Genetically encoded sensors for monitoring intracellular redox health of the pathogenic fungus Cryptococcus neoformans

Abstract

Redox sensing and regulation are critical to both the survival and virulence strategies used by the pathogenic fungus Cryptococcus neoformans to evade host immunity and establish infection. However, the precise genetic and biochemical mechanisms driving these redox regulation systems in the context of fungal virulence are unclear. To address this limitation, we designed genetically encoded redox sensors optimized for expression in C. neoformans, and linked these sensors to cryptococcal redox proteins for real time monitoring of intracellular redox status. Using these sensors, we established several fluorescence-based techniques for monitoring dose-responsive changes in the intracellular oxidation status of C. neoformans under stress. Specifically, we demonstrated sensor responsiveness to non-toxic doses of peroxide stress and during different stages of cell growth, and we verified sensor responsiveness in a mutant with known sensitivity to oxidative stress. This approach provides a framework for developing and deploying biosensors in pathogenic fungi and in basidiomycetes – a group of microorganisms with relatively few sophisticated genetic tools for molecular and synthetic biology. Overall, our sensors enable real time insights into the key redox mechanisms driving growth and survival of a globally important pathogen and pave the way for tool development in other fungi.

Graphical Abstract

C. neoformans is an opportunistic pathogen of global significance that has emerged as a valuable model for studying infectious diseases^1–4. Recently designated as a fungal pathogen of critical importance by the World Health Organization, C. neoformans can cause life-threatening meningoencephalitis and pulmonary disease if left untreated^5,6. Though the incidence of cryptococcosis has declined in Europe and North America since the introduction of antiretroviral therapy, the global burden of disease is still substantial, especially among countries with a high prevalence of HIV/AIDS^7–9. Despite its global impact, little is known about the molecular mechanisms that drive cryptococcosis – a disease with limited treatment options and no vaccine^5,6. Advances in genetic approaches, including the sequencing of C. neoformans genomes from several isolates, have enhanced our understanding of cryptococcal gene function and elaboration of key virulence factors required for infection^10,11. Development of selection markers, fluorescent reporters, and genomic ‘safe havens’ for gene integration have also improved the genetic tractability of this organism. However, as with other basidiomycetes – a group of fungi that includes mushrooms, plant pathogens, and wood-rotting fungi – the molecular toolkit for C. neoformans has lagged behind that of model systems such as Escherichia coli and Saccharomyces cerevisiae. New systems for genetic manipulation, regulation, and monitoring of C. neoformans are therefore needed to better elucidate the molecular mechanisms that drive cryptococcal pathogenesis.

For fungal pathogens of mammals, mitigation of external reactive oxygen species (ROS) is essential for survival and proliferation in the host, where innate immunity depends on the generation of harmful free radicals by phagocytic cells to eliminate invading microbes^12,13. Cells must also tightly regulate internally generated ROS, which participate in cell signaling events for various metabolic pathways^14–16. Hydrogen peroxide (H₂O₂), for example, participates in the reversible oxidation of reactive protein cysteine residues, which can trigger signalling cascades involved in regulating cellular redox, cell growth and metabolism, virulence, and apoptosis^17–21. Additionally, antimicrobial tolerance and fungal persistence under stress may be linked to oxidative stress response mechanisms, and could represent potential targets for antifungal drug development^22–24. Our previous work demonstrated that redox metabolism is a key determinant of cryptococcal virulence as it contributes to both survival in the hostile host environment and the formation of melanin, a critical virulence factor of C. neoformans²³. Specifically, we discovered that genetic perturbation of the GSH biosynthetic pathway – a key redox system across all kingdoms of life – results in metabolic dysregulation that ultimately impacts extracellular acidity, virulence factor formation, and disease progression in a murine model of cryptococcosis²³. Other studies on C. neoformans have identified key roles for redox metabolism in melanin formation and regulation of phagosomal pH^25,26, iron homeostasis^27,28, resistance to oxidative and nitrosative stressors^29,30, immune evasion³¹, and virulence^3,23,30–32. Thus, understanding the complex mechanisms driving redox homeostasis is crucial for characterizing the strategies used by pathogenic fungi to adapt to the host environment, proliferate, and cause disease.

Genetically encoded redox sensors have been fine-tuned for real-time measurements of redox homeostasis in several model systems including budding and fission yeasts, plants, bacteria, and mammalian cells^33–37. For instance, HyPer – a yellow fluorescent protein based on the bacterial OxyR transcription factor – has been exploited for the development of an intracellular H₂O₂ sensor³⁸. Other sensors, including redox-sensitive green fluorescent protein (roGFP) and redox-sensitive mCherry (roCherry), are sensitive to changes in whole cell or organellar redox homeostasis, and show dramatically improved sensitivity when paired with specific redox proteins including glutaredoxins (Grxs) or thiol peroxidases (e.g., peroxiredoxins (Prxs) and glutathione peroxidases (Gpxs))^33,35. Such fusions enable dynamic quantification of the redox potential for specific thiol redox couples (e.g., GSH vs. GSSG)^33–35,37. Though ROS are produced through normal respiration and contribute to cellular signalling cascades, excess accumulation of intracellular and/or extracellular ROS can have deleterious effects on the cell³⁹. Thus, redox-based probes should be sufficiently sensitive to detect pre-toxic doses of exogenous and endogenous ROS to prevent undue damage to the cell³⁷.

