Cryptococcus neoformans melanization incorporates multiple catecholamines to produce polytypic melanin

Melanin is a major virulence factor in pathogenic fungi that enhances the ability of fungal cells to resist immune clearance. Cryptococcus neoformans is an important human pathogenic fungus that synthesizes melanin from exogenous tissue catecholamine precursors during infection, but the type of melanin made in cryptococcal meningoencephalitis is unknown. We analyzed the efficacy of various catecholamines found in brain tissue in supporting melanization using animal brain tissue and synthetic catecholamine mixtures reflecting brain tissue proportions. Solid-state NMR spectra of the melanin pigment produced from such mixtures yielded more melanin than expected if only the preferred constituent dopamine had been incorporated, suggesting uptake of additional catecholamines. Probing the biosynthesis of melanin using radiolabeled catecholamines revealed that C. neoformans melanization simultaneously incorporated more than one catecholamine, implying that the pigment was polytypic in nature. Nonetheless, melanin derived from individual or mixed catecholamines had comparable ability to protect C. neoformans against ultraviolet light and oxidants. Our results indicate that melanin produced during infection differs depending on the catecholamine composition of tissue and that melanin pigment synthesized in vivo is likely to accrue from the polymerization of a mixture of precursors. From a practical standpoint, our results strongly suggest that using dopamine as a polymerization precursor is capable of producing melanin pigment comparable to that produced during infection. On a more fundamental level, our findings uncover additional structural complexity for natural cryptococcal melanin by demonstrating that pigment produced during human infection is likely to be composed of polymerized moieties derived from chemically different precursors.

