BALB/c (stock # 000651) and C57BL/6J (stock #:000664) background mice were purchased from Jackson Laboratory (Bar Harbor, ME) and housed in a specific pathogen-free environment at the Ann Arbor Veterans Affairs Medical Center. Male and female mice were infected at 8–15 weeks of age and were humanely euthanized via CO₂ inhalation. All experiments in this study were performed per National Institutes of Health guidelines and the Guide for the Care and Use of Laboratory Animals and approved by the Veterans Affairs Institutional Animal Care and Use Committee (protocol 1702709, 1597346). For mortality experiments, mice were euthanized at 80% of their body weight after infection. Mice were also monitored for neurological symptoms although none were observed.

C. neoformans strain H99 (ATCC 208821, WT), tps1-gene deleted mutant strain H99 (tps1Δ), and reconstituted mutant strain tps1Δ, with TPS1 gene restored (tps1Δ:TPS1) (Petzold et al., 2006), were grown for 3 days in Sabouraud dextrose broth (Difco) at 30°C and then moved to 37°C overnight after recovery from 10% glycerol frozen stocks stored at −80°C. Yeast cells were washed in PBS, counted on a hemocytometer with trypan blue, and adjusted to a concentration of 1*10⁷ cell/0.03 mL (high, Figures 4 , 5 and ex vivo data for Figure 7 ) or 1*10⁵ cells/0.03 mL (normal/low, Figure 1 ) prior to the infection. Serial dilutions of the inocula were plated on Sabouraud dextrose agar (SDA) plates to confirm the CFU number of each inoculum. For the intratracheal infections, mice were anesthetized via intraperitoneal injections with ketamine/xylazine (100/6.8 mg/kg body weight) prior to infection (Osterholzer et al., 2009b; Stempinski et al., 2022). For the disseminated mode ( Figures 2 , 3 ), yeast cells were washed and counted as above and adjusted to a concentration of 1 × 10⁵ cells/0.2 mL just prior to infection. Male and female mice were infected via retroorbital intravenous injection under inhaled isoflurane anesthesia (Neal et al., 2017; Goughenour et al., 2021).

Organs were dissected using sterile instruments and homogenized in 2 mL of sterile distilled water. Fungal burden for each organ (lungs, brains, and spleens as indicated in the figures) was quantified by seven-step 10-fold serial dilution of organ homogenate in sterile distilled water, followed by plating of 10 μL of homogenate at each dilution level (undiluted—1:10⁷ dilution) on SDA. CFU per organ was determined by averaging two replicate platings. When required to determine clearance, all 2 mL of organ homogenate was plated on SDA.

Per our labs’ established T-cell depletion method (Neal et al., 2017), mice were dosed with either an anti-CD4 monoclonal antibody (Bio X Cell) on day −3 (pre-infection), day 0 (day of infection), and days 7, 14, 21, 28, and 35 postinfection. Animals in the treatment group received 300 µg of anti-CD4 antibody (clone GK1.5) in 200 μL of sterile PBS, whereas those in the control group received 200 mL of sterile PBS. This depletion protocol was performed in both the intratracheal ( Figure 1C ) and disseminated ( Figure 3 ) infection context.

Lungs IT infected with H99, tps1Δ, or tps1Δ:TPS1 at 1*10⁷ cells (high) were removed and fixed inflated with 10% neutral buffered formalin on 1 and 3 days postinfection. They were then paraffin embedded and cut into 5-mm sections for H&E staining. The sections were visualized by light microscopy with microphotographs taken by the Digital Microphotography system DFX1200 with ACT-1 software (Nikon, Tokyo, Japan).

Lungs from uninfected mice we digested as per our established protocol (Osterholzer et al., 2009b). In brief, the lungs were removed, minced with scissors, and digested enzymatically at 37°C for 30 min in digestion buffer [RPMI, 5% fetal calf serum, antibiotics, and 1 mg/mL collagenase A (Boehringer Mannheim Biochemical, Chicago, IL), and 30 μg/mL DNase (Sigma-Aldrich, St Louis, MO)]. The digested lung was dispersed by repeated aspiration through the bore of a 10-mL syringe. The cells were then exposed to a red blood cell lysis buffer (0.829% NH₄Cl, 0.1% KHCO₃, 0.0372% Na₂EDTA, and pH 7.4) for 3 min and then quenched by the addition of 10-fold excess of RPMI. Cells were resuspended and filtered through a sterile 100-μm nylon filter. A 20% Percoll (Sigma) gradient enriched leukocytes from cell debris and epithelial cells. The digested lung cells were enumerated on a hemocytometer following dilution in Trypan blue (Sigma). Cells were diluted in culture buffer (RPMI, 5% fetal calf serum, antibiotics) to a concentration of 2 × 10⁶ cells/mL (10:1 MOI) or 2 × 10⁷ cells/mL (100:1 MOI) and plated at 100 µL in 96-well plates.

