Introduction

The basidiomycetous yeast Cryptococcus neoformans is an important fungal pathogen clinically and causes systemic infections in immunocompromised hosts such as patients with AIDS, resulting in approximate 600,000 deaths annually (Park et al. 2009). As a model system for the study of fungal pathogenesis, several virulence-associated traits have clearly been identified, including the formation of polysaccharide capsule, the production of antioxidant melanin and the ability to grow at 37 °C (Zhu and Williamson 2003; Janbon 2004; Zaragoza et al. 2009).

Several reports have described the importance of copper (Cu) homeostasis in virulence of C. neoformans. As an essential trace nutrient, copper can easily transit between two oxidation states, Cu+ and Cu2+ (Van Ho et al. 2002), providing a suitable chemical environment for various biological processes. Production of the pigment melanin depends on the copper-containing laccase. Endogenous copper induces laccase expression and melanin production in C. neoformans (Zhu and Williamson 2004; Jiang et al. 2009). Disruption of the chloride channel (CLC) also reduces laccase activity, whereas exogenous copper supplementation restores the phenotype (Jiang et al. 2009). Moreover, a mutant of the secretory compartment copper pump (CCC2) also showed reduced virulence and copper deficiency phenotypes (Walton et al. 2005). A copper-dependent transcription factor Cuf1 has been defined to be required for copper induction of laccase in C. neoformans (Jiang et al. 2011). Deletion of CUF1 leads to defects in growth and reduced expression of virulence factors in low copper conditions (Waterman et al. 2007). Cuf1 plays dual roles in both copper detoxification in response to Cu excess and high-affinity uptake under Cu limitation (Ding et al. 2011; Jiang et al. 2011). High-affinity copper uptake is mediated by two plasma membrane transporters encoded by CTR1 and CTR4 in this yeast (Sun et al. 2014). Expression of cryptococcal CTR1 or CTR4 both rescues the cell growth defect of a S. cerevisiae ctr1Δ/ctr3Δ strain, indicating the independent role of these two copper transporters (Ding et al. 2011; Sun et al. 2014).

As a participant in the sophisticated mechanisms of copper homeostasis, Ctr4 was the first high-affinity copper transporter identified in C. neoformans (Ory et al. 2004). Surprisingly, conflicting observations have been reported regarding its role in the virulence of C. neoformans (Waterman et al. 2012; Sun et al. 2014). Thus, we decided to make a deletion mutant of CTR4, attempting to clarify this issue and to gain a better understanding of Ctr4-mediated physiological processes as a copper transporter. By flame atomic absorption spectrometry, we first confirmed its crucial function in copper uptake. We found several undescribed phenotypic outcomes of CTR4 deletion in extreme temperature tolerance and osmotic resistance. In particular, we observed a decreased virulence for the mutant in two different mouse cryptococcosis models.

Materials and methods

Strains and culture conditions

C. neoformans var. grubii H99 is the wild-type strain. H99FOA, a uracil auxotrophic mutant derived from H99, was employed as the recipient strain for target gene deletion (Toffaletti et al. 1993; Zhu and Williamson 2004; Jiang et al. 2009). Yeast strains were routinely cultured in YPD medium (2 % glucose, 1 % yeast extract, 2 % Bactopeptone, and pH 6.0). Asparagine salt medium (0.1 % glucose, 0.1 % asparagine, 0.3 % KH2PO4, pH 5.2, Asn) containing 100 mg/l nor-epinephrine (NE) and 0.05 % MgSO4·7H2O was used to test the laccase-mediated melanin production. Synthetic minimal medium, YNB (0.17 % amino acid-free yeast nitrogen, 0.5 % ammonium sulphate, and 2 % glucose) was employed for growth test. Strains were incubated at 28 °C unless indicated.

Disruption and complementation of CTR4

To generate the targeted null mutant for CTR4 (CNAG_00979), a 2.3-kb genomic fragment containing the whole length of CTR4 gene was PCR amplified and cloned into the vector pBluescript SK (+). Then, a 1.9-kb PCR-amplified fragment of the URA5 gene was inserted into the CTR4 fragment at the SacII site. C. neoformans strain H99FOA was then transformed by electroporation (Jiang et al. 2009). Candidate transformants were primarily selected on YNB medium and further screened by PCR amplification. The deletion candidates were confirmed by Southern blot analysis. To reconstitute the wild-type allele, a 2.0-kb CTR4 fragment was amplified from the genome of H99 using primers CTR4-C1/CTR4-C2 (5′-CCGTCTAGAGAT GTCCCATTGCGCTGATA/5′-GTACTCGAGGCGGACGCATTCAAGGTTCA) and inserted into the plasmid pBS-hyg, which contained a hygromycin resistance cassette for selection. The complemented strain was also confirmed by PCR amplification and Southern hybridization.