Melanin is a major virulence factor in pathogenic fungi that enhances the ability of fungal cells to resist immune clearance. Cryptococcus neoformans is an important human pathogenic fungus that synthesizes melanin from exogenous tissue catecholamine precursors during infection, but the type of melanin made in cryptococcal meningoencephalitis is unknown. We analyzed the efficacy of various catecholamines found in brain tissue in supporting melanization using animal brain tissue and synthetic catecholamine mixtures reflecting brain tissue proportions. Solid-state NMR spectra of the melanin pigment produced from such mixtures yielded more melanin than expected if only the preferred constituent dopamine had been incorporated, suggesting uptake of additional catecholamines. Probing the biosynthesis of melanin using radiolabeled catecholamines revealed that C. neoformans melanization simultaneously incorporated more than one catecholamine, implying that the pigment was polytypic in nature. Nonetheless, melanin derived from individual or mixed catecholamines had comparable ability to protect C. neoformans against ultraviolet light and oxidants. Our results indicate that melanin produced during infection differs depending on the catecholamine composition of tissue and that melanin pigment synthesized in vivo is likely to accrue from the polymerization of a mixture of precursors. From a practical standpoint, our results strongly suggest that using dopamine as a polymerization precursor is capable of producing melanin pigment comparable to that produced during infection. On a more fundamental level, our findings uncover additional structural complexity for natural cryptococcal melanin by demonstrating that pigment produced during human infection is likely to be composed of polymerized moieties derived from chemically different precursors.
Cryptococcus neoformans is an encapsulated, free-living fungal organism that can establish opportunistic infections in both plant and animal hosts (1). C. neoformans is broadly prevalent in the environment and colonizes a variety of ecological niches, which notably include soil and bird guano (2). Despite ubiquitous exposure of the human population, C. neoformans rarely causes disease in individuals with intact immunity as the infection is either cleared or becomes asymptomatically latent (3). However, C. neoformans poses a serious threat to those with compromised immunity resulting from disease or medical treatment, especially AIDS patients, and is estimated to cause 180,000 deaths annually (4). Cryptococcal infection commences with the inhalation of desiccated fungal cells or spores that become deposited in the lungs, which can result in cryptococcal pneumonia for immunocompromised individuals (5). Progression of the primary infection can occur through extrapulmonary dissemination as yeast cells dispersed through the bloodstream are able to cross the blood-brain barrier and take up residence in brain tissue (6,7). The ensuing meningoencephalitis that is typical of this aggressive stage of cryptococcosis, if left untreated, is almost always fatal (8). Although concerted efforts are underway to develop an effective vaccine (9), there is currently no reliable means of preventing infection and even the most current treatment strategies that employ a combined antifungal regimen only decrease 10-week mortality rates to 24% (10). Thus, an improved understanding of factors that contribute to fungal virulence is vital to our efforts in combating this harmful pathogen.
Among the adaptations that safeguard C. neoformans from harsh conditions both in the environment and during host infection is the deposition of a layer of melanin in the fungal cell wall. Melanins are a diverse group of pigments, typically black or dark brown in color, which are formed by the oxidative polymerization of phenolic or indolic precursors (11). They are notorious for their extreme recalcitrance, structural complexity, and unique biophysical properties, such as the ability to absorb a wide spectrum of UV light. Melanin pigments are widely employed throughout the biosphere, serving such diverse functions as camouflage, thermal modulation, and protection from radiation (12); in pathogenic fungi, they are linked to increased virulence (13) and drug resistance (14). In C. neoformans cells, melanin is synthesized from phenolic compounds in small vesicles called melanosomes that are transported and subsequently deposited within the cell wall (15). There, the melanin particles assemble into larger granules that become tightly interwoven with the constituents comprising the cell wall, which consequently thickens over time (16). In the environment, melanization is credited with protecting C. neoformans from electromagnetic radiation, temperature stresses, and amoeba (17,18). During infection, melanization helps shield C. neoformans cells from engulfment and killing by host macrophages (19)(20)(21) and contributes to the severity of infection by promoting neurotropism (22,23). As melanization is one of two primary virulence factors for C. neoformans, the disruption of this process, perhaps by inhibiting the laccase enzyme responsible for catalyzing melanin synthesis, has been widely recognized as a potential therapeutic target (24).
A defining feature of the C. neoformans laccase enzyme is its inability to produce melanin pigments from endogenously synthesized compounds such as tyrosine, the precursor most commonly used by other melanotic organisms (25). Instead, C. neoformans draws on exogenous sources of substrates, producing melanin pigments from a wide variety of phenolic compounds found in natural habitats and in animal hosts, as evidenced by the isolation of melanized cells from both natural isolates (26) and infected host tissues (27,28). Indeed, the animal nervous system serves as a rich source of precursors in the form of catecholamines, nitrogen-containing diphenolic compounds, which include the neurotransmitters dopamine, epinephrine, and norepinephrine. C. neoformans melanization has been studied in vitro by culturing cells in media supplemented with each of these compounds and more commonly, with 3,4-Dihydroxy-L-phenylalanine (L-DOPA), the immediate biosynthetic precursor of dopamine. Considerable insight into the properties of C. neoformans melanin has been gained by applying biophysical techniques such as solid-state nuclear magnetic resonance (NMR) and electron paramagnetic resonance (EPR) to analyze melanin containing particles derived from such cultures (29)(30)(31). These analyses have indicated that the pigments produced by C. neoformans from each of these precursors are indeed melanins. They have also revealed distinct differences in molecular structure and EPR signal intensity, but the implications of these differences with respect to virulence remain unclear.
Whereas previous studies of C. neoformans melanin have largely employed L-DOPA as the precursor (reviewed in (32)), the low abundance of this compound in host tissues prompted us to question how faithfully L-DOPA melanin produced in culture reflects the structure and function of "natural" melanin produced during the course of C. neoformans infection. The progression of cryptococcal disease ultimately results in brain dissemination, which provides a rich and varied source of catecholamines, thereby presenting the potential for simultaneous incorporation of multiple precursors. Although acid-resistant melanin particles can be isolated from infected human and animal tissue (28,33), the limited quantity of melanized material that can be recovered from animal models of infection has precluded extensive biophysical analyses. To circumvent this obstacle, we sought to develop a cell culture system that would approximate the repertoire of catecholamine precursors present in the human brain. Drawing on previous neurochemical studies that quantified catecholamines isolated from postmortem human brains (34,35), we melanized C. neoformans in a mixture of three catecholamines according to their relative brain proportions of 60% dopamine, 33% norepinephrine, and 7% epinephrine. Here we describe the isolation and analysis of melanin produced from this brain catecholamine mixture (BCM) and provide evidence that C. neoformans can incorporate multiple precursors simultaneously during melanin synthesis. BCM is found to be comparable to L-DOPA melanin in its ability to protect C. neoformans from sources of cellular damage such as UV radiation and free oxygen radicals. Despite these functional similarities, the ability of C. neoformans to utilize a mixture of chemically distinct precursors during melanin synthesis implies architectural heterogeneity for this polymer depending on the local abundance of precursors and significantly complicates its structural characterization.