Intriguingly, wild-type C. neoformans can tolerate H₂O₂ treatment up to ~1 mM without deleterious impacts on cell growth⁴⁰. Such resilience likely evolved in response to environmental predation, and coincidently confers virulence and survival advantages within host alveolar macrophages that are encountered during the initial stages of infection in mammalian hosts^41,42. However, to our knowledge, genetically encoded sensors for monitoring intracellular redox health have not been successfully developed in basidiomycetes, a group of fungi that are often recalcitrant to genetic manipulation⁴³. Developing such tools for monitoring intracellular redox conditions has become increasingly important to understand complex metabolic circuits in different environmental contexts. This is particularly true for pathogenic organisms, which have evolved complex attack and defense systems to rapidly adapt to stressful conditions of their hosts^44–46. Among these systems, defense mechanisms to maintain redox homeostasis are critical for pathogen survival, host immune evasion, and dissemination of disease^12,47–49. For example, C. neoformans has evolved complex mechanisms for combating oxidative agents encountered in the host and can manipulate the host environment to favor conditions required for virulence factor elaboration^23,25. Although redox-dependent signaling pathways/processes in C. neoformans are known to influence the expression of stress response genes and elaboration of key virulence factors (e.g., melanin formation), the mechanistic and spatiotemporal aspects of these processes are less clear. Moreover, redox exchanges often occur rapidly due to the reactive and transient nature of radical intermediate species, limiting their detection by omics- and other time-point-based approaches⁵⁰. Current approaches for quantifying oxidative stress in C. neoformans depend on the use of fluorescent dyes (e.g., 2’,7’-dichlorodihydrofluorescein diacetate, DCFDA) which, although effective, are only useful for single timepoint measurements due to the irreversible conversion of dyes upon oxidation⁵⁰. Additionally, chemical dyes can interact with other metabolites in the cell resulting in shifts in cellular redox homeostasis or cell toxicity; the sensitivity of dyes may also vary between experiments depending on the level of dye penetration into the cell^50,51.

We therefore sought to engineer genetically encoded roGFPs optimized for expression in C. neoformans to enable real-time monitoring of cryptococcal redox metabolism. To do this, we adapted the genetically encoded redox sensor roGFP2 for use in the fungus and linked this sensor to two cryptococcal redox-active proteins. Using these sensors, we developed techniques and parameters for quantifying the dose-responsive intracellular oxidation status of C. neoformans under stress and we demonstrated sensor response in a cir1∆ mutant with known sensitivity to H₂O₂. Overall, the development of these sensors provides a framework for sensor development in other pathogenic fungi and in basidiomycetes, which currently lack probes for monitoring intracellular redox health. Specifically in C. neoformans, our sensors can help elucidate redox mechanisms that govern the stress response of this fungus, and overcome limitations in studying the survival and virulence of C. neoformans.

EXPERIMENTAL SECTION

Strains and growth media.

A list of strains, plasmids, and oligonucleotides used in this study are provided in Supplementary Table S1. C. neoformans var. grubii (serotype A) strain H99 was used as the wild type strain and for expression of roGFP2 fusion redox sensors. Wild type strains were maintained on solid yeast peptone dextrose (YPD, Difco; 1% yeast extract, 2% peptone, 2% dextrose, 2% agar) and transformants were maintained on solid YPD containing 200 μg ml⁻¹ neomycin. For sensor expression assays, single colonies were inoculated in 5 mL liquid YPD and grown overnight at 30°C (with shaking at 220 rpm) and were then harvested or transferred to yeast nitrogen base minimal medium (YNB, Difco; with amino acids, supplemented with 2% glucose, pH 6.0) for further experimentation.

Sensor and strain construction.

All redox sensors were constructed using a C. neoformans codon-optimized roGFP2 construct (Bio Basic) integrated into plasmid pSDMA57, which contains a neomycin resistance cassette and homology arms to a cryptococcal intergenic safe haven region for ease of genomic integration⁵². Genes encoding Gpx2 and Tsa1∆C_R were synthesized and integrated downstream of the roGFP2 construct in pSDMA57. The roGFP2 and Gpx2/Tsa1∆C_R fragments were separated by a 30 amino acid glycine-rich linker domain. All roGFP2 fusion proteins were integrated chromosomally into the C. neoformans strain H99 or the cir1∆ mutant via biolistic transformation of AscI-linearized recombinant pSDMA57 plasmids, as described previously^52,53, and positive transformants were selected on solid YPD containing 200 μg ml⁻¹ neomycin. Fusion protein constructs were placed under the control of the C. neoformans TEF1 promoter to ensure constitutive expression.

Fluorescence measurement of roGFP2 sensor expression.

