The effect of L-DOPA on Cryptococcus neoformans growth and gene expression

Abstract

Cryptococcus neoformans is unusual among melanotic fungi in that it requires an exogenous supply of precursor to synthesize melanin. C. neoformans melanizes during mammalian infection in a process that presumably uses host-supplied compounds such as catecholamines. L-3,4-dihydroxyphenylalanine (L-DOPA) is a natural catecholamine that is frequently used to induce melanization in C. neoformans and L-DOPA-melanized cryptococci manifest resistance to radiation, phagocytosis, detergents and heavy metals. Given that C. neoformans needs exogenous substrate for melanization one question in the field is the extent to which melanin-associated phenotypes reflect the presence of melanin or metabolic changes in response to substrates. In this study we analyze the response of C. neoformans to L-DOPA with respect to melanization, gene expression and metabolic incorporation. Increasing the concentration of L-DOPA promotes melanin formation up to concentrations > 1 mM, after which toxicity is apparent as manifested by reduced growth. The timing of C. neoformans cells to melanization is affected by growth phase and cell density. Remarkably, growth of C. neoformans in the presence of L-DOPA results in the induction of relatively few genes, most of which could be related to stress metabolism. We interpret these results to suggest that the biological effects associated with melanization after growth in L-DOPA are largely due to the presence of the pigment. This in turn provides strong support for the view that melanin contributes to virulence directly through its presence in the cell wall. 

Introduction

Many pathogenic fungi produce melanin in the cell wall. Melanin enhances the virulence of fungi in various ways. In the rice blast fungus, Magnaporthe grisea, melanin strengthens the cell wall, allowing for the buildup of high turgor pressure that aids in invasion of plant tissues.1 In Paracoccidioides brasiliensis, melanin interferes with normal macrophage function.2,3 In Exophiala dermatitidis, melanin contributes to fungal resistance to oxidative killing by neutrophils.4 In addition melanin increases the resistance of fungi to environmental damage, as in the melanized fungi isolated from the radioactively contaminated Chernobyl reactors.5 Melanin is one of the major virulence factors in Cryptococcus neoformans, a fungus that is a relatively frequent cause of life-threatening meningoencephalitis in immunocompromised individuals.6 Cryptococcosis is one of the more common secondary infections in AIDS patients. There are estimated to be nearly 1 million cases of the disease each year, the majority of which are in sub-Saharan Africa.7

In contrast to most other melanotic fungi that are able to make melanin from endogenous precursors, the synthesis of melanin by C. neoformans requires the addition of exogenous substrate in the medium, such as L-DOPA.8 The initial chemical step in the synthesis of melanin is the oxidation of L-DOPA to dopaquinone and is catalyzed by laccase.9 Subsequent steps are believed to occur spontaneously, producing dihydroxyindole or dihydroxyindole-2-carboxylic acid intermediates that finally polymerize into melanin.10,11 Melanins are amorphous polymers composed of crosslinked phenolic and/or indolic subunits.12 Since the exact composition of subunits in the polymer has not been determined, the definition of a compound as a melanin is based on physical characteristics, including dark color, insolubility in most solvents, resistance to acid, oxidation by hydrogen peroxide and degradation by alkali.13 In C. neoformans, melanin has a complex architecture with multiple layers and a granular surface.14

Despite its importance to pathogenesis, many details of the cell biology of C. neoformans melanization remain unknown. In fact, the inability of C. neoformans to produce its own melanin without exogenous substrate makes this organism an ideal system to study melanization since it is possible to vary conditions to better understand this process. In this study we have explored the conditions that lead to optimum pigment expression and the response of the organism to the presence of substrate. The results provide new insights into the process of cryptococcal melanization that are relevant to understanding the role of melanin in pathogenesis.

Results

Dose-dependent responses of melanization and growth to L-DOPA.

