The structural unit of melanin in the cell wall of the fungal pathogen Cryptococcus neoformans

Melanins are synthesized macromolecules that are found in all biological kingdoms. These pigments have a myriad of roles that range from microbial virulence to key components of the innate immune response in invertebrates. Melanins also exhibit unique properties with potential applications in physics and material sciences, ranging from electrical batteries to novel therapeutics. In the fungi, melanins such as eumelanins, are components of the cell wall that provide protection against biotic and abiotic elements. Elucidation of the smallest fungal cell wall-asociated melanin unit that serves as a building block is critical to understand the architecture of these polymers, its interaction with surrounding components, and their functional versatility. In this study, we used isopycnic gradient sedimentation, NMR, EPR, high-resolution microscopy, and proteomics to analyze the melanin in the cell wall of the human pathogenic fungus Cryptococcus neoformans. We observed that melanin is assembled into the cryptococcal cell wall in spherical structures of ∼200 nm in diameter, termed melanin granules, which are in turn composed of nanospheres of ∼30 nm in diameter, the fungal melanosomes. We noted that melanin granules are closely associated with proteins that may play critical roles in the fungal melanogenesis and the supramolecular structure of this polymer. Using this structural information, we propose a model for C. neoformans melanization that is similar to the process used in animal melanization and is consistent with the phylogenetic relatedness of the fungal and animal kingdoms.


INTRODUCTION
Melanins (Greek melano, meaning dark or black) are perhaps the most enigmatic biopolymers in the biosphere despite being ubiquitous in nature (1). These pigments carry out a vast array of functions in all biological kingdoms including thermoregulation, radical scavenging, energy transduction, camouflage, invertebrate immunity and sexual display (2)(3)(4). In the microbial world melanins contribute to the virulence of many pathogenic microbes. Melanins are produced through the oxidation and polymerization of phenolic/indolic precursors; notably, these polymers are characterized by a strong negative charge, high molecular weight, and hydrophobic nature (5). In the fungal field, black-to-brown insoluble pigments designated as eumelanins confer greater fitness to melanotic species in vivo, thereby contributing to their virulence (6). In Cryptococcus neoformans, a facultative intracellular pathogen, melanized cells are able to modulate the immune response of the host through multiple mechanisms e.g. altering cytokine profiles (7), reducing antibodymediated phagocytosis, and decreasing the toxicity of microbial peptides, reactive oxygen species (ROS), and antifungal drugs (8,9). Recent synthesis of a soluble melanin (homogentesic acid and L-DOPA at a 1:1 ratio), demonstrated that it suppresses the production of cytokines and ROS under stimulation of fungal components thus verifying melanin's ability to modulate the immune system (10).
C. neoformans is a unique system for the study of melanin biology because unlike other melanotic fungi, only exogenously provided catecholamines or diphenolic substrates, such as dihydroxyphenylalanine (L-DOPA), are utilized for melanin synthesis and not as an energy source (11)(12)(13). Consequently, it is possible to selectively label the melanin polymer with NMRor radioactive isotopes, a property that has been exploited for melanin structural studies (14)(15)(16)(17)(18)(19). Melanization of C. neoformans results in the deposition of the polymer at the innermost surface of the cell wall, which becomes filled throughout over time (20). The pigment is composed of closely packed spherical particles ranging from 40 to 130 nm in diameter, which are found arranged in concentric layers in the cell wall (21). Early ultrastructural observations in Agaricus bisporus (22), Fonsecaea pedrosoi (23), Cladosporium carrionii and Hormoconis resinae (24) suggested compartmentation of fungal melanins into organelles akin to mammalian melanosomes (25). Evidence for trans-cell-wall vesicular transport in fungi associated with virulence (26), supported the hypothesis that fungal melanization may occur within specialized vesicles (27)(28)(29)(30).
Studies using isotopically-labeled precursors in conjunction with high resolution solid-state nuclear magnetic resonance (ssNMR), revealed that the melanin polymer is likely to be covalently bonded to cell-wall chitin and also found strongly associated with other non-pigment cellular moieties including polysaccharides such as chitosan and vesicle and/or plasma membrane derived lipids (16,19). These components serve as the scaffold for melanin synthesis. Indeed, C. neoformans strains with aberrant chitin and/or chitosan biosynthesis are unable to retain the pigment within the cell wall and display a leakymelanin phenotype (31)(32)(33)(34), whereas boosting cell-wall chitin/chitosan content increases melanin deposition and assembly (14). The association between fungal melanin and chitin was initially reported in Aspergillus nidulans (35), followed by work done on Exophilia dermatitides (36) and Candida albicans (29). The concept of a fibrillary matrix upon which melanin is deposited extends to other species including human eumelanin, where its known to also play a critical role in absorbing toxic oxidative melanin intermediates and provide the elliptical shape to melanosomes (37).
In addition to the relevance of melanin to all biota and its great impact on fungal pathogenesis, this polymer has numerous potential applications in biophysics, material sciences, and even in the cosmetics and health care industries (38)(39)(40). The remarkable potential of melanin derives both from its chemical composition on the molecular level and its versatility as a component of the supramolecular structure within a cellular milieu. Investigation of the melanin structure is a notoriously challenging task due to the insoluble, amorphous, and heterogeneous nature of this architecturally complex polymer (41). Despite these challenges, over the past two decades a wealth of molecular level structural information has been uncovered using multiple nondestructive methodologies among them: X-ray scattering, ssNMR, electron microscopy, and atomic force microscopy (16,(42)(43)(44)(45)(46). Current views of eumelanin molecular architecture posit that these pigments are composed of a common chemical motif and arranged as graphite-like πstacked planar sheets of locally ordered monomers, mainly derivatives of 5,6dihydroxyindole (DHI) (47,48), which give rise to subunit nanoparticles that then undergo spontaneous aggregation until submicron particles are formed (45,46). Nonetheless, natural eumelanins differ in structure from synthetic melanin and neuromelanin (44,45), suggesting that the mode of synthesis can affect the final structure. A graphical schematic highlighting the current views on eumelanin synthesis and the relationship of melanin to the fungal cell wall, capsule and cell body shows that melanin resides in the cell wall (Figure 1).
Previous studies from our group have shown that the budding of melanized C. neoformans cells involves a decrease in the melanin thickness of the pigmented cell wall near the bud-site (49,50). We have hypothesized that the pigmented nanoparticles released into the extracellular space as part of the normal cell-wall remodeling process are the smallest structural units of melanin found anchored to the cell wall. Hence, we have now investigated the physical properties, supramolecular structure, and protein content of these secreted melanin nanoparticles in the context of current knowledge of the cell-wall associated fungal melanin.