C. neoformans KN99α variant achieves robust melanization in conditions mimicking the human brain environment
We sought to approximate the melanization process during infection by growing C. neoformans in a mixture of catecholamines representative of human brain proportions (BCM). A key consideration in this effort was the selection of a laboratory strain variant that would be capable of robust melanin production from BCM at the human body temperature of 37 C. Several variants derived from the widely used H99 strain of C. neoformans var. grubii (H99O), including H99E, H99C, and H99W, display reduced virulence in mice and decreased melanization on L-DOPA agar as a result of laboratory passage (36). Conversely, variants H99S, H99F, and KN99α/a that were derived from passage of H99 through the rabbit central nervous system model are more virulent and produce more melanin when grown on L-DOPA (36). We compared melanization by H99O and KN99α on agar supplemented with BCM as well as the individual precursors DA, NE, and L-DOPA at 30 C and 37 C (Fig. 1A). The well-characterized H99-derived ΔLAC1 strain deficient in Laccase-1, the primary enzyme required for melanization, was included as a negative control. C. neoformans laccase production and activity has been shown previously to decrease with increasing temperature (37,38). Accordingly, we observed a slower rate of melanin production and a diminished amount of overall pigmentation for both H99O and KN99α grown at 37 C compared with 30 C, regardless of which precursors were provided. Whereas H99O melanization was delayed compared with that of KN99α at 30 C, similar degrees of pigmentation were observed for the two strains after incubation for 48 h (Fig. 1A, left panel). However, H99O cells showed significantly less pigmentation than KN99α after 48 h at 37 C (Fig. 1A, right panel; Fig. 1B) and only achieved equivalent melanization levels after an additional 2 days of growth (data not shown). Melanization by KN99α on BCM-supplemented agar was comparable to the pigmentation observed with each of the individual catecholamines at 30 C (Fig. 1C, left graph) and was more efficient than NE alone at 37 C (Fig. 1C, right graph). Given the increased melanin production of KN99α compared to H99O when grown in the presence of BCM at the human body temperature of 37 C, the KN99α variant of C. neoformans was chosen for all subsequent analyses.
Melanin-containing "ghosts" can be isolated from BCMsupplemented cultures and brain catecholamine extract plates To further characterize the melanin produced by C. neoformans when provided with a mixture of catecholamine precursors, KN99α was melanized in liquid culture for 10 days at 30 C, which is the temperature of optimal growth, in MM supplemented with BCM (0.6 mM DA/0.33 mM NE/ 0.07 mM E) or either 1 mM DA or 1 mM NE. Pigmentation intensity increased more rapidly for cells grown in BCM compared with DA initially but both cultures reached maximal melanization levels after 5 days ( Fig. 2A). The culture grown in NE melanized more slowly but eventually reached an intensity comparable to the other two cultures after 10 days ( Fig. 2A). Melanization was also achieved for cultures grown in MM supplemented with BCM, DA, or NE at the human body temperature of 37 C, although pigment production occurred more slowly and at a comparable rate for all three cultures (Fig. 2B).
To obtain melanin for biophysical analysis, cells from each culture were harvested and subjected to enzymatic cell wall digestion, protein denaturation and hydrolysis, lipid extraction, and finally boiling in concentrated HCl. In each case, acidresistant material was recovered that, upon examination by light microscopy, was verified to be dark, yeast-shaped and yeast-sized silhouettes referred to as melanin "ghosts" (Fig. 2C). These particles were also confirmed to be nonviable since plating >1 × 10 7 particles on SAB agar yielded no colonies. Microscopic examination revealed differences in the physical appearance of particles derived from each culture. The melanin ghost particles isolated from the DA control culture grown at 30 C were darkly pigmented ( Fig. 2C) and similar in appearance to those isolated previously from cultures melanized in L-DOPA at 30 C (39). Fewer intact particles were recovered from the corresponding NE control culture, and they tended to be small and pale (Fig. 2C); there was an overall lower yield of acid-resistant material (70 mg/l) compared with that of DA (130 mg/l). The BCM melanin ghosts from the culture grown at 30 C were intermediate between the two controls both in appearance, with a mixture of intact dark particles and more fragile, pale particles ( Fig. 2C), and in terms of the yield of material (100 mg/l). Ultrastructural examination of melanin ghost particles using transmission electron microscopy revealed "empty shells" with no evidence of membrane-bound organelles or polysaccharide capsule that were visible in the corresponding pretreatment samples (Fig. 2D).
Melanin ghost particles isolated from cultures melanized at 37 C were smaller and reduced in yield by several-fold (37 mg/l for DA and 14 mg/l for BCM and NE) compared with those isolated at 30 C. The reduced yield is consistent with lower expression of laccase enzyme at 37 C (38). An even smaller yield of intact acid-resistant particles (<1 in 10 6 of the total number of cells) was recovered from KN99α cells grown at 30 C on agar Statistical significance for this and all subsequent analyses was calculated using an ordinary one-way ANOVA (ns = not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). BCM, brain catecholamine mixture; MM, minimal media.
plates containing brain tissue extract (Fig. 2E). In light of this limitation, the robust pigment production achieved in our largescale BCM-supplemented culture system, especially at 30 C, supports its utility for acquiring melanin that approximates that produced in human brain tissue but in sufficient quantities to permit structural characterization.
Solid-state NMR suggests that cells provided BCM incorporate more than one precursor into the melanized cell wall architecture Given the strong resemblance of the particles derived from C. neoformans grown in BCM to melanin ghosts, we sought to characterize the molecular profile of this pigmented material.
