04565/2013; UIDB/04565/2020 and UID/QUI/00100/2013) and to the project LactoCan (PTDC/BIA-MIC/31515/2017) is acknowledged by sustaining the activities of the group in the study of new therapeutic approaches for the treatment of infections caused by Candida. Programa Operacional Regional de Lisboa 2020 is also acknowledged for its financial support to iBB (project no. 007317). The doctoral programs Biotecnico and Applied and Environment Microbiology are also acknowledged for their support to PhD students SBS and NAP (grants PD/BD/135139/2017 and PD/BD/143026/2018).

Conflicts of Interest:
The authors declare no conflict of interest.