C. neoformans strain H99 (ATCC 208821) and mutant strain H99 tps1Δ strains (Petzold et al., 2006) were grown for 2 days in Sabouraud dextrose broth (Difco) at 30°C. Cells were enumerated on a hemocytometer following dilution in Trypan blue (Sigma) to a concentration of 2 × 10⁵ cells/mL. 100 µL was added to the wells containing dispersed lung cells. The wells were cocultured for 4 h at 37°C in 5% CO₂. Sterile water was used to lyse the lung cells to end the coculture, and then fungal survival for each well was quantified by a seven-step 10-fold serial dilution in sterile distilled water, followed by plating of 10 μL of homogenate at each dilution level (undiluted 1 × 10⁻⁷dilution) on SDA. CFU per well was determined by averaging two replicate platings.

WT (H99), tps1 KO (tps1Δ), or complemented (tps1Δ:TPS1) C. neoformans strains were incubated with murine serum for 4 h at 37°C in 5% CO₂. Fungal cells were enumerated at time 0 h, 2 h, and 4 h by a seven-step 10-fold serial dilution in sterile distilled water, followed by plating of 10 μL of each dilution on SDA.

H99, tps1Δ, and tps1Δ:TPS1 C. neoformans strains were grown for 2 days in Sabouraud dextrose broth (Difco) at 30°C. Cells were enumerated on a hemocytometer following dilution in Trypan blue (Sigma) to a concentration of 1 × 10⁷ cells/mL. 10-µL spots were plated on Christensen’s urea agar (Sigma 27048-500G-F) for visual determination of urease production (pink color indicator).

Values are reported as means with variance indicated. GraphPad Prism v8 software (GraphPad Software, San Diego, CA) was used to perform statistical analyses. Mantel–Cox tests were used for survival experiments ( Figures 1B , 2A , 5A ). Unpaired Student t test (two-tailed test) was used for comparisons of individual means ( Figures 1C , 2C , 3 , 5A, C , 6A ); Welch’s t-test ( Figure 1B ) and one-way ANOVA were used for multiple comparisons ( Figure 7C ). Data in all figures are the means ± SEM (unless otherwise indicated). Values that are statistically significant are indicated in the figures as follows: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

The requirement of cryptococcal tps1 gene expression for virulence of C. neoformans was demonstrated in several invertebrate and vertebrate models of infection (Petzold et al., 2006). However, little is known about the role of this pathway for fungal interactions with the host, including its ability to survive long-term in the mammalian host and its contribution to dissemination. To establish these parameters, we evaluated the role of tps1 in a group of murine models of mouse infection. Our first study tested whether tps1Δ would be sufficient to render C. neoformans virulent in the highly susceptible C57Bl/6 mice as compared with BALB/c mice, which show improved resistance to pulmonary infection with C. neoformans (Zaragoza et al., 2007; Chen et al., 2008). WT (H99) or tps1 KO (tps1Δ) C. neoformans (1 × 10⁵ cells/mL) were instilled intratracheally, as described earlier (Chen et al., 2008) to perform parallel inoculations in BALB/c and C57Bl/6 mice, and mouse survival was evaluated ( Figure 1A ). While more sensitive C57BL/6 succumbed to infection with WT C. neoformans (reached endpoint criteria between 14 dpi and 21 dpi) consistent with our previous report (Osterholzer et al., 2009b), all tps1Δ-infected mice survived until day 28 with no remaining fungus in the lung at this time. In this respect, deletion of cryptococcal tps1 had a similar effect in C57BL6 mice as in more resistant BALB/c, which survive up to 25 days with the H99 WT strain (Zhang et al., 2009; Eastman et al., 2015) and completely withstand tps1Δ ( Figure 1A ).

