Introduction

The predominant species of Cryptococcus involved with lethal infections in humans are Cryptococcus neoformans and Cryptococcus gattii [1]. The disease caused by these pathogens is called cryptococcosis, which usually begins after inhalation of infectious propagules that colonize the lung resulting in clinical or sub-clinical symptoms and signs of pneumonia [1, 2]. The lack of ability to control fungal growth or/and to eliminate the fungus in the lung can be followed by its dissemination to other organs and tissues [2,3,4]. In immunocompromised patients, C. neoformans frequently disseminates to the brain, causing meningoencephalitis, which is one of the major causes of death in HIV patients [1, 5]. Global incidence of meningoencephalitis due to Cryptococcus spp. was estimated in approximately 200,000 cases per year [6]. Current antifungal therapy to combat cryptococcosis is limited to a combination of drugs that comprise amphotericin B, azoles, and 5-flucytosine (5-FC) [7, 8]. However, treatment is frequently accompanied by pitfalls that cause failure or interruption of drug administration [8, 9]. These drawbacks include primarily adverse side effects and the emergence of resistance [8,9,10]. Furthermore, anticryptococcal therapeutic protocols are considerably long and very expensive [8]. Drug-drug interactions are frequent and high toxicity is common [8, 9]. Considering that treatment failure during cryptococcosis results in high rates of mortality and morbidity, new strategies for treatment or prevention are imperative. There are simply no vaccines in the clinic against any fungal infections. The research and development in this area has been scarce, most likely because it is difficult to envision a protective immunity induced by a fungal vaccine when fungal infections mostly occur in the condition of immunodeficiency. However, recent investigations of new vaccine formulations to combat fungal infections have re-emerged in the last decade, mainly because a better understanding of the antifungal immune response prompts the elaboration of novel vaccination protocols and immunotherapeutic strategies [11,12,13,14]. The major focus of this review is to discuss the most recent published work using lipids as tools to induce, and perhaps preserve, a robust host response that would protect the host from cryptococcosis (and perhaps other fungal infections) even in the condition of immunodeficiency.

Antifungal Vaccines

Vaccination has been successfully used to prevent human diseases caused by several infectious agents, mostly bacteria and virus [15,16,17]. However, after decades of research, only a few vaccine formulations have reached clinical trials to combat fungal infections, and so far, none of them was approved to be used in humans [14, 18]. This scenario can change considerably in the next years, considering the number of promising formulations that have been developed recently [14]. Purified and recombinant antigens have been tested, especially using combinations of proteins and polysaccharides found at the fungal cell wall [14, 19, 20]. In addition, killed organisms and attenuated strains were reported as protective in mice models of invasive fungal infections [21, 22]. Other immunotherapy protocols have also been suggested, including adoptive and antibody-based therapies [23]. In this manuscript, we will limit the discussion to antigen preparations based on the immunological role of fungal lipids.

Very little is known about the use of fungal lipids as immunogens. Nami and colleagues have recently listed and discussed studies where the activity of antifungal vaccines was investigated. Remarkably, out of 61 vaccine prototypes, only one formulation was prepared testing lipids as antigens [14]. Using murine CpG oligodeoxynucleotide or dendritic cells as adjuvants, Bozza and colleagues compared the protective effect of different Aspergillus fumigatus antigens, including membrane glycosphingolipids (GSL) glycosylinositolphosphoceramide and the GPI-anchored lipophosphogalactomannan [24]. Although immunization with GSL induced a Th17 response, usually associated with protection, this response was accompanied by a concomitant Th2 activation and no protection was observed for both models.

Fungal Lipids: Structure and Function for Antifungal Targets

Overall, lipids such as fatty acids, phospholipids, triacylglycerol, and sterols are usually associated with membrane architecture, permeability, fluidity control, and signaling [25,26,27,28]. More complex lipids, including sphingolipids and its derivatives, are also typical plasma membrane-enriched molecules. However, in recent years, they gained attention as potent immunogenic and regulators of fungal virulence [28, 29•, 30, 31]. Importantly, immunogenicity, pathogenic profile, and membrane physiology are all interconnected. For instance, changes in membrane asymmetry through deletion of Apt1, a flippase responsible for aminophospholipid translocation, not only directly affected the synthesis of phospholipids, glucosylceramide, and ergosterylglycoside but also impaired the synthesis and secretion of glucuronoxylomannan (GXM) by C. neoformans and resulted in reduced pathogenic potential [32]. Other fungal lipids were functionally implicated in processes of morphogenesis and pathogenesis. For instance, farnesol is a sesquiterpene produced by C. albicans and released extracellularly as a quorum-sensing molecule that inhibited yeast-to-hypha differentiation [33]. The response to farnesol by other species is diverse. Farnesol reverses the deficiency in conidiation that occurs in Neurospora crassa after deletion of genes related with frequence or white collar [34]. In addition, farnesol induces apoptosis in Aspergillus nidulans [35] and inhibits Histoplasma capsulatum, C. neoformans, and C. gattii growth in vitro [36, 37].