Solid-state NMR (ssNMR) analysis was implemented to examine its molecular architecture and compare it with that of melanin ghost material isolated from cells grown with either of the two major constituent precursors, DA or NE. The 1D 13 C cross-polarization magic-angle spinning (CPMAS) spectra of each of these three samples clearly displayed all of the spectral features that are characteristic of melanin ghosts ( Fig. 3) (29)(30)(31), notably the group of broad, overlapping peaks located between 110 and 160 ppm. These peaks correspond to the various aromatic carbons that comprise the inherently heterogeneous melanin polymer: although many of the carbons are chemically similar, they differ in both their bonding patterns and local chemical environments and consequently resonate at slightly different frequencies. This results in a multitude of signals that overlap in the aromatic-carbon spectral region, forming what visually appears as a single, broad peak and serves as a telltale indicator of the presence of melanin pigments. Other characteristic features of the cellular components of melanin ghosts were also observed in all three sample spectra: each of the spectral regions between each 10 to 40 ppm and 50 to 110 ppm displayed groups of peaks, which are attributable to long-chain fatty acids within triacylglycerols and polysaccharide-ring carbons, respectively (40). The exceptional recalcitrance of melanin pigments, which in C. neoformans are deposited and intimately interwoven throughout the cell wall, protects this structure from the harsh treatments performed during melanin isolation. Hence, this process yields melanized cell walls (i.e., melanin "ghosts"), which contain remnants of the polysaccharide scaffold in addition to trapped neutral lipids, rather than only the purified fungal pigments. Melanin ghost spectra therefore characteristically display ssNMR signals that correspond to nonpigment cellular remnants in addition to those that correspond to the melanin pigment itself. The fact that these signals were observed in all three sample spectra confirmed that the black particles recovered from C. neoformans cell cultures grown with BCM are indeed melanin ghosts.
Our ssNMR analysis also provided insight into whether C. neoformans is able to incorporate more than one melanization precursor into the pigment architecture when provided with multiple substrates. The heterogenous nature of the melanin polymer, which results in multiple broad, overlapping peaks, precludes the identification of specific signals that are unique to a melanin synthesized from DA or from NE. Through visual inspection alone it was evident that the DA, NE, and BCM melanin ghost samples all share common spectral features, but the relative intensities of the groups of peaks corresponding to pigments, lipids, and polysaccharides clearly differed. Notably, the aromatic-carbon signals attributed to the pigment contributed substantially less to the total signal intensity in the NE-only melanin ghost spectrum in comparison to the DA-only spectrum, even though both samples were isolated from cell cultures that contained 1 mM concentrations of the respective precursors. This finding was anticipated since prior studies had demonstrated that precursors such as DA or L-DOPA result in more robust melanin production in comparison to others such as NE or E (29)(30)(31). This, in turn, results in melanin ghosts that have a greater pigment content relative to the other cellular remnants present such as polysaccharide and lipids. The spectrum of the BCM melanin ghosts, which were isolated from a cell culture containing 0.6 mM DA, 0.33 mM NE, and 0.07 mM E, exhibited pigment contributions with respect to other cellular constituents that exceeded expectations based on utilization of only the robust DA precursor, suggesting incorporation of the additional catecholamines. Quantitatively reliable direct polarization (DPMAS) spectra reinforced the trend of higherthan-expected pigment contributions in the melanin ghosts (Fig. S1). Moreover, this trend was also observed in both the CPMAS and DPMAS spectra of melanin ghost samples isolated from C. neoformans cells grown at 37 C despite the overall reduction in melanin production that occurs at this temperature (Fig. S2). Thus, even without detailed discrimination of the pigment structures synthesized from different precursors, our ssNMR analysis suggested that C. neoformans can use at least two different melanization precursors when they are provided simultaneously.
Radiolabeled norepinephrine and dopamine are incorporated into C. neoformans melanin As our structural data provided highly suggestive evidence for the incorporation of multiple precursors, we extended our analysis by employing radiolabeled catecholamines as a sensitive and quantitative means of determining whether both DA and NE can be incorporated into the melanin pigment at the same time. Individual C. neoformans KN99α cultures were grown in the presence of a low concentration (0.0001 mM) of high specific activity 3 H-labeled NE with and without the addition of unlabeled NE or DA. A control culture deficient in the primary laccase enzyme (ΔLAC1) was also supplemented with the same concentration of 3 H-NE as a negative control. Even though the molar concentration of NE in the culture was very low, we detected significantly more radioactivity retained in the cell pellet from the KN99α cell culture grown without any additional unlabeled precursors compared with that of the ΔLAC1 control culture (Fig. 4A, graph). The addition of excess unlabeled NE at a concentration of either 0.1 or 1 mM resulted in a noticeable, concentration-dependent increase in pigmentation (Fig. 4A, lower panels) and a concomitant increase in the incorporation efficiency of 3 H-NE (Fig. 4A, graph). We attribute these increases to the established dose dependence of melanization for DA and NE precursors, possibly resulting from induction of the laccase enzyme (41). Notably, a comparable increase in 3 H-NE incorporation was also detected when cultures were supplemented with excess (1 mM) of unlabeled DA (Fig. 4A, graph), indicating that C. neoformans can utilize more than one type of precursor simultaneously.
To confirm that the radioactivity associated with C. neoformans cell pellets resulted from the incorporation of the radiolabeled precursors into melanin, we heated cells that had been melanized in mixtures of radiolabeled and unlabeled precursors in concentrated hydrochloric acid to generate melanin ghost particles. Approximately 30 to 40% of the radioactivity was retained in melanin from cells labeled with either 3 H-NE or 3 H-DA (Fig. 4B). The efficiency of incorporation of either radioactively labeled precursor into melanin was not significantly different when melanization was supported by an excess of a single unlabeled catecholamine or a 1:1 mixture of NE and DA (Fig. 4B) since in either case, a substantial proportion of the 3 H radioactivity was retained after HCl treatment.

