Introduction

The pathogenic Cryptococcus species complex consists of two major lineages: Cryptococcus neoformans and Cryptococcus gattii1. C. neoformans causes life-threatening meningoencephalitis, mainly in the immunocompromised population, whereas C. gattii causes diseases in both immunocompetent and immunocompromised individuals, as shown by the outbreak in British Columbia2. The C. neoformans lineage is now speciated into C. neoformans (molecular type VNI/AFLP1, VNII/AFLP1A/AFP1B, and VNB/AFLP1; serotype A; formerly C. neoformans var. grubii), C. deneoformans (VNIV/AFLP2; serotype D; formerly C. neoformans var. neoformans), and C. deneoformans and C. neoformans hybrids (VNIII/AFLP3, serotype AD hybrid)3,4,5,6,7. C. neoformans strains are generally more virulent and more frequently found worldwide compared with C. deneoformans strains1. The C. gattii lineage is currently delineated as five species, C. gattii (VGI/AFLP4; serotype B), C. deuterogattii (VGII/AFLP6; serotype B), C. bacillisporus (VGIII/AFLP5; serotype B and C), C. tetragattii (VGIV/AFLP7; serotype C), and C. decagattii (VGIV/AFLP10; serotype C)3,4,5,6,7. C. gattii including strain WM276 is a species commonly isolated from natural environments and patients; C. deuterogattii including strain R265 is predominant in the Vancouver outbreak8,9.

The C. gattii lineage exhibits distinct biochemical, ecological, and pathological features compared to the C. neoformans lineage. The two lineages have different capabilities to utilize nitrogen and carbon sources. Although creatinine is used as the source of nitrogen in both species, creatinine deiminase, which is required for creatinine metabolism, is regulated differently. Its expression is suppressed by ammonia in C. neoformans, but not in C. gattii10. Glycine is utilized as both a carbon and nitrogen source in C. gattii, but only as a nitrogen source in C. neoformans11. Given that most C. gattii strains, but none of C. neoformans, are resistant to L-canavanine, CGB (canavanine-glycine-bromothymol blue) medium is widely used to distinguish the two species12. Pigeon guano, a well-known ecological niche of C. neoformans, supports the growth of both species. On solid pigeon guano-containing medium, both C. neoformans and C. gattii produce melanin; however, spore production is more robust in C. neoformans than in C. gattii13. The mating efficiency of C. gattii is increased on media containing plant-derived substrates such as those of eucalyptus and Arabidopsis14. C. neoformans primarily causes meningitis through central nervous system infection following pulmonary infection, and C. gattii commonly causes pulmonary infection15,16. During murine infection, the primary target organ of C. neoformans is different from that of C. gattii17: After infection through pulmonary route, C. gattii grows faster in the lung and causes pneumonia, while C. neoformans moves faster and efficiently to the brain and causes meningoencephalitis17.

In spite of such pathobiological differences between C. neoformans and C. gattii species complexes, signalling pathways governing their virulence have been more extensively studied in C. neoformans than in C. gattii, partly because the C. neoformans H99 strain has been more widely used as a reference strain for pathogenic Cryptococcus species. Nevertheless, several evolutionarily conserved signalling pathways have been shown to be critical for the pathobiology of both species2. For example, the cAMP/PKA pathway is involved in the production of two major virulence factors: capsule and melanin18,19,20. The calmodulin/calcineurin pathway plays conserved roles in regulating thermotolerance, virulence, and cell wall/membrane integrity in both species21,22,23.

We recently reported that the unfolded protein response (UPR) pathway is one of the critical virulence-regulating signalling pathways in C. neoformans24. In C. neoformans, it consists of the evolutionarily conserved Ire1 kinase/endonuclease and its downstream bZIP transcription factor Hxl1. Notably, Hxl1 is structurally and functionally divergent from other previously reported Ire1 downstream transcription factors, Hac1 in Saccharomyces cerevisiae and Xbp1 in humans. Upon ER stress, the activated Ire1 eliminates the intron of the unspliced HXL1 mRNA (HXL1u) in a splicesome-independent manner, thereby increasing spliced HXL1 mRNA (HXL1s) transcripts that can be translated to the active Hxl1 transcription factor required for induction of UPR target genes. Therefore, perturbation of the UPR pathway, e.g. using tunicamycin (TM) and dithiothreitol (DTT) treatment, renders C. neoformans extremely susceptible to ER stress and cell wall-destabilizing agents, e.g. to calcofluor white (CFW) and Congo red (CR)24. Most notably, both ire1∆ and hxl1∆ mutants are unable to grow at the mammalian body temperature of 37 °C and are thereby completely avirulent in a murine model of systemic cryptococcosis24, suggesting that the UPR pathway, particularly Hxl1, could be a promising antifungal drug target. We also reported that the UPR pathway regulates Kar2, an ER-resident molecular chaperone, to counteract ER and cell wall stress and control thermotolerance and azole drug resistance25. Although Hxl1 is a bona fide transcription factor downstream of Ire1, Ire1 plays both Hxl1-dependent and -independent functions in modulating stress responses, virulence, and differentiation24,26. For instance, Ire1, but not Hxl1, regulates opposite- and unisexual mating and capsule production in a Rim101-dependent manner in C. neoformans26.

Despite the critical role of the UPR pathway in the virulence of the C. neoformans lineage, it remains unknown whether the function of the UPR pathway is conserved in the C. gattii lineage, how it is regulated, and whether it is critical for the virulence of the other pathogenic Cryptococcus species. To address these questions, we have identified IRE1 and HXL1 genes in C. deuterogattii and characterized their in vitro and in vivo roles. Here, we demonstrate that the UPR pathway plays evolutionarily conserved but different roles in C. deuterogattii and C. neoformans; however, it is still absolutely necessary in the virulence of C. deuterogattii.

Results

Identification of the Ire1 ortholog in C. deuterogattii

To identify the Ire1 ortholog in the C. deuterogattii strain R265, we performed BLASTp searches using serotype A Ire1 as the query in the C. gattii genome database (https://www.broadinstitute.org/scientific-community/science/projects/fungal-genome-initiative/cryptococcus-gattii-genome-project-0). C. deuterogattii R265 contains CNBG_0073.2, which is orthologous to C. neoformans Ire1 (score 1571.6, e-value: 0). Notably, the C. deuterogattii R265 Ire1 ortholog is 152 and 153 amino acids shorter than the C. neoformans strain H99 and C. deneoformans JEC21 Ire1 orthologs, respectively (Fig. 1a). Based on protein domain analysis by Pfam (http://pfam.sanger.ac.uk/), the C. deuterogattii Ire1 ortholog is predicted to contain kinase and endonuclease domains like the C. neoformans lre1 ortholog (Fig. 1a).