The lipid structures and their biosynthesis pathways are highly conserved between mammalian and fungal organisms. However, key differences have been identified and particular lipids and enzymes have been explored as targets of antifungal drugs [9]. The broad-spectrum drug amphotericin B (AmB), for instance, was described in the 1950s and until now is the principal drug choice for cryptococcosis treatment [7]. This amphipathic polyene binds to ergosterol in fungal membranes compromising their function, leading to pore formation and consequent fungal death [9]. Additional studies suggested that AmB stimulates the oxidative response of phagocytes, contributing to the anticryptococcal activity [38, 39]. The ergosterol biosynthesis pathway is also a target for azoles [9]. This class of antifungal drugs inhibits the lanosterol 14 alpha-demethylase, reducing ergosterol production and causing accumulation of toxic precursors. As a consequence, the fungal membrane loses stability and vital functions. Other lipid biosynthesis inhibitors (e.g., myriocin, fumonisin, australifungin) displayed potent antifungal activity. However, the use of these molecules in humans was impaired by the toxic effects or due to their limited effect against molds (e.g., aureobasidin) [9, 13]. Recently, Mor and colleagues identified a new class of antifungal molecules derived from hydrazides [40]. These compounds indirectly inhibited the synthesis of fungal glucosylceramide (GlcCer) and were effective in vitro and in vivo against a variety of fungal pathogens, including C. neoformans, C. albicans, and Pneumocystis and displayed reduced toxicity to mammalian cells. Hydrazide derivatives will be discussed in more detail later in this review. It is clear, in summary, that the search for new drugs that bind to or impair lipid biosynthesis pathway in fungal organisms is highly promising.

Lipid Immunogens Produced by Cryptococcus

Initially considered as a structural component of fungal membranes, GlcCer (Fig. 1A) was first characterized as an antigen and a component of C. neoformans cell wall by Rodrigues and colleagues in 2000 [30]. A few years later, GlcCer was reported as a pathogenicity regulator in C. neoformans [29•]. It was established that the absence of GlcCer (Δgcs1 strain, Fig. 1B) in C. neoformans affects cell cycle progression and growth in neutral/alkaline pH at physiologically CO2 concentrations. Mutants lacking GlcCer lost the ability to cause cryptococcosis in mice challenged intranasally with lethal doses of C. neoformans. These studies were followed by an extensive investigation on the relationship between structure and function of GlcCer in C. neoformans and other fungal pathogens [31]. By knocking out the enzymes required for the GlcCer biosynthesis pathway, Del Poeta’s group has demonstrated that changes in GlcCer structure profoundly impact the C. neoformans surface, interfering with membrane physical properties and virulence [29•, 41, 42]. For instance, when the C. neoformans ability to methylate the sphingosine backbone is blocked through the deletion of the smt1 gene, yeasts accumulate demethylated sphingosine and demethylated GlcCer (Fig. 1B) [41]. Membrane permeability and rigidity were strongly affected in Δsmt1 strain, which seemed to impact adaptation to neutral/alkaline environment found in the lung alveoli. The Δsmt1 phenotype was similar to that observed in the Δgcs1 strain, linking a unique GlcCer lipid structure (methylated GlcCer) to a phenotype. Furthermore, virulence of the Δsmt1 mutant was also reduced when this strain was administered intranasally in mice. Yeasts of the Δsmt1 strain were trapped in lung granulome with a consequent decrease in dissemination and virulence. Changes on membrane permeability also occurred when the enzyme Sld8, a sphingolipid desaturase, was knocked out in C. neoformans (Δsld8 strain, Fig. 1B) [42]. Membrane permeability of the Δsld8 strain was similar to that of the Δsmt1 mutant and significantly reduced when compared with the WT strain. In addition, the Δsld8 mutant was more sensitive to ionic and non-ionic detergents. Differently from the Δgcs1 and Δsmt1 mutants, the Δsld8 strain was able to grow in vitro under acidic or neutral/alkaline pH at 37 °C and 5% CO2. Yet, a reduced intracellular growth inside macrophages was observed when compared with WT and the reconstituted strain. Despite its ability to grow under different pH levels, the mutant strain lost completely its virulence when administered intranasally in mice.