Melanin derived from different precursors provides similar protection from UV radiation and oxidative damage
Melanization of C. neoformans in L-DOPA was previously demonstrated to afford protection from the fungicidal effects of UV radiation (42). To investigate the protective capacity of the brain mixture of catecholamines, KN99α cells were melanized for 7 days in BCM, DA, NE, or L-DOPA and compared with nonmelanized control cells for survival after UV exposure. Cells melanized under all conditions showed significantly higher percentages of survival compared with nonmelanized cells, although melanization in NE was less protective than with the other three precursors (Fig. 5A). Melanization in BCM showed no statistical difference from DA or L-DOPA in shielding cells from UV radiation (p values of 0.8107 and 0.6413, respectively).
Since melanin is a key virulence factor that contributes to C. neoformans survival inside human host cells, we sought to explore the ability of "natural" melanin to safeguard against damaging free radicals generated inside macrophages. An in vitro oxidative system that produces oxygen-derived free radicals (43) was used to treat KN99α cells that were nonmelanized or had been melanized for 7 days in the BCM, DA, NE, or L-DOPA. Melanization in the presence of each of these precursors, with the exception of NE, significantly increased the percentage of cells that survived exposure to oxidative damage at 37 C compared with nonmelanized cells (Fig. 5B).

Discussion
Melanization by C. neoformans is a prominent virulence factor that facilitates evasion from the host immune system. Melanin produced by C. neoformans grown in defined culture media supplemented with the phenolic precursor, L-DOPA, has been well characterized, but much less is known about the protective layer of pigment deposited during infection. In the brain, L-DOPA is readily converted to dopamine, which is in turn a precursor of norepinephrine and epinephrine. Hence, C. neoformans cells that disseminate to brain tissue are  expected to utilize one or more of these three catecholamines for melanin synthesis. We cultured C. neoformans in a mixture of dopamine, norepinephrine and epinephrine reflective of their relative proportions in the human brain and found that the resulting melanin was similar to L-DOPA melanin in its ability to protect cells from UV radiation and oxidative damage. Structural analysis using solid-state NMR revealed a spectrum consistent with the simultaneous incorporation of more than one precursor. Melanization experiments employing radiolabeled catecholamines provided additional evidence that C. neoformans can use multiple precursors simultaneously and suggested that melanin made during infection is likely synthesized from a mixture of the available substrates. Initial evidence for C. neoformans melanization during infection was provided by the isolation of acid-resistant melanin ghost particles from infected animal and human tissue and from cells grown on tissue homogenate agar plates (28,33). We isolated similar particles from cells grown on agar plates infused with a lamb brain catecholamine extract, but only a very small fraction of the total cells was recovered as melanin ghosts. Neurotransmitters such as dopamine in brain tissue are contained within vesicles and degraded by monoamine oxidase, an enzyme found in glial cells (44). Consequently, several factors could have contributed to the small yield when using brain tissue as the source of substrate, including the relatively lower concentrations of catecholamines in brain tissue compared with the concentrations achievable in culture, low bioavailability of the precursors sequestered in neurotransmitter vesicles, and degradation of catecholamines during the extraction procedure when cells are lysed. In light of these limitations, we opted for culturing C. neoformans in a mixture of catecholamines proportional to their relative concentrations present in the human brain and achieved robust melanization on both solid agar substrates and in liquid cultures. We also observed more efficient melanization by the KN99α variant compared with its parental H99O strain, especially at the human body temperature of 37 C, and this is consistent with the correlation between melanization and virulence (36). Although C. neoformans is able to produce melanin pigments from a variety of aromatic substrates, certain precursors, e.g., L-DOPA and dopamine, promote more robust melanization compared with norepinephrine or epinephrine, as evidenced by a greater yield by mass melanin ghosts that have relatively thicker "shells" and a larger stable free-radical population (39). In our prior work, ssNMR analysis was implemented to determine whether C. neoformans is able to distinguish between different melanization precursors when two are provided at the same time (31). We found that C. neoformans will selectively use a more favorable precursor (L-DOPA) over a less favorable one (norepinephrine) when each is provided at a 1 mM concentration. Although we had determined that C. neoformans will preferentially utilize the more favorable L-DOPA precursor, we did not determine whether this occurred with the exclusion of norepinephrine utilization, or if the outcome would differ if the total concentration of both precursors was 1 mM. In our current work, we used ssNMR to examine the melanin ghosts isolated from cells grown in a mixture of catecholamines in the proportions found in the human brain (60% dopamine, 33% norepinephrine, and 7% epinephrine (34,35)) and provided at a total precursor concentration of 1 mM. As observed for ghosts isolated from cultures containing 1 mM of either dopamine or norepinephrine, the 13 C CPMAS spectrum of these "brain catecholamine mix" (BCM) ghosts displayed all the hallmark features of C. neoformans melanized cell walls, including signals that correspond to pigments, polysaccharides, and lipids (Fig. 3). However, the relative intensities of the signals attributed to each of these constituents, which are indicative of their approximate relative quantities, differed among the three ghost spectra. The analysis of quantitatively reliable 13  spectra verified this observation. As anticipated, the ghosts prepared from the culture containing 1 mM of the lessfavorable norepinephrine precursor had the lowest relative pigment content, whereas those prepared from the 1 mM dopamine culture had the highest content. Remarkably, the ghosts isolated from the BCM culture, which only contained 0.6 mM dopamine, were found to have a pigment content nearly equivalent to that of ghosts derived from a culture provided with 1 mM dopamine, and thus suggested that one or both of the other precursors (norepinephrine and epinephrine) present in the mixture were being utilized for pigment formation in addition to dopamine.