Sensor expression and responsiveness to oxidation were quantified as described previously, with slight modifications⁵⁴. Briefly, recombinant strains stably expressing roGFP2 fusion proteins were grown to stationary phase in 5 mL YPD overnight at 30°C with shaking at 220 rpm. Cells were then harvested via centrifugation at 12,000 rpm. for 2 min, washed twice with sterile water, and subcultured in 25 mL fresh YNB at an initial OD₆₀₀ of 0.03 for 18 h at 30°C with shaking at 220 rpm. Wild type strains were used as a control to account for cell autofluorescence. After 18 h of growth, cells were harvested via centrifugation at 2,500 rpm., washed twice with sterile water, and resuspended in 100 mM 2-(N-morpholino)ethanesulfonic acid (MES) buffer (pH: 6.0) at a final concentration of OD₆₀₀ = 8. A 180 µL aliquot of resuspended cells was then transferred to a flat-bottom 96-well plate per treatment for each strain, and 180–200 µL 100 mM MES buffer was aliquoted as a blank control with or without 20 µL treatment. For oxidized and reduced controls, 20 µL of 1 M H₂O₂ or 1 M DTT were added to resuspended cells, respectively, and to 180 µL aliquots of 100 mM MES buffer as blank controls. Suspensions were then mixed briefly by orbital shaking and measured immediately via fluorescence plate reader or fluorescence microscopy. Fluorescence plate reader measurements were obtained using a BioTek Synergy H1 microplate reader (Agilent) with BioTek Gen5 software (v.3.15.15) using excitation wavelengths of 405 and 480 nm and an emission wavelength of 510 nm. Fluorescence values for all experiments were calculated by subtracting the background fluorescence of wild type C. neoformans cells for each treatment from the fluorescence of sensor-expressing recombinant cells. Overall sensor oxidation (OxD) for fluorescence plate measurements was calculated as described previously^34,55, using the following equation:

$$\mathrm{O}\mathrm{x}\mathrm{D}=\frac{I_{405}\times I_{480\left(\mathrm{r}\mathrm{e}\mathrm{d}\right)}-I_{405\left(\mathrm{r}\mathrm{e}\mathrm{d}\right)}\times I_{480}}{I_{405}\times I_{480\left(\mathrm{r}\mathrm{e}\mathrm{d}\right)}-I_{405}\times I_{480\left(\mathrm{o}\mathrm{x}\right)}+I_{405\left(\mathrm{o}\mathrm{x}\right)}\times I_{480}-I_{405\left(\mathrm{r}\mathrm{e}\mathrm{d}\right)}\times I_{480}}$$

Where I_405/480 is the background-corrected fluorescence intensity at a given timepoint, I_405/480(red) is the background-corrected fluorescence intensity of the fully reduced control at a given timepoint, and I_405/480(ox) is the background-corrected fluorescence intensity of the fully oxidized control at a given timepoint.

Fluorescence microscopy was performed using a Zeiss Axioplan 2 microscope equipped with a Plan-Apochromat 100x/1.46 objective lens and an ORCA-Flash4.0 LT CMOS camera (Hamamatsu Photonics). roGFP2 fusion proteins were excited by 399 and 488 nm lasers and emissions were detected with a 510 nm filter. Zeiss Zen 2 Blue edition (v.2.3) and ImageJ (v.2.14.0)⁵⁶ were used for image collection and measurement of fluorescence intensity, respectively. Analysis of microscopy images was conducted as described previously⁵⁷. Briefly, images were exported to ImageJ software (v.2.14.0)⁵⁶ as 32-bit TIFF images for fluorescence quantification. Background fluorescence was subtracted from each image and fluorescence ratio images were constructed by dividing the 405 nm image by the 488 nm image. Image thresholds were set to avoid artifacts from ratiometric analysis. Fluorescence ratios and mean fluorescence intensity were quantified for each image, and fluorescence ratios were normalized to the ratios of fully oxidized and fully reduced controls using the following equation, as described previously⁵⁸:

$${\mathrm{R}}_{\mathrm{n}\mathrm{o}\mathrm{r}\mathrm{m}}=0.9\times \frac{({\mathrm{R}}_{\mathrm{t}}-{\mathrm{R}}_{\mathrm{t}\left(\mathrm{r}\mathrm{e}\mathrm{d}\right)})}{({\mathrm{R}}_{\mathrm{t}\left(\mathrm{o}\mathrm{x}\right)}-{\mathrm{R}}_{\mathrm{t}\left(\mathrm{r}\mathrm{e}\mathrm{d}\right)})}+0.1$$

Where R_norm is the normalized 399/488 nm ratio, R_t is the 399/488 nm ratio at a given timepoint, R_t(red) is the 399/488 nm ratio of the fully reduced control at a given timepoint, and R_t(ox) is the 399/488 nm ratio of the fully oxidized control at a given timepoint.

Flow cytometry.

Samples for flow cytometry were prepared as described in the previous section for fluorescence measurement of sensor expression. After 18 h growth, cells were harvested via centrifugation at 2,500 rpm., washed twice with sterile water, and resuspended in 100 mM MES buffer at a final concentration of OD₆₀₀ = 1. Cells exposed to oxidative stress were treated with H₂O₂ 5–10 min prior to analysis. Sensor expression was quantified using a CytoFLEX S flow cytometry (Beckman Coulter) with lasers at wavelengths of 405 nm (violet), 488 nm (blue), 561 nm (yellow), and 633 nm (red). The results were gated to single yeast cells and PBS and wild type cells were used as blank and negative controls, respectively. Fluorescence of sensor-expressing strains was measured using the FITC-GFP and KO525 channels, and data were acquired and analysed using the CytExpert cytometry analysis software (Beckman Coulter).

Protein extraction and immunoblotting.

