Monoclonal Antibodies as Tools to Combat Fungal Infections

Antibodies represent an important element in the adaptive immune response and a major tool to eliminate microbial pathogens. For many bacterial and viral infections, efficient vaccines exist, but not for fungal pathogens. For a long time, antibodies have been assumed to be of minor importance for a successful clearance of fungal infections; however this perception has been challenged by a large number of studies over the last three decades. In this review, we focus on the potential therapeutic and prophylactic use of monoclonal antibodies. Since systemic mycoses normally occur in severely immunocompromised patients, a passive immunization using monoclonal antibodies is a promising approach to directly attack the fungal pathogen and/or to activate and strengthen the residual antifungal immune response in these patients.


Introduction
Fungal pathogens represent a major threat for immunocompromised individuals [1]. Mortality rates associated with deep mycoses are generally high, reflecting shortcomings in diagnostics as well as limited and often insufficient treatment options. Apart from the development of novel antifungal agents, it is a promising approach to activate antimicrobial mechanisms employed by the immune system to eliminate microbial intruders. Antibodies represent a major tool to mark and combat microbes. Moreover, monoclonal antibodies (mAbs) are highly specific reagents that opened new avenues for the treatment of cancer and other diseases. This review provides an overview on studies in which mAbs have been used to combat experimental fungal infections caused by pathogenic yeasts, (Candida, Cryptococcus), dimorphic fungi (Histoplasma, Paracoccidioides, Sporothrix), or molds (Aspergillus, Rhizopus, Scedosporium).

Elimination of Microbial Pathogens by Antibody-Dependent Mechanisms
The antibody-antigen binding is a highly specific interaction that can directly modulate the biological activity of a target molecule, e.g., by neutralization of a toxin. Apart from secreted molecules, antibodies can also inhibit microbial surface proteins, such as adhesins or surface-bound enzymes. Surface-reactive antibodies can furthermore act as opsonins and thereby mark microbes out for destruction. Fcγ receptors reside in the cytoplasmic membrane of phagocytes and recognize bound IgGs. In concert with the parallel recognition of conserved microbial structures by dedicated pattern recognition receptors, this boosts phagocytosis, enhances phagosome-lysosome fusion, and results in a more efficient microbial killing [2]. Bound IgM or IgG can furthermore recruit complement proteins to activate this part of the innate immune response resulting in an enhanced C3 receptor-mediated phagocytosis. Moreover, antibodies can have a catalytic activity; as exemplarily shown by Bowen et al. [3]: Two mAbs directed against glucuronoxylomannan (GXM) the major

Protective Antibodies against Cryptococcus neoformans
Cryptococcus neoformans is a major yeast pathogen that is unique among medically important fungi in its possession of a polysaccharide capsule. While infections of healthy individuals usually remain asymptomatic, hosts with a severely impaired cellular immunity can develop life-threatening, disseminated infections and meningitis. In contrast to C. albicans, Cryptococcus does not form hyphae during infection making it a seemingly easier target for an antibody-based therapy.
The Cryptococcus polysaccharide capsule is a crucial virulence determinant with GXM being its major component. As for certain bacteria, the capsule prevents recognition by pathogen recognition receptors and thereby protects the fungus from phagocytes. However, as for capsulated bacteria, this can be overcome by antibody-mediated opsonization.
The first report of a mAb providing protection against experimental cryptococcosis dates back to 1987 [4]. In this pioneering study, Dromer and co-workers used a GXM-specific IgG 1. Several years later, a similar protective activity was reported for a GXM-specific IgM [5]. Several studies directly compared GXM-specific mAbs belonging to different (sub)classes (Table 1A). The IgG 3 subtype turned out to be less protective or even deleterious, whereas mice immunized by administration of IgG 1 , IgA or IgM antibodies showed an improved outcome [6,7]. In vitro experiments revealed no difference in the opsonizing activity of the different isotypes [8]. Yuan et al. provided evidence that the IgG 1 -mediated protection and the deleterious effect of IgG 3 depend on CD4 + -and CD8 + -T cells, respectively [9]. A non-protective IgG 3 could be converted into a protective IgG 1 by isotype switching, indicating that the IgG 3 subclass is a crucial determinant in this context [10]. Further studies implicated distinct Fcγ-receptor functions [6], the genetic background of the infected mice [11], and distinct catalytic activities [3] in the strikingly different biological activities of these GMX-specific IgG 1 and IgG 3 switch variants.