Figure 1
figure 1

The evolutionarily conserved UPR pathway in the pathogenic Cryptococcus species complex. (a) The conserved domain structure of the C. neoformans H99 strain, C. deneoformans JEC21 strain, and C. deuterogattii R265 strain Ire1. (b) RT-PCR analysis of HXL1 splicing in the pathogenic Cryptococcus species complex during ER stress. The cDNA was synthesized from total RNA from Cryptococcus cells treated with or without TM (0.3 μg/mL) for 1 h as the template. The grouping of gels is cropped from different parts of the same gel and the original images are shown in the Supplementary Fig. 1. (c) Quantitative RT-PCR analysis of KAR2 in the pathogenic Cryptococcus species complex during ER stress. The qRT-PCR was performed with cDNA from wild-type strains and ire1Δ mutants treated or not treated with TM (0.3 μg/mL) for 1 h. Duplicate technical experiments with two independent biological samples were performed. Representative images from independent experiments for KAR2 gene are shown. Error bars indicate standard deviations. Asterisks indicate statistical significance of differences in expression levels of each gene (***P < 0.001; NS, not significant).

Full size image

In our previous study, we identified the Hxl1 ortholog in C. deuterogattii R265 (CNBG_4842.2) and confirmed that ER stress-induced unconventional splicing of HXL1 mRNA occurs24. To test whether the unconventional splicing of HXL1 mRNA is mediated by Ire1 in C. deuterogattii, we constructed the ire1∆ mutant of the R265 strain and then monitored HXL1 splicing in response to the ER stress-inducing agent tunicamycin (TM). Similar to C. neoformans, C. deneoformans and C. deuterogattii strains exhibited an Ire1-dependent HXL1 splicing event during ER stress (Fig. 1b and Supplementary Fig. 1). During ER stress, the molecular chaperone KAR2 is induced in the Ire1-Hxl1-dependent UPR pathway in C. neoformans24. To confirm that KAR2 induction is regulated by Ire1, we monitored the expression level of KAR2 in both R265 strain and the ire1∆ mutant when cells were exposed to TM. In accordance with the KAR2 induction in C. neoformans, expression of KAR2 was upregulated in C. deuterogattii, but not in the ire1∆ mutant (Fig. 1c). These data indicate that Ire1-mediated UPR induction is widely conserved in the pathogenic Cryptococcus species complex.

Ire1 and Hxl1 play evolutionarily conserved roles in ER stress response and adaptation

To address whether UPR activation is required for ER stress response and adaptation in C. deuterogattii, we disrupted the HXL1 gene in the R265 strain. Expectedly, C. deuterogattii ire1Δ and hxl1Δ mutants displayed severe growth defects in response to ER stress inducers such as TM and DTT, whereas their complemented strains (ire1Δ + IRE1 and hxl1Δ + HXL1) exhibited levels of ER stress resistance restored to the wild-type level (Fig. 2a). Similar to the phenomenon observed in C. neoformans UPR mutants, the ire1Δ mutant was more sensitive to ER stress than the hxl1Δ mutant of the R265 strain, indicating that Ire1 plays a role in modulating ER stress in both Hxl1-dependent and -independent manners in C. deuterogattii. ER stress causes induction of KAR2 expression through Ire1-mediated HXL1 splicing in the pathogenic Cryptococcus species complex (Fig. 1b and c). Supporting the finding that the UPR pathway is required for counteracting ER stress in C. deuterogattii, deletion of IRE1 or HXL1 abolished the induction of KAR2 in response to ER stress in the R265 strain (Fig. 2b).

Figure 2
figure 2

The evolutionarily conserved functions of the pathogenic Cryptococcus UPR pathway in response to ER stress, cell wall stress, and diamide resistance. (a,c and d) Cryptococcus cells were cultured overnight at 30 °C and washed with sterilized water. Cells were serially diluted (dilutions from 1 to 10−4) and then spotted onto YPD medium containing the indicated concentrations of ER stress inducers (TM or DTT), cell wall destabilizers (CFW or CR), or diamide. The plates were incubated at 30 °C and photographed daily for 2–4 days. The two images split by a horizontal white line in each spot assay were obtained from the same plate. (b) C. deuterogattii strains were grown in liquid YPD medium to mid-log phase (OD600 = 0.6), exposed to TM (0.3 μg/mL) for 1 h, and harvested. The cDNA was synthesized from total RNA in cells treated or not treated with TM. Duplicate technical experiments with two biological samples were performed. Representative images from independent experiments for the KAR2, DER1, ALG7, and ERV29 genes are shown. Error bars indicate standard deviations. Asterisks indicate the statistical significance of differences in expression levels of each gene. (***P < 0.001; *P < 0.05; NS, not significant) Strain information: WT (H99, JEC21, and R265), ire1Δ mutant (YSB552, YSB2886, and YSB1889), hxl1Δ mutant (YSB723, YSB2030, and YSB1891), ire1Δ + IRE1 complemented strains (YSB3158), and hxl1Δ + HXL1 complemented strains (YSB2957).

Full size image

Our previous study reported that some C. neoformans UPR-target genes, including KAR2, harbor the UPR element (UPRE) and UPRE-like sequence in their promoter and their expression is regulated in response to ER stress and temperature upshift24. These include genes involved in protein secretion and lipid linked N- and O-glycosylation24. Therefore, we examined whether the other known C. neoformans UPR target genes are also regulated by the C. deuterogattii UPR pathway. First, we confirmed the presence of UPRE and UPRE-like sequences in the promoters of each putative UPR target (Supplementary Fig. S2). Similar to C. neoformans, expression levels of DER1 (ER-associated degradation), ALG7 (lipid-linked N-oligosaccharyltransferase), and ERV29 (ER to Golgi vesicle-mediated transport) were significantly increased in the R265 strain, but not in the ire1Δ or hxl1Δ mutants (Fig. 2b), suggesting that that they are ER stress-dependent UPR target genes in C. deuterogattii. However, expression changes of other putative UPR downstream genes such as PPS1, SOD2, PMT1, PMT4, CHS2, SEC61, OST1, and WBP1 were not significantly different between wild- type and UPR mutants in both C. neoformans24 and C. deuterogattii (Supplementary Fig. S3).

Our previous study revealed that Ire1 has an Hxl1-independent function in modulating resistance to diamide, which is a thiol-specific oxidant inducing unnatural disulfide formation in a protein or between proteins and triggering protein misfolding, in C. neoformans24. To examine whether this phenomenon is also evolutionarily conserved, we monitored growth of UPR mutants under diamide treatment in other pathogenic Cryptococcus species. Similar to the case in C. neoformans, the C. deneoformans ire1Δ mutant, but not hxl1Δ mutant, was more susceptible to diamide than wild-type strain (Fig. 2c). In C. deuterogattii, however, the ire1∆ mutant was as sensitive to diamide as the hxl1∆ mutants, suggesting that Ire1 controls diamide sensitivity in an Hxl1-dependent manner in C. deuterogattii. Another notable phenotype observed in the UPR mutants is the increased susceptibility to cell wall-damaging agents24. C. deneoformans ire1Δ and hxl1Δ mutants showed increased susceptibility, albeit to a lesser extent compared to C. neoformans ire1Δ and hxl1Δ mutants, to CFW and CR (Fig. 2d). C. deuterogattii ire1Δ and hxl1Δ mutants also exhibited increased susceptibility to CFW and CR (Fig. 2d). Notably, however, C. deuterogattii ire1Δ mutants displayed higher susceptibility to CFW and CR than the hxl1Δ mutants (Fig. 2d). In fact, the C. deuterogattii hxl1Δ mutants showed only a weakly increased susceptibility to CFW and CR (Fig. 2d). Taken together, all these data suggest that the regulatory mechanism of the C. deuterogattii UPR pathway for ER stress response and adaptation is generally conserved, albeit with some modification, compared to the C. neoformans UPR pathway.