Overnight cultures were transferred 1:10 to YPD at a final volume of 50 mL and grown for 6 h to log phase at 30°C with shaking at 220 rpm. After the 6 h growth period, all 50 mL of cells were harvested for each sample via centrifugation at 3,500 rpm. for 5 min at 4°C. For fully oxidized samples, 100 mM H₂O₂ was added to cell culture 10 min prior to harvesting. Cells were kept on ice, washed twice with sterile cold water, and flash frozen with liquid nitrogen. Cell pellets were then homogenized mechanically by grinding using a mortar and pestle, and the resulting powder extract was transferred to a 1.5 mL microfuge tube. We then added 300 µL cell lysis buffer (50 mM Tris-HCl, pH: 7.5, 5 mM EDTA, 100 mM NaCl, 1% Triton X-100, 1X cOmplete mini EDTA-free protease inhibitor cocktail (Roche 11836170001)) to powder extracts and vortexed the samples at 4°C for 5 min. The mixtures were then placed on ice for 5 min, followed by centrifugation at 13,000 rpm. for 15 min at 4°C to collect cell lysates in the supernatant. The resulting supernatant was transferred to a 1.5 mL microfuge tube and stored at −20°C until immunoblotting analysis. For immunoblotting, 25µL of cell lysate was combined with 1X protein loading dye and boiled for 10 min. Samples were separated via SDS-PAGE using a 10% SDS gel and transferred to a nitrocellulose membrane for 2 h at 4°C. Membranes were then blocked in Tris-buffered saline (TBS-T) with 0.1% Tween 20 and 5% skim milk for 1 h at room temperature, rinsed twice for 10 min each with TBS-T, and incubated with 1:1000 anti-mouse GFP monoclonal antibody (B-2) conjugated to HRP (Santa Cruz Biotechnology) in TBS overnight at 4°C with gentle shaking. Following incubation, membranes were rinsed twice for 5 min each with TBS-T, incubated for 5 min in protein detection reagent (100 mM Tris-HCl, pH: 8.5, 1.25 mM luminol, 0.198 mM p-coumaric acid, 0.93% H₂O₂), and visualized using chemiluminescence using a Bio-Rad ChemiDoc XRS imaging system.

Statistical analysis.

Statistical analyses were performed using GraphPad Prism 8 (v.8.2.1). Distribution of data was evaluated using the Shapiro-Wilk test for normality. Statistical significance was calculated using the Kruskal-Wallis or two-way analysis of variance (ANOVA) tests for non-parametric and parametric data, respectively, and corrected for multiple comparisons using the methods indicated. The threshold for statistical significance was defined as α = 0.05.

RESULTS AND DISCUSSION

Design of roGFP2-based sensors for C. neoformans.

As with other basidiomycete fungi, C. neoformans lacks genetic tools for real-time characterization of the redox-based mechanisms that support survival and proliferation of cells under stress. We therefore sought to design and optimize expression of the genetically encoded redox sensor, roGFP2, in C. neoformans. Since the excitation behaviour of roGFP-based probes varies depending on oxidation status, sensor activity can be quantified by ratiometric analysis of fluorescence intensity under different conditions. Specifically, roGFP2 sensors exhibit two excitation maxima when monitored at a fluorescence emission of 510 nm; fully reduced sensor conformations have excitation maxima ranging from 475–490 nm, while fully oxidized sensor conformations have excitation maxima from 400–410 nm^33,34. Therefore, the relative status of the cellular redox environment can be inferred from the degree of sensor oxidation upon treatment with various oxidizing/reducing agents.

Here, we generated two redox protein-coupled roGFP2 fusions to establish real-time measurements of the cryptococcal redox environment under stress. To do this, we first generated an roGFP2 construct codon optimized for expression in C. neoformans using the approach described by Huang et al., 2022⁵⁹. Next, we identified candidate oxidoreductase proteins in C. neoformans that could equilibrate the codon-optimized roGFP2 sensor to specific redox pairs of interest (e.g., GSH and GSSG) and/or oxidative stress agents (e.g., H₂O₂). The fusion of redox-active proteins increases the responsiveness and sensitivity of roGFP2-based sensors^33,34. For instance, roGFP2 fusions with the glutaredoxin Grx1, a GSH-dependent thiol-disulfide oxidoreductase, or the GSH peroxidase Orp1, have been successfully employed in yeast, plant, and mammalian cells^33,36,57. More recently, fusions with the thioredoxin peroxidases Tsa2 and Tpx1 in the model yeasts S. cerevisae and Schizosaccharomyces pombe, respectively, have achieved sensitivity to redox perturbations in the low-nanomolar to micromolar range upon mutating the resolving cysteine in each protein (C_R)^34,37. We therefore searched for candidate proteins in C. neoformans with high sequence similarity to redox proteins with demonstrated efficacy as redox probe fusions in model yeasts, including thiol peroxidases (e.g., Tpx1, Tsa1, Tsa2) and glutathione peroxidases (e.g., Orp1).