Our interpretation of the ssNMR quantitative analysis assumes that 100% of the melanization precursors provided in culture media are utilized for melanin formation. But if the complement of provided precursors influences the efficiency with which a particular precursor is utilized, our results could also be interpreted to suggest increased relative dopamine incorporation for cells provided with the brain catecholamine mixture compared with dopamine alone. To address more directly whether C. neoformans can incorporate more than one catecholamine simultaneously, we cultured cells in the presence of radiolabeled catecholamines to follow their incorporation into melanin. A significantly higher level of radioactivity was detected in the cell pellet of the wild-type KN99α strain compared with a laccase-deficient mutant, ΔLAC1, which suggests that the labeled precursor was used to support melanization. However, the efficiency of incorporation by KN99α was very low at approximately 5%, which is not unexpected given the low molar concentration of 0.0001 mM for the labeled precursor. C. neoformans melanization in culture was previously shown to have a dose dependence for L-DOPA and norepinephrine substrates, with an optimal concentration of 1 mM (41). In this instance, supplementing cultures of KN99α with 0.1 and 1 mM unlabeled norepinephrine in combination with 0.0001 mM 3 H-labeled norepinephrine resulted in a concentration-dependent increase in both visible pigmentation and incorporation of 3 H-NE. Remarkably, melanization with 1 mM unlabeled DA also significantly increased the incorporation of 3 H-labeled norepinephrine, although not to the same extent as 1 mM norepinephrine, possibly reflecting a slight incorporation bias for dopamine. However, the increased efficiency of 3 H-labeled norepinephrine incorporation in the presence of a 10,000-fold molar excess of dopamine supports the conclusion that C. neoformans does not use dopamine to the exclusion of norepinephrine. To confirm that the radiolabeled precursors were being incorporated into melanin pigment and not just peripherally associated with cells, we isolated acid-resistant melanin ghosts from radiolabeled cultures. Using either 3 Hlabeled norepinephrine or 3 H-labeled dopamine, we were able to isolate radiolabeled melanin when cells were melanized in either individual unlabeled catecholamines or an equimolar mixture of the two precursors. Together, these results suggest that the most readily abundant substrate is used to establish a melanin scaffold that is then built upon using any and all available precursors, thereby generating local structural diversity for the melanin particles.
The study of C. neoformans melanization has used predominantly L-DOPA as a substrate, but this precursor would not be the major substrate available in host tissues. C. neoformans infections are initially established in the lungs where they encounter alveolar macrophages. As these and other immune cells are known to produce and release catecholamines (45), it is possible that melanization supported by lung resident catecholamines could enhance virulence and promote brain dissemination. The ability of C. neoformans to incorporate multiple catecholamines simultaneously suggests that melanin synthesized in brain tissue may differ in distinct anatomical regions as the relative proportions of these neurotransmitters have been found to vary widely from one area of the brain to another (34,35). Intriguingly, C. neoformans infections are often found to be concentrated in the basal ganglia region of the brain where dopamine levels are highest (46,47), a finding that could reflect resistance of melanized cells to clearance by host immune mechanisms. Basal ganglia preference is also consistent with the notion that melanin produced by C. neoformans in the brain is primarily dopamine-derived. Consequently, we suggest that future studies of the role of melanin in C. neoformans pathogenesis are more likely to reflect brain infection conditions if dopamine is used as the melanization substrate. However, L-DOPA may be a more suitable substrate to study the role of cryptococcal melanization in the environment where L-DOPA is a common precursor of plant compounds and accumulates in soils (48). We note that although the type of melanin produced by C. neoformans is dependent on the available precursor, our prior ssNMR studies revealed considerable structural similarity among different melanins (29)(30)(31).
An improved understanding of the composition and structure of "natural" melanin produced by C. neoformans during infection will help elucidate its function in virulence. We have shown that C. neoformans can incorporate multiple precursors, even when their molar ratios differ by several orders of magnitude. Given that melanin synthesis is driven by the accretion of highly reactive intermediates into a solid material, it is expected that C. neoformans provided with a mixture of precursors produces polytypic melanin. Furthermore, depending on the relative availability of specific catecholamines, structurally diverse forms of melanin may be produced in different anatomical regions during the course of infection. Despite this added level of complexity, melanins produced in culture from single or mixed substrates were found to be functionally comparable in their capacity to protect C. neoformans from UV radiation and oxidative damage, but whether specific forms of melanin could imbue C. neoformans with a greater survival advantage during infection remains unanswered. Whereas the present study has shed light on initial polymerization events, the resulting melanin compound is a highly complex, amorphous substance for which detailed structural elucidation remains beyond our analytical horizon. Culture media with relative proportions of catecholamines in the human brain To study melanization under conditions representative of the human brain, we used the total cortical and subcortical measurements of the three brain catecholamines reported previously (34,35)