Ire1, but not Hxl1, is critical for the survival of Cryptococcus deuterogattii at mammalian body temperature

In our previous study, C. neoformans and C. deneoformans ire1Δ and hxl1Δ mutants exhibited growth defects at 37 °C: these mutations are believed to be the most critical factors accounting for their avirulence24,26. Similarly, the C. deuterogattii ire1Δ mutant exhibited severe growth defects at high temperature (37 °C and 39 °C), compared to the wild-type and its complemented strains (Fig. 3a), indicating that the role of Ire1 in thermotolerance is evolutionarily conserved in the pathogenic Cryptococcus species complex. Strikingly, however, the C. deuterogattii hxl1Δ mutant showed almost wild-type levels of thermotolerance at 37 °C (Fig. 3a). At higher temperature (39 °C), the C. deuterogattii hxl1Δ mutant showed subtly increased sensitivity. This is in stark contrast to C. neoformans and C. deneoformans hxl1Δ mutants that exhibited even higher thermosensitivity than the ire1Δ mutants (Fig. 3a). This phenomenon suggests that C. deuterogattii R265 Ire1 regulates thermotolerance through downstream transcription factors other than Hxl1.

Figure 3
figure 3

The conserved and distinct functions of the UPR pathway in the thermotolerance of the pathogenic Cryptococcus species complex. (a) For the thermotolerance test, cells were cultured in YPD liquid medium at 30 °C for 16 h. Each cell was serially diluted (dilutions from 1 to 10−4) and spotted onto YPD medium. Strains were further incubated at 30 °C, 37 °C, and 39 °C. The two images split by a horizontal white line in each spot assay were obtained from the same plate. (b) RT-PCR analysis of HXL1 splicing in high temperature conditions. The cDNA was synthesized from the total RNA extracted from cells exposed to temperature upshift for 1 h. The specific primers for the amplification of HXL1 and ACT1 are described in Supplementary Table 3. The grouping of gels is cropped from different parts of the same gel and the original images are shown in the Supplementary Fig. 4. (c) KAR2 was induced under temperature upshift. The qRT-PCR was performed with cDNA from wild-type strains upon temperature upshift (from 30 °C to 37 °C and from 30 °C to 39 °C) for 1 h. (d and e) Quantitative reverse transcription-PCR (qRT-PCR) analysis of putative UPR genes in C. deuterogattii during thermal shock. The cDNA was synthesized with total RNA from the wild-type R265 strain and ire1Δ and hxl1Δ mutants upon temperature upshift (from 30 °C to 37 °C and from 30 °C to 39 °C). Duplicate technical experiments with two or more independent biological samples were performed. Representative data from independent experiments for UPR putative gene are presented. Error bars indicate standard deviations. Asterisks indicate statistical significance of differences in expression levels of each gene (*P < 0.05; **P < 0.01; ***P < 0.001).

Full size image

Given that C. deuterogattii Hxl1 has a minor role in thermosensitivity, we wondered whether an HXL1 splicing event occurs in C. deuterogattii during temperature upshift. In C. neoformans, temperature upshift induced splicing of HXL1 mRNA (Fig. 3b and Supplementary Fig. 4). Compared to the splicing event of HXL1 mRNA during ER stress, unspliced HXL1 mRNA was still observed at 37 °C and 39 °C, although spliced HXL1 mRNA was dominant. In C. deneoformans, however, spliced HXL1 mRNA was significantly increased upon temperature upshift. Similar to C. deneoformans, the spliced HXL1 mRNA was dominant at 37 °C and 39 °C in C. deuterogattii (Fig. 3b). Previous studies reported that the expression level of KAR2 is up-regulated during thermal shock24. The constitutive KAR2 overexpression partially rescued thermotolerance in C. neoformans ire1Δ and hxl1Δ mutants25. In line with the previous results, KAR2 was induced upon temperature upshift in C. deuterogattii (Fig. 3c). To demonstrate whether the difference in thermosensitivity between C. deuterogattii ire1Δ and hxl1Δ mutants was attributable to the level of KAR2 expression, we monitored its expression levels during the temperature upshift. However, the expression patterns of KAR2 appeared to be indistinguishable between R265 strain and UPR mutants (Fig. 3d), suggesting that KAR2 induction by thermal stress is independent of the UPR pathway in C. deuterogattii and not responsible for thermosensitivity of the ire1Δ mutant.

We monitored expression levels of other putative UPR target genes upon temperature upshift in C. deuterogattii. In contrast to KAR2, ALG7 expression was upregulated upon temperature upshift in Ire1- and Hxl1-dependent manners (Fig. 3d). The expression of OST1 (lipid-linked N-oligosaccharyltransferase), DER1 and WBP1 (lipid-linked N-oligosaccharyltransferase) were decreased in both ire1∆ and hxl1∆ mutants at 39 °C (Fig. 3e). In the case of OST1, its expression was also decreased in the hxl1∆ mutants at 37 °C. Notably, expression of SEC. 61 (translocation of misfolded proteins out of ER) was decreased in both ire1∆ and hxl1∆ mutants at 37 °C and 39 °C (Fig. 3e). The expression patterns of other putative UPR downstream genes including PPS1, SOD2, PMT1, ERV29, CHS2, and PMT4 in the R265 strain were similar to those in both ire1∆ and hxl1∆ mutants upon temperature upshift (Supplementary Fig. 5). Based on the fact that the C. deuterogattii hxl1Δ mutant exhibited only a minor growth defect at high temperature, Alg7, Ost1, Der1, Wbp1, and Sec61 play a minor role in thermotolerance and other UPR target(s) may have a major role in the UPR-dependent-thermotolerance of C. deuterogattii.