Our analysis identified two candidate proteins in C. neoformans with high sequence homology to their respective orthologs in yeast (Figure 1a). Specifically, C. neoformans Gpx2 showed 63% amino acid sequence identity to S. cerevisiae Orp1, and the thioredoxin peroxidase Tsa1 showed high sequence identity to both Tpx1 in S. pombe (61%) and Tsa2 in S. cerevisiae (51%) – both of which are exquisitely sensitive to H₂O₂ stress when paired with roGFP2^34,37. Importantly, C. neoformans Gpx2 fully conserves the key active site residues C³⁶, Q⁷⁰, and W¹²⁵ with S. cerevisiae Orp1 that form the enzyme catalytic triad, as well as the F³⁸, N¹²⁶, and F¹²⁷ residues that support formation of the active site pocket⁶⁰ (Figure 1a). Similarly, C. neoformans Tsa1 maintains the conserved catalytic residues T⁴⁵, C⁴⁸ (C_P), and R¹²⁴ with S. pombe Tpx1 (S⁴⁵, C⁴⁸, and R¹²⁴ in S. cerevisiae Tsa2) that form the catalytic triad responsible for nucleophilic attack on peroxide substrates, and C¹⁶⁹ (C_R) which facilitates disulfide bond formation and subsequent reduction of oxidized C_P by an appropriate electron donor⁶¹. Moreover, C. neoformans Gpx2 and Tsa1 show high predicted structural similarity in key catalytic regions when compared with Orp1 or Tpx1 and Tsa2, respectively, using AlphaFold (Figure 1b). The predicted structures are presented in Supplementary Figure 1, and we focused on the regions with high confidence scores to avoid spurious predictions in regions with low scores. The genes for Gpx2 and Tsa1 in C. neoformans also show transcriptional upregulation in response to peroxide stress, and knockout mutants lacking TSA1 show considerable sensitivity when grown on H₂O₂-containing medium^29,31, suggesting that these proteins participate in redox reactions and could effectively equilibrate roGFP2. We therefore designed codon optimized roGFP2 fusion proteins containing C. neoformans Gpx2 or Tsa1∆C_R and placed each fusion construct under the control of the cryptococcal TEF1 constitutive promoter to enhance sensor abundance in transformed cells (Figure 1c). Constructs were biolistically transformed in cryptococcal cells and integrated into an established intergenic Safe Haven locus in the cryptococcal genome (Figure 1d)⁵². Cells expressing the fusion proteins were then isolated and monitored for fluorescence activity. We note that the GPX2 and TSA1 gene constructs fused to our codon optimized roGFP2 contained both introns and exons, and that we were unable to obtain transformants with the roGFP2-TSA1∆C_R construct when TSA1∆C_R was derived from complementary DNA (cDNA), suggesting that expression required intron splicing⁶³. This finding is consistent with previous studies demonstrating the importance of introns for mRNA accumulation in C. neoformans, a species with an abundance of introns^{11, 63}. The importance of introns can potentially inform tool development in other basidiomycete fungi.

Figure 1.

Analysis of the cryptococcal redox proteins Gpx2 and Tsa1 for use in genetically encoded redox sensors. (a) FoldScript (https://foldscript.ibcp.fr) sequence alignment of the glutathione peroxidases Gpx2 from C. neoformans and Orp1 from S. cerevisiae (top), and of the thioredoxin peroxidases Tsa1 from C. neoformans and Tpx1 from S. pombe (bottom). Residue letters written in white on a red background represent exact identity (100%); red letters on a yellow background represent significant amino acid similarity (70–99%); black letters represent low similarity (< 70%). PDB secondary structure features (helices, sheets) are shown above amino acid residues. Key catalytic regions are indicated by red triangles, and blue lines beneath amino acid residues represent key protein domains. Strict α- and β-turns are represented by the letters TTT and TT, respectively. C_P = peroxidatic cysteine residue; C_R = resolving cysteine residue. (b) AlphaFold protein structure prediction for Gpx2 and Tsa1 from C. neoformans, Orp1 and Tsa2 from S. cerevisiae, and Tpx1 from S. pombe. Colors represent distinct structural domains as classified by The Encyclopedia of Domains⁶². (c) Design of the roGFP2-Gpx2 and roGFP2-Tsa1∆CR fusion constructs (top) and integration into a vector backbone containing homology arms to the Safe Haven 1 (SH1) region of the C. neoformans genome (middle). Once integrated, roGFP-containing vectors are linearized and transformed into C. neoformans (bottom). pCnTEF1 = C. neoformans TEF1 promoter; pCnACT1 = C. neoformans ACT1 promoter; pAmpR = ampicillin promoter; tCnGal7 = C. neoformans GAL7 terminator; ori = E. coli origin of replication; NeoR = neomycin resistance cassette; AmpR = ampicillin resistance cassette; (d) AlphaFold protein structure prediction for the C. neoformans optimized roGFP2-Gpx2 and roGFP2-Tsa1∆CR fusion proteins. Green coloration represents the roGFP2 moiety and pink/red coloration represents the Gpx2 or Tsa1∆CR redox protein.

The C. neoformans thiol peroxidases Tsa1 and Gpx2 can oxidize roGFP2.

C. neoformans transformants containing the sensor constructs were identified via colony PCR as described by Huang et al., 2022⁵⁹. Next, we confirmed expression of the roGFP2-Gpx2 and roGFP2-Tsa1∆C_R fusion proteins in positive transformants under stress and non-stress conditions via immunoblot analysis (Supplementary Figure S2a). Once protein expression was confirmed for both proteins, we determined the excitation wavelength spectra of the fully reduced (roGFP2_red) and fully oxidized (roGFP2_ox) conformations of each sensor. Consistent with the excitation spectra of roGFPs in other organisms, we observed distinct peaks at 405 nm and 480 nm under oxidizing and reducing conditions, respectively (Supplementary Figure S2b). These excitation parameters ± 10 nm were used for all experiments described in this study.