IgM antibodies to GMX can be either protective or non-protective, which depends on a variety of factors, e.g., the route of infection, the size of the inoculum, the amount of mAbs administered, and the ability of these antibodies to promote phagocytosis [12]. Further studies showed that protection requires binding to certain GXM epitopes [13][14][15]. Shapiro et al. showed that protection mediated by GXM-specific IgM antibodies is independent of complement component C3 indicating that complement fixation is not required [16].
Remarkably, high GXM-specific titers can also cause deleterious effects in mice and this was attributed to the formation of antibody-antigen complexes. Depending on the antibody titer and the inoculum, this antibody-mediated acute lethal toxicity (ALT) can be induced by different IgG subclasses [17]. Cryptoccoccus infections are often chronic and released capsular polysaccharides can accumulate to very high levels in tissue and serum. Antibody induced ALT depends on the antigen concentration in the blood and the isotype of the antibody. Data of two groups indicate that IgG 1 , IgG 2a , and IgG 2b can be deleterious, while IgG 3 , IgM, and IgA lack this harmful activity [18,19]. ALT is triggered by the murine IgG 1 2H1, but is not induced by a mouse-human chimeric IgG 2 derived from 2H1 suggesting that Fcγ-receptor binding is crucial for this toxic effect [20].

Protective Antibodies against Candida albicans
The first evidence that antibodies are important during candidiasis came from the finding that patients who survived systemic infections developed strong antibody responses to certain C. albicans proteins, whereas patients who succumbed to infection had no, minor or fading responses [21]. An immunodominant 45 kDa polypeptide was described and later on identified as a fragment of the heat shock protein Hsp90 [22]. Hsp90 is a highly conserved ATP-dependent molecular chaperone that stabilizes other molecules, governs morphogenesis, and is regarded as a key regulator of Candida virulence traits [23]. An IgG raised against C. albicans Hsp90 was the first mAb that was successfully tested in a murine model of systemic candidiasis [24]. Based on this murine immunoglobulin, a humanized, single chain antibody was developed, initially designated Efungumab, but later on renamed to Mycograb. This recombinant antibody is assumed to inhibit Hsp90 activity by binding to a central domain of Hsp90 that is responsible for the conformational change triggered by ATP binding [25]. Hsp90 is normally cytoplasmic, but to a certain extent, also a surface-bound protein [26]. It plays an important role in several stress responses including those triggered by antifungals. Accordingly, in vitro studies demonstrated a synergistic activity of Mycograb and antifungals, such as fluconazole, caspofungin, and amphotericin B. A clinical trial revealed that Mycograb plus lipid-associated amphotericin B produced significant clinical improvement for patients suffering from invasive candidiasis [27], but despite these promising results, marketing authorization was disapproved by the European Medicines Agency in 2017 based on concerns that the benefits of this treatment do not outweigh its risks.
A different approach was taken by Torrosantucci et al. [28], who coupled the β-glucan laminarin to the diphtheria toxoid and obtained polyclonal antibodies that defended mice against infections caused by C. albicans or A. fumigatus. A monoclonal β-glucan specific IgG2b antibody obtained by this approach and designated 2G8 turned out to be protective against C. albicans, A. fumigatus, and C. neoformans infections [29,30]. Remarkably, an IgM harboring the same complementarity-determining region as 2G8 was not protective [30]. Further experiments revealed that the IgG2b was highly specific for β-1,3-glucan and showed a much stronger reactivity with β-glucan molecules that are released by C. albicans than the corresponding IgM. These distinct specificities may explain the strikingly different protective potential of both antibodies. Interestingly, the IgG 2b was also reported to inhibit growth of C. albicans and A. fumigatus in vitro, but the precise mode of action was not determined. More recently, a mouse-human chimera and a scFv-Fc derived from 2G8 were shown to promote killing of C. albicans by isolated neutrophils and to protect mice in a vulvovaginal model of infection [31]. In 2019, Matveev et al. reported that an IgG 1 specific for β-1,3-glucan delayed germination of A. fumigatus conidia and improved survival of mice infected with C. albicans via the intravenous route [32].