Results
Prolonged acid hydrolysis of melanin ghosts yields melanin nanoparticles. Prior studies have shown that the solid yeast-shaped particles lacking all internal structures designated as "melanin ghosts", are composed of smaller spherical structures ranging in diameter from 40 to 130 nm (21,49). We hypothesized that these particles were held together by non-pigment cellular components, a supposition supported by NMR studies showing that C. neoformans ghosts contain tightly associated lipid moieties and cellwall polysaccharides such as chitin (16,19). Since polysaccharides are typically susceptible to acid hydrolysis, we reasoned that prolonging the hydrochloric acid (HCl) incubation conducted during the melanin ghost isolation protocol would free the smaller spherical melanin particles for further analysis. Consistent with these expectations, increasing the incubation time of the HCl treatment led to significant morphological changes in the melanin ghosts samples. SEM micrographs showed that ghosts, subjected to the prolonged acid incubation, initially appeared to be deformed as the relatively larger melanin particles, located close to the external surface, are release; over time, they took on a "hollowed out" appearance as the smaller melanin nanoparticles were subsequently liberated (Figures 2A-C). In addition, the roughness of the ghost surface increased with the prolonged acid treatment, suggesting that the removal of acid-labile non-pigment material exposes more of the underlying melanin particles. SEM visualization of material released during 4 h of HCl boiling revealed melanin aggregates between 60 and 160 nm in diameter size, with the majority measuring around 60 to 80 nm, that were surrounded by smaller granule-like melanin particles with diameters less than 60 nm ( Figures  2D-E). Cross sectional imaging of this suspended material showed a mix of granular nanoparticles ~30 nm in diameter with variable electron density and degrees of consolidation ( Figure 2F). Further TEM analyses of these particles using negative staining, which involves coating the sample with a heavy metal stain that protects its structure from collapsing during drying and thus provides contrast for the detection of structural features, revealed images consistent with the notion that these nanospheres form tightly packed melanin aggregates (Figure 2G-I). Altogether, these results established that melanin ghosts of C. neoformans are composed of aggregated melanin nanoparticles arranged as small to large sized spherical particles, which are found embedded within an acid-labile matrix that produces a smooth appearance for these cell-sized structures.
C. neoformans-derived melanin granules differ from auto-polymerized L-DOPA. L-DOPA auto-polymerization occurs both in the presence of C. neoformans cells or in cell-free conditions (17). However, in the various cellular systems where melanogenesis has been studied, melanin synthesis occurs within specialized vesicles known as melanosomes (24,(27)(28)(29)51). During the budding process of melanized fungal yeasts, melanosomes in the cell wall are either degraded and/or laterally displaced so that the daughter cell can emerge (49,50,52). Whereas in synthetic systems, melanin formation is due to the auto-polymerization of precursor molecules, such as L-DOPA. Consequently, we hypothesized that the released melanosomes would be found in the supernatants of C. neoformans melanized cultures, possibly in association with extracellular vesicles. The culture supernatants from a C. neoformans wild type strain (grown with or without L-DOPA) and a strain defective for melanin synthesis, LAC1,2 mutant, (Figures   3A-D,G) were centrifuged at 100,000 x g and the resulting pellets were analyzed. The solid material from each of the cell cultures, herein refered as crude material, was found to contain particles mostly around 100 nm in diameter as measured by Dynamic Light Scattering (DLS) (Figures 3B,E,H). As a control, we also analyzed the pelleted material from a culture containing L-DOPA and no fungal inoculum, observing a population of larger particles from 200 to 300 nm in diameter ( Figure 3K) that we attributed to melanin aggregates from L-DOPA autopolymerization. Characterization of these materials by TEM using negative staining was also performed. The sample from C. neoformans grown with L-DOPA revealed the presence of electron-dense aggregated granular structures, henceforth called melanin granules ( Figure 3C). Similar aggregated structures were seen in the ∆LAC1,2 strain sample ( Figure 3I), while they were absent in the wild type without L-DOPA ( Figure 3F). In the L-DOPA sample without C. neoformans cells we observed irregular and nonspherical aggregates that did not resemble the melanin granules observed for the C. neoformans cultures (Figure 3L), and which we attribute to L-DOPA auto-polymerization.
Melanin has the unique ability to absorb almost every wavelength of light and in contrast to most other natural chromophores, has a broadband monotonic absorption spectrum (53,54). Thus, we used UV-Vis absorbance spectroscopy to verify whether the aggregated granular particles visualized by TEM are indeed melanin pigments. A monotonic curve was observed for the pelleted cellular material collected from C. neoformans cells grown in MM with L-DOPA (melanin granules and vesicles), but not for ∆LAC1,2 strain grown in MM with L-DOPA (L-DOPA aggregates and vesicles) nor for the fungal cells grown in MM lacking L-DOPA (vesicles) (Figure 4A). DLS also yields a dispersity index, which is a measure of the dispersion (or spread) in the estimated hydrodynamic sizes of colloidal particles in a solution (15,16). We observed that melanin granules had lower dispersity than autopolymerized L-DOPA ( Figure 4B). Furthermore, negative staining TEM allowed us to visualize spherical nanoparticles ~200 nm in diameter, which were isolated from the supernatant of C. neoformans melanized cultures (Figures 5A,B) whereas, melanin resulting from autopolymerized L-DOPA particles, showed minuscule structures with no specific pattern for their association (Figures 5C,D). These results demonstrated that C. neoformans extracellular vesicles along with melanin granules are found in the culture supernatant, while aggregated material seen in the ∆LAC1,2 strain have similar aggregation properties to auto-polymerized L-DOPA. Thus, melanin granules have unique colloidal characteristics; and these particles are distinct structures in size and morphology distinguishable from melanin resulting from auto-polymerized L-DOPA Divalent cations promote aggregation of C. neoformans melanin granules. We studied the hydrodynamic size of melanin granules suspensions in the presence of mono (Na + ) and divalent (Ca 2+ ) salt concentrations to gain insight into their intramolecular interactions. The crude material recovered from the culture supernatants of C. neoformans cells grown with L-DOPA containing both melanin granules and vesicles (hereinafter referred to as crude melanin granules) aggregated at >0.01 M CaCl2, >0.1 M NaCl, and >10X PBS (Figure 6A-C). These results indicate that divalent cations promote melanin aggregation; or fusion events at lower concentrations than monovalent cations. Alternatively, divalent cations likely promote particle aggregation by neutralizing negative charges on the surface of melanin (55).

