Potential Roles of Fungal Extracellular Vesicles during Infection

Extracellular vesicles (EVs) are produced by virtually all cell types. Within the past few years, work in this field has revealed more information about fungal EVs. Fungal EVs have been shown to carry proteins, lipids, pigments, polysaccharides, and RNA; these components are known virulence factors, a fact which supports the hypothesis that fungal EVs concentrate pathogenic determinants.

induced robust immune responses in BALB/c mice that led to higher survival rates in animals challenged with the pathogen (12). Similar results were observed with Mycobacterium tuberculosis EVs (32). Mice immunized with mycobacterial EVs induced strong T helper type 1 cell responses (Th1), elicited antibody production, and reduced bacterial burden (32).

PROTOZOAN VESICLES
Parasite EVs were first described more than 20 years ago, although their relevance for secretory mechanisms has been realized only recently (33). In Leishmania spp. and Trypanosoma cruzi, EVs play important roles in protein export pathways, macrophage communication, and inflammatory responses (34-36). Proteomic analysis in T. cruzi has shown the presence of a great number of proteins containing nucleic acid-binding sites and ribosomal proteins within EVs (37). Different types of small RNAs were also detected in parasite EVs (38). Additionally, there were found to be differences between small RNAs packaged in EVs in each parasite developmental form (38). It has been demonstrated that T. cruzi (39) and Plasmodium falciparum (40) can transfer genetic information between parasites and from parasites to mammalian host cells. More recently, Fernandez-Calero and colleagues suggested that, under conditions of nutri-tional stress, a specific type of small RNA is released into EVs that may play a role in parasite-host interaction (41). In Leishmania spp., EV release is considered the most important mechanism of protein secretion and mediates delivery of parasite proteins into macrophages to cause production of interleukin-8 (IL-8) (35). Accordingly, EVs may increase parasite virulence; thus, pretreatment of mice with T. cruzi EVs followed by intraperitoneal parasite inoculation resulted in mortality indices that were higher than those seen with untreated mice (34). Moreover, EVs induced increased heart parasitism and inflammation through enhanced IL-10 and IL-4 production (34). These data suggested that T. cruzi EVs could facilitate parasite dissemination and pathogenic mechanisms (34). More recently, Szempruch and colleagues demonstrated that African trypanosomes release EVs through flagellum-derived nanotube formation, resulting in vesicular fusion with mammalian erythrocytes, membrane disruption, and anemia (42). Together, these results indicate that parasite EVs might participate in mechanisms of either virulence or immune response modulation in parasitic infections.