A dose-response assay was carried out with L-DOPA to ascertain the effects on growth and melanization and determine the optimal concentration for the other experiments in the study (Fig. 1). Cells were plated in agar containing 0.1, 1.0 and 10 mM L-DOPA and examined visually for growth and melanization effects over several days. The highest concentration of L-DOPA inhibited growth of cells, producing smaller colonies. No growth differences were observed for cells grown in low or intermediate L-DOPA concentrations. Yeast grown on media with 1 mM L-DOPA produced the darkest colonies. At the highest concentration, less melanization occurred. A concentration of 1 mM L-DOPA was used in the subsequent experiments because it produced the most pigment without compromising growth of cells. Toxicity of the catecholamine norepinephrine was also tested. At a concentration of 10 mM, growth of C. neoformans was completely inhibited.

Timing of melanization.

When cells were cultured in liquid medium, we noted that it often took several days for the cells to turn black, depending on the strain. We reasoned that this delay in melanization could have several explanations. For example, the melanization lag could have indicated that L-DOPA was incorporated very slowly into melanin. Alternatively, it could have indicated that the melanization process took several days to commence, but occurred rapidly once begun. To distinguish between these possibilities, cells were grown for seven days in minimal medium without L-DOPA, and then L-DOPA was added to a concentration of 1 mM. After one or two days, the “old” cultures melanized (Fig. 2). In contrast, when cells were cultured for two days in “new” media containing L-DOPA, they did not melanize. This result suggested that the rate of melanization in culture was limited by the initiation of melanization, not the production of melanin.

Density of cells influenced melanization.

The results with old cultures suggested that cell density and/or age impacted the rate of melanization, since older, denser cultures melanized rapidly and new cultures did not. Cell density effects were tested by concentrating an overnight culture 10- and 100-fold. L-DOPA was added and the cultures were incubated overnight. Only the densest culture turned black; no color change was observed for the less dense cultures (Fig. 3A).

Similar to growth in liquid, the timing at which melanization of C. neoformans cells grown in agar became visually apparent was variable. To test if the timing of melanization in agar was related to density, L-DOPA agar was inoculated with variable amounts of C. neoformans. When grown in agar, melanization occurred faster when more CFUs were plated per culture dish (Fig. 3B).

Genes induced by L-DOPA.

Gene expression in melanizing cells was analyzed. L-DOPA was added to a culture of C. neoformans grown to a cell density of 4 × 10⁷ CFU/ml. After four hours of incubation with L-DOPA, RNA was extracted from the cells and purified. Microarray analysis was carried out comparing cells grown in the presence of L-DOPA to those without L-DOPA. Of the approximately 6,500 genes in C. neoformans, eight were identified with increased expression in the presence of L-DOPA (Table 2). Real-time PCR was used to confirm the induction of the genes. Induction of five of the eight genes was confirmed by real-time PCR analysis (CNB04110, CNK02300, CND03820, CNG04630 and CNB04240). To determine the putative functions of the induced genes, homology searches were carried out by entering the predicted protein sequence into the NCBI BLAST search webpage (www.ncbi.nlm.nih.gov/BLAST). Among these genes, several were putative enzymes with redox functions (Table 2).

To test the relationship of these genes to melanization, we determined their expression in the presence of L-DOPA under non-melanizing conditions. To do so, gene induction was measured by real-time PCR in a C. neoformans strain without laccase (LAC1 deletion strain 2E-TU and the complemented strain, 2E-TUC, of a related genetic background). The same five genes with confirmed increase in gene expression had little or no induction in the laccase deletion strain. The complemented strain had variable expression, perhaps due to differences in laccase expression between this strain and the wild-type (Table 2).

Next, cultures were incubated with L-DOPA for several days and changes in gene expression were evaluated over time (Fig. 4). The maximum levels of induction were considerably higher at these longer times in comparison to the previous experiments in which cultures were incubated with L-DOPA for 4 h (Table 2). For example, gene CNB04110 had a maximum induction of 350-fold after three days of incubation with L-DOPA. A similar pattern of induction was observed for most of the genes, with low expression on day 1, a peak of gene expression on day 3, and gradual decrease over time. This peak of gene expression coincided with the cultures reaching a density at which melanization occurs. Gene induction was generally not observed in the laccase mutant strain.

Uptake of L-DOPA by C. neoformans cells.