Evolutionarily conserved and divergent roles of Ire1 and Hxl1 in antifungal drug resistance

Our previous studies showed that C. neoformans ire1Δ and hxl1Δ mutants exhibit increased susceptibility to azole drugs24. Their azole susceptibility is partially suppressed when KAR2 is constitutively overexpressed25. Similarly, the C. deuterogattii ire1Δ and hxl1Δ mutants displayed growth defects in response to fluconazole, itraconazole, and ketoconazole (Fig. 4a). Notably, the deletion of IRE1 made the R265 strain more susceptible to azoles than that of HXL1, which was in contrast to the finding that the hxl1Δ mutants were more susceptible to azoles than the ire1Δ mutants in C. neoformans and C. deneoformans (Fig. 4a). This indicates that the UPR components play conserved but different roles in azole drug resistance among the pathogenic Cryptococcus species complex.

Figure 4
figure 4

The pathogenic Cryptococcus UPR pathway plays a role in antifungal drug resistance, affecting membrane stability. (a and c) Strains were incubated in liquid YPD medium at 30 °C overnight. Cells were serially diluted and spotted onto YPD medium containing the indicated concentrations of azole drugs (FCZ (fluconazole, 5 μg/mL), ICZ (itraconazole, 0.025 μg/mL), and KCZ (ketoconazole, 0.1 μg/mL)) or cell membrane destabilizers (FDX (1 μg/mL) and SDS (0.03%)). Plates were further incubated at 30 °C for 2–4 days and photographed daily. The two images split by a horizontal white line in each spot assay were obtained from the same plate. (b) The qRT-PCR analysis for measuring ERG11 expressions in wild-type and ire1Δ and hxl1Δ mutants of the pathogenic Cryptococcus species complex. ERG11 induction is not required for the UPR pathway in the pathogenic Cryptococcus species complex. The cDNA was synthesized from total RNA in cells treated or not treated with fluconazole (10 μg/mL). Duplicate technical experiments with two biological samples were performed. Representative images from independent experiments for ERG11 gene are shown. Error bars indicate standard deviations. Asterisks indicate the statistical significance of differences in expression levels of ERG11 (*P < 0.05; **P < 0.01; ***P < 0.001).

Full size image

Fluconazole treatment induces expression of ergosterol biosynthesis genes to counteract the inhibition of ergosterol biosynthesis in C. neoformans27. To determine whether the UPR pathway affects azole drug resistance by controlling the induction of ERG genes, we measured the expression levels of ERG genes in the pathogenic Cryptococcus UPR mutants after fluconazole treatment. We found that ERG3 and ERG11 were normally induced by fluconazole treatment in UPR mutants (Fig. 4b and Supplementary Fig. 6), indicating that the UPR pathway is not likely to be directly involved in ergosterol biosynthesis. As Erg11 inhibition by azoles allow toxic sterol intermediates to be incorporated into the cell membrane and then prevents cell proliferation28, we hypothesized that the UPR pathway could be involved in azole drug resistance by regulating cell membrane stability. Similar to the C. neoformans ire1Δ mutant, the C. deuterogattii ire1Δ mutant exhibited growth retardation in the presence of SDS, an ionic detergent that disrupts cell membrane integrity (Fig. 4c). In contrast, the C. deuterogattii hxl1Δ mutant exhibited wild-type levels of SDS resistance (Fig. 4c). Fludioxonil, which induces hyperactivation of the HOG pathway and thereby increases intracellular turgor pressure and affects cell membrane stability, inhibits the growth of C. neoformans, whereas it does not affect C. deneoformans29. C. deuterogattii UPR mutants were also susceptible to fludioxonil like C. neoformans UPR mutants (Fig. 4c). Notably, the ire1Δ mutant was more susceptible to fludioxonil than the hxl1Δ mutant in C. deuterogattii, which is in contrast to the finding that the hxl1Δ mutant was more susceptible to fludioxonil than the ire1Δ mutant in C. neoformans (Fig. 4c). Therefore, the results indicate that the role of the UPR pathway in cell membrane stability could contribute to azole resistance of the pathogenic Cryptococcus species. However, other factors may also indirectly contribute to the role of the UPR pathway in azole resistance.

The Cryptococcus UPR pathway plays a minor role in capsule and melanin production

The pathogenic Cryptococcus species complex has several virulence factors responsible for survival and proliferation within the host30. Among these, melanin has antiphagocytic and antioxidant activities and the polysaccharide capsule protects the cells from being phagocytosed by host phagocytic cells31,32. To address the role of the UPR pathway in the formation of virulence factors, we examined the capability of the UPR mutants to produce melanin and capsules. We found that the C. neoformans ire1Δ mutant, but not the hxl1Δ mutant, produced melanin in a delayed manner, but eventually produced wild-type levels of melanin (Fig. 5a). Similarly, the C. deuterogattii ire1Δ mutant, but not the hxl1Δ mutant, showed delayed melanin synthesis (Fig. 5a). In C. deneoformans, however, both ire1Δ and hxl1Δ mutants exhibited decreased levels of melanin production (Fig. 5a).

Figure 5
figure 5

The UPR pathway affects capsule and melanin production in a species-dependent manner. (a) Each strain was cultured at 30 °C overnight and spotted onto solid Niger seed agar medium containing the indicated concentration of glucose (0.1%). Cells were photographed with a digital camera for 1–2 days. (b) Cells were incubated in solid-agar based DME medium (upper panel) or liquid Littman medium (lower panel) at 30 °C for 2 days. After incubation, capsules were stained with India ink and visualized using an Olympus BX51 microscope equipped with a SPOT insight digital camera. The white bar indicates 10 μm. (c) Quantitative measurement of the relative capsule diameter in the UPR mutants. The scatter plot indicates relative capsule diameter of strains incubated in liquid Littman medium. A total of 30 cells in each strain were measured for capsule production. P-values were determined by one-way analysis of variance with Bonferroni’s multiple-comparison test. Error bars indicate standard deviations. NS, not significant.

Full size image

We previously reported that the C. neoformans ire1Δ mutant, but not the hxl1Δ mutant, displayed capsule defects in Dulbecco’s modified Eagle’s (DME) medium24. In contrast, both C. deneoformans and C. deuterogattii UPR mutants produced wild-type levels of capsules in DME medium (Fig. 5b). As capsule production levels in C. deuterogattii were not evident in DME medium, we used Littman medium, which is known to induce capsule formation much more strongly than DME medium33. C. deuterogattii UPR mutants incubated in Littman medium exhibited wild-type levels of capsule (Fig. 5b). Interestingly, as opposed to the reduced capsule production of C. neoformans ire1Δ mutant in DME medium, wild-type levels of capsules were observed in the ire1Δ mutants in Littman medium (Fig. 5b). Quantitative measurements of relative capsule size from cells cultured in Littman medium also confirmed that all UPR mutants produced wild-type levels of capsule (Fig. 5c). Thus, the UPR pathway appears to play a minor role in melanin and capsule production in the pathogenic Cryptococcus species complex.