To test whether the C. neoformans-optimized sensors were responsive to oxidizing and reducing agents and whether oxidation was reversible, we challenged cells with 100 mM H₂O₂ followed by treatment with 100 mM of the reducing agent dithiothreitol (DTT) and monitored changes in relative fluorescence. After treatment with H₂O₂, we observed an abrupt increase in the 405/480 nm ratio of both the roGFP2-Gpx2 and roGFP2-Tsa1∆C_R sensors (Supplementary Figure S2c). Both sensors reached peak oxidation approximately 2–3 minutes after H₂O₂ treatment and returned to a fully reducing conformation 3–4 minutes after DTT treatment, indicating that sensor oxidation was fully reversible (Supplementary Figure S2c). Thus, we concluded that our sensors showed sufficient dynamic range to be used for monitoring intracellular redox health of C. neoformans.

Next, we further examined the sensitivity and kinetics of our sensors by challenging cells with various concentrations of H₂O₂ and measuring intracellular redox potential over time using a fluorescence plate reader. Both the roGFP2-Gpx2 and roGFP2-Tsa1∆C_R sensors showed complete oxidation or reduction upon treatment with 100 mM H₂O₂ or 100 mM DTT, respectively, and these conditions were used as the fully oxidized or fully reduced controls throughout this study (Figure 2a). Our dose-response analysis showed that the roGFP2-Tsa1∆C_R sensor had a greater dynamic range between the maximum and minimum 405/480-nm excitation ratios compared with roGFP2-Gpx2 and was responsive to H₂O₂ concentrations as low as 10 µM (Figure 2a, Supplementary Figure 3, Supplementary Figure 4). In contrast, the roGFP2-Gpx2 sensor was sensitive to concentrations as low as 100 µM H₂O₂ (Figure 2a). Despite the reduced sensitivity of roGFP2-Gpx2, this sensor was responsive to non-toxic doses of H₂O₂ (< 1 mM) which cause no discernible growth defects for C. neoformans⁴⁰. Consistent with the behaviours of the S. pombe roGFP2-Tpx1.C169S and S. cerevisiae roGFP2-Tsa2∆C_R sensors, our roGFP2-Tsa1∆C_R sensor maintained an oxidized conformation for a longer duration and at lower H₂O₂ concentrations than the roGFP2-Gpx2 sensor, suggesting that the absence of the Tsa1 resolving cysteine residue (C_R) impairs formation of the disulfide bond required for probe reduction, thereby limiting the intracellular H₂O₂ consumption/turnover rate³⁴. Specifically, the degree of oxidation (OxD) of the roGFP2-Tsa1∆C_R sensor was consistent over the 60-minute treatment period when treated with H₂O₂ concentrations of 1 mM or greater, and showed only gradual reduction over time in samples treated H₂O₂ concentrations less than 1 mM (Figure 2b). Conversely, cells expressing roGFP2-Gpx2 treated with 1 mM H₂O₂ showed significant reduction in OxD values after the 60-minute interval (Figure 2b). Moreover, cells treated with 100 µM H₂O₂ reverted to a reducing conformation within 20–25 minutes of H₂O₂ treatment, suggesting a greater rate of sensor reduction and H₂O₂ turnover than in cells expressing the roGFP2-Tsa1∆C_R sensor (Figure 2b). Treatments with less than 100 µM H₂O₂ resulted in no discernible impact on the roGFP2-Gpx2 sensor (Figure 2a,b).

Figure 2.

The C. neoformans codon optimized roGFP2-Gpx2 and roGFP2-Tsa1∆C_R sensors are sensitive to exogenous stress. (a–b) The response of the roGFP2-Gpx2 and roGFP2-Tsa1∆C_R sensors to DTT and various concentrations (0–100 mM) of H₂O₂ (n = 3 biological replicates for each strain and treatment) via fluorescence plate reader. Values represent the mean fluorescence ratio (405 nm/480 nm), a) or degree of sensor oxidation (OxD, b) after excitation at 405 nm and 480 nm and emission at 510 nm ± S.D. Values are corrected using the WT strain autofluorescence for each condition.

roGFP2-based sensors enable imaging of intracellular redox changes in C. neoformans.

To further investigate sensor responsiveness, expression, and localization within the cell, we performed fluorescence microscopy with strains expressing the roGFP2-Gpx2 and roGFP2-Tsa1∆C_R sensors under oxidizing and reducing conditions. Consistent with our fluorescence plate reader data, we observed changes in the oxidation of both sensors upon treatment with H₂O₂ at various concentrations (Figure 3a,b). Specifically, the roGFP2-Gpx2 and roGFP2-Tsa1∆C_R fusion proteins were observed abundantly in the cytosol and showed strong oxidative responses to all concentrations of H₂O₂ tested (Figure 3a,b). Treatment with 100 mM H₂O₂ or 100 mM DTT resulted in complete oxidation or reduction of the roGFP2-Gpx2 and roGFP2-Tsa1∆C_R sensors, respectively (Figure 3a,b), and the observations were used to normalize the ratiometric signal of other treatments for each experiment on a scale of 0 (most reduced) to 1 (most oxidized). Although abundantly expressed in the cytosol, the roGFP2-Tsa1∆C_R protein partially accumulated in the center of recombinant cells (Figure 3a). Though we initially suspected that sensor aggregation resulted from partial cleavage of the roGFP2 moiety from the fusion protein construct, immunoblot analysis revealed that the roGFP2-Tsa1∆C_R fusion protein remained intact and was of the expected molecular weight (~51 kDa) (Supplementary Figure S2a). Intriguingly, Morgan et al. found that the roGFP2-Tsa2∆C_R fusion protein in S. cerevisiae can self-assemble or co-assemble with endogenous Tsa1 and Tsa2 proteins, suggesting that the aggregation of roGFP2-Tsa1∆C_R observed in C. neoformans may result from oligomerization of fusion protein constructs. Overall, these observations demonstrate that the sensors can be employed to effectively monitor intracellular redox changes in C. neoformans by fluorimetry and microscopy.