DNA preparation and Southern blotting

Genomic DNA was prepared as described before (Li et al. 2012). Digestion of wild-type genome with SalI would produce a 2.9-kb signal band, when hybridizing with the CTR4 probe. In contrast, the genome of the mutant treated in the same way would generate two signal bands (2.4 and 1.1 kb) (Fig. 1) because of the existence of two SalI restriction sites within the URA5 marker. Apart from the two bands identified in the ctr4Δ mutant, the reconstituted strain would produce a third band resulting from gene complementation (Fig. 1). The same fragment of CTR4 used on the complemented cassette was applied as probe (2.0 kb) and labelled with [α-32P]-dCTP using the Random Primer DNA Labelling Kit (TaKaRa Biotech Co. Ltd. Dalian, China). Digested DNA was subject to 0.8 % agarose gel separation and transferred to XH395 Magaprobe Nylon Transfer Membrane (GE OsmonicsInc, MN, USA).

Fig. 1
figure 1

Disruption and Southern blot analysis of the ctr4Δ mutant. The ORF of CTR4 was replaced by URA5. Genomic DNA from wild-type H99, ctr4Δ and the complement was digested with SalI, and probed with a 2.0-kb CTR4 fragment. Digestion of the wild-type genome yielded a 2.9-kb band, whereas the mutant had two bands, 1.1 and 2.4 kb in length. The reconstituted strain exhibited the third band resulted from the gene complement

Full size image

Quantification of intracellular copper by flame atomic absorption spectrometry

To determine the quantity of intracellular copper, flame atomic absorption spectrometry (FAAS) was employed as described previously (Rae et al. 1999; Jiang et al. 2011). Briefly, yeast cells were collected and inoculated to 25 ml Asn liquid medium (2 % glucose) supplemented with 0, 5 or 50 μM CuSO4. The concentration of each culture was diluted to OD600 of 0.1. After shaking at 28 °C for 45 min, 2 × 108 cells were taken out and washed three times with sterile ddH2O, which had been treated beforehand with Chelex-100 (Bio-Rad Laboratories Inc., Hercules, CA, USA). Cells were lyophilized and re-suspended in 2 ml nitric acid, and then the suspension was heated at 85 °C till the solution became clear. The quantity of copper was determined by Hitachi 180-80 AAS machine according to the manufacturer’s instructions (Hitachi Ltd, Japan). Wild-type H99 and the complement were used as controls in parallel. The assay was performed in triplicate, and errors were expressed as the standard deviation.

Virulence test in mouse cryptococcosis models

Virulence studies were conducted using a previously described mouse meningoencephalitis model (Yang et al. 2010). Yeast cells were incubated overnight in liquid YPD medium, washed and suspended in physiological saline. Cell numbers were estimated by haemocytometer and confirmed by colony formation on YPD agars. For the intravenous model, 6–10-week-old BALB/c female mice were anaesthetized and inoculated with 1 × 104 cells of each yeast suspensions in 0.1 ml via the lateral tail vein. For the pulmonary model, C57BL/6 mice were anaesthetized and inoculated intranasally with 3 × 107 cells in 50 μl physiological saline. Groups of ten mice were monitored daily for survival. Mice unable to reach food or water were sacrificed and counted as dying that day. For fungal burden assays, organs (lung and brain) were removed under sterile conditions from sacrificed mice, weighed and homogenized in sterile PBS buffer. Serial dilutions of the suspensions were plated on YPD agars at 28 °C for 3 days, and colonies were counted after incubation.

Results

Targeted disruption of CTR4 gene

To further characterize the role of CTR4 in C. neoformans, we made a targeted deletion of the gene. The ORF of CTR4 was disrupted by URA5 via homologous recombination as demonstrated by Southern blotting (Fig. 1). By screening with PCR amplification, the correct mutant, ctr4Δ, was obtained. Mutant strain ctr4Δ was then complemented with a 2.0-kb genomic fragment of the wild-type CTR4 gene and a hygromycin resistance cassette. The correctness of the ctr4Δ mutant and the complemented strain CTR4C was verified by Southern blotting (Fig. 1). The reconstituted strain restored the expression of CRT4 (data not shown).