Evolutionarily conserved role of the UPR pathway in Cryptococcus species complex virulence

Our previous study reported that both Ire1 and Hxl1 of the UPR pathway are indispensable for the virulence of C. neoformans24. We initially hypothesized that avirulence of both ire1Δ and hxl1Δ mutants in C. neoformans results from their severe growth defects at host physiological temperature (37 °C). However, it is still possible that a non-thermotolerance-related phenotype of the UPR mutants may contribute to their virulence defects. To address this question, the wax moth Galleria mellonella model for Cryptococcus infection provides a unique opportunity because the host can be incubated at 25 °C or 37 °C for virulence testing34. Similar to our previous data in the murine model of systemic cryptococcosis24, both C. neoformans ire1Δ and hxl1Δ mutants were avirulent in G. mellonella when the infected insect was incubated at 37 °C (Fig. 6a). Notably, however, the C. neoformans ire1Δ mutant, but not the hxl1Δ mutant, exhibited a complete lack of virulence in the insect incubated at 25 °C (Fig. 6b). This finding strongly suggests that thermotolerance-independent phenotypic traits significantly contribute to the avirulence of the C. neoformans ire1Δ mutant. In contrast to the C. neoformans strains, the G. mellonella model could be not used for testing the virulence of C. deneoformans and C. deuterogattii UPR mutants because wild-type JEC21 and R265 strains could not kill the insect even at 37 °C (data not shown).

Figure 6
figure 6

The UPR pathways governs the virulence of the pathogenic Cryptococcus species complex. (a and b) Ire1 plays a role in regulating virulence in an Hxl1-independent manner. G. mellonella in the final instar stage were infected with each C. neoformans strain and further incubated at 37 °C (a) or 25 °C (b). Percentage of survival (%) was monitored for 2 weeks post-infection. The survival curve was statistically analysed by log-rank (Mantel-Cox) test. At 25 °C, P = 0.003 for WT vs. ire1Δ mutant (YSB552), and P > 0.05 for WT vs. hxl1Δ mutant (YSB723). At 37 °C, P < 0.0001 for H99 vs. ire1Δ mutant and for H99 vs. hxl1Δ mutant. Strain information: WT (H99), ire1Δ mutant (YSB552), ire1Δ + IRE1 (YSB1000), hxl1Δ mutant (YSB723), and hxl1Δ + HXL1 (YSB762) (c) A/Jcr mice were infected with 5 × 104 cells of WT (R265), ire1Δ mutant (YSB1889), ire1Δ + IRE1 strain (YSB3158), hxl1Δ mutant (YSB1891), and hxl1Δ + HXL1 strain (YSB2957) by intranasal instillation. Survival was monitored for 60 days post-infection. The survival curve was statistically analysed by log-rank (Mantel-Cox) test. P = 0.0001 for WT vs. ire1Δ mutant and WT vs. hxl1Δ mutant; P = 0.8404 for WT vs. ire1Δ + IRE1; P = 0.5435 for WT vs. hxl1Δ + HXL1; P = 0.0002 for ire1Δ vs. ire1Δ + IRE1; and P = 0.0013 for hxl1Δ vs. hxl1Δ + HXL1. (d) The colony forming unit (CFU) count of Cryptococcus cells per gram of lung tissue was determined from sacrificed mice infected with WT (R265), ire1Δ mutant (YSB1889), ire1Δ + IRE1 strain (YSB3158), hxl1Δ mutant (YSB1891), and hxl1Δ + HXL1 strain (YSB2957) by intranasal instillation 14 days post-infection.

Full size image

To address the role of the UPR pathway in the virulence of C. deuterogattii, we employed a murine model of systemic cryptococcosis, as previously published21. Based on our observations from the insect killing assay with C. neoformans UPR mutants, we hypothesized that the C. deuterogattii ire1Δ mutant, but not the hxl1Δ mutant, might exhibit avirulence. As expected, the mice infected with the C. deuterogattii ire1Δ mutant did not show any signs of illness (monitored up to 60 days), whereas those infected with the wild-type R265 strain and the complemented strains became moribund within 20 days of infection (Fig. 6c). Surprisingly, however, the C. deuterogattii hxl1Δ mutants also did not cause illness in any mice (Fig. 6c), indicating that Hxl1 is still required for the virulence of C. deuterogattii.

To examine whether the avirulence of C. deuterogattii ire1Δ and hxl1Δ mutants is caused by reduced cell survival similar to C. neoformans ire1Δ and hxl1Δ mutants24, we measured fungal burden in the lungs and brains recovered from mice at day 14 post infection. The lung tissue recovered from mice infected with the C. deuterogattii ire1Δ mutant showed a very reduced burden (less than 400 CFU/g of tissue) compared to that from mice infected with wild-type and the complemented strain (more than 107 CFU/g of tissue) (Fig. 6d). This is not a surprising result because the C. deuterogattii ire1Δ mutant cannot grow at 37 °C. Notably, mice infected with the C. deuterogattii hxl1Δ mutant, which did not show significant growth defects at 37 °C, also exhibited decreased fungal burden (less than 12,000 CFU/g of tissue) in the lungs (Fig. 6d). Even at day 64 post infection, the lungs of mice infected with the hxl1Δ mutant still showed low fungal burden (less than 33,000 CFU/g of tissue; Supplementary Table 1). As previously reported, a significant fungal burden was not observed in the brain tissue recovered from all infected mice (less than 3,000 CFU/g of tissue), owing to the lungs being the major target tissue of C. deuterogattii infection17.

Histopathological analysis of the infected lung and brain tissues further supported the observations made in the survival and fungal burden assays (Fig. 7). Each of the mice within the groups had similar histopathological findings. Mice infected with the wild-type R265 strain had marked expansion of alveolar spaces with numerous fungi (Fig. 7, upper panel in R265 lane). These fungus-laden regions of distended pulmonary parenchyma were associated with severe inflammatory changes consisting of angiocentric and bronchocentric accumulation of a pleocellular inflammatory infiltrate (Fig. 7, lower panel in R265 lane). Airway inflammation was severe within areas of high fungal burden, and bronchiolar epithelial hypertrophy/hyperplasia was marked. Neutrophils were present in the epithelium of large airways, and fungi and copious mucus were present within the lumens. No significant lesions or fungi were noted in the brain sections, supporting the finding that no significant fungal burden of the R265 strain was observed in the brain (data not shown). The ire1Δ + IRE1 and hxl1Δ + HXL1 complemented strains exhibited similar findings to the wild-type strain. In stark contrast to wild-type and complemented strains, no ire1Δ mutant cells were noted in mucicarmine-stained brain or lung sections (Fig. 7, upper panel in ire1Δ mutant lane; data not shown), and the lungs were free of inflammatory changes (Fig. 7, lower panel in ire1Δ mutant lane). In the case of mice infected with the hxl1Δ mutant, there were small focal areas where there were multiple contiguous alveolar spaces that contained fungi (Fig. 7, upper panel in hxl1Δ mutant lane). Pulmonary inflammation in these areas was mild and consisted of pleocellular inflammatory infiltrates, mild perivascular lymphoid cuffing, and mild bronchiolar epithelial hypertrophy (Fig. 7, lower panel in hxl1Δ mutant lane). Taken together, the results indicate that both Ire1 and Hxl1 of the UPR pathway are essential for virulence in C. deuterogattii.