The gene expression studies using the LAC1 deletion strain suggested that melanization was required for induction of the genes in the presence of L-DOPA. To further address the issue of whether the gene expression changes were related to melanization, the ability of cells to incorporate L-DOPA was tested. Cells were incubated with ¹⁴C-labeled L-DOPA and the level of incorporation into the cells was determined by liquid scintillation counting. Initial experiments showed a low level of incorporation for all cells (data not shown). To increase the level of L-DOPA incorporation, cells were incubated in starvation media prior to adding ¹⁴C-L-DOPA, a condition known to induce melanization.15 Incorporation was measured at 5 min, 1 h, 4 h and 23 h of incubation. After overnight incubation, the wild-type and laccase complemented strains accumulated much of the ¹⁴C-L-DOPA (77% and 70%, respectively), while the laccase deletion strain and heat-killed cells incorporated a minimal amount (8% and 5%, respectively) (Fig. 5). At all time points except for 5 min, cells of the JEC21 and laccase-complemented strains had significantly greater radiation incorporation than both the laccase deletion and heat killed cells (p < 0.001). These results indicate that only melanizing cells accumulated L-DOPA.

Discussion

Melanization is critical to the virulence of C. neoformans and other pathogenic fungi. However, the cellular events that occur in melanizing cells are not well understood. The purpose of this study was to gain insight into the metabolism of L-DOPA into melanin by examining the response of C. neoformans to incubation with L-DOPA with regard to the timing of melanization and the changes in gene regulation.

Melanization of C. neoformans is affected by cell density. When grown in liquid medium with L-DOPA, C. neoformans cells require several days to a week to turn black. However, when cells are grown for several days prior to the addition of L-DOPA, melanization occurs rapidly after substrate addition. In general, the appearance of dark color in liquid culture coincides with cell densities of approximately 4 × 10⁷ CFU/ml. Cell density effects are similarly apparent when cells are plated on solid medium. Melanization occurs more rapidly as the number of colonies per unit area increases. Together, these results suggest that the rate of melanization is density-dependent. This dependence of melanization on cell density has echoes of a quorum sensing effect. Indeed, these results add to considerable work that has been done in our lab16,17 suggesting that C. neoformans melanization is controlled by quorum sensing. In this regard, a quorum sensing-like phenomenon has been described for C. neoformans mediated by a peptide.18 Although this peptide has not been studied for its effect on the rate of melanization, it has protean changes in C. neoformans metabolism that include effects on melanization19 and it is conceivable that a related mechanism is responsible for our results reported in this manuscript.

Both melanization and growth are inhibited in the presence of a high concentration of L-DOPA or norepinephrine (10 mM). The inhibition of melanization may be due to autopolymerization of the substrate in the medium making it unavailable for use in melanin formation. Examination of the plates shows that the agar surrounding the colonies is dark at this concentration. The inhibition of growth suggests that L-DOPA, or a metabolic product of L-DOPA, presents some toxicity to C. neoformans. This hypothesis is supported by the gene expression data, as discussed below.

These data elucidate the optimal conditions for in vitro melanization of C. neoformans. The relationship between in vitro and in vivo melanization is an important topic about which very little is known. Melanized cryptococcal cells have been observed in the brains of patients as well as in experimental animals.20,21 The in vivo substrate has yet to be identified, but it is hypothesized to be one or a combination of catecholamines such as dopamine, norepinephrine, homovanillic acid and L-DOPA. Several of these compounds are normally present at concentrations of µg/g in brain tissue22,23 and nM in the CNS.24 The distribution of catecholamines in the brain is not uniform. For example, dopamine concentration is particularly high in the basal ganglia,25 a region where C. neoformans is often found.26 The levels of these catecholamines in cryptococcosis patients are not known. Likewise, cell density effects on in vivo melanization are hard to determine. In vivo, cryptococcal cells may not be evenly distributed as in a liquid culture, but may be inside host macrophages27 or a cryptococcoma.28 The effects of immune attack and hypoxia on C. neoformans further complicate the scenario.