Melanization assay on agar plates
To prepare melanization assay plates, 4× MM with no additives or with 1.0 mM DA, 1.0 mM NE, 1.0 mM 3,4-Dihydroxy-Lphenylalanine (L-DOPA, MilliporeSigma D9628), or BCM was mixed 1:3 with cooled 2% agar (Difco 214530) and poured into individual wells of 24-well plates (2 ml per well). Plates were allowed to cool at RT for 5 h, then 1 × 10 6 PBS-washed KN99α, H99O, or H99-ΔLAC1 cells were pipetted into each well, and plates were incubated at either 30 C or 37 C. After 24 and 48 h, the plates that had been incubated at each temperature were photographed in a single field using a 12-megapixel camera. Images were converted to grayscale in Adobe Photoshop (version 22.0.1) and intensities were quantified using Image Studio Lite software (version 5.2.5, Li-Cor Biosciences). The signal measured for the MM control was applied as the image background and intensity measurements for the H99-ΔLAC1 negative control were subtracted from those of KN99α and H99O to derive pigmentation levels above background.

Isolation of melanin particles
Cultures of KN99α were grown in MM supplemented with BCM, 1.0 mM DA, or 1.0 mM NE at 30 C or 37 C. Volumes of 0.5 l for BCM and DA and 1 l for NE for cultures grown at 30 C were doubled for growth at 37 C. Melanization of the cultures was monitored by photographing and quantifying pigment intensities (as described above) for 2 ml samples that had been transferred to the wells of a transparent 12-well plate (Celltreat, 229112). After incubation for 10 days at 30 C or 16 days at 37 C, acid-resistant melanin "ghost" particles were isolated from cultures as described in detail previously (29,33). Briefly, cells were treated with a 10 mg/ml solution of Trichoderma harzianum lysing enzymes (MilliporeSigma L1412) in 1.0 M sorbitol and 0.1 M sodium citrate for 24 h at 30 C to remove the fungal cell wall. The resulting protoplasts were subjected to protein denaturation with 4.0 M guanidinium thiocyanate for 18 h at RT and then proteolysis with a 20 mg/ ml solution of proteinase K (MilliporeSigma 03115801001) in 10 mM Tris-HCl (pH 8), 5.0 mM CaCl 2 , and 0.5% sodium dodecyl sulfate for 4 h at 65 C. Following lipid removal by three successive extractions with a 4:8:3 ratio of methanolchloroform-saline (49) samples were heated at 90 C in 6 M HCl for 1 h. Melanin-containing particles were recovered by centrifugation and dialyzed for several days against pure H 2 O prior to lyophilization. Dry material was weighed using an analytical balance (Mettler Toledo AT261 DeltaRange) to determine yield.

Melanization in brain tissue extracts
The tissue from three lamb brains (227 g) was homogenized in 1 ml/g ice cold buffer comprised of 0.1 M perchloric acid, 1.0 mM sodium metabisulfite, and 0.1 mM EDTA using a 30 ml Potter-Elvehjem tissue grinder (Wheaton 358011) attached to a drill press. Following centrifugation at 5000 rpm for 20 min at 4 C, the pH of the supernatant was adjusted to 5 by dropwise addition of sodium hydroxide and then passed through a 0.22 μm filter (Millipore SCGPS02RE). The filtrate was rapidly frozen on dry ice and then lyophilized for 72 h. The freeze-dried extract was resuspended with 75 ml 2× MM containing 1% Penicillin-Streptomycin, mixed with 75 ml cooled 3% agar, and poured into six 100 ml petri dishes. After 24 h at RT, plates were spotted with PBS-washed KN99α and incubated at 30 C for 26 days. Melanin ghost particles were isolated from cells melanized on plates as described previously.

Light and transmission electron microscopy
Melanized cells and melanin ghost particles were imaged under bright field illumination with an Olympus AX70 microscope (Olympus) using an oil immersion 100× objective. Images were captured using QCapture-Pro 6.0 software with a Retiga 1300 digital CCD camera. (Teledyne Photometrics). For transmission electron microscopy (TEM), samples were fixed in a buffer comprised of 2.5% glutaraldehyde, 3 mM MgCl 2 , and 0.1 M sodium cacodylate (pH 7.2) overnight at 4 C. After buffer rinse, samples were postfixed in 1% osmium tetroxide, 1.25% potassium ferrocyanide in 0.1 M sodium cacodylate for at least 1 h (no more than two) on ice in the dark. After the fixing step, samples were rinsed in dH 2 O, followed by uranyl acetate (2%, aq.) (0.22 μm filtered, 2.5 h, dark), dehydrated in a graded series of ethanol and embedded in Spurrs (Electron Microscopy Sciences) resin. Samples were polymerized at 60 C overnight. Thin sections, 60 to 90 nm, were cut with a diamond knife on a Leica UCT ultramicrotome and picked up with 2 × 1 mm Formvar copper slot grids. Grids were stained with 2% aqueous uranyl acetate followed by lead citrate and observed with a Hitachi 7600 TEM at 80 kV. Images were captured with an AMT CCD XR80 (8-megapixel side mount AMT XR80 high-resolution high-speed camera).