Intracellular copper deficiency in ctr4Δ mutant

To ascertain the effect of CTR4 deletion on copper uptake, we tested the quantity of intracellular copper in the ctr4Δ mutant by FAAS. Wild-type H99 and the complement served as control. The intracellular levels of copper in ctr4Δ fell significantly to 0.66 ± 0.02 μg in 2 × 108 cells, while the value of H99 and the complemented strain were 2.01 ± 0.04 and 1.41 ± 0.03 μg respectively (Fig. 2). Thus, disruption of CTR4 leads to a defect in copper uptake in the ctr4Δ mutant, suggesting a role for Ctr4 as a copper transporter in C. neoformans.

Fig. 2
figure 2

Intracellular copper deficiency in ctr4Δ. The copper quantity in 2 × 108 cells collected from copper starvation conditions was determined by FAAS. The intracellular copper in ctr4Δ was 0.66 ± 0.02 μg per 2 × 108 cells, whereas the amount for H99 and the complemented strain was 2.01 ± 0.03 and 1.41 ± 0.06 μg, respectively

Full size image

As one of the crucial trace nutrients, we wondered whether the decreased copper content had marked effects on the activities of copper-containing elements in ctr4Δ. Thus, two of the essential indexes, Cu/Zn-SOD activity and cytochrome c oxidase activity were assayed to test the copper-requiring processes. On YPD, the ctr4Δ mutant exhibited smaller colonies compared with the wild-type H99 (Fig. 3a). However, on 3 μg/ml of menadione, a superoxide generator that affects Cu/Zn-SOD activity (Cox et al. 2003), the ctr4Δ mutant showed a discernible growth delay (Fig. 3b) taking into account of its retarded growth on YPD (Fig. 3a). Similarly, respiratory carbon sources, as the index for cytochrome c oxidase activity, had modest effects on growth rates of ctr4Δ (Fig. 3c), consistent with the decreased intracellular level of copper. This result is consistent with a previous report (Ding et al. 2011).

Fig. 3
figure 3

Effects of CTR4 deletion on phenotype. Starting with 2 × 106 cells/ml, yeast cells were diluted with tenfold serial dilution and 3 μl of the diluted suspension spotted on agars as indicated at 28 °C for 3 days. H99 and the complement served as control. a Growth rate of ctr4Δ mutant on YPD agar at 28 °C. b Sensitivity of ctr4Δ to 3 μg/ml menadione. c Growth rate of ctr4Δ mutant on non-fermentable carbon sources

Full size image

Critical roles of CTR4 in extreme temperatures

We found that ctr4Δ is sensitive to elevated temperature, as the growth rate of ctr4Δ at 37 °C on YPD agar was investigated. Deletion of CTR4 resulted in slight delays in growth as compared with H99 and the complemented strain under 28 °C condition (Fig. 4a). When the mutant was grown at 37 °C, the defect became severer (Fig. 4a). Reconstitution of Ctr4 restored the growth at high temperature. Interestingly, the ctr4Δ mutant exhibited severe growth defects at low temperatures, such as 4 °C (Fig. 4a). Exogenous copper could not restore the temperature tolerance defect (Fig. 4b).

Fig. 4
figure 4

Requirement of CTR4 for thermotolerance. Starting with 2 × 106 cells/ml, 3 μl of the serially diluted suspension was dropped on YPD agars for 3 days as indicated. H99 and the complement served as control. a Critical roles of CTR4 in temperature adaptability. b Exogenous copper could not restore the temperature tolerance

Full size image

To further confirm the growth defect of the ctr4Δ mutant, we constructed a growth curve under various test conditions. As shown in Fig. 6a, it was apparent that the ctr4Δ mutant exhibited a decreased growth rate at 37 °C. The result was consistent with the results on YPD agars. Unfortunately, due to the severe growth defect for the mutant at 4 °C, we were unable to obtain a growth curve assay in this condition. These results imply a relationship between copper uptake and temperature adaptability.

Critical roles of CTR4 in osmolality stress tolerance

Further findings include the tolerance roles of CTR4 to other stresses. We found that, compared with H99 and the complemented strain, the mutant displayed a high sensitivity to 2 M KCl or NaCl. The reconstituted cells restored growth on plates with high concentrations of salt (Fig. 5a), which suggested a critical role of CTR4 in response to osmolality stress. Again, to ensure that the ctr4Δ mutant failed to resist the stresses imposed by 2 M KCl and NaCl, we made growth curves under high osmotic pressure and observed a significant delay in the growth rate of ctr4Δ (Fig. 6b). We also noticed that high concentrations of salt in the medium apparently increased the lag phase, for both the wild-type H99 and ctr4Δ mutant. The above results indicate that C. neoformans CTR4 plays a vital role in response to osmolality stress.