In a series of papers, Han, Cutler, and colleagues characterized two IgM mAbs directed against different C. albicans cell wall components. In immunofluorescence, both antibodies recognized the yeast, but not the hyphal form. B6.1, which is specific for β-1,2-linked mannotriose, stained Candida yeast cells more homogenously than B6 [33]. Both B6.1 and B6 protected mice against vaginal C. albicans infections [34], whereas only B6.1 provided protection in a model of disseminated candidiasis [35]. An IgG 3 mAb also recognizing β-1,2-linked mannotriose was later on shown to be protective in both the disseminated and the vaginal infection model and this was attributed to the strong complement binding mediated by IgG 3 immunoglobulins [36].
Using a mannan-specific humanized IgG 1 antibody, Zhang et al. observed enhanced phagocytosis of C. albicans by murine macrophages, increased deposition of complement component C3, and protection of mice from an otherwise lethal dose of C. albicans yeast cells [37]. Moreover, generation of recombinant switch variants of this antibody revealed that an IgG 2 variant was less protective than the corresponding IgG 1 , IgG 3 , or IgG 4 immunoglobulins [38].
A fully humanized IgG 1 specific for β-1,6-linked poly-N-acetyl-D-glucosamine, a capsular antigen of several bacterial pathogens, cross-reacts with C. albicans and protected mice in a Candida keratitis model [39]. Kavishwar and Shukla described another protective antibody that belongs to the IgA isotype and binds to glycosyl moieties of C. albicans proteins [40].
Several other studies analyzed the impact of antibodies directed against different Candida surface proteins. In these experiments, IgG 1 , IgG 3 , and IgM mAbs provided protection in different models of infection [41][42][43][44]. One of these mAbs, designated C7, is directed against the Als3 mannoprotein, which has multiple functions, e.g., as an adhesin and invasin [45]. In vitro studies with this IgM demonstrated a direct growth inhibitory activity [46] that was later on attributed to an antibody-mediated inhibition of fungal iron acquisition [47].
Polyclonal rabbit antibodies directed against the surface-bound Hyr1 protein of C. albicans were also shown to be protective in a murine model of infection [48] and more recently, Rudkin et al. characterized several recombinant IgG 1 mAbs specific for Hyr1 and other C. albicans surface molecules that were derived from B cells isolated from human patients [49]. These mAbs enhanced phagocytosis of C. albicans yeasts and short hyphae by murine macrophages and protected mice from a systemic C. albicans infection. Probably, these human IgG 1 antibodies interact with murine Fcγ receptors and thereby boost the antifungal activities of macrophages and neutrophils. While the Hyr1-specific antibodies recognized exclusively C. albicans, the mAbs directed against other surface proteins were also reactive with other members of the genus Candida and may therefore possess a broader therapeutic potential [49].
The minor role of secreted proteins in fungal virulence in mammals has been mentioned above. However, de Bernardis et al. showed that antibodies against the secreted aspartic protease Sap2 are protective in a rat model of vaginitis, thereby demonstrating that protection is not restricted to mAbs directed to surface antigens [41]. Since Sap2 plays an important role in vaginal infections caused by C. albicans [50], it is conceivable that the protective activity of this mAb is due to inhibition of the proteolytic activity of Sap2.
Killer toxins (KTs) have been described for Saccharomyces spp., Pichia spp., and other non-pathogenic yeasts. These short, secreted proteins bind to the surface of sensitive fungi and kill them through different effector mechanisms [51]. According to the concept of anti-idiotypic antibodies, an antibody raised against an immunoglobulin specific for the active site of a particular enzyme can possess the enzymatic activity of this enzyme. Using this approach, anti-idiotypic recombinant antibodies were generated that mimic the antifungal activity of a KT derived from Wyckerhamomyces anomalus (formerly Pichia anomala). This recombinant antibody killed C. albicans in vitro and provided protection in a rat model of C. albicans vaginitis [52]. Interesting features of KT-like antibodies are their direct antifungal activity and their target structures that are often conserved in many fungal pathogens.  [80] In In summary, most antibodies that are protective against C. albicans or C. neoformans infections recognize different glycostructures or surface proteins (Table 1A,B). They are either IgMs or belong to one of the four IgG subclasses. Many protective antibodies were shown to enhance the phagocytic uptake and stimulate the phagolysosomal maturation process (Table 2). In all C. albicans protection experiments that mimic a systemic infection (Table 1B), the yeast form was injected intravenously and in most cases, the antibodies were given prior to infection. Hence, the yeast cells are immediately opsonized and rapidly eliminated by phagocytes in the blood stream. However, this setting does not reflect the normal sequence of events associated with a systemic C. albicans infection. The yeast form is often present in the blood stream, but it is normally not the dominant morphotype during infection; hyphae are more abundant and spread in the infected tissue. Whether the protective antibodies described so far are also able to attack C. albicans hyphae and thereby to provide protection in naturally acquired cases of invasive candidiasis remains to be determined. Moreover, for those antibodies that were shown to inhibit hyphal growth in vitro, it is, in most cases, unclear how this growth repression is achieved. More research is clearly required to address these issues.