Figure 7
figure 7

Histopathological analysis of C. deuterogattii UPR mutants in a murine model of systemic cryptococcosis. Mucicarmine-stained sections from wild-type R265 strain-infected mice showed extensive accumulation of fungi expanding and distending alveolar spaces. H&E-stained sections from R265-infected mice showed marked inflammatory changes. Photomicrographs of lung sections from ire1Δ mutant strain-infected mice displayed a lack of mutant cells and no inflammation in either airways or the parenchyma. Photomicrographs of sections from ire1Δ + IRE1 or hxl1Δ + HXL1 complemented strain-infected mice showed identical patterns of fungi and inflammation to those of R265-infected mouse. Photomicrographs of lung sections from mice infected with the hxl1Δ mutant, stained with mucicarmine, exhibited focal aggregations of fungi. H&E-stained sections from mice infected with the hxl1Δ mutant showed mild inflammatory changes. The arrow indicates C. deuterogattii strains in the lung tissue stained with mucicarmine or H&E. The scale bar indicates 200 μm.

Full size image

C. deuterogattii Hxl1 is involved in organic peroxide stress resistance in an Ire1-independent manner

The animal study revealed that Hxl1 contributes to the virulence of C. deuterogattii in a thermotolerance-independent manner. As Cryptococcus is exposed to reactive oxygen species (ROS) generated by the host immune system, we examined the possibility that Hxl1 may contribute to oxidative stress resistance in C. deuterogattii. To address this question, we examined the susceptibility of C. deuterogattii UPR mutants to the following oxidative stress agents: hydrogen peroxide (H2O2), tert-butyl hydroperoxide (tBOOH; organic peroxide), and menadione (a superoxide generator). Notably, both C. deuterogattii ire1∆ and hxl1∆ mutants were as resistant to the oxidative stresses as the wild-type strain at 25 °C (Fig. 8). At 37 °C, however, the hxl1∆ mutant exhibited very weakly increased susceptibility to tBOOH, but not to H2O2 or menadione (Fig. 8). As the C. deuterogattii ire1∆ mutant did not show any increased susceptibility to tBOOH at 37 °C compared to control (Fig. 8), the role of Hxl1 in the organic peroxide resistance appeared to be Ire1-independent in C. deuterogattii. This phenomenon suggests that the role of Hxl1 in resistance to certain oxidative stress, in addition to its role in ER stress response and adaptation, might be one of contributing factors for the avirulence of the C. deuterogattii hxl1∆ mutant.

Figure 8
figure 8

C. deuterogattii controls organic peroxide stress in an Hxl1-dependent manner at host physiological temperature. C. deuterogattii strains were incubated in liquid YPD medium at 30 °C overnight. Cells were serially diluted and spotted onto YPD medium containing the indicated concentrations of oxidative stress inducers [tBOOH (tert-butyl hydroperoxide, 0.5 mM), H2O2 (hydrogen peroxide, 3.5 mM), and menadione (0.03 mM)]. Plates were further incubated at 25 °C or 37 °C for 1–3 days and photographed daily.

Full size image

Discussion

The UPR pathway has been characterized in diverse fungi including Saccharomyces cerevisiae, Cryptococcus neoformans, Candida albicans, Candida glabrata, Aspergillus fumigatus, Alternaria brassicicola, and Ustilago maydis24,26,35,36,37,38,39. In this study, we functionally characterized in vitro and in vivo roles of the UPR pathway in C. deuterogattii and compared them with those in C. neoformans (Fig. 9). Here, we demonstrated that the UPR pathway has evolutionarily conserved roles in the pathogenic Cryptococcus species complex. The most evident conserved role is ER stress response and adaptation. In response to ER stress, the unconventional, spliceosome-independent splicing event of HXL1 mRNA and induction of KAR2 occurred in Ire1-dependent manners in all studied species. Other conserved roles include azole drug resistance, cell wall integrity, thermotolerance, and virulence. Therefore, it seems clear that the UPR pathway has evolved as a core signalling pathway for stress response and virulence of the pathogenic Cryptococcus species complex.

Figure 9
figure 9

The proposed model for UPR pathways of the pathogenic Cryptococcus species complex. (a) C. neoformans Ire1 controls environmental stress responses, differentiation, and virulence in Hxl1-dependent and -independent manners. The Ire1-Hxl1-dependent UPR pathway governs responses to ER and cell wall/membrane stress, azole drug resistance, and thermotolerance through controlling expression levels of KAR2. In capsule production, the transcription factor Rim101 acts as a downstream factor of Ire1. The unidentified downstream factors of Ire1 regulate DNA damage response, sexual differentiation, and melanin production. (b) C. deuterogattii (R265) Ire1 regulates environmental stress responses and virulence in Hxl1-dependent and -independent manners. The Ire1-Hxl1-dependent UPR pathway mainly controls cell wall/membrane integrity, azole drug resistance, and ER stress. Ire1 mostly regulates thermotolerance, and melanin synthesis. ERS, ER stress; CWI, cell wall/membrane integrity; THT, thermotolerance; VIR, virulence; DDR, DNA damage response; AZR, azole drug resistance; CAP, capsule production; SEX, sexual differentiation; MEL, melanin synthesis; OXR, oxidative stress response.

Full size image

Regardless of its conserved roles, several notable functional divergences of Ire1 and Hxl1 were observed. The most notable difference was the different roles of Ire1 and Hxl1 in the thermotolerance of C. deuterogattii. Deletion of HXL1 did not significantly affect the growth of C. deuterogattii at 37 °C, whereas the ire1Δ mutant could not grow at the host physiological temperature. This is in contrast to the finding that deletion of HXL1 causes even more severe growth defects at 37 °C than IRE1 in C. neoformans. These findings suggest that the C. deuterogattii UPR pathway might control thermotolerance in a fashion different than it does in C. neoformans. Supporting this, we found that the expression level of PMT1 was similar among R265 and UPR mutants upon temperature upshift, although it is significantly downregulated in the C. neoformans hxl1Δ mutant24. Ire1 appears to mainly regulate the thermotolerance of C. deuterogattii through activation of other unknown transcription factors. In fact, we previously observed that complementation of the ire1Δ mutant with a spliced version of HXL1 (induced active form) only partially restores the growth of C. neoformans at 37 °C24, suggesting that other transcription factors along with Hxl1 mediate its Ire1-dependent thermotolerance. It is probable that such transcription factors could play a major role in the UPR pathway-mediated thermotolerance in C. deuterogattii. Supporting this, the expression of KAR2 is positively regulated by Ire1 and Hxl1 in C. neoformans upon temperature upshift24. However, we found that C. deuterogattii Ire1 and Hxl1 do not regulate KAR2 induction upon temperature upshift. In addition to thermotolerance, azole drug resistance patterns of UPR mutants are different within the pathogenic Cryptococcus species complex. In both C. neoformans and C. deneoformans, hxl1∆ mutants are more susceptible to azole drugs than ire1∆ mutants, which is opposite to the phenotype of C. deuterogattii UPR mutants. In addition to Hxl1, other transcription factors could also modulate the UPR-mediated azole drug resistance in C. deuterogattii. Such Ire1-dependent transcription factors involved in thermotolerance and azole drug resistance remain to be further identified and characterized in future studies.