C. neoformans melanin granules can be isolated
based on their buoyancy. Ultracentrifugation of C. neoformans cell culture supernatants leads to the recovery of a mixture of vesicles and protein aggregates (56). We hypothesized that in addition to these constituents, the cellular material recovered from melanized cell cultures would contain tightly packed melanin granules of high molecular weight and low volume that could be separated from the vesicles and protein aggregates by exploiting their buoyancy. Thus, we analyzed the pelleted crude material from the supernatants of C. neoformans non-melanized and melanized cultures using density gradient ultracentrifugation. A distinct dark band visible by examination with the naked eye was apparent towards the bottom of the tube for the melanized sample, corresponding to fraction 5 ( Figure 7A). TEM imaging of the collected fractions revealed vesicles primarily in fractions 1 and 2, for both samples ( Figure 7B); aggregated electron dense nanoparticles ~30 nm in diameter, corresponding to the dimensions and appearance of melanin granules, were also apparent in fraction 5 of the melanized sample ( Figure 7B).
Melanin particles from ghosts and culture supernatants are similar. We entertained the hypothesis that melanin granules in C. neoformans culture supernatants were similar to those recovered from prolonged acid hydrolysis of melanin ghosts. Consequently, we sought to mechanically break down the melanin ghosts isolated from C. neoformans by extended ultrasonication, avoiding the chemical exposure. We found that ultrasonication of melanin ghosts for up to 12 min resulted in the release of particles in the range 0-200 nm measured by DLS ( Figure  8A). SEM visualization revealed that an ultrasonic pulse as short as a 30 s led to fracture of the melanin ghosts characterized by jagged edges, which then turned into smaller round structures exhibiting a coarse and rough appearance as the process continued ( Figure 8B). The spherical melanin particles that resulted from extensive physical disruption of C. neoformans melanin ghosts showed a similar size distribution to the crude melanin granules in the culture supernatant. These data support the proposal that C. neoformans cell-wall melanin is composed of tightly packed melanin granules with a common substructure of melanin nanoparticles, which are also release to the extracellular environment. To gain further insight into the structure of C. neoformans melanin granules we used cryoEM. Micrographs of fungal melanin granules isolated by buoyancy revealed a pattern of similarly sized, round beads on a string ( Figure 9A). Each spherical nanoparticle, ~200 nm in diameter, resembles a multilayered structure with welldefined edges but not distinguishable lipid bilayer, suggesting that these structures may result from stepwise aggregation of smaller particles ( Figure 9B). We considered whether the structures visualized by cryo-EM images were water ice particles. We cannot unequivocally distinguish between these possibilities by this form of microscopy. Nevertheless, we noted that our putative melanin structures are relatively uniform, have dimensions comparable to those obtained melanin by other EM methods ( Figure  S1), are translucent, less electron dense than ice, and do not express the geometric structures associated with water ice. Hence, we are confident that these are melanin granules.