Fig. 1
figure 1

Structures of GlcCer, GlcCer intermediates, and sterylglucoside produced by C. neoformans. a Initial steps for C. neoformans GlcCer biosynthesis (pathway conserved between humans and fungi). b Final steps of GlcCer biosynthesis (specific for fungi and mutant strains discussed here). c Sterylglucoside hydrolysis by sterylglucosidase

Full size image

GlcCer has been also considered a target for new antifungal drugs [9]. Cerezyme, an enzyme that metabolizes GlcCer, promoted deficiencies in C. neoformans that were similar to those observed in the Δgcs1 strain, including membrane integrity defects, and reduced its ability to grow in physiological CO2 atmosphere. Treatment with Cerezyme prolonged the survival of intranasally infected mice [43]. More recently, a new class of a synthetic drug able to reduce synthesis of fungal GlcCer was reported by Mor and colleagues [40]. The hydrazide derivatives D0 and BHBM were efficient in vivo against several fungal pathogens, including C. neoformans. Both drugs affected fungal cell morphology promoting the accumulation of intracellular vesicles. According to this study, proteins involved with vesicular transport and cell cycle progression were targeted. These studies were followed by the generation of acylhydrazones derivatives with even lower toxicity to mammalian cells and a higher anticryptococcal activity in vitro and in vivo [44].

The results discussed above characterize GlcCer as a virulence regulator that participates intensively during membrane and cell surface organization. In addition, GlcCer is conserved in a list of fungal pathogens [45, 46]. Thus, if immunogenic, GlcCer would be an interesting target for the development of a vaccine formulation. A set of data validates this hypothesis. Almost two decades ago, Rodrigues and colleagues demonstrated that antibodies against GlcCer are produced by patients with cryptococcosis and other fungal infections, including histoplasmosis, aspergillosis, and paracoccidioidomycosis [30]. The anti-GlcCer reactivity was concentrated at the C. neoformans cell wall and distinctively enriched at budding sites. Interestingly, localization of GlcCer at the cell surface was considerably higher under neutral pH and high CO2, conditions regularly found by yeasts of C. neoformans during infection [43]. Furthermore, incubation of C. neoformans with anti-GlcCer resulted in inhibition of cell budding and fungal growth in vitro [30]. In a subsequent study, passive immunization of mice with anti-GlcCer followed by a lethal inoculum with C. neoformans prolonged the survival of A/J mice [47]. Intriguingly, at the first week of infection, the antibody administration did not reduce the CFU number in the lungs. In contrast, anti-GlcCer reduced significantly the inflammatory response when compared with control systems. The mechanisms by which anti-GlcCer controls inflammation are not clear, but it could be related to neutralization of GlcCer-containing extracellular vesicles (EVs). In fact, in the initial investigation showing the localization of GlcCer in C. neoformans, an association with vesicles was suggested [30]. This hypothesis was confirmed years later when C. neoformans EVs were isolated for the first time [48]. Analysis of lipid composition of C. neoformans EVs revealed GlcCer and sterols as their major neutral lipids. Since EVs produced by C. neoformans were associated with an increase in blood-brain barrier permeability and inflammatory response, treatment with anti-GlcCer antibodies could have a direct effect in disease control [49]. In fact, monoclonal antibodies can change the protein loading and modulatory activity of EVs released by Histoplasma capsulatum [50, 51]. It is possible then that anti-GlcCer impairs EV release or neutralizes the EVs already released inside the host.

The confirmation that GlcCer could indeed be used in a vaccine formulation to prevent infection with C. neoformans was published by Mor and colleagues [52]. GlcCer was administered intraperitoneally in different mice strains followed by a lethal dose of C. neoformans. Administration of GlcCer induced production of anti-GlcCer antibodies and prevented fungal dissemination. A significant reduction of fungal burden was observed. Interestingly, protection was observed even in the absence of adjuvants. Furthermore, histopathological analysis revealed absence of fungi in the brain and no major abnormalities when the mice were immunized with GlcCer. Additionally, normal structures were visualized in the lung of immunized mice, contrasting with high inflammatory response and massive infiltration by C. neoformans yeasts in the control. Importantly, when GlcCer was administered alone, no toxicity or damage was observed. These results confirm that fungal GlcCer can be exploited alone or in combination with other structures in vaccine formulations.