This type of functional divergence in conserved signalling components appears to be extensive in the pathogenic Cryptococcus species complex. One example of this is found in the catalytic subunits Pka1 and Pka2 of PKA in the cAMP pathway. In C. neoformans, Pka1 regulates melanin and capsule production, whereas Pka2 is not involved in virulence factor formation18. In C. deneoformans, however, Pka2, but not Pka1, plays a role in virulence factor production19. In C. deuterogattii, Pka1 and Pka2 share roles in capsule production and mating responses, and Pka2, but not Pka1, retains its function in melanin biosynthesis20. As another example, C. neoformans and C. deuterogattii calcineurin orthologs have divergent roles in controlling Ca2+ and Li2+ homeostasis and fluconazole tolerance21. This functional divergence of evolutionarily conserved signalling components in several core virulence-regulating signalling pathways may contribute to the competitive fitness of the pathogenic Cryptococcus species complex in diverse environmental and host niches and to different infectivity and virulence potentials.

In this study, we found that the role of the UPR pathway in azole drug resistance is conserved in the pathogenic Cryptococcus species complex, but in a way different from other human fungal pathogens such as A. fumigatus and C. glabrata36,37,38. Askew et al. reported that UPR components such as HacA and IreA play critical roles in antifungal drug resistance of A. fumigatus37,38. Transcriptome analysis revealed that expression levels of many ergosterol biosynthesis genes including ERG11 are reduced in both hacA and ireA mutants. Supporting this, the ergosterol content itself is also decreased in the A. fumigatus UPR mutant38. In contrast, basal and fluconazole-mediated induced expression levels of ERG11 in all UPR mutants of the pathogenic Cryptococcus species complex were not different from wild-type levels, indicating that the UPR pathway does not directly regulate ergosterol biosynthesis. Instead, the UPR-mediated cell membrane stability appears to contribute to azole drug resistance of the Cryptococcus species at least partly, because azole treatment eventually perturbs cell membrane integrity. Similar to the case in Cryptococcus, perturbation of IRE1 alone does not affect azole drug susceptibility in C. glabrata36, indicating that the UPR pathway is not a direct regulator of ergosterol biosynthesis in these pathogenic yeasts.

As the most important finding in this study, we showed that ER stress response and adaptation itself are critical for the virulence of the pathogenic Cryptococcus species. This has been a challenging question because impairment of the UPR pathway leads to pleiotropic effects, including the inability to grow at 37 °C, which is the most obvious virulence determinant for C. neoformans. In this study, we provided two pieces of evidence showing that the UPR pathway contributes to the virulence of the pathogenic Cryptococcus species in a thermotolerance-independent manner. First, the C. neoformans ire1Δ mutant did not kill the wax moth G. mellonella when the insect was incubated at 25 °C. The finding that the C. neoformans hxl1Δ mutant exhibited wild-type levels of virulence in the G. mellonella model at 25 °C suggests that proliferation regulated by Hxl1 at the host physiological temperature is an essential factor contributing to virulence of C. neoformans. Second, the C. deuterogattii hxl1Δ mutant, which did not show significant growth defects at 37 °C, is avirulent in the murine model of systemic cryptococcosis. Although the C. deuterogattii hxl1Δ mutant showed a minor growth defect in the presence of tBOOH at 37 °C, the most notable phenotype of the hxl1Δ mutant was high susceptibility to ER stress agents such as TM (an N-glycosylation inhibitor) and DTT. Therefore it is likely that ER stress response and adaption itself are critical for the virulence of C. deuterogattii. Previous studies revealed that N-glycosylation, a posttranslational modification, occurs in the ER and affects fungal pathogenesis40,41. Furthermore, O-mannosylation, which is initiated in the ER, also contributes to fungal virulence42,43. Specifically, the deletion of O-mannosyltransferases PMT1 and PMT4 attenuates virulence in C. neoformans44,45. Our finding that expression of lipid-linked N-oligosaccharyltransferase genes including ALG7, WBP1, and OST1 is regulated by the UPR pathway supports that Hxl1-mediated ER stress response and adaptation is important for the virulence of C. deuterogattii.

In conclusion, our data suggest that the UPR pathway in C. deuterogattii plays evolutionarily conserved and divergent roles in stress responses and virulence compared to C. neoformans.

Methods

Ethics statement

Animal studies were conducted in the Division of Lab Animal Research at Duke University according to guidelines set forth by the United States Animal Welfare Act and the Duke Institutional Animal Care and Use Committee. The methods were approved by the Duke IACUC under protocol A171-16-08.

Strains and growth conditions

C. neoformans, C. deneoformans, and C. deuterogattii strains used in the study are listed in Supplementary Table 2 and were cultured on yeast extract-peptone-dextrose (YPD) medium. For capsule production assay, the agar-based Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen, Carlsbad, CA) and Littman’s medium were used19,33,46. For melanin biosynthesis, Niger seed medium containing the indicated concentration of glucose was prepared as previously described46.

Construction of C. deuterogattii R265 strain ire1Δ and hxl1Δ mutants

For disruption of IRE1 or HXL1 genes in the C. deuterogattii R265 strain, information regarding the IRE1 or HXL1 genomic structure and sequences was obtained from the C. deuterogattii genomic database (http://www.broadinstitute.org/annotation/genome/cryptococcus_neoformans_b/MultiHome.html). Primers used for the amplification of 5′- and 3′-flanking regions of IRE1 or HXL1 genes in the C. deuterogattii R265 strain are listed in Supplementary Table 3. M13Fe and M13Re primers were used for amplifying the Narr dominant selectable marker. The IRE1 and HXL1 gene disruption cassettes were generated by double-joint PCR (DJ-PCR), as previously described47. Each gel-extracted gene disruption cassette was biolistically inserted into the C. deuterogattii R265 strain. Stable transformants were selected on YPD medium containing nourseothricin and screened by diagnostic PCR. To verify correct genotypes of ire1Δ and hxl1Δ mutants in R265 strains, Southern blot analysis was performed (Supplementary Fig. 7)48.