Melanin granules from C. neoformans exhibit a characteristic EPR signal for eumelanins.
A key feature of melanins is the presence of a stable free-radical population that can be detected by EPR (57), which serves as a molecular fingerprint for natural materials such as charcoal (58). Eumelanins can be identified based on the modification of their EPR signals by a number of agents including light, pH, redox agents, and diamagnetic multivalent metal ions (40). The crude melanin granules recovered from the C. neoformans culture supernatant yielded a single slightly asymmetric EPR trace similar to the spectrum previously described for melanin ghosts (59) ( Figure 10A). Irradiation of fungal crude melanin granules with a 250 W LED white light enhanced the EPR signal intensity 1.2 times, while doping with zinc ions increased it 5.3-fold. Alkaline pH had little or no effect on the EPR signal of crude melanin granules, but low pH was found to attenuate their EPR signal intensity, possibly due to changes in protonation of accessible ionizable groups. The melanin ghosts EPR signal was insensitive to pH ( Figure 10B). Thus, these results demonstrated that melanin granules found in the culture media of C. neoformans, in presence of L-DOPA, retain the paramagnetic properties characteristic of eumelanins.
C. neoformans melanin is molecularly associated with non-polysaccharide cellular constituents. Previous ssNMR studies by our group have revealed that melanin ghosts isolated from C. neoformans cells are an assembly of the L-DOPA-derived melanin pigment and the cellular "scaffolding" moieties on which it the pigment is deposited (16). Here we again used ssNMR to determine whether the melanin granules that are present in the pelleted crude material from C. neoformans culture supernatants are also associated with non-pigment cellular constituents. To obtain a sample mass sufficient for ssNMR analysis, we took advantage of the previously reported C. neoformans mutant strain ST211A (GenBank Accession Number MK609896), which displays a "leaky melanin" phenotype; "leaky melanin" strain cells are able to synthesize melanin, but the cell wall is unable to the retain the pigments and thus they leak into the extracellular space (33). The 13 C crosspolarization magic angle spinning (CPMAS) spectrum of the crude melanin granules obtained from the supernatant of a C. neoformans ST211A cell culture containing uniformly 13 C-enriched glucose as the sole carbon source and L-DOPA (at natural abundance) is compared to that of isolated melanin ghosts in Figure 11 (A,B); this isotopic labeling scheme allows us to selectively observe melanin-associated cellular components without interfering signals from the melanin pigment. Whereas the most prominent peaks in the melanin ghosts CPMAS spectrum are centered at ~30 ppm, likely arising from the aliphatic "tails" of vesicle-derived lipids functionalized during melanization, the crude melanin granules spectrum is dominated by signals in the region between ~55 and 105 ppm that correspond to various types of polysaccharide ring carbons. The strong presence of polysaccharides in this crude preparation is unsurprising since our research group has previously demonstrated that capsular polysaccharides are synthesized intracellularly and transported to the extracellular space within vesicles (26). To test whether these polysaccharides are indeed capsular components that are not associated with the melanin granules, the crude sample was reexamined after 30 minutes of incubation in hot HCl ( Figure 11C). This relatively short treatment was sufficient to drastically diminish the polysaccharide content so that the signals in the ~55-105 ppm region were of lesser relatively intensity than observed for melanin ghosts, which had been subjected to an hour of HCl boiling in addition to other enzymatic and chemical treatments during sample preparation.
Further insight into the molecular identity of the cellular constituents that associate with melanin granules was obtained from 2D 13 C-13 C DARR spectra that display correlations between proximal carbon pairs equivalent to approximately 1-2 bond-lengths in distance. In analogy to our 13 C CPMAS data, the DARR plot of the crude melanin granules displays a multitude of prominent cross-peaks corresponding to correlations between spatiallyclose polysaccharide ring carbons (~55-105 ppm), whereas this region in the melanin ghosts 2D plot only contains a handful of correlations that we have previously attributed to the cell-wall scaffolding moieties (e.g., chitin) on which the pigment is deposited (15) (Figure 11E,F). No cross-peaks attributable to polysaccharides were observed for the crude melanin granules after the 30-minute incubation in hot HCl, supporting our hypothesis that the polysaccharides were indeed superficially bound capsular components and that melanin granules released into the extracellular space are not tightly associated with cell-wall constituents as for melanin ghosts ( Figure 11G). Moreover, all of the spectral features displayed in the 2D DARR plot of the HCl-treated melanin granules were also observed in the melanin ghosts spectrum. We interpreted this spectroscopic similarity to indicate that the 30minute HCl incubation was adequate to hydrolyze all cellular moieties that were not intimately associated with the melanin granules. The major difference between the two datasets is the presence of cross-peaks attributable to the polysaccharide scaffolding constituents in the melanin ghost DARR spectrum. To test the robustness of the association between these residual polysaccharides and the melanin pigment, we subjected the melanin ghosts to a series of treatments in boiling HCl using incrementally longer incubation times and monitored the peak intensities of the non-pigment constituents using 13 C CPMAS experiments ( Figure S2). Increasing the incubation time from 2 h to 4 h resulted in only minimal spectral differences, and thus we chose to further examine the 4 h HCl-treated melanin ghosts by acquiring a 2D DARR spectrum ( Figure 11H). Although the additional 4 h HCl treatment was effective in diminishing the signals corresponding to polysaccharides to undetectable levels (Figures  11 F,H), the majority of other cross-peaks were preserved; furthermore, these cross-peaks were also retained in the DARR spectrum of the crude melanin granules after 30 min of HCl treatment ( Figure 11G). This comparison indicates that the two samples contain many of the same cellular constituents, which are likely strongly associated with the pigment and could play a role in melanin ultrastructure. Interestingly, the correlations common to both sample spectra suggest the presence of lipids and likely also protein remnants: the cluster of signals centered around ~53 x 175 ppm is consistent with correlations between the alpha-and carbonyl-carbons of amino acids within a peptide bond. However, this later assignment is provisional and requires further testing.
Melanin-associated proteins are identified from C. neoformans melanin granules. To assess the possibility that C. neoformans melanin nanoparticles are associated with, or possibly covalently bonded to, melanization-specific proteins, we performed a proteomic analysis on the isolated melanin granules comprising the visibly dark band (fraction 5) after density gradient centrifugation. To distinguish between proteins that are vesicle-associated and/or loosely bound from those chemically functionalized by the melanin nanoparticles, we included as controls fraction 5 samples collected from a C. neoformans culture grown in MM without L-DOPA as well as auto-polymerized L-DOPA. To ensure that sufficient material was available for LC-MS/MS analysis, 2-3 density gradients procedures were performed for each sample and their respective fractions 5 were pooled together. Melanin granule-associated proteins were solubilized, reduced, carboxymethylated, and digested before LC-MS/MS measurements. Data analysis yielded was the consistent identification of four proteins exclusively associated with C. neoformans melanin granules ( Table 1). A bioinformatics analysis to examine the presence of predicted signal peptides and GPI anchors showed that three proteins had an N-terminal signal peptide, two of which (Cig1 and Blp1), were also found in extracellular vesicles (60). Only one protein, CNAG_05312, had both an Nterminal signal peptide and a GPI-anchor indicative of plasma membrane association. These secreted peptides were previously reported to play critical roles in C. neoformans virulence (61)(62)(63)(64)(65). Taken together, these results revealed that C. neoformans melanin granules are indeed intimately associated with proteins that employ both a conventional and non-conventional secretory pathway for regulation, and which may be involved with structural and/or functional properties of the cryptococcal eumelanin.
Based on the results obtained in this study, we propose a model of C. neoformans cell wall melanization that begins intracellularly and requires melanosome transport to the cell wall for their consolidation into melanin granules ( Figure  12). Melanin synthesis occurs within melanosomes (nanospheres 10-20 nm in diameter) enclosed in multivesicular bodies (MVBs) (~450 nm in diameter) ( Figure 12A) and laccase-containing vesicles (~80 nm in diameter) ( Figure 12B, upper inset). These organelles may be signaled by the Qsp1 mature peptide to activate melanogenesis (64). Ongoing melanization is evidenced by the presence of melanosomes with variable degrees of aggregation and electron density within these double-layered compartments. The melanosomes aggregate into ~60 nm particles ( Figure 12B, lower inset), both within MVBs and in the cell cytoplasm, and are eventually transferred into the cell wall by events that are facilitated by the presence of invaginations in the plasma membrane (Figures 12C,D). These structures are characterized by finger-like protrusions that "scoop" melanosomes from the cytosol to transfer them to the cell wall. Further aggregation of the melanosomes into melanin granules (up to 200 nm in diameter) and anchoring within the cell wall could be augmented by vesicle materials, such as laccase and Blp1, respectively. Some of these vesicles are delivered to the cell wall within MVBs that fuse with the plasmatic membrane ( Figures 12E). Melanin granules' concentric arrangement within the cell wall uses as a scaffold membranous sheets shown as multilaminated structures of cryptococcal cell walls ( Figures  12F).
Finally, fungal melanosomes are held together by and anchored to cell-wall constituents (i.e. chitin, chitosan, glucans) due to diverse interactions: i) associative or covalent interactions between melaninassociated proteins and cell-wall components, and ii) electrostatic interactions of oppositely charged melanin-chitosan pairs. As the fungal cell replicates, cell wall remodeling is required to allow budding; thus the melanized scaffold is locally modified by secreted enzymes and allows the release of melanin granules (Figure 12G) into the surrounding environment. A summary of C. neoformans melanin particles analyzed with multiple microscopic techniques is shown in Figure 13.