Figure 3.

The roGFP2-Gpx2 and roGFP2-Tsa1∆C_R sensors enable real time imaging of cryptococcal intracellular redox status. (a) C. neoformans cells expressing either the roGFP2-Gpx2 (left) or roGFP2-Tsa1∆C_R (right) sensor with or without treatment after excitation at 399 nm or 488 nm and emission at 510 nm. Cells were treated with either 100 mM DTT or 100 mM H₂O₂ and imaged. Scale bars = 5 µM. Images represent cells from n = 3 independent experiments for each strain and treatment. The color key indicates the 399 nm/488 nm ratio of mean fluorescence intensity (MFI), where a ratio of 1.0 indicates complete sensor oxidation. (b) MFI values of microscopy images from (a) after treatment with 100 mM DTT or H₂O₂ at the indicated concentrations and subtraction of background fluorescence (left). Ratio of MFI values at 399 nm and 488 nm with or without treatment (middle). Correlation of H₂O₂ concentration with MFI ratio values of H₂O₂-treated cells after normalization to fully oxidized (100 mM H₂O₂ = ratio of 1) or fully reduced (100 mM DTT = ratio of 0) controls (right). Significance values were calculated via Kruskal-Wallis (b, left column) or two-way ANOVA (b, centre column) with Dunn’s or Bonferroni’s correction for multiple comparisons, respectively, relative to the untreated control (****P < 0.0001, **P < 0.01).

C. neoformans roGFP2-based sensors are active during different growth phases.

In a recent study, Ke et al. identified differences in antifungal sensitivity between early exponential phase and stationary phase cryptococcal cells to treatment with Amphotericin B (AmpB)²² – a polyene antifungal known to induce ROS, which contribute to its fungicidal properties⁶⁴. Specifically, stationary phase cells showed enhanced AmpB persistence relative to exponential phase cells, and this was conserved among evolutionarily distant fungal pathogens including various Cryptococcus and Candida species. We suspected that a similar phenomenon may occur in sensor-expressing strains at different growth stages due to changes in physiological and metabolic activity over time. To quantify the responsiveness of the roGFP2-Gpx2 and roGFP2-Tsa1∆C_R sensors during different stages of cell growth, we measured the fluorescence of H₂O₂-treated recombinant strains during log or stationary phase growth using flow cytometry. Interestingly, we found that the proportion of cells responsive to H₂O₂ stress was similar during exponential and stationary phase for both sensors (Figure 4). Specifically, ≥ 95% of roGFP2-Gpx2-expressing cells at either exponential or stationary phase responded to H₂O₂ treatment (Figure 4a). We did observe marginal differences between the responsiveness of the roGFP2-Tsa1∆C_R sensor during exponential (~99% responsive) or stationary (~85% responsive) phase growth, but most cells were responsive to H₂O₂ treatment under either condition (Figure 4b). In summary, the vast majority of recombinant C. neoformans cells express either roGFP2-Gpx2 or roGFP2-Tsa1∆C_R during exponential or stationary growth phases. Additionally, both sensors are responsive to exogenous oxidative agents during different stages of growth, suggesting that the sensor may be used to monitor intracellular redox health during various stages of cell development.

Figure 4.

Cryptococcal roGFP2 redox sensors are responsive to stress during different stages of cell growth. (a–b) Flow cytometry of the roGFP2-Gpx2 (a) and roGFP2-Tsa1∆C_R (b) sensors with or without 100 mM H₂O₂ treatment at either exponential phase (EP) or stationary phase (SP) growth. Fluorescence after excitation at 405 nm and 488 nm was measured at 510 nm using the KO525 or FITC filters, respectively. Treatment with H₂O₂ results in quenching of the FITC/GFP signal. Data represent single cells from n = 3 independent experiments.

A cir1∆ mutant expressing roGFP2-Tsa1∆CR shows elevated sensitivity to oxidative stress.