Considering that tps1Δ was equally required for virulence in both strains of mice, we analyzed the effect of cryptococcal tps1Δ on fungal clearance rate from the infected lungs. We performed infections with H99, tps1Δ, or tps1Δ:TPS1 (complement) strains in BALB/c mice ( Figure 1B ) and analyzed pulmonary fungal burdens at 1, 3, and 7 days postinfection (dpi). The tps1Δ C. neoformans significantly reduced fungal burdens in the lungs as soon as 24 h postinfection, continuing to decrease at day 3, and completely cleared the fungus from the lungs by 7 dpi. In contrast, the H99 and tps1Δ:TPS1 cryptococcus strain established a persistent infection in the lungs. The rapid clearance of the tps1Δ strain suggested that the rapid fungal clearance of tps1Δ C. neoformans likely occurred independently of the adaptive immune response. To test this, we performed a CD4-T-cell depletion with anti-CD4 antibody in BALB/c mice, as reported previously (Neal et al., 2017), and then infected with tps1Δ or H99 ( Figure 1C ). As expected, the T-cell depletion at this early time point of infection did not result in any changes in the expansion of the WT strain H99 but also did not affect the trajectory of tps1Δ lung burden. Thus, cryptococcal tps1 resulted in C. neoformans failure to develop persistent infection in the lungs and enabled the host to rapidly clear the lung infection without the help from CD4+ T cells.

Due to the rapid and complete elimination of the tps1Δ C. neoformans from the infected lungs in a pulmonary infection model, we could not evaluate the role of tps1Δ in other organs to which the fully virulent fungus disseminates. As such, we used our disseminated infection model achieved via an intravenous infection (Neal et al., 2017) that bypasses the pulmonary defenses. The WT, tps1Δ, or TPS1-complemented C. neoformans 10⁵ cells were injected IV into BALB/c mice, and mouse survival was evaluated ( Figure 2A ). All WT and complemented C. neoformans-infected mice reached endpoint criteria between 8 and 18 DPI, whereas all tps1Δ-infected mice survived for an observed 21 days. To determine the specific requirement of tps1Δ for C. neoformans to establish infection at specific target organs ( Figure 2B ), we quantified fungal burdens in the lungs (gray), brains (red), and spleens (tan) at 1 dpi, 2 dpi, 3 dpi, 7 dpi, 14 dpi, and 21dpi infected with tps1Δ C. neoformans. The tps1Δ C. neoformans strain is unable to reach more than a transient low level (<100 cells) of infection in the lungs and spleens. WT H99 C. neoformans establishes a progressive infection in these tissues (fungal burdens reported in Supplementary Figure 1 ) reaching ~10⁵ CFU/lungs and spleens by the time of death. Even at the peak of infection (time of death for WT and 14 dpi for tps1Δ), the tps1Δ strain still shows severe attenuation compared with the WT ( Figure 2C ). Interestingly, despite this severe attenuation, the tps1Δ C. neoformans is capable of entering the CNS compartment and establishing a low-level infection in the brain rather than be cleared ( Figure 2B ). The tps1Δ strain increased from low-level brain infection, from 10¹ CFU at 7 dpi to around 10⁴ CFU/brain at 14 dpi to a slight taper off on 21 dpi, a timing consistent with the development of the adaptive response in the brain in this model (Neal et al., 2017). However, our follow-up experiment excluded the role of CD4+ T cells, in this delayed control of tps1Δ strain. Mice treated with either anti-CD4 depleting antibody or received control IP injections (PBS) showed comparable fungal burdens on day 14 and day 21 postinfection in the lungs ( Figure 3A ), brains ( Figure 3B ), and spleens ( Figure 3C ). Thus, cryptococcal tps1Δ results in an improved control of the fungus in multiple organs, with somewhat less effective control in the brain, but in all cases independent of T-cell-mediated responses.

The very rapid suppression of the tps1Δ C. neoformans in the lungs infected with lower inoculum, typically used in most studies, would make the direct comparison of lung pathology very challenging as the tps1Δ strain would not rise to the sufficiently large fungal burdens to be analyzed by histology. Thus, we increased the inoculums to 10⁷ cells and performed follow-up IT-infection studies. Histological analysis of BALB/c lungs infected with the WT, tps1Δ, or complemented C. neoformans-infected mice revealed that all three strains were present in the alveolar space and induced a robust inflammatory response by 1DPI ( Figure 4 ). In the lungs infected with any of the three strains we observe abundant leukocytes that migrated to the alveolar airspace, especially in peribranchial areas of the lungs. However, by 3 DPI, the mice infected with the tps1Δ strain show few visible C. neoformans cells compared with the tps1-sufficient strains in the airspace and much lower level of inflammation. The WT and complemented-strain-infected lungs show a progressing infection and an ongoing inflammatory response, characteristic for cryptococcal pneumonia (Osterholzer et al., 2009b). Consistent with the histology data, we found a roughly one-log difference in fungal burden at day 1 between the tps1Δ strain and tps1-sufficient strains. This progressed to a 2-log difference in CFU on day 3 postinfection between the tps1-sufficient and tpsΔ strains in the infected lungs ( Figure 5A ). The follow-up survival studies showed rapid mortality of the WT-strain-infected mice, which all succumbed within 6 DPI, whereas the tps1Δ-infected mice showed no apparent symptoms during 15 DPI of observation and a 100% survival ( Figure 5B ). Our lung CFU analysis at the conclusion of the survival study showed 10⁸ CFU at the time of death in mice infected with the WT and a substantial clearance of tps1Δ from the infected lungs down to 10⁴ CFU/lung on day 14 ( Figure 5C ). This clearance kinetics was slower compared with the low-dose tps1Δ-infected mice, which cleared the infection by day 7, but ultimately, we saw the complete clearance of the mutant from the lungs by day 21 ( Figure 5C ). Together, these data show that while the resident lung defenses could be temporarily overwhelmed by the extremely high inoculum of tps1Δ C. neoformans, this strain remained avirulent, unable to cause severe lung pathology, and was ultimately cleared by the infected lungs.