Discussion
Elucidation of the smallest fungal cell-wall melanin particle that retains its characteristics from C. neoformans is essential to understanding how the pigment is assembled and interacts with other cellular components. Similarly, an appreciation of melanin supramolecular structure (40) is essential to understand how this material confers upon the fungal cell with unique biological properties beyond microbial virulence, such as photoprotection, energy harvesting, heat gain, thermoregulation, and metal scavenging, among others (3).
Previous structural studies on the molecular architecture of fungal eumelanins in the cell wall have focused on the so-called melanin "ghosts", which are macromolecular structures that retain cellular morphology after hot acid digestion of melanized cells (16,27,49). Although such studies have been extremely informative in revealing chemical aspects of melanin composition, they focus on structures where the pigment remains organized into a cellular carcass of what was once a melanized cell. As such, this melanin polymer includes non-melanin components that serve to hold the pigment together into structures with cell-like dimensions. In this study, we sought to go deeper in searching for the building block unit of the cell-wall melanin through complementary approaches that included prolonged acid digestion of melanin ghosts as well as to investigate the presence of such structures in culture supernatants. Both approaches produced much smaller melanin particles that were comparable in size and form. We then applied multiple techniques to study these particles: SEM, TEM, and cryoEM were used to examine the surface and cross-sectional images of the melanin nanoparticles. DLS was used to analyze the size distribution of melanin particles in suspension and to probe their colloidal properties. Isopycnic centrifugation was used to isolate melanin granules based on their buoyancy. Spectroscopic methodologies (optical absorbance, EPR, and ssNMR) were used to probe the molecular properties of fungal eumelanins and investigate their interaction with cell-wall components. Lastly, proteomic analysis was conducted to identify proteins that maintained an intimate association with melanin granules. The results demonstrate the existence of relatively monodisperse melanin granules ranging from 40 to 200 nm in diameter, depending on the approach, that manifested a monotonic broadband absorption curve and stable EPR signal consistent with melanin. These particles are in turn composed of melanosomes, substructural nanospheres serving as the building unit of the fungal cell wall, functionally comparable to those described in mammalian melanization systems (66).
The small melanin particles derived from "ghosts" or culture supernatants are different from those obtained by auto-polymerization of L-DOPA into melanin, implying differences between the products of the biological catalytic and chemical reactions. Melanin "ghosts" can be broken down to smaller particles by either acid digestion or sonic cavitation generating particles that are comparable in size to those released into the culture supernatant during C. neoformans growth under melanizing conditions. The differences between C. neoformans derived particles and those from auto-polymerized L-DOPA imply different synthesis, as fungal particles showed a unique electron microscopic appearance and size. We have shown previously that C. neoformans and synthetic melanins shared similar chemical groups in the aromatic core (18); however, differences between these materials could stem from the fact that the production of fungal particles is vesicle-associated and catalyzed by a laccase, whereas autopolymerization is a spontaneous oxidative process that occurs in solution. Moreover, C. neoformans melanin granules isolated from the extracellular media are spherical shaped and biologically synthesized particles of up to 200 nm in diameter, similar to those isolated from natural eumelanins such as Sepia officinales (43,67), and are found covalently bound to proteins as also reported for mammalian melanin (68).
The fungal melanin granules demonstrated a close association with four proteins relevant to C. neoformans virulence: Cig1, Blp1, Qsp1, and CNI3590 macrophage-activating glycoprotein ( Table  1). The mannoprotein Cig1 (CNAG_01653) plays a role in iron acquisition and it is essential for virulence in a mutant lacking the ferroxidase for high-affinity uptake (69); no defect in melanin production is reported for the C. neoformans cig1∆ mutant (61). The hypothetical protein containing a Barwin-like domain (Blp1) (CNAG_06346) is involved in capsule-independent anti-phagocytosis activity and is critical for successful colonization of mammalian lungs by C. neoformans. Blp1 is not essential for capsule formation and melanin production (63), but this domain is reported to show binding affinity for chitin oligomers (70), suggesting a possible association of this protein with the cell-wall chitin. The quorum sensing peptide (Qsp1) (CNAG_03012) is a secreted central signaling molecule that acts intracellularly to control virulence in C. neoformans and promote cell wall function at high cell densities (64,71). QSP1 mutant strains display hypomelanization at 37˚C but no capsule reduction (64). The melanization process requires the intracellular released of the mature Qsp1 peptide (64), which could explain the finding of this unique sequence in our proteomic analyses of melanin granules. CNI3590 (CNAG_05312) is a GPI-anchored protein and also a secreted factor associated with virulence that contains a carbohydrate-binding domain (62,63), suggesting its interaction with the plasma membrane and cell-wall components such as glucans, chitin and chitosan. Furthermore, Cig1 and CNAG_05312 are positively regulated by the pH-responsive transcription factor Rim101, which in turns is induced by Pka1, a catalytic subunit of the cyclic-AMP/Protein Kinase A (cAMP/PKA) pathway (62,72). Pka1 is known to modulate the secretion of proteins associated with virulence and survival of C. neoformans in the host, as well as the expression of cell-wall related genes (62,73). Similarly, Blp1, Qsp1 and CNAG_05312 are direct targets of three transcription factors required for C. neoformans virulence: Gat201, Gat204 and Liv3 (63,64,74). It is noteworthy that pka∆ mutant strains fail to produce melanin or capsule (75), whereas deletion of GAT201 reduces the capsule size and severely attenuates virulence in a murine model but shows a hypermelanized colony phenotype (74). In contrast, neither gat204∆ (63) nor liv3∆ (74) mutant strains exhibit defects in melanin or capsule production. Altogether, these data strongly support the notion that C. neoformans synthesis of melanin granules is a highly regulated process that leads to particles with unique morphological and compositional features, whose biogenesis might be subject to highly redundant gene networks. Although we cannot rule out that these proteins become associated with the melanin particles in culture supernatants, covalent association of melanin to proteins has been extensively demonstrated in mammalian melanogenesis (68). The association between C. neoformans melanin granules and these proteins may serve complex regulatory roles during the biopolymer synthesis and structural functions within the granular particle or the fungal cell wall milieu.
C. neoformans melanin granules visualized by negative staining TEM and cryoEM most likely are formed from the aggregation of melanosomes (~30 nm in diameter), resembling the supramolecular structure described in Sepia melanin (43) and biomimetic eumelanin (46). Analysis of particles resulting from eumelanin synthesis via enzyme immobilization prompted the proposal of a mechanism of hierarchical buildup of melanin by the gradual aggregation of building block particles (type-A particles), which led to the formation of particles 200 nm in diameter (type-B particles) (46). Prior studies from our group noted that melanin ghosts are composed of smaller melanin granules arranged in concentric layers embedded within the fungal cell wall (49). Ghost structures revealed that the smallest melanin nanospheres with smooth appearance are seen on the surface of early buds. We propose that as the fungal cell ages, aggregated melanosomes giving rise to melanin granules (up to 200 nm in diameter) are embedded into the cell wall. These melanin particles display a layered concentric arrangement guided by membranous sheets within the cryptococcal cell wall, which becomes evident by the removal of non-pigment acidlabile cell-wall constituents. This supramolecular structure is also consistent with the findings of previous studies using NMR cryoporometry, which indicated that C. neoformans melanin layers have pores that become smaller with increased cell age and progressive melanization (49).
Moreover, our micrographs showed a close association between "finger-like" plasma membrane invaginations and melanin granules, suggesting that these membranes structures could be involved in transport from the cytosol to the fungal cell wall. In Saccharomyces cerevisiae, "finger-like" invaginations are sphingolipid-rich lateral domains termed MCP (membrane compartment of Pma1) (76)(77)(78). These domains are characterized by cortical patches of fibrous actin involved with vesicular traffic (79), most likely associated with an "inverted macropinocytosis" mechanism required for extracellular vesicles transport across the plasma membrane and into the cell wall (80). Previous vesicles studies in C. neoformans have reported invaginations of the plasma membrane (81) and revealed an intimate association with sphingolipids and sterols (26). The exact nature of lateral membrane microdomains involved in the vesicular transport of C. neoformans and the role of lipids in this context remains to be elucidated.
Mammalian melanin synthesis is enzymatically driven and occurs in membraneenclosed compartments called melanosomes, which prevents the exposure of most cellular components to toxic or deleterious reactants produced during the process (82). These specialized intracellular organelles are characterized by different developmental stages (I-IV) (25) that progressively incorporate pigment cell-specific components such as Pmel17 (83), "lipid rafts" critical for sorting of tyrosinase and other melanogenic proteins (84). Structurally, eumelanosomes (most pigment content corresponds to eumelanin) are aggregated structures (type-B particles) of 95-307 nm in diameter, depending on the source, formed due to the previous assembly of 10-60 nm subunits of small nanoparticles (type-A particles) (45). Analogous to this system, fungal melanogenesis is also compartmentalized within a sphingolipidenriched vesicle, depends on the laccase enzyme, and leads to the supramolecular buildup of cryptococcal eumelanin particles that are retained within the cell wall, where they provide protection against environmental stressors (3,85,86) and interfere with the host immune responses (20,87).
Our proposed model for melanization in C. neoformans is in consonance with previous observations by our group and others. We have identified organelles showing that pigment synthesis starts intracellularly within melanosomes, enclosed in double-layered compartments, which are then transferred to the cell wall to consolidate as melanin granules anchored within the cell wall. Several pieces of evidence are presented, which can be organized into a testable model . First, laccase is expressed within MVBs as well as in the cell wall (88,89). C. neoformans melanized vesicles measuring between 320 to 380 nm in diameter are shown to incorporate L-DOPA as an electron-dense material (27), suggesting an explanation to the increased number of multivesicular protrusions seen in melanized versus non-melanized cells (60) (Figure 12A). Next, aggregated melanosomes ~60 nm in diameter ( Figure 12B, lower inset) may rely on MCP-like regions of the plasma membrane that generate finger-like protrusions that mediate melanosomes' traffic to the cell wall (Figures 12C,D). A previous study revealed that the major capsular component (GXM, glucoronoxylomannan) was found in association with endosome-like membranous structures possibly involved in GXM synthesis/transport (90). The plasma membrane is a dynamic organelle compartmentalized into two lateral microdomains with specific functions and composition (91). One of these highly conserved domains in S. cerevisiae is closely related to an unconventional mechanism for vesicle secretion termed inverted macropinocytosis (79,80). Lastly, ongoing aggregation of melanin granules within the cell wall is supported by eumelanin biomimetic studies demonstrating that consolidation of type-B particles is due to high amounts of the laccase enzyme (46), which is in agreement with increased laccase localization at the outer edge of the cell wall versus the inner edge (92). Melanin layered deposition on the cell wall follows the existence of membranous sheets described in early ultrastructural studies of the cryptococcal cell wall (93,94) ( Figure 12F); however, as we have demonstrated in this study melanin particles are loosely associated with an acid-labile polysaccharide matrix that makes this layout within the melanized cell wall difficult to demonstrate. Nonetheless, the concentric layered arrangement of these granular particles ~75 nm in diameter was exposed when analyzing melanin ghosts (49,95). The covalent binding of melanin to aliphatic moieties (polysaccharides, chitin/chitosan and lipids) has been shown by ss-NMR analyses (16,19), possibly favored via electrostatic interactions between chitin/chitosan and melanin. Indeed, alterations in the cell-wall chitin/chitosan induced a "leaky melanin" phenotype (31,32,34) or an enhanced deposition of the pigment (14). The recovery of melanin granules in the supernatant of the culture media is explained by normal cell-wall remodeling required for cellular budding (Figures 5A,B;  9A,B), which involves local modifications of the melanin scaffold by secreted peptidases (96), chitinases (97), and glucanases (62) that lead to melanin granules detachment from the cell wall. The model does not address the source of strongly associated lipids with melanin granules, but rather provides new pieces of data that should encourage experiments to continue dissecting the phenomenon of fungal cell wall melanization.
In summary, we note remarkable parallels between cryptococcal and mammalian melanization. We have assembled the available facts into a model for the cryptococcal melanogenesis process that can be tested in future studies. The resemblances in these two systems, are perhaps not surprising given that animals and fungi are close relatives in the tree of life and that the chemistry of melanin synthesis is similar. In both systems, the melanization reaction is contained within a vesicular structure, which protects the host cell from the cytotoxicity of reaction intermediates. The isolation of melanin particles from culture supernatants promises to be a new source of melanin material for studies that further unlock the details of pigment formation and for raw materials in industrial applications include radioprotection, light harvesting, and drug delivery, among various uses.