Previous work from our group identified Cir1 as an iron-responsive transcription factor and master regulator of iron homeostasis in C. neoformans⁶⁵. Moreover, Cir1 is contributes to elaboration of key virulence factors including melanin, which is critical for the antioxidant response of C. neoformans during infection^2,65. Intriguingly, mutants lacking Cir1 are avirulent, show enhanced melanin production, and have increased sensitivity to H₂O₂ stress – the latter of which likely results from the influence of Cir1 on mitochondrial functions and dysregulation of iron homeostasis⁶⁶. Given the importance of this gene for iron regulation and virulence in mice, and the enhanced sensitivity of cir1∆ mutants to oxidative stress, we decided to confirm the elevated H₂O₂ sensitivity of this strain using our roGFP2-Tsa1∆C_R sensor, which showed greater dynamic range than our roGFP2-Gpx2 sensor. To introduce the roGFP2-Tsa1∆C_R sensor into the cir1∆ mutant background, the sensor construct was biolistically transformed into an established intergenic Safe Haven locus of cir1∆ mutant cells⁵². Mutant cells expressing the roGFP2-Tsa1∆C_R sensor were then isolated and monitored for fluorescence activity. Consistent with previous findings showing higher DCFDA staining of cir1∆ mutants compared to the WT, we observed greater oxidation of the roGFP2-Tsa1∆C_R sensor in the cir1∆ mutant background versus the WT strain upon H₂O₂ treatment (Fig. 5a,b). Moreover, sensor-expressing cir1∆ cells showed a more dramatic increase in sensor oxidation upon H₂O₂ treatment than sensor-expressing WT cells (Fig. 5a,b), corroborating the elevated sensitivity of cir1∆ mutants described previously^65,66. Additionally, we observed significantly higher initial oxidation level of the roGFP2-Tsa1∆C_R sensor in untreated cir1∆ cells compared to sensor expression in the WT background (Fig. 5c). Changes in fluorescence of sensor-expressing mutant cells upon H₂O₂ treatment were confirmed using both fluorescence microscopy and flow cytometry (Fig. 5a,d). Taken together, these experiments with the cir1∆ mutant demonstrate the utility of the sensor in evaluating the impact of specific functions on redox homeostasis in C. neoformans.

Figure 5.

Mutants lacking the iron-responsive transcription factor Cir1 show elevated H₂O₂ sensitivity via roGFP2-Tsa1∆C_R sensor expression. (a) C. neoformans cells expressing the roGFP2-Tsa1∆C_R sensor in either the WT (roGFP2-Tsa1∆C_R, left) or cir1∆ mutant (cir1∆::roGFP2-Tsa1∆C_R, right) background. Images are taken with or without 100 mM H₂O₂ treatment after excitation at 399 nm or 488 nm and emission at 510 nm. Scale bars = 5 µM. Images represent cells from n = 3 independent experiments for each strain and treatment. The color key indicates the 399 nm/488 nm ratio of mean fluorescence intensity (MFI), where a ratio of 1.0 indicates complete sensor oxidation. (b–c) Ratio of MFI values at 399 nm and 488 nm with or without treatment calculated from fluorescence microscopy images in (a). (d) MFI of untreated WT and cir1∆ sensor-expressing strains quantified via flow cytometry with or without 100 mM H₂O₂ treatment. Significance was calculated via two-way ANOVA with Bonferroni’s correction for multiple comparisons relative to the untreated condition for each excitation wavelength (****P < 0.0001).

CONCLUSIONS

Pathogenic fungi including C. neoformans lack extensive genetic toolkits for real-time characterization of the redox-based mechanisms driving growth and virulence during infection.

Here, we designed and implemented two genetically encoded redox sensors for real-time monitoring of cellular redox status in C. neoformans. Excitingly, our roGFP2-Tsa1∆C_R and roGFP2-Gpx2 sensors were sensitive to H₂O₂ concentrations as low as 10 µM and 100 µM, respectively, suggesting that our sensors can be used to monitor pre-toxic endogenous or exogenous stressors encountered by C. neoformans in the environment or during infection. Additionally, because the roGFP2-Tsa1∆C_R sensor maintained a greater OxD over the treatment period, this sensor can be employed without rapid scavenging of H₂O₂ and disrupt intracellular redox homeostasis (Figure 2b). Thus, sensor output more precisely reflects intracellular redox changes in response to stress without artificially interfering with redox processes. We also demonstrated that sensor expression can be quantified using a variety of fluorescence-based approaches including fluorescence plate readers (Figure 2), fluorescence microscopy (Figure 3), and flow cytometry (Figure 4), and confirmed elevated sensor oxidation in a cir1∆ mutant with known ROS sensitivity (Figure 5). Moreover, sensor activity is maintained during both exponential and stationary growth phases (Figure 4), suggesting that sensor expression is maintained over extensive periods for length experimental protocols. These findings demonstrate the utility of our sensor for a variety of experimental applications including mutant analysis. Finally, our study is the first application of a genetically encoded fluorescent probe in a basidiomycete, a group of organisms historically recalcitrant to genetic manipulation⁴³, and provides a framework for developing genetically encoded redox sensors for other fungal pathogens and for basidiomycetes. Overall, the genetically encoded fluorescent sensors developed in this study expand the synthetic toolkit of C. neoformans to enhance our understanding of its genetics, ability to grow and survive in human hosts, and the progression of cryptococcal disease. These tools can be used in concert with other genetic and biochemical approaches to create a high-resolution understanding of redox mechanisms governing key aspects of cryptococcal virulence, and may ultimately guide the development of novel antifungal therapeutics.

Supplementary Material

Supplementary Materials

Table S1. Strains, plasmids and oligonucleotides

Fig. S1. Confidence scores for AlphaFold predictions

Fig. S2. Expression and responsiveness of sensors

Fig. S3. Mean fluorescence intensities for roGFP2-Gpx2 (reduced/oxidized)

Fig. S4. Mean fluorescence intensities for roGFP2-Tsa1∆C_R (reduced/oxidized)