Fig. 5
figure 5

Roles of CTR4 in stress resistance. Starting with 2 × 106 cells/ml, 3 μl of the serially diluted suspension (1:10) was spotted on agars supplemented with testing agents at 28 °C for 3 days. The mutant strain ctr4Δ was highly sensitive to 2 M NaCl and KCl (a) but not to H2O2 (0.15 %) and NaNO2 (5 mM) (b). ctr4Δ mutant exhibited mild sensitivity to Cong red (1 %) and SDS (0.01 %) (c)

Full size image
Fig. 6
figure 6

Defects of ctr4Δ mutant in resistance to temperature and osmolality stresses by growth curve experiments. Overnight cultures were diluted with YPD liquid medium to an OD600 of 0.05, and the optical density was measured every 2 h until the growth curve of wild-type H99 entered into stationary phase. The results shown were the averages of three measurements. a Growth curves of ctr4Δ mutant in 28 °C and 37 °C conditions. The growth curves from top to bottom represent H99 in 28 °C, H99 in 37 °C, ctr4Δ in 28 °C and ctr4Δ in 37 °C, respectively. b Growth curves of ctr4Δ mutant in mediums containing 2 M NaCl and 2 M KCl. The growth curves from top to bottom represent H99 in 2 M KCl, H99 in 2 M NaCl, ctr4Δ in 2 M KCl and ctr4Δ in 2 M NaCl, respectively

Full size image

We also examined the resistance of ctr4Δ to oxidative stresses created by H2O2 and NaNO2, and found that sensitivity of ctr4Δ to reactive oxygen and nitrogen species was similar to that of wild type (Fig. 5b) if taking into consideration of its retarded growth on YPD (Fig. 3a). However, the CTR4 deficient strain exhibited obvious sensitivity to 0.01 % SDS and 1 % Congo red (Fig. 5c), suggesting defects in the cell membrane and cell wall integrity of the ctr4Δ mutant. Moreover, we confirmed a severe growth defect of ctr4Δ under nutrient limitation conditions with YNB minimal medium (data not shown), which was also observed by Waterman et al. (2012).

Effect of CTR4 deletion on survival and fungal burden in mouse cryptococcosis models

The role of CTR4 in the virulence of C. neoformans remains puzzling, as inconsistent results have been obtained by different studies (Waterman et al. 2012; Sun et al. 2014). Thus, we carried out virulence testing for the H99, ctr4Δ and the complemented strains in mouse cryptococcosis models. Initially, we used the intravenous model to test the virulence of the indicated strains though tail inoculations. As shown in Fig. 7a, infections with H99 and the complemented strain resulted in complete mortality after 20th and 22nd day post-infection respectively. However, the ctr4Δ strain-infected mouse survived for more than 40 days after infection (P < 0.01), exhibiting apparent attenuated virulence compared with the wild type and reconstituted strains.

Fig. 7
figure 7

The attenuated virulence of ctr4Δ in mouse cryptococcosis models. Mice were inoculated with C. neoformans intravenously and intranasally. Survival time and fungal burden were analysed. a BALB/c female mice were infected with H99, ctr4Δ and CTR4C cells via the lateral tail vein. b C57BL/6 female mice were infected intranasally with H99, ctr4Δ and CTR4C cells. c, d Fungal burden of the infected C57BL/6 mice in the lung and brain. CFUs were normalized to tissue weight. Significant differences are shown as indicated (***P < 0.001, **P < 0.05)

Full size image

Subsequently, we conducted virulence testing using a pulmonary model. Data of mortality exhibited a parallel result of attenuated virulence resulting from deletion of CTR4 as in the intravenous model (Fig. 7b). In the pulmonary model, the ctr4Δ strain-infected mouse survived to the 29th date after infection (P < 0.01), while infections with H99 and the complement resulted in complete mortality after the 19th and 24th day post-infection, respectively. An additional fungal burden analysis of lungs and brains was performed in the pulmonary model. Interestingly, we detected striking increases in fungal burden in the lungs of mice infected with ctr4Δ mutant (Fig. 7c), suggesting that deletion of CTR4 favours the yeast to survive in the high-copper lung. Notably the fungal burden in the brains decreased significantly (Fig. 7d). Cumulatively, environmental availability of copper (host organs) appears to play a critical role for the pathogen to colonize and thus virulence of the pathogen.