Strains
In this study, the following strains were used: C. neoformans strain H99 (serotype A), C. neoformans ∆LAC1,2 (H99 background), and C. neoformans ST211A (H99 background) kindly provided by Alexander Idnurm. Each strain was kept in 20% glycerol stocks at -80˚C, and thawed on ice when needed for replication.

Extended acid-hydrolysis of melanin ghosts
Melanin ghosts were isolated following the protocol previously described by Chatterjee et al (16). Briefly, cells grown in MM with L-DOPA were collected by centrifugation and washed twice with phosphate buffer (pH 7.4). Melanized cells were then subjected to a multiple-step enzymatic and chemical treatment to remove cellwall, proteins, and lipid components associated with the melanin polymer. Consequently, the end product is a hollowed particle coated by the acidresistant pigment called a "melanin ghost" (59). The acid treatment step consists of incubation of the melanized particles in 6 N HCl at 80˚C for 1 h. For the extended acid-hydrolysis, melanin ghosts were incubated for up to 4 days including periodic addition of fresh acid solution. Aliquots at 1 h, 24 h and 4 d were analyzed by scanning electron microscopy (SEM). The supernatant from the acid-hydrolysis was also collected, visualized by transmission electron microscopy (TEM) and SEM. Its particle population size was estimated by measuring the diameter using SEM micrographs.

Transmission Electron Microscopy (TEM)
C. neoformans yeast cells or melanin particles were fixed in 2.5% (v/v) glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.3) for 24 h at 4°C. Samples were encapsulated in 3% (w/v) lowmelting-point agarose prior to being processed in Spurr resin following a 24-h schedule on a Lynx tissue processor (secondary 1% osmium tetroxide fixation; 1% uranyl acetate contrasting; ethanol dehydration and infiltration with acetone/Spurr resin). Additional infiltration was conducted under vacuum at 60°C before the samples were embedded in TAAB capsules (TAAB Laboratories, Amersham, United Kingdom) and polymerized at 60°C for 48 h. Semithin survey sections 0.5 μM thick were stained with 1% toluidine blue to identify the areas with the best cell density. Ultrathin sections (60 to 90 nm) were prepared with a Diatome diamond knife on a Leica UC6 ultramicrotome and stained with 2% uranyl acetate in 50% methanol, followed by lead citrate for examination with a JEOL 1200EX (Analytical Imaging Facility at Albert Einstein College of Medicine) or a Philips CM120 at 80 kV (Microscopy Facility at the Johns Hopkins School of Medicine). Negative staining of the fractions from OptiPrep density gradients was performed by adsorbing 10 μL of each fraction to glow-discharged 400 mesh ultra-thin carbon coated grids (EMS CF400-CU-UL) for two minutes, followed by 3 quick rinses of TBS and staining with 1% uranyl acetate with 0.05% Tylose. Images were captured with an Gatan Orlus Digital 2K x 2K CCD camera or an AMT XR80 high-resolution (16-bit) 8 Megapixel camera.