The extremely rapid and effective suppression of the tps1Δ C. neoformans in the lungs but not in the brains led us to test whether there were lung-specific/resident defenses that restrained the trehalose-deficient strain. We analyzed the short-term effect on C. neoformans survival in a coculture with dispersed lung cellular components, which would exclude the effects of recruited immune components. CFUs of WT and tps1Δ cryptococci were evaluated at 4 h post-incubation compared with the fungi incubated in media alone to calculate percent inhibition ( Figure 6A ). Even after this short-term incubation, the tps1Δ yeasts were suppressed by lung cells (20%–25% inhibition) whereas WT cells were uninhibited. To account for possible effects of humoral defenses (such as complement or antibodies) we performed similar incubation of the WT or tps1Δ cryptococci in mouse serum ( Figure 6B ). No differences in fungal survival were detected during the 4 h of incubation, indicating that tps1Δ strain did not render the fungus susceptible to the humoral serum components present in the naive mouse serum. Thus, trehalose biosynthesis provides protection against resident lung defenses present in the lungs at the time of infection, but not against the serum humoral components.

While TPS1 enzyme is unlikely to directly interfere with host defenses, it has been proposed to be an upstream element required for synthesis of other known virulence factors, which in turn affect the host defenses. Some effects of tps1Δ for several cryptococcal virulence factors activities were studied (Petzold et al., 2006; Ngamskulrungroj et al., 2009), but the strong early lung-specific survival defect in tps1Δ prompted us to reevaluate factors that allow the fungus to be better adapted to pulmonary defenses. Urease production by Cryptococcus is one of fungal pulmonary virulence factors we studied previously (Osterholzer et al., 2009b), and thus we evaluated the urease production in WT, tps1Δ, or complemented Cryptococcus strains on Christensen’s urea agar ( Figure 7A ). All three C. neoformans strains, including tps1Δ, showed the robust and equivalent production of urease indicated by pink color conversion of the colonies ( Figure 7A ). We next explored the effect of tps1Δ on cryptococcal capsule formation, as this virulence factor is important for the yeast to efficiently withstand the innate host defenses (Casadevall et al., 2019). While tps1Δ C. neoformans strain was reported to produce capsule in vitro (Petzold et al., 2006; Ngamskulrungroj et al., 2009), the effect of tps1 on capsule size in culture and capsule formation in the lung environment in vivo has not been investigated. We cultured WT, tps1Δ, and complemented strains in vitro using capsule stimulating conditions, in addition to recovering fungal cells from the lungs from the infected animals at 1 DPI and 3 DPI, we performed India ink stains to visualize the capsule ( Figure 7B ). All three strains produced visible capsule compared with the acapsular cap60Δ control (Chang and Kwon-Chung, 1998). However, the capsule produced by the tps1Δ strain was visibly less robust than that produced by WT cells and this difference became particularly apparent in the lungs at 3 DPI. The capsule size measurements in vitro and ex vivo showed highly significant reduction of the capsule size in tps1Δ strain ( Figure 7C ). The capsule of the tps1Δ Cryptococcus was at least half the size of that of the WT. Thus, whereas tps1Δ Cryptococcus strain can generate a capsule both in capsule stimulating conditions in vitro and expands its size over time in the murine lungs, the capsule generated is not as robust, providing a potential explanation for the increased susceptibility of tps1Δ strain to the resident and innate mechanisms of host defenses, especially in the lungs.