Ergosterol is another classical component of fungal membranes required for a number of physiological events. In Cryptococcus species, at least 13 different species of sterol derivatives were characterized, with a predominance of ergosterol [28]. As discussed previously, ergosterol is the major target for the currently antifungal drugs [9]. Besides that, additional functions have been described for this lipid in plants and mammals. In plants, this lipid is considered as a microbe-associated molecular pattern (MAMP) [53]. For instance, ergosterol reduces the activity of H+ ATPase in sugar beet and induces a rapid and transient increase in H+ flux, an early marker of elicitor action in plants [54]. Koselny and colleagues [55] recently investigated the role of ergosterol during pyroptosis in macrophages. They demonstrated that ergosterol levels and localization at the fungal cell surface are correlated with the ability of C. neoformans, C. albicans and Saccharomyces cerevisiae to initiate pyroptosis. The participation of ergosterol in this process was confirmed using liposomes. Ergosterol-containing liposomes were able to induce pyroptosis and mediated macrophage lysis. Together, these data suggest that ergosterol could be better explored as components of antifungal vaccines. Importantly, previous attempts to generate anti-ergosterol antibodies have failed. Rabbits were immunized with BSA-conjugated ergosterol and the development of antibodies was monitored. Although high titers of antibodies were observed and binding was inhibited by free ergosterol, the serum was not reactive to the membrane of fungi [56].

In several organisms, including animals, plants, and fungi, sterol derivatives can be modified by glycosyltransferases to produce sterylglycosides (SGs) (Fig. 1C) [57,58,59]. Little is known about the functions of SG in animals and fungi, but a number of studies have reported the involvement of SG during plant stress [58]. In addition, the effect of plant SGs in animals has also been investigated and anti-inflammatory, anti-tumoral, immunomodulatory, and neurodegenerative activities were reported [60,61,62]. Considering that plant SGs are able to stimulate a Th1 response [63], Lee and colleagues investigated whether this glycolipid displayed a protective effect against C. albicans in vivo [61]. Pre-treatment of mice with sitosterylglucoside stimulated the production of IFN-γ and IL-2 by splenocytes. During infection, pre-treatment with sitosterylglucoside reduced the fungal burden in the kidney of mice infected intravenously with C. albicans. Protection was not observed when the CD4+ T cells were depleted and the adoptive transference with sitosterylglucoside CD4+ T treated cells reversed the resistance, confirming the relevance of Th1 response during protection [61].

Although bacteria lack the enzymatic machinery to produce sterols, SGs can be produced using cholesterol from host cells [58]. For instance, Helicobacter pylori extracts cholesterol from epithelial membranes of the host and converts it into SGs. SGs produced by H. pylori are able to reduce the bacterial uptake by macrophages [64]. These SGs can be subsequently acylated or phosphorylated in the C6′-primary hydroxy group of glucose [63]. Importantly, acylated SGs are presented by CD1d and recognized by murine and human invariant TCR-bearing NKT cells [65]. Borrelia burgdorferi is another bacterial species that glycosylates host cholesterol [66]. Production of cholesteryl 6′-O-acyl β-galactoside during Lyme disease resulted in antibody production, demonstrating that acylated SGs are indeed immunogenic [67]. Thus, glycosylation of sterols can modify their functions and improve their ability to activate the immune response, including the development of specific antibodies.

The immunomodulatory activity of SGs in C. neoformans was also investigated. Rella and colleagues explored the potential use of fungal SG as a vaccine strategy during cryptococcosis [68]. By deleting the sterylglucosidase (Sgl1) in C. neoformans, an attenuated strain with higher SG content was generated and tested as an anticryptococcal vaccine. Remarkably, all mice pre-treated with the attenuated strain Δslg1 survived to a lethal challenge with C. neoformans or C. gattii. Vaccination using the Δslg1 strain was efficient even in immunosuppressed mice where CD4+ T cells were depleted.