Construction of IRE1 and HXL1 complemented strains of the C. deuterogattii R265 strain

To verify the phenotypes observed in R265 ire1Δ and hxl1Δ mutants, we constructed corresponding complemented strains as follows. For construction of IRE1 and HXL1 complemented strains of the R265 strain, PCR fragments containing 5′- and 3′-flanking regions and open reading frames (ORFs) of IRE1 or HXL1 genes were amplified with primers B5808/B5809 and B5803/B5804, respectively. The amplified PCR products were cloned into the plasmid pTOP-V2 to produce the plasmids pTOP-IRE1 (YSBE542) and pTOP-HXL1 (YSBE543). After confirming the DNA sequence with no errors, the cloned genes were subcloned into the plasmid pJAF12 to produce the plasmids pJAF12-IRE1 (YSBE544) and pJAF12-HXL1 (YSBE545). The designated restriction enzyme-digested plasmids (NdeI for pJAF12-IRE1 (YSBE544) and PmlI for pJAF12-HXL1 (YSBE545)) were linearized and biolistically inserted into the ire1Δ (YSB1889) and hxl1Δ (YSB1891) mutants, respectively. The re-integration of IRE1 and HXL1 into their native loci was confirmed by diagnostic PCR with the primer pairs B4644/B4646 and B4638/B4640, respectively.

Stress sensitivity test

Each pathogenic Cryptococcus strain was incubated overnight at 30 °C in liquid YPD medium, washed, serially diluted (dilutions from 1 to 10−4) with dH2O, and spotted (3 μL) onto solid YPD medium containing the indicated concentrations of stress inducers. To test ER and cell wall/membrane stress, cells were spotted onto a solid YPD medium containing the indicated concentration of the ER stress inducers tunicamycin (TM) or dithiothreitol (DTT), cell wall stress inducers calcofluor white (CFW) or Congo red (CR), or cell membrane destabilizers sodium dodecyl sulphate (SDS) or fludioxonil (FDX). To determine thermotolerance, plates were incubated at 30 °C, 37 °C, and 39 °C. For the antifungal drug resistance test, cells were spotted onto a solid YPD medium containing the indicated concentrations of the azole drugs fluconazole (FCZ), ketoconazole (KCZ), and itraconazole (ICZ). Cells were further incubated at 30 °C for 4 or 5 days and photographed daily. For oxidative stress, C. deuterogattii cells were cultured at 30 °C in liquid YPD medium, washed, serially diluted (dilutions from 1 to 10−4) with dH2O, and spotted (3 μL) onto solid YPD medium containing the indicated concentration of oxidative stress inducers. Cells were further incubated at 25 °C or 37 °C for 1–3 days and photographed daily.

Total RNA isolation, RT-analysis, and quantitative RT-PCR

Total RNA was isolated with the TRIzol reagent as previously described49. For cDNA synthesis, the total RNA amount was adjusted to 5 μg with DEPC-treated water, to which 1 μL of 5 μg/μL oligo (dT)-pdN6 was added, and the mixture was incubated at 70 °C for 10 min and then placed on ice for 10 min. Next, 15 μL of the cDNA synthesis mixture including reverse transcriptase (Thermo Fisher Scientific, Waltham, MA), RNase inhibitor (Intron), and 10 mM dNTP was added and further incubated at 42 °C for 2 h. After incubation, the mixture was heat-inactivated at 65 °C for 15 min. To monitor HXL1 splicing events, RT-PCR of HXL1 and ACT1 was performed with gene-specific primers listed in Supplementary Table 3. For quantitative RT-PCR analysis, expression levels of ERG11, ERG3, KAR2, DER1, ALG7, PMT1, PMT4, WBP1, OST1, SEC61, ERV29, SOD2, PPS1, and CHS2 were measured with gene-specific primers listed in Supplementary Table 3 on a MyiQ2 Real-Time PCR detection system. For qRT-PCR, target gene expression levels were normalized using ACT1 expression levels as controls.

Melanin and capsule test

For the capsule and melanin assay, cells were incubated overnight (16 h) at 30 °C in a liquid YPD medium. For melanin assay, 3 μL of cells were spotted on agar-based Niger seed medium, which contained the indicated concentration of glucose (0.1%). The plates were incubated at 30 °C, monitored daily, and photographed using SPOT insight digital camera (Diagnostic Instruments Inc., Sterling Heights, MI). For capsule production assay, cells were spotted onto agar-based DMEM or cultured in liquid Littman’s medium at 30 °C for 2 days. For visualization of capsule synthesis, cells scraped from agar-based DMEM or cultured in liquid Littman’s medium were resuspended in PBS, stained with India ink (BACTIDROPTM; Remel, San Diego, CA), and observed under the Olympus BX51 microscope equipped with a SPOT Insight digital camera (Diagnostic Instruments Inc.). For quantitative measurement of capsule, cells were collected from liquid Littman medium. The relative capsule diameter was determined by the following equation (1): (D w –D c ) × D w , where D w and D c indicate the diameter of the whole cell body and the diameter of the cell body, respectively46.

Galleria mellonella infection assay and virulence assay

The G. mellonella infection assay was performed by following the previously described method with minor modifications34. Fifteen G. mellonella caterpillars in the final instar stage (Vanderhorst, Inc., St Marys Ohio USA) were used per each group. C. neoformans strains were grown in liquid YPD media for 16 h at 30 °C and then washed three times with PBS. Cells were resuspended in PBS and counted using a haemocytometer. The caterpillars were inoculated with 4 μL of 1 × 106 cells/mL using Hamilton syringes equipped with 10-μL needles. PBS was injected as a negative control. After injection, caterpillars were incubated at 25 °C or 37 °C and monitored daily. The survival curve was illustrated by Prism 6 (GraphPad Software, La Jolla, CA) and statistically analysed by log-rank (Mantel-Cox) test.

Mouse study

Five- to six-week-old female A/Jcr mice (Charles River, Wilmington, MA) were used for this study. Strains were grown in YPD broth for 24 h at 24 °C and washed twice in sterile PBS (Sigma-Aldrich, St. Louis, MO). Cells were counted using a haemocytometer, and the final inoculum was adjusted to 50000 CFU/50 μL of PBS. This volume was administered to each mouse. Groups of 10 mice for each strain were anesthetized with isoflurane and infected via intranasal instillation. Mice were monitored each day as described in the approved protocol and euthanized by CO2 inhalation upon reaching humane endpoints. The survival curve was created using GraphPad Prism software, and P values of < 0.05 were considered significant. Fungal burden analysis of brains and lungs was performed at 14 days post-infection. Tissue samples were homogenized using a Biospec bead beater and steel beads (Qiagen, Hilden, Germany). Homogenates were serially diluted, and 100 μL of the homogenates were plated on YPD plates with chloramphenicol and ampicillin. Plates were incubated at 24 °C for 48–72 h, after which the burden (CFU/g of tissue) was measured. For histopathology, each organ was cut in half and fixed in 10% buffered formalin solution (Thermo Fisher Scientific), embedded in paraffin and sectioned at 5 μm. Histopathological changes were assessed on H&E-stained sections by a board-certified veterinary pathologist without knowledge of the group. Mucicarmine-stained sections were evaluated to visualize fungi.