Protective Antibodies against Dimorphic Fungi
Dimorphic fungi are a family of six fungal pathogens of humans mainly found in the Americas that show a unique temperature-induced morphological transition: They grow in their filamentous form in the environment but switch to the yeast morphotype during infection. Several studies explored the therapeutic use of antibodies in infections caused by Histoplasma capsulatum, Paracoccidioides brasiliensis, and Sporothrix schenckii. Opsonization of H. capsulatum by an IgM directed to a surface-bound, histone 2b-like protein promoted the anti-fungal activity of macrophages and resulted in a faster maturation and stronger acidification of their phagosomes [82]. Moreover, administration of this antibody to Histoplasma-infected mice reduced the fungal burden, decreased pulmonary inflammation, and prolonged survival [59]. The heat shock protein Hsp60 is a major and protective H. capsulatum T cell antigen. Hsp60-specific antibodies of the IgG 1 and IgG 2a , but not of the IgG 2b subclass reduced the intracellular survival in macrophages, increased phagolysosomal fusion, and prolonged the lives of infected mice [60]. In contrast, an IgG 1 directed against a 70 kDa surface protein of H. caspulatum surprisingly increased the intracellular fungal growth and reduced macrophage nitric oxide release in vitro but had no effect on fungal burden or survival in a murine model of infection [61]. Another promising surface protein is the so-called M antigen of H. capsulatum. Opsonization with three M antigen-specific mAbs (one IgM and two IgG2a) resulted in enhanced phagocytosis and provided full protection in experimental murine histoplasmosis [62].
In a more recent study, Liedke et al. generated a chitin-specific, recombinant antibody-chimera consisting of the chitin-binding domain of the lectin wheat germ agglutinin (WGA) and the Fc portion of a murine IgG 2a . Only 10 µg of WGA-Fc were sufficient to elicit full protection in mice that received a normally lethal dose of H. capsulatum [81]. In vitro, WGA-Fc triggered increased phagocytosis and complement deposition and thereby promoted an efficient elimination of the pathogen. Remarkably, WGA-Fc also bound to C. albicans and C. neoformans and sparked an enhanced killing of these pathogens by murine macrophages. Due to this cross-reactivity, WGA-Fc is a promising candidate for the development of a pan-fungal therapeutic [81].
In P. brasiliensis, several surface glycoproteins are well-known diagnostic antigens. Passive transfer of mAbs directed against gp70, gp43, and a 75 kDa secreted phosphatase proved to be protective in murine models of infection [63][64][65] and the same applies to mAbs directed against the heat shock protein 60 of P. lutzii [67]. A different approach was taken by Ferreira et al., who constructed a single-chain variable fragment (scFv) antibody derived from the antiidiotypic antibody 7.B12 [66]. This recombinant construct resembles the internal image of gp43 and thereby served as a substitute for this antigen. When expressed in dendritic cells that were administered to mice, it triggered an enhanced T cell response, elevated levels of anti-gp43 antibodies, and a dramatic reduction in the number of viable fungi. In a subsequent study, the same group demonstrated that the protective effect could be further enhanced if the scFv molecules were incorporated into poly(lactide-co-glycolic) acid nanoparticles [85].
Another approach that targeted glycoproteins was undertaken with Sporothrix schenckii, a fungus causing chronic subcutaneous mycosis in humans and animals. An IgG 1 raised against a 70 kDa glycoprotein and putative adhesin protected mice from this pathogen when administered either before, during, or even three days after infection [68,69]. A humanized version of this IgG 1 , given 3d post infection, also reduced the fungal burden in the spleens, but not in livers of infected mice [70].