The accumulation of SGs in the C. neoformans Δsgl1 strain was accompanied by formation of a thicker and highly branched capsule, suggesting that changes in SG metabolism could also impact capsule architecture and/or GXM composition and release [69•]. In fact, protection was lost when GXM was absent in condition of SG accumulation. Since GXM is a potent immunomodulator, changes in its structure could be also attributed to the protection conferred by the Δsgl1 strain. This hypothesis has been addressed very recently by using a double knockout strain lacking the abilities to synthesize capsule and to degrade SGs. In summary, CAP59, a putative regulator of capsule export, was deleted in the Δsgl1 strain [69•]. As expected, the double mutant strain, Δsgl1/Δcap59, accumulated SGs and had no capsule. However, this strain was unable to promote protection when used as an attenuated vaccine formulation. To investigate the mechanism of protection generated by the Δsgl1 strain, parameters of the immune response and fungal burden were compared after infection with WT, Δsgl1, Δcap59, and Δsgl1/Δcap59 strains. In the first days post infection, a higher number of monocytes, eosinophils, and dendritic cells were found in the lung of mice treated with the Δsgl1 strain. The initial cytokine production stimulated by the WT and Δsgl1 strains was similar, but the WT infection sustained this response during the infection, suggesting a hyper-inflammatory reaction that could culminate with mice death. The response to the Δsgl1/Δcap59 strain was similar to that induced by Δcap59 cells and apparently not sufficient to promote protection. The vaccination/infection models showed the presence of yeasts only in systems where WT and Δcap59 strains were used. However, only the WT strain was lethal. During the re-challenge with the WT strain, an increase in dendritic cells and CD4+ cells was observed only in mice previously infected with the Δsgl1 strain. Also, the inflammatory response of Δsgl1-treated mice displayed the same kinetic, with an initial acute response followed by a decrease in the cytokine production. The results discussed here suggest that an association between higher levels of SG and GXM is required for the protection promoted by the Δsgl1 strain.

EVs: a Platform Carrying Immunogenic Lipids of C. neoformans

The extracellular release of EVs is a conserved strategy developed by fungal cells to export lipids, proteins, polysaccharides, pigments, and nucleic acids [48, 70,70,71,72,73,74,75,77]. These compartments have been associated with a number of processes in C. neoformans and other fungal pathogens, including host cell activation, pathogenesis, virulence transference, and resistance to antifungal drugs [70, 78,78,79,80,82].

Some of the most effective vaccines used to prevent microbial infections are multi-antigenic, especially those using inactivated and attenuated fungal strains [14, 22] [69•]. Organisms with attenuated virulence can induce a stronger response than inactivated cells, but their application is limited since the target population is immunocompromised. However, they represent efficient models to investigate the mechanisms of protection and to inspire the development of safer vaccine formulations. EVs have been successfully explored in vaccine formulations against other diseases, including bacterial meningitis [83,83,84,85,86,87,89]. A major advantage of the use of vaccinal EVs is the use of cell-free preparations containing multiple structures in their native form. As discussed above, the protection given by the Δsgl1 strain of C. neoformans required a combination of SGs and GXM [69•]. On the basis that EVs from the Δsgl1 strain are enriched in SGs and in the immunogenic structures GlcCer and GXM, these compartments were tested in a Galleria mellonella model of cryptococcosis. Pre-treatment of G. mellonella larvae with EVs from the Δsgl1 strain increased the survival of the animals after a lethal challenge with C. neoformans. Importantly, EVs from the Δcap59 and Δsgl1/Δcap59 strains showed no effects, while EVs produced by the WT strain accelerated killing of infected larvae. Protection by EVs was previously reported in a candidiasis model of G. mellonella [70]. These data suggested that fungal EVs can be explored as vaccination prototypes. One major limitation of the use of EVs in vaccination models is the low yields of the protocols traditionally used for EV isolation [72]. However, the recent establishment of an improved protocol of isolation of fungal EVs with high yields has the potential to change this scenario (Reis et al., in press). Alternatively, one could envision the production of giant synthetic vesicles containing GXM and various lipids, such as the immunogenic SG and GlcCer.

Concluding Remarks and Future Vaccine Formulations for Cryptococcosis

The urgent need of alternatives for prevention of fungal diseases associated with the recent advances in the understanding of the immunological mechanisms controlling fungal infections revitalized the vaccine field in medical mycology. New vaccinal formulations exploiting virulence regulators and genetically modified attenuated strains were described in the last decades. Apart from other promising immunogens, lipids consist of molecular species with highly complex biosynthesis pathways and consequently huge structural diversity. Remarkably, minimal changes in lipid structures or relative concentration can modify completely their functions and impact fungal physiology and virulence. GlcCer, SGs, and lipid-containing EVs have been recently tested in vaccine formulations and their anticryptococcal efficacy was confirmed in vivo. Together, the data discussed here encourage the use of fungal lipids in anticryptococcal vaccinal strategies. One major challenge in the field is to generate vaccines to prevent cryptococcosis in immunosuppressed patients. In addition, considering that cryptococcosis mainly affects neglected populations, it is important that such preparations will be affordable in underdeveloped and developing countries.