One of the unanswered questions in the field is the extent to which growth of C. neoformans in media containing substrate affects fungal metabolism independently of the capacity of substrate to induce melanization. Published studies in melanization have identified genes required for C. neoformans to produce pigment, resulting in a white phenotype when mutated. These include genes in cell wall synthesis, metal ion transport and virulence regulators.29,30 In contrast, our study identifies novel genes that are upregulated in the presence of melanization substrate. Gene induction is only observed in cells capable of melanization, suggesting the genes either have a role in this process or are induced as a consequence of melanization. In addition, the density at which maximal gene expression occurs is similar to the density at which melanization becomes visible in liquid culture. Furthermore, analysis of L-DOPA incorporation by C. neoformans cells shows that only wild-type cells capable of melanization incorporate the substrate. This suggests that L-DOPA does not accumulate inside cells and is unlikely to have an effect on gene expression that is unrelated to melanization. Together, these data support the hypothesis that the identified genes play a role in melanization.

We hypothesize that the genes identified function in regulating melanization and/or protecting cells from toxic by-products produced during melanization. In general, melanin and its synthetic intermediates are highly reactive molecules. For example, the intermediate dopaquinone can react with amine and thiol groups, such as in proteins, resulting in crosslinking of melanin and other molecules.31 Thus, it seems logical that melanization is regulated in some way. The toxicity hypothesis is supported by published data showing that two of the genes, CNB04110 and CNB04240, are also induced by nitric oxide stress.32 Furthermore, homology searches also provided information consistent with the toxicity hypothesis. CNG04630 has homology to permeases of the major facilitator family, proteins involved in removal of toxic substances from the cell, whereas CNK02300 has homology to glutathione transferases, proteins involved in metabolizing toxins inside cells.

Perhaps the most surprising aspect of the gene expression analysis is the paucity of genes that are differentially regulated in the melanized state. Given that melanization is associated with diverse new attributes for melanized cells such as resistance to radiation,33 heavy metals,34 oxidants,35,36 enzymatic degradation,37 antifungal drugs38,39 and defensins,40 one might have expected melanin synthesis to be associated with more global changes in gene expression. In fact, the relative paucity of gene expression changes implies that melanization in C. neoformans is a relatively simple system from a genetic point of view and this, in turn, supports the association of new cellular properties in melanized cells with the formation of pigment rather than substrate-induced metabolic changes. The paucity of gene responses is also consistent with our recent proposal that melanin is synthesized in vesicles that are subsequently exported to the cell wall41 since this model involves pigment synthesis from highly reactive and toxic intermediates of L-DOPA oxidation in a confined membrane enclosed compartment that shields cellular processes. A similar strategy for avoiding melanin precursor toxicity is used by mammalian cells, which make melanins within melanosomes. In this regard, defective melanosomes have been associated with necrosis in melanoma cells.42 Nevertheless, the observation that most differentially regulated genes are associated with redox functions suggests that C. neoformans growth in the presence of L-DOPA is a stressful condition for this microbe. In summary, our results show that melanization in C. neoformans is a cell density dependent process that is associated with the induction of very few genes relative to the protean effects conferred by pigmentation that reduce fungal cell susceptibility to numerous insults. This in turn provides strong evidence for the view that the pigment itself is responsible for many protective effects associated with melanization and further strengthens the connection between melanin synthesis and virulence.

Materials and Methods

Fungal strains.

JEC21 is a serotype D, MATα strain that was produced by crossing an environmental isolate with a clinical isolate.43 The serotype A, MATα strain H99 is a clinical isolate.44 The serotype D LAC1 deletion strain (2E-TU) contains a partial deletion of LAC1 near the 5′ end of the gene, and the complemented strain (2E-TUC) contains an integrated copy of LAC1.45 The laccase deletion and complemented strains are related to JEC21.

Growth and melanization assays.

C. neoformans cells were grown in chemically defined minimal medium (15 mM dextrose, 10 mM MgSO₄, 29.4 mM KH₂PO₄, 13 mM glycine and 3 µM thiamine, pH 5.5) with or without 1 mM L-DOPA (Sigma-Aldrich) and incubated at 30°C, 150 rotations per minute (RPM) in the dark. Cultures were examined daily to monitor growth and pigment production.