FUNGAL VESICLES
EVs produced by fungal cells are peculiar because, like bacterial EVs, fungal EVs must traverse a cell wall in order to be released. The mechanisms of EV release across the complex molecular network of the fungal cell wall are still unknown. Wolf and colleagues have utilized electron microscopy techniques to suggest that EVs interact with cell wall components (43). They showed that single and multiple vesicles simultaneously gain access to the extracellular milieu by crossing of the cell wall (43). The cellular origin of fungal EVs remains unknown. Rodrigues and colleagues suggested that EVs could originate from cytoplasmic subtractions (44). Other studies indicated that multivesicular bodies and membrane budding may also participate in EV formation (43, 45). These mechanisms are compatible with the presence of cytoplasmic proteins lacking secretory signals in EVs (44). The mechanisms of passage through the cell wall are similarly unknown. Proteomic analysis of EVs from Saccharomyces cerevisiae, Histoplasma capsulatum, Paracoccidioides brasiliensis, Candida albicans, and Cryptococcus neoformans revealed the presence of cell wall-degrading enzymes, suggesting that EV-associated molecules could be present in vesicular membranes and thus exposed, catalyzing cell wall crossing by hydrolysis of structural components (13,(46)(47)(48)(49).
EVs transport virulence-associated components to the extracellular medium, suggesting that they are required for fungal pathogenesis (2,13,47,48,50). In contrast, EVs have been shown to induce protection in experimental models of fungal infection (13). In the following sections, possible roles of fungal EVs as "virulence bags" and their potential as vaccine tools are discussed.
Complex carbohydrates, such as GXM, are also exported in fungal EVs (48). In P. brasiliensis, EVs contain antigenic ␣-galactopyranosyl epitopes (50). Peres da Silva and colleagues have shown that residues of mannose and N-acetylglucosamine are present at the surface of P. brasiliensis EVs, where they mediate recognition by innate immune receptors (14). These observations might suggest a role for P. brasiliensis EVs in pathogen-host communication with the innate immune system, but it remains to be confirmed (14).
Fungal pigments are also found in and exported extracellularly by fungal EVs (48,52). Melanin granules have dimensions that are comparable to those of EVs (52). In addition, melanin ghosts stained with lipophilic dyes suggested the presence of associated lipids. In fact, purified C. neoformans EVs contain electron-dense and dark granules (48) and are able to catalyze melanin synthesis in the presence of L-3,4dihydroxyphenylalanine (L-DOPA) (52), strongly suggesting that melanin synthesis occurs inside fungal vesicles for further transport to the cell wall (52).
RNA-containing vesicles have been characterized in C. neoformans, C. albicans, S. cerevisiae, and P. brasiliensis (51). In these organisms, EVs carry small RNA molecules that are protected by EV membranes against exogenous RNase degradation. Mature tRNAs and mRNAs and noncoding RNAs are also present in fungal EVs (51). Although little is known about the function of RNAs in fungal EVs, it has been suggested that these molecules could participate in cell-to-cell communication, including host cells (51). This is supported by studies in T. cruzi, in which tRNA induces changes in host cells, making them more susceptible to infection (58).
Proteomic analysis of fungal EVs revealed the presence of a complex protein composition with multiple functions, including sugar metabolism, cell wall architecture, cell signaling, lipid metabolism, cell growth/division, and virulence (13,(46)(47)(48)(49). Although many of these proteins have been shown to be shared in C. neoformans, P. brasiliensis, C. albicans, H. capsulatum, and S. cerevisiae EVs, species-specific protein molecules have also been abundantly observed (13,(46)(47)(48)(49). For instance, enzymes essential to glucuronic acid metabolism were found in C. neoformans EVs but not in those of other species (48). EVs produced by fungi contain proteins such as plasma membrane ATPases, cytoskeleton proteins, heat shock proteins (HSP70 and HSP90), adhesins, and antioxidant proteins, including superoxide dismutase and catalase B (2,13,47,48). The activity of urease and phosphatase was also detected in C. neoformans EVs (48). Urease is known to be an important virulence factor of C. neoformans in enhancing the invasion of host central nervous system (59). Phosphatases, which are surface components in different fungal species (60)(61)(62), were detected in other fungal EVs (48,49).
In the context of all the EV cargo mentioned above, some examples of the involvement of mutants in vesicle formation can help understanding of the importance of fungal EVs. In C. neoformans, a Sec6 RNA interference (RNAi) mutant accumulates EVs inside the cells and the levels of some virulence factors, such as those involved in laccase, urease, and polysaccharide secretion, are significantly decreased. As a consequence, its virulence is attenuated (63). This was the first evidence correlating fungal EVs with virulence. Also, in C. neoformans, a secretion mutant called sav1, lacking a Sec4 GTPase homolog protein of post-Golgi secretion, accumulated EVs inside the cells and showed reduced secretion of proteins (64). In C. albicans, phosphatidylserine synthase (CHO1) and phosphatidylserine decarboxylase (PSD1 and PSD2) mutants also showed differences in the structures and constitution of EVs (65), but whether this phenotype affects Candida virulence is not known. There are some secretion-defective mutants that have been reported in the literature, but the correlation of defects in secretion of EVs with fungal virulence has been studied by our group.