Scanning electron microscopy (SEM)
Melanin particles were suspended in PBS and incubated on coverslips coated with 0.01% polylysine and fixed with 2.5% glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.3) for 24 h at 4°C. Cells and melanin particles were postfixed with 1% OsO4; dehydrated in 70%, 90%, 95%, and 100% ethanol; and critical-point dried in CO2. The cells were sputter coated with gold using an EMitech K550 Sputter Coater and viewed in a LEO/Zeiss Field-emission scanning electron microscope at a voltage of 10 kV.

Dynamic Light Scattering (DLS)
Dynamic light scattering (DLS) techniques gives an estimate of the size and heterogeneity of a sample by measuring the intensity fluctuations of scattered light by particles in solution. Measurement of melanin particles by DLS was performed with a 90Plus instrument (Brookhaven Instruments). To assess the size of melanin ghosts after different sonication times, five hundred microliters of a melanin ghost suspension in distilled water (1 mg/ml) was sonicated with a horned sonicator at amplitude 5 for up to 12 minutes. Particle size of the suspension was measured at different time points. Data are expressed as the average of 10 runs of 1 min data collection each. To examine the colloidal properties as a function of salt concentration, 1M, 0.1 M, 0.01 M, and 0.001 M solutions of Sodium Chloride (NaCl) and Calcium Chloride (CaCl2) and 10X, 1X, 0.1X and 0.01X Phosphate Buffered Saline solutions were used to suspend 1 l of vesicles/granules in 100 l of the respective salt solution. Average hydrodynamic diameter was obtained as above mentioned.

Isolation of melanin granules
The isolation of melanin granules from C. neoformans was performed in two phases: 1) Isolation a mixture of extracellular vesicles and melanin granules from the culture supernatant herein called crude melanin granules, and 2) Isolation of melanin granules by density gradient. For the first phase, yeast cells were pelleted at 5,000 rpm (SLA rotor, Sorvall) for 30 min at 4˚C. The supernatant was filtered through a 0.22 µm pore filter (Millipore Sigma, USA) and pelleted at 100,000 x g for 1 hour at 4 ˚C, with a slow break (SW28 rotor, Beckman Coulter). The pellet was suspended in 1X Dulbecco's Phosphate-Buffer Saline (DPBS) without calcium and magnesium (Corning, USA) and washed twice with the same buffer. The pellet (crude melanin granules) was suspended in approximately 500 l of 0.1X DPBS and stored at -80C until further analysis. For the second phase, OptiPrep solution (60%) dilutions were made in 10 mM HEPES, 0.85% NaCl at pH 7.4. The crude melanin granules were mixed with undiluted OptiPrep to make a 45% solution that was layered at the bottom of the gradient. The gradient was comprised of 45%, 35%, 30%, 25%, 20% and 15% layers in the ratio 0.4:3:3:2:2:1 for a total volume of 3 ml. The gradient was spun at 50,000 rpm for 10 hours (SW Ti55 rotor, Beckman Coulter) to allow the vesicles and granules to equilibrate at their respective densities. Five equal volume fractions were collected from each tube.

Spectroscopic studies UV-Visible analysis
The absorption properties of crude vesicle samples were analyzed with UV-Vis spectroscopy using a SpectraMax M2 microplate reader (Molecular Devices, USA). The spectra were recorded in the wavelength range of 300 to 800 nm.

Electron Paramagnetic Resonance (EPR)
Suspensions of melanin samples (~2 mg) in 150 µl distilled water were sonicated for 1 min using a horned sonicator. Equal volumes of appropriated buffer (distilled water at pH 2.0, 7.0, 12.0 or 0.1 M ZnCl2) was added. Using a glass long-tip transfer pipette, each sample was placed in a 4-mm precision-bore quartz EPR tubes and frozen by immersion in liquid nitrogen. The EPR spectra were recorded at 77 K using a Bruker EMX EPR spectrometer operating in X-band mode. Experimental parameters were as follows: modulation amplitude, 10 G; microwave power, 6.362 mW; modulation frequency, 100 kHz; microwave frequency, 9.43 GHz; attenuation, 15 dB; number of scans averaged, 20. Data acquisition was performed using WinEPR software (Bruker). All spectra were obtained under identical instrumental conditions. Light irradiation of samples was carried out at room temperature for 20 min using a 250 W cool white light operating at 4000K.

Solid-State Nuclear Magnetic Resonance (ss-NMR)
Solid-state NMR experiments were carried out on a 600 MHz Varian (Agilent) DirectDrive2 (DD2) spectrometer using a 1.6 mm FastMAS probe (Agilent Technologies, Santa Clara, CA, USA) with a magic-angle spinning (MAS) rate of 15.00±0.02 kHz and a nominal spectrometer-set temperature of 25 ˚C. All spectra were acquired on 13 C-enriched samples obtained from a C. neoformans ST211A cell culture containing [U-13 C6-glucose] and natural abundance L-DOPA as the sole carbon source and obligatory melanization precursor, respectively. Average post-lyophilization sample masses post lyophilization were ~1-3 mg for the pelleted cellular material containing melanin granules and ~5-7 mg for the melanin ghosts. The typical 90 degree pulse_lengths for the 13C cross-polarization (CPMAS) experiments were 1.65 μs for 1 H and 2.25 μs for 13 C, respectively. As described previously (18), 1 H decoupling using the small phase incremental alternation pulse sequence (SPINAL) with a field strength of ~180 kHz was applied during signal acquisition. The 2D 13 C-13 C through-space dipolar assisted rotational resonance (DARR) correlation experiments were collected using a 50 ms mixing time with similar 1 H and 13 C 90 degree pulselengths and decoupling field strength as described for the 13 C CPMAS experiments.