Cell density assays.

Cell densities were determined by hemocytometer count and/or plating on agar. For the plate assay, the indicated numbers of CFUs were plated on chemically defined minimal medium agar containing 1 mM L-DOPA. For the liquid assay, an overnight culture grown in chemically defined minimal medium at a density of 3 × 10⁵ CFU/ml was concentrated 10- and 100-fold. L-DOPA (1 mM) was added to the concentrated cultures and they were incubated for 18 h before photographing. All incubations were done at 30°C, 150 RPM in the dark.

Microarray analysis of genes induced by L-DOPA.

C. neoformans cells were grown in minimal medium without substrate at 30°C, 150 RPM to a density of about 4 × 10⁷ CFU/ml. L-DOPA was then added at a concentration of 1 mM and the cells were allowed to grow for an additional four hours. After the incubation, cells grown in the presence of L-DOPA were slightly grayish in color, indicating that melanization was occurring in the cells. Two independent biological replicates were performed. Total RNA was isolated from the cultures using the RNeasy kit (Qiagen) and incubated with DNase (GenHunter) to remove contaminating DNA. The purified RNA was sent to Washington University Genome Sequencing Center for microarray analysis with the JEC21 genomic microarray as follows. Comparisons were done with both RNA pools and a Cy3–Cy5 dye swap was done between the sample with and without L-DOPA. Immediately after hybridization, the microarray slides were scanned on a ScanArray Express HT Scanner (Perkin Elmer) to detect fluorescence. Gridding and analysis of images was performed with ScanArray Software Express V2.0 (Perkin Elmer). Gene expression data were averaged across the two RNA pools and analyzed using the GeneSpring 7.2 software (Agilent) to identify genes in which the mean of the replicates had >2-fold change and p < 0.05. The Benjamini & Hochberg false discovery rate for multiple testing corrections was then done for all genes showing changes in gene expression.46

Real-time PCR analysis of identified genes.

L-DOPA was added to C. neoformans cells of strain JEC21, 2E-TU or 2E-TUC at densities of between 2 and 5 × 10⁷ CFU/ml and RNA was isolated after four hours as described above. cDNA was made by reverse transcription of two independent pools of RNA using the Quantitect Reverse Transcription kit (Qiagen). For the time course analysis, L-DOPA was added to cells at a density of 1 × 10⁴ CFU/ml (day 0) and RNA was isolated days 1, 3, 5 and 8. Primers were designed based on the JEC21 genome sequences to amplify approximately 100 base pairs (Table 1). PCR products were amplified with SYBR^® Green PCR Master Mix (Applied Biosystems) in an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). Fold change in gene expression was determined relative to untreated cells of the same strain according the method of Pfaffl.47 The gene encoding glyceraldehyde-3-phosphate dehydrogenase was used as the reference.

[¹⁴C]L-DOPA incorporation analysis.

Cells of C. neoformans strains JEC21, 2E-TU and 2E-TUC were cultured for two days in chemically defined minimal medium at 30°C, 150 RPM. Cells were collected by centrifugation, washed and suspended in starvation medium (0.2 g/l, K₂HPO₄, 0.1 g/l KH₂PO₄). The cells were then incubated overnight at 30°C and 150 RPM. Before incubation with the labeled substrate, cells were collected by centrifugation again and suspended in starvation medium to a final density of 1–2 × 10⁸ CFU/ml. Heat-killed cells were incubated at 65°C for 1 h. Cell suspensions were incubated with L-3,4-dihydroxyphenyl[3-¹⁴C]alanine [4 µCi of a 54 mCi/mmol solution (GE Healthcare)] such that the concentration of L-DOPA was approximately 0.005 mM. At indicated time points, aliquots were removed and briefly centrifuged to pellet cells. The pellets were washed twice with PBS to remove unincorporated label. Liquid scintillation counting was used to determine the total counts per minute (CPM) in the supernatants and pellets (LKB Wallac 1217 Rackbeta, PerkinElmer). Percent incorporation was calculated using the formula: CPM pellet ÷ (CPM pellet + CPM supernatant) × 100. A multivariate analysis of variance (MANOVA) was used to test for incorporation differences between C. neoformans strain and time. Simple contrasts were then done to determine significant differences among strains at a particular time.