FUNGAL EVs AS POTENTIAL VACCINES
There are many practical advantages to the general use of EVs as potential vaccine tools. As discussed by Schorey and coworkers (7), EVs offer more-stable conformational conditions for protein components, are able to circulate in body fluids, and show efficient association with antigen-presenting cells (7). However, fungal EVs manifest important particularities. For instance, C. neoformans EVs are unstable in the presence of host serum proteins (66).
Studies using EVs as vaccine tools initially included dendritic cell vesicles used for tumor control (67). Recent studies have demonstrated that EV prophylaxis induced increased survival and slower tumor development in mice (68). Development of vaccines using pathogen-derived EVs first included Gram-negative OMVs produced by Neisseria meningitidis serogroup B (69)(70)(71). Four licensed OMV vaccines are now available (70)(71)(72), each of them having been developed for a specific outbreak. Meningococcal group B vaccine (Bexsero), the most recently licensed OMV vaccine, was designed to provide broad protection through combining multiple antigens (72). These EV vaccines were associated with high effectiveness in regions where the circulating and vaccine strains were the same (22,70,71,73). Promising studies focusing on other bacterial species are available in the current literature (20,21,74,75).
Studies on the immunobiological activity of fungal EVs have been performed more recently. In 2011, it was shown that EVs in Malassezia sympodialis carry antigens and allergens that modulate the immune system through in vivo stimulation of IL-4 and tumor necrosis factor alpha (TNF-␣) (76), suggesting an involvement of EVs with atopic eczema (76). Oliveira and colleagues showed that C. neoformans EVs are biologically active in vitro (77): they stimulate nitric oxide, cytokine (TNF-␣, IL-10, and transforming growth factor ␤ [TGF-␤]), and fungicidal activity in macrophages (77). This observation might be related to the detection of key immunogens in C. neoformans EVs, including heat-shock proteins and GlcCer (78). In fact, human antibodies against C. neoformans GlcCer were demonstrated to be fungistatic (45,78). Similarly, P. brasiliensis EVs also stimulated cytokine production in macrophages (14).
Although some reports suggest a protective role of C. neoformans EVs, other reports indicate that these compartments can enhance pathogenesis of this fungus. Huang and colleagues have demonstrated that C. neoformans EVs are able to cross the brain-blood barrier and accumulate in lesion sites of brain infection (79). In an in vitro model of C. neoformans infection, EVs facilitated adhesion and transcytosis of C. neoformans during interaction with human brain microvascular endothelial cells (79), leading to increased brain infectivity. Finally, pretreatment of mice with C. neoformans EVs rendered the mice more susceptible to cryptococcosis (79). Therefore, it is still unclear how C. neoformans EVs interfere with either fungal virulence or host stimulation.
There are no fungal vaccines in the clinic, although some animal studies suggest that a fungal vaccine in the clinic may be closer than ever. Wormley and colleagues have suggested the potential protection of using a gamma interferon (IFN-␥)producing C. neoformans strain in a mouse pulmonary infection model (80). Mice were able to resolve the primary infection and showed complete protection against second pulmonary challenge with the C. neoformans strain (80). This finding, which was based on utilization of engineered fungi to produce host cytokines in order to induce a protective host immune response, was an important advance in the vaccine field. Another example of a potential fungal vaccine is the C. neoformans knockout strain lacking sterylglucosidase 1 (Sgl1), an enzyme responsible for degradation of sterylglucosides. The ⌬sgl1 strain was nonpathogenic in an infection model and, interestingly, acted as a vaccine strain, protecting mice when administered prior to challenge with wild-type C. neoformans or C. gattii (81). This protection may be ascribed to a dramatic accumulation of sterol glucosides (SGs) in the ⌬sgl1 mutant (81), and these fungal SGs may act as stronger immunostimulators than plant SGs (82)(83)(84)(85). The most interesting observation was that the host was protected against primary and secondary infection even when CD4 ϩ T cells were deficient. These data suggest the potential protective effect of an attenuated fungal strain in immunodeficient hosts, such as HIV-positive patients, where cryptococcosis most frequently occurs. Interestingly, if SGs are found in the vesicles of the ⌬sgl1 strain, this not only might significantly contribute to the protective effect but also could increase interest in using EVs instead of the live attenuated mutant as a vaccine formulation.
Recently, Vargas and colleagues demonstrated that EVs obtained from C. albicans cultures are immunologically active (13). Exposure of macrophages and dendritic cells to purified EVs resulted in nitric oxide production and release of IL-12, TGF-␤, and IL-10 in macrophages and of IL-10 and TNF-␣ in dendritic cells (13). The ability of C. albicans EVs to stimulate host cells depended on lipid composition, since EVs from phospholipid synthase mutants differentially stimulated cytokine production in macrophages (65). Treatment of Galleria mellonella with EVs before infection with C. albicans reduced fungal burden (13). These data indicate the ability of C. albicans EVs to modulate the innate immune response and the potential of EVs to interfere with fungal pathogenesis in vivo in favor of infection control (13). Together, these results strongly suggest that fungal EVs activate host immunity by multiple mechanisms that may favor the host during fungal infections, reinforcing their potential use as a vaccination strategy.

CONCLUSION
Fungal diseases are a global public health problem, especially for immunocompromised individuals, such as those affected by HIV infection and cancer (86) (http:// www.cdc.gov/fungal/global/index.html). It is estimated that millions of people die every year due to invasive fungal infection (86)(87)(88), with a mortality rate comparable to that of malaria (89). The current available antifungal therapies have significant disadvantages: polyene are toxic, azoles induce the development of resistance and are notorious for drug-drug interactions, and echinocandins have a narrow spectrum of activity (86,90). Thus, there is a need for research and development (R&D) of new antifungals as well as new, alternative preventive strategies to combat fungal diseases. As discussed above, fungal EVs may represent this new, alternative strategy once we fully understand their cargo and their role as virulence bags. Although there are a variety of examples of the role of fungal EVs in carrying many virulence factors, the relevance of their cargo in vivo is incompletely understood. On the other hand, fungal EVs are promising in vitro activators of the host immune system against pathogens that operate by inducing cytokine release by innate immune cells, suggesting that they may be a potent tool for vaccine development. Nevertheless, their role in in vivo protection is still under study. In bacteria, exciting results that have shown some examples of successful EV vaccines support the idea that fungal EVs could also be used as a protective tool against the most deadly fungal infections. In addition, the concept of therapeutic fungal vaccines is gaining more traction, since there are currently fungal proteins in preclinical phase I studies for vaccine development (91,92). More studies and more effort are clearly necessary for assessing the role of fungal EVs in the fungus-host interaction and to decipher the mechanisms by which they might stimulate a protective host response.

ACKNOWLEDGMENTS
This work was supported by grants from Brazilian agencies Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ), by NIH grants AI56168, AI100631, and AI116420, and by Merit Review grant I01BX002624 from the Veterans Affairs Program in Biomed- The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
We are very grateful to Rodrigo Pinheiro and Arielle Bryan for valuable suggestions and English proofreading.