Protective Antibodies against Molds
Molds are a heterogeneous group of soil-dwelling fungi that share a common lifestyle. Their asexual spores are efficiently spread in the environment, whereas hyphae, their vegetative morphotype, grow in the soil and other habitats. Aspergillus fumigatus is currently the most frequent mold causing severe mycoses, but infections caused by Mucorales and other filamentous fungi are recognized with increasing frequencies. As mentioned above, the filamentous growth of these pathogens during infection represents a particular challenge for the immune system.
In a first study, Frosco et al. analyzed five mAbs specific for a so-called elastase of A. fumigatus that all turned out to be non-protective [71]. Cenci et al. reported a first successful passive immunization experiment with A. fumigatus using an antiidiotypic mAb representing the internal image of yeast killer toxin [72]. This mAb also inhibited the hyphal growth in in vitro experiments. A similar growth inhibition and protection was later on reported for 2G8, a mAb specific for β-1,3-glucan [30]. As for Candida, a corresponding IgM sharing an identical binding site with 2G8 was non-protective. Another abundant and homogenously distributed glycostructure present on Aspergillus hyphae is galactomannan, but a galactomannan-specific IgM failed to provide protection in mice infected intravenously [75]. These data fit well to the more recent finding that an efficient killing of A. fumigatus hyphae by neutrophils requires antibody-mediated opsonization and activation of Fcγ-receptors through binding of suitable IgG antibodies [86].
However, other studies provided evidence that IgM can be protective against A. fumigatus infections. An IgM initially raised against sialyl-lacto-N-tetraose of B group streptococci was shown to recognize a glycoantigen present on A. fumigatus conidia and hyphae. After passive transfer, this mAb protected mice infected with A. fumigatus via the intravenous or intratracheal route [76]. The elimination of A. fumigatus conidia and germ tubes by human neutrophils was previously shown to depend on antibody-mediated complement activation [87]; as IgM binds complement factors, protection is most likely established by activation of the classical complement pathway.
Another IgM directed against enolase, an enzyme of the glycolytic pathway, strongly inhibited Aspergillus hyphal growth and prolonged survival of intravenously infected mice [78]. Enolase is one of the so-called moonlighting proteins; it normally resides in the cytoplasm, but some molecules are also found on the cell surface. In vitro experiments showed that the enolase-specific IgM had a striking growth inhibitory activity on Aspergillus hyphae [78], but the underlying mechanisms have not been defined yet. Appel et al. coupled an IgM recognizing a cell wall antigen of A. fumigatus to alliinase, an enzyme that converts the harmless garlic compound allicin to alliin, a substance with a broad antifungal activity. If administered together with allicin, this conjugate was able to protect mice from A. fumigatus infections [74].
Two further studies analyzed mAbs directed against protein antigens. An IgG 1 recognizing an A. fumigatus cell wall glycoprotein inhibited the growth and even killed A. fumigatus hyphae in vitro. Moreover, this antibody substantially increased the survival times in a murine model of infection [73]. However, due to the lack of follow-up studies, the identity of the antigen and the antifungal mode of action employed by this antibody remained undefined. Chauvin and co-workers generated a humanized IgG 1 antibody directed against Crf1, an Aspergillus cell wall enzyme with transglycosylase activity [77]. The Crf1 protein is a prominent T cell antigen providing striking cross-protection against A. fumigatus and C. albicans [88]. The IgG 1 antibody detected Crf1 on the hyphal surface, both in vitro and in vivo, inhibited the enzymatic activity of Crf1 and caused a slight growth retardation of A. fumigatus hyphae in vitro. However, when tested in a rat model of infection, this antibody failed to provide protection [77].
Mucorales are a group of non-septated, filamentous molds representing another severe threat for immunocompromised patients. So far, mucormycoses are less frequent than Aspergillus infections, but the numbers have clearly increased in recent years and these rapidly progressing infections are particular difficult to treat [89]. The CotH3 protein of Rhizopus delemar resides on the fungal surface and its interaction with the human glucose-regulated protein (GRP) 78 represents a key event in the hyphal invasion of endothelial cells. Loss of CotH3 results in attenuated virulence [84] and polyclonal CotH3-specific antibodies were shown to block the interaction between CotH3 and GRP78 and thereby reduce invasion of an endothelial layer. These antibodies were furthermore able to inhibit the growth of R. delemar in in vitro experiments. The monoclonal anti-CotH3 antibody designated C2 had similar activities and was successfully used in protection experiments with intratracheally infected mice [79]. Protection was mediated by binding of the Fc part of the C2 IgG 1 immunoglobulin to the corresponding Fcγ-receptor. This interaction triggered enhanced opsonophagocytosis and thereby limited the infection. Application of the antibody in combination with either posaconazole or amphotericin B amplified the protective effect and saved all infected animals [79].