Figures and Tables

Figure 1 Melanization dose response of C. neoformans to L-DOPA and Norepinephrine. Cells of strains JEC 21 (top row) and H99 (bottom row) were serially diluted and plated on media with or without the indicated concentration of substrate (mM). Plates were incubated for four (L-DOPA) or seven (Norepinephrine) days before photographing.

Figure 2 Effect of culture age on the rate of L-DOPA-mediated melanization. “New” indicates L-DOPA was added to cultures at the time of inoculation and incubated two days. “Old” indicates L-DOPA was added to a seven-day culture of C. neoformans and incubated for one or two more days. Left tube, strain JEC 21; right tube, strain H99.

Figure 3 Effect of cell density on melanization. (A) C. neoformans strain JEC 21 was grown overnight in chemically defined minimal medium and concentrated before adding L-DOPA. Flasks were photographed after overnight incubation with L-DOPA. Cell densities of the cultures before adding L-DOPA were 6.6 × 10⁵, 4.6 × 10⁶ and 3.6 × 10⁷ CFU/ml. (B) L-DOPA agar was inoculated with 12, 135 or 1,250 CFUs of C. neoformans strain JEC 21 (top) or 14, 90 and 1,400 CFU of strain H99 (bottom) and incubated seven days at 30°C.

Figure 4 Fold changes in gene expression over time. C. neoformans wild-type (top) and laccase deletion (bottom) cultures were incubated with or without L-DOPA for 1–8 days. RNA was isolated from the cells at the indicated times for gene expression analysis by real-time PC R. The experiment was repeated twice. The results of one representative time course are shown. The fold changes in gene expression in the presence of L-DOPA are indicated for each gene.

Figure 5 Incorporation of L-DOPA by cells. Assays were performed as described in the Methods and the percentage of L-DOPA incorporated calculated at indicated time intervals. Strains used were JEC 21 (WT), 2E-TU (Δlac1), 2E-TUC (COMPL.) and heated killed JEC 21 (HK).

Table 1 Primers used for real-time PCR amplification

Locus	5′ primer	3′ Primer
CNB04110	GTC GTC ATT TCG GCC ATT	GAC CAG GGA TGC TGA TTT CT
CNK02300	AAC ACT GGA TTG ATC CGA CA	CTT GAT GTG GTC GAA ATT GG
CND02380	GCT GAC CGA GTC GGC GTC	GAC GAG AGT CAC CGG TAG TG
CND03820	TGC GTT TGT TAC TGA AAC CC	GCA TCC TCC TCT TCC TCT TG
CNG04630	GAG CGC TAC ATT CC T GAT GA	GAA GCG AAT GCT AGT GAG TAG G
CNB04240	GTT GGC AAG TTT GTT TCC CT	ACC AAC AGA GGG CTC AGA GT

Table 2 Fold changes in gene expression upon addition of L-DOPA to cultures

Locus	Putative function	Microarray		Real-Time PCR
		Wild-type	Wild-type	LAC1 deletion	Complemented
CNB04110	Isoflavone reductase	12.9	16.6	0.85	4.2
CNK02300	Glutathione transferase	3.9	4.2	2.2	40
CND02380	NADH flavin oxidoreductase	2.7	1.6	1.5	4
CND03820	Myosin I binding protein	2.1	2.5	0.77	1
CNG04630	Permease	2.1	4.8	0.86	1.4
CNB04240	Benzoquinone oxidoreductase	2.0	2.1	0.87	1.6
CND00360	Unknown	6.7		Undetectable
CNM01700	Unknown	2.0		Not determined

Acknowledgments

This work was supported (in part) by a grant from The City University of New York PSC-CUNY Research Award Program to H. Eisenman. A. Casadevall is supported by NIH grants HL059842, AI033774, AI033142 and AI052733. In addition, we thank the Washington University Genome Sequencing Center for performing the microarray analysis.