CryoEM
Five microliters of purified C. neoformans melanin granules (~3 mg) was applied to a 300 mesh copper grid coated with a holey carbon film; and vitreously frozen in liquid nitrogencooled ethane using an FEI Vitrobot Mark IV device. The sample was transferred to a Gatan 626/70° cryo-transfer holder; and observed in an FEI Talos S200C transmission electron microscope operating at 200 kV. Images were collected with a Falcon 2 direct-electron detector camera; and processed in Adobe Photoshop CS6 with only linear adjustments in brightness and contrast.
Proteomic analysis using mass spectrometry C. neoformans isolated melanin granules and the equivalent fraction from non-melanized material (~3 mg of each sample) were treated following a protocol described previously (98). Briefly, samples were subjected to in-solution digestion by resuspending in a solubilization buffer [8 M urea, 10% acetonitrile in 100 mM ammonium bicarbonate (pH 8.5)], flash frozen and thawed in a water bath at 30 ˚C for three rounds. Each sample was sonicated briefly on ice (10 s, horned sonicator at maximum amplitude) and mixed using a vortexer. Next, samples were reduced with 10 mM dithiothreitol (DTT) (Sigma) at 51˚C for 1 h and carboxyamido-methylated with 20 mM iodoacetamide (Sigma) for 45 min in the dark at room temperature. Subsequent digestions with Endo-Protease Lys-C (Roche Diagnostics, USA) and trypsin (1:20 enzyme/protein ratio) for 6 h and 10 h, respectively, were performed at room temperature. The reaction was stopped via addition of 10% trifluoroacetic acid (TFA) and incubation for 30 min at 37 ˚C. Lastly, samples were centrifuged for 15 min at 20,800 g, supernatants were individually transferred to low-protein binding 1.5 ml tubes (Eppendorf) and completely dried in acid-resistant CentriVap vacuum concentrator (Labconco). Peptides were resuspended in 40 µl of 0.1% TFA and stored at -80˚C until the mass spectrometric analysis could be performed. Digested samples were cleaned up on an Oasis HLB micro-elution plate and evaporated to dryness. The samples were taken up in 2% acetonitrile, 0.1% formic acid then vortexed and centrifuged before placing into a 96 well plate. The samples were injected using an EasyLC into a QE-Plus (Thermo Scientific) mass spectrometer and eluted over a 90 minute gradient from 100% buffer A (2% acetonitrile, 0.1% formic acid) to 100% buffer B (90% acetonitrile, 0.1% formic acid) with an MS resolution of 70,000 and an MS2 resolution of 35,000 running a top 15 DDA method. Target values for MS and MS2 were set at 3e6 and 1e5 with 100 and 150 milliseconds, respectively and a normalized collision energy of 28. Data were searched against the Cryptococcus neoformans database (Broad Institute, November 2016) using PEAKS version 7 (Bioinformatics Solutions Inc., Waterloo, ON, Canada) using a peptide tolerance of 5 ppm and a fragment ion tolerance of 0.02 daltons. Enzyme specificity was set to semi specific trypsin with 2 missed cleavages allowed and variable carbamidomethyl C, oxidation M, and deamidation N,Q. Search results were reported at a 1% FDR for all peptides.

Fig. 1 Current views about the synthesis of fungal eumelanins. A)
Catecholamine precursors like L-DOPA are oxidized by the polyphenol oxidase laccase to form dopaquinone, that undergoes further oxidation to form DHI or DHICA. DHI and DHICA form locally-ordered oligomers that in turn form planar stacks, presumably with stacking distances of about 4.5 Å (44). The higher order of structure of eumelanin is considered 'disordered', since different planar structures are oriented in diverse geometries to one another resulting from diversity of DHI/DHICA oligomers stabilized by hydrogen bonding, cation-pi and Van der Waals interactions (48). The aggregation of these simpler units lead to formation of the granular structure that we refer to as melanin granules, which have functional groups that interact with cell wall and cell membrane components such as lipids [(CH2)n, depicted in green] proteins and polysaccharides (C=C, depicted in pink) (18). B) Model representing the architecture of the plasma membrane represented by a lipid bilayer, that interact with melanin nanoparticles that aggregate to form larger melanin granules (21). Melanin granules are anchored to chitin, chitosan and beta-glucan components of the cell wall components that underlie the polysaccharide capsule of C. neoformans.          Fig. 11 ssNMR analyses of C. neoformans crude melanin granules demonstrated a weak association to polysaccharides and suggested the presence of peptides or proteins strongly "functionalized" during melanization. A,B) 13 C cross-polarization magic-angle spinning (CPMAS) spectra of crude melanin granules and melanin ghosts isolated from the leaky-melanin strain C. neoformans ST211A, showing dominant signals for polysaccharides (~55-105 ppm) or lipids (~30 ppm), respectively. C,D) 13 C CPMAS spectra of crude melanin granules and melanin ghosts post-acid treatment revealed that lipid signal (~30 ppm) corresponding to fatty acids is intimately associated with melanin particles. E,F) 2D 13 C-13 C DARR spectra of crude melanin granules confirmed augmented polysaccharide content (region ~55-105 ppm) whereas in melanin ghosts these constituents are less prominent. G,H) 2D 13 C-13 C DARR spectra of crude melanin granules and melanin ghosts post-acid treatment revealed that both samples are composed of the same cellular constituents and suggest the presence of lipids and possibly protein remnants (~53 x175 ppm) that might be strongly associated with the melanin polymer. Fig. 12 Illustrative model of melanization in C. neoformans. Melanogenesis in C. neoformans is characterized by multiple events that can be grouped into four phases: synthesis, transport, aggregation-deposition, and remodeling. Intracellular biosynthesis of the polymer occurs within melanosomes (nanospheres 10-20 nm in diameter) enclosed in multivesicular bodies (MVBs) (60,81) (A) and laccase-containing vesicles (89) (B, upper inset). These organelles could be signaled to start the biopolymer production via incorporation of the mature Qsp1 peptide (64). Active melanization is evidenced by the presence of melanosomes with variable degrees of pigmentation within these lipid bilayer compartments. Melanosomes forming aggregates ~60 nm in diameter are also noticeable in the cell cytoplasm (B, lower inset). Melanosomes are transported to the cell wall using an unconventional mechanism that involves invaginations of the plasma membrane (79,80,90). Finger-like protrusions of the plasma membrane scoop up melanosomes to transport them from the cytosol to the cell wall (C,D, white arrow heads). Once in the cell wall, aggregation of melanosomes into up to 200 nm melanin granules might be promoted by laccase enzyme delivered to the cell wall via MVB fusion with the plasma membrane (46,81,89) (E), while their arrangement within the fungal cell wall may use as a scaffold concentric membranous sheets reported in basidiomycetes (49,93,94) (F, white arrows). Melanosomes anchor to the cell wall via multiple interactions: i) associative and covalent interactions between them and cell-wall constituents (chitin, chitosan, glucans, lipids), possibly mediated by Blp1-chitin and CNAG_05312glucans/chitin/chitosan (16,18,19), and ii) electrostatic interactions modulated by chitosan-melanin charges (14,31,32,34). Regular cell growth requires the cell-wall remodeling to allow budding inducing the melanin scaffold degradation/alteration via secretion of of peptidases (96), chitinases (97), and glucanase (62) thus releasing melanin granules and strung-like granules (G) into the extracellular environment. Removal of non-pigmented acid-labile cellwall components exposes the concentric layered arrangement of the melanin granules within the cryptococcal cell wall from melanin ghosts (48,94) (H). Scale bars, 100 nm (main micrographs), and 50 nm (insets).