A remarkable example for a mAb causing an exacerbated infection was reported for Scedosporium proliferans. This IgG 1 directed to surface-bound peptidorhamnomannan enhanced fungal germination, impaired phagocytosis by macrophages, and reduced the survival time of infected mice [80]. The authors speculated that binding of this mAb modifies certain activities of the fungus and thereby enhances its virulence.

Conclusions
A large number of studies have provided evidence that the passive transfer of antibodies can protect animals from fungal infections. However, these studies summarized in Table 1A-D are difficult to compare for several reasons: (i) The different pathogenic fungi have a variety of distinct features that are decisive for the respective infections and this can influence the biological impact of therapeutic antibodies, e.g., shedding of capsular polysaccharides by C. neoformans is the reason for ALT. (ii) The biological properties of immunoglobulins differ significantly, e.g., their ability to interact with Fcγ-receptors or to fix complement depend on their (sub)class and the animal species they are derived from. (iii) The routes of experimental infections differ and do not always reflect the natural infection processes. (iv) In a patient, therapeutic antibodies will be given when the infection process has already progressed to a level causing clinical symptoms, but in most studies, antibodies were given prior to infection, which reflects a prophylactic rather than a therapeutic use. (v) The amount of antibody is a critical factor and varies in murine studies at 10-1000 µg per animal. Up to now, only few studies compared different amounts of a given antibody to determine an optimal dosage.
The protective impact of a certain antibody clearly depends on the mechanisms exerted to eliminate the fungal pathogen. The major mechanisms employed by antibodies in order to harm fungal pathogens are schematically depicted in Figure 1. Opsonization can result in an efficient elimination of small and predominantly unicellular fungi, and it can boost other antimicrobial effector mechanisms, e.g., by attraction and activation of neutrophils. Another mechanism reported by several studies is the antibody-mediated inhibition of fungal growth, but our knowledge about suitable antigen/antibody combinations and the underlying molecular processes is still in its infancy. Antibodies with a direct and deleterious impact on the fungus represent a particularly promising option, since they act independently of other immune molecules and cells, which is an obvious advantage in a severely immunocompromised host. Antibodies may inhibit the biological function of surface proteins and thereby reduce the ability of the target cell to adapt to certain stress situations (e.g., anti Hsp90 mAbs). Alternatively, antibodies may interfere with transport channels, uptake systems, or proteins that are required for the maintenance and reorganization of the cell wall. Anti-idiotypic antibodies employ a direct antifungal mode of action or may alternatively act as a substitute for the original antigen; this can boost an immune response directed towards this antigen and thereby provide protection. Secreted proteins are in principle attractive targets, but up to now, only one mAb specific for C. albicans Sap2 was shown to be protective in models of Candida vaginitis. Antibodies against conserved cell wall glycostructures can bind to a range of fungal pathogens. The use of chimeric molecules, such as the lectin domain-containing WGA-Fc construct, can extend the repertoire of suitable molecules, but a potential drawback of this strategy is that the lectin domain may trigger a strong immune response that could prevent a prolonged application.
The serious threat posed by invasive fungal infections is a persisting problem and therefore new therapeutic options are clearly required. Monoclonal antibodies are now widely used in modern medicine, but we are just beginning to explore their potential in the context of fungal infections. The data available so far that are summarized here strongly suggest that mAbs are promising prophylactic tools, but further studies are clearly required to determine whether the same applies to a therapeutic use in the setting of an already established fungal infection. The serious threat posed by invasive fungal infections is a persisting problem and therefore new therapeutic options are clearly required. Monoclonal antibodies are now widely used in modern medicine, but we are just beginning to explore their potential in the context of fungal infections. The data available so far that are summarized here strongly suggest that mAbs are promising prophylactic tools, but further studies are clearly required to determine whether the same applies to a therapeutic use in the setting of an already established fungal infection.

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