Discussion

The trace element copper has been demonstrated to play critical roles in the virulence of C. neoformans (Waterman et al. 2007; Jiang et al. 2009; Ding et al. 2011). Most of the knowledge about copper acquisition and storage comes from studies on non-pathogenic fungi (Peña et al. 2000; Ioannoni et al. 2010). Copper homeostasis as a virulence factor in C. neoformans has also been studied recently. Unfortunately, the role of the copper-affinity transporter CTR4 in the pathogenesis remains contradictory based on previous reports (Waterman et al. 2012; Sun et al. 2014). In this report, we confirmed that the high-affinity copper transporter Ctr4 in C. neoformans is required for virulence in cryptococcosis mouse models, and additionally for copper acquisition, temperature adaptability and osmotic stress resistance. Several reports demonstrate that expression of CTR4 is copper-dependent whose transcription is activated by copper deficiency and inhibited under copper excess, which was also observed by us (data not shown). In this paper, we provide FAAS data to show that disruption of CTR4 resulted in marked reductions in intracellular copper level, suggesting its role in copper uptake (Fig. 2). Contrary to our expectations, the impaired copper absorption did not completely abolish activities of copper-containing elements, including the laccase, Cu/Zn-SOD and cytochrome c oxidase. Taking into account the fact that S. cerevisiae ctr1Δ/ctr3Δ mutants are unable to grow in media containing non-fermentable carbon sources (Puig and Thiele 2002), we believe that other pathways such as CTR1 may play overlapping functions in copper uptake (Sun et al. 2014).

As to the roles of CTR4 in virulence, discrepancies in observation have been reported by two groups (Waterman et al. 2012; Sun et al. 2014). We demonstrated a remarkable attenuated virulence in this study for the mutant strain ctr4Δ in both pulmonary and intravenous models (Fig. 7). The consistent results we obtained with two infection models were in a general agreement with the observation by Waterman et al. (2012), however, in a striking discordance with Sun’s result (2014) which presented a hypervirulent phenotype for their ctr4Δ. We noticed that the means used in construction of ctr4Δ by the groups were distinct. In our study and the Waterman report, CTR4 was disrupted with the native cryptococcal URA5 gene as selection marker, although Waterman’s ctr4Δ contained a 255-bp deletion in the CTR4 ORF, while our mutant resulted from an insertion of the URA5 marker at the SacII site. Sun’s group created their mutant with the drug resistant gene NEO which encodes a phosphorylase. We speculate that the selection marker NEO gene may boost the virulence for C. neoformans in the background of CTR4 deletion. Unfortunately, this issue was not addressed by the authors but warrants further investigation.

The decreased virulence of the ctr4Δ mutant can be explained by the involvement of CTR4 in tolerance for the yeast to various stress conditions. An interesting finding of this work is the role of CTR4 in response to temperature variation (Fig. 4). Deletion of CTR4 resulted in a hypersensitivity to low temperature 4 °C (Fig. 4). Unfortunately, little attention has been paid to the growth of this yeast under low temperature conditions. We also found retarded growth of the mutant at elevated temperature 37 °C (Fig. 4). We noticed that a different result as to the role of CTR4 in elevated temperature tolerance was reported by Waterman et al. (2012) based on a spot plate assay. Thus, we further conducted the growth curve experiments and confirmed the growth defects of ctr4Δ mutant at high temperature (37 °C, Fig. 6). This difference may result from the inoculum and dilution factors they used which differed from ours. As the authors (Sun et al.) did not carry out a quantification test with growth curves, it is not easy to determine the exact cause for the discrepancy. In H99, a variety of elements have been shown to play critical roles in high-temperature tolerance, such as the Ras–Cdc24 signal pathway, the HOG pathway and a vacuolar H+-ATPase gene VPH1 (Erickson et al. 2001; Bahn et al. 2005; Nichols et al. 2007). Importantly, cell wall integrity is critical for heat stress resistance (Kraus et al. 2003). Since deletion of CTR4 increased the sensitivity of yeast to 1 % Congo red and 0.01 % SDS (Fig. 5c), we believe that defects in cell wall/membrane integrity in our mutant is perhaps responsible for elevated-temperature sensitivity. Moreover, novel roles of CTR4 in resistance to 2 M NaCl and KCl have been characterized (Fig. 5a). As it has been established that the high osmolality glycerol (HOG) pathway controls the resistance to osmotic stress in C. neoformans (Bahn et al. 2005), it seems that the Ctr4 transporter may interact with the HOG pathway. Taking together, the results indicate that as a membrane protein, Ctr4 may play various roles in addition to copper transporter in C. neoformans, such as signal transduction and maintenance of cell wall/membrane integrity. Our work, along with previous reports, provides further insight into the physiological roles of CTR4 in C. neoformans.