Solid-state NMR of melanin particles
Solid-state NMR measurements were carried out on a Varian (Agilent) DirectDrive2 (DD2) spectrometer operating at a 1 H frequency of 600 MHz. All data were acquired on 8 to 10 mg of lyophilized material using a 1.6-mm T3 HXY fast-MAS probe (Agilent Technologies) with a magic-angle spinning (MAS) rate of 15.00 ± 0.02 kHz and a spectrometer-set temperature of 25 C. The 1D 13 C cross-polarization (CPMAS) experiments were conducted with 90 pulse lengths of 1.5 and 1.7 μs for 1 H and 13 C, respectively, a 10% linearly ramped amplitude for 1 H and a 1 ms contact time for 1 H- 13 C polarization transfer. The 1D 13 C direct polarization (DPMAS) experiments were conducted with 90 pulse lengths of 1.2 and 1.4 μs for 1 H and 13 C, respectively, and long recycle delays (50 s) were used in order to obtain quantitatively reliable signal intensities. Heteronuclear decoupling was applied during signal acquisition using the small phase incremental alternation pulse sequence (SPINAL) (50) and a radiofrequency field strength of 104 kHz.

Radiolabeled norepinephrine incorporation analysis
KN99α was grown in MM for 2 days at 30 C, after which cells were washed and resuspended in fresh media at a cell density of 2 × 10 8 cells/ml. Cultures were labeled with 2 μCi/ ml (0.0001 mM) norepinephrine DL-[7-3H] hydrochloride (3H-NE, American Radiolabeled Chemicals, Inc ART-0172) for 7 days at 30 C in the absence or presence of 0.1 or 1.0 mM unlabeled NE or 0.1 or 1.0 mM unlabeled DA. H99-ΔLAC1 was also labeled with 3H-NE as a negative control. Color photographs were taken of each culture to document relative pigmentation intensities. Labeled cells were collected by centrifugation at 16,000g for 5 min, washed twice with MM, then resuspended in MM. Samples of media, wash buffer, and resuspended pellets were mixed with UniverSol-ES liquid scintillation cocktail (MP Biomedicals 01882480), and decays per minute (DPM) were quantified using an LS 6500 liquid scintillation counter (Beckman). Percent incorporation was calculated by dividing the DPM measured for the pellet sample by the total DPM of the pellet, media, and wash samples.

Isolation of radiolabeled melanin particles
KN99α was grown as described above in MM supplemented with 2 μCi/ml 3H-NE and 1.0 mM unlabeled NE or a mixture of 1.0 mM unlabeled NE and DA or with 4 μCi/ml (0.0001 mM) Dihydroxyphenylethylamine 3,4-[Ring-2,5,6-3 H] ( 3 H-DA, PerkinElmer NET673250UC) and 1.0 mM unlabeled DA or 1.0 mM unlabeled NE and DA for 10 days at 30 C. Labeled cells were collected by centrifugation at 16,000g for 10 min and then treated with 6 M hydrochloric acid in a 90 C water bath for 1 h. Acid-resistant melanin ghost material was collected by centrifugation at 16,000g for 10 min. DPM were measured as described above for samples taken before and after acid treatment and the percent radioactivity retained in the melanin ghost particles compared with pretreatment samples was calculated.

Exposure to ultraviolet radiation
Equal numbers of nonmelanized and 7-days melanized KN99α were plated on Sabouraud dextrose agar (Difco 210950) and then exposed to 15,000 μJ/cm 2 UV radiation (254 nm) using a Stratalinker 1800 (Stratagene). Percent survival was calculated from CFU counts of treated compared with untreated control plates after incubation at 30 C for 3 days.

Exposure to oxygen-derived oxidants
Cultures of KN99α were grown at 30 C for 7 days in MM (control) or MM supplemented with BCM, 1.0 mM DA, 1.0 mM NE, or 1.0 mM L-DOPA. After washing with PBS, 2.5 × 10 4 cells were treated for 5 h at 37 C with an oxidative system that uses epinephrine as an electron donor, H 2 O 2 as an electron acceptor, and Fe 3+ as a transition metal catalyst (43); reaction among these three components results in the formation of Fe(III)-epinephrine complexes that are a primary source of oxidative damage in biological systems and have been linked to various stress-related pathologies in mammals (51). A fresh stock solution of 5.0 mM Fe 3+ was prepared by mixing 10 mM ferric ammonium sulfate, NH 4 Fe(SO 4 ) 2 , 1:1 with 1.0 mM sulfuric acid. The oxidative treatment solution consisted of 0.5 mM Fe 3+ , 1.0 mM H 2 O 2 , and 1.0 mM epinephrine bitartrate (MilliporeSigma E4375) in a buffer comprised of 50 mM sodium acetate (pH 5.5) and 1.0 mM MgSO 4 . Percent survival was calculated from counts of colony-forming units (CFUs) for each condition compared with untreated controls.

Statistical analysis
Biological replicates were performed for experiments as indicated, and statistical significance was determined using an ordinary one-way analysis of variance (ANOVA) with Sidak's multiple comparisons test (GraphPad Prism 9 software).

Data availability
All data are contained within the article.
Supporting information-This article contains supporting information.