key: cord-0312408-7dkp5bt3
authors: Normile, Tyler G.; Del Poeta, Maurizio
title: Three models of vaccination strategies against cryptococcosis in immunocompromised hosts using heat-killed Cryptococcus neoformans Δsgl1
date: 2022-02-01
journal: bioRxiv
DOI: 10.1101/2022.01.31.478598
sha: 92c2b6c1449ee2a0d3c8f784ea53e9f0d8143a71
doc_id: 312408
cord_uid: 7dkp5bt3

Vaccines are one of the greatest medical accomplishments to date, yet no fungal vaccines are currently available in humans mainly because opportunistic mycoses generally occur during immunodeficiencies necessary for vaccine protection. In previous studies, a live, attenuated Cryptococcus neoformans Δsgl1 mutant accumulating sterylglucosides was found to be avirulent and protected mice from a subsequent lethal infection even in absence of CD4+ T cells, a condition most associated with cryptococcosis (e.g., HIV). Here, we tested three strategies of vaccination against cryptococcosis. First, in our preventative model, protection was achieved even after a 3-fold increase of the vaccination window. Second, because live C. neoformans Δsgl1-vaccinated mice challenged more than once with WT strain had a significant decrease in lung fungal burden, we tested C. neoformans Δsgl1 as an immunotherapeutic. We found that therapeutic administrations of HK C. neoformans Δsgl1 subsequent to WT challenge significantly improve the lung fungal burden. Similarly, therapeutic administration of HK C. neoformans Δsgl1 post WT challenge resulted in 100% or 70% survival depending on the time of vaccine administration, suggesting that HK Δsgl1 is a robust immunotherapeutic option. Third, we investigated a novel model of vaccination in preventing reactivation from lung granuloma using C. neoformans Δgcs1. Remarkably, we show that administration of HK Δsgl1 prevents mice from reactivating Δgcs1 upon inducing immunosuppression with corticosteroids or by depleting CD4+ T cells. Our results suggest that HK Δsgl1 represents a clinically relevant, efficacious vaccine that confers robust host protection in three models of vaccination against cryptococcosis even during CD4-deficiency. Importance Cryptococcosis results in ∼180,000 global deaths per year in immunocompromised individuals. Current antifungal treatment options are potentially toxic, lacking in areas of need, and exhibit limited efficacy. In addition to these lackluster therapeutic options, no fungal vaccines are currently available for clinical use. Due to the increasing rate of immunocompromised individuals, there is a dire need for the development of improved antifungal therapeutics. Presently, we have demonstrated the high efficacy of a clinically relevant heat-killed mutant strain of Cryptococcus neoformans in inducing advantageous host protection in three models of vaccination against cryptococcosis during immunodeficiencies most associated with this disease.

Abstract: Vaccines are one of the greatest medical accomplishments to date, yet no fungal 13 vaccines are currently available in humans mainly because opportunistic mycoses generally 14 occur during immunodeficiencies necessary for vaccine protection. In previous studies, a live, 15

attenuated Cryptococcus neoformans Δsgl1 mutant accumulating sterylglucosides was found to 16 be avirulent and protected mice from a subsequent lethal infection even in absence of CD4 + T 17 cells, a condition most associated with cryptococcosis (e.g., HIV). Here, we tested three 18 strategies of vaccination against cryptococcosis. First, in our preventative model, protection was 19 achieved even after a 3-fold increase of the vaccination window. Second, because live C. 20 neoformans Δsgl1-vaccinated mice challenged more than once with WT strain had a significant 21 decrease in lung fungal burden, we tested C. neoformans Δsgl1 as an immunotherapeutic. We 22

found that therapeutic administrations of HK C. neoformans Δsgl1 subsequent to WT challenge 23 significantly improve the lung fungal burden. Similarly, therapeutic administration of HK C. 24 neoformans Δsgl1 post WT challenge resulted in 100% or 70% survival depending on the time of 25 vaccine administration, suggesting that HK Δsgl1 is a robust immunotherapeutic option. Third, 26

we investigated a novel model of vaccination in preventing reactivation from lung granuloma 27 using C. neoformans Δgcs1. Remarkably, we show that administration of HK Δsgl1 prevents 28 mice from reactivating Δgcs1 upon inducing immunosuppression with corticosteroids or by 29 depleting CD4 + T cells. Our results suggest that HK Δsgl1 represents a clinically relevant, 30 efficacious vaccine that confers robust host protection in three models of vaccination against 31 cryptococcosis even during CD4-deficiency. 32 33 Importance: Cryptococcosis results in ~180,000 global deaths per year in immunocompromised 34

Invasive fungal infections are primarily caused by environmental fungi that mainly infect 48 immunocompromised individuals resulting in ~1.5 million deaths a year that account for ~50% 49 of all AIDS-related deaths [1, 2] . Individuals most at risk include HIV/AIDS patients [3] [4] [5] , 50 cancer patients receiving chemotherapy [6, 7] , solid organ transplant recipients [8] [9] [10] , or 51 patients taking medication to control chronic diseases [11] [12] [13] [14] . Unlike most fungi that do not 52 infect humans, the pathogenicity of invasive fungal species begins with the ability to grow and 53 replicate at human body temperature [15, 16] , which suggests that climate change, particularly 54 global warming, may play a role in increasing infections from environmental fungi in more 55 temperate climates [17] [18] [19] . The incidence of invasive fungal infections is expected to further 56 increase as the global immunocompromised population continues to rise due to novel 57 immunosuppressive therapies or comorbidities, such as the current COVID-19 pandemic [20-58 unvaccinated mice (Fig. 1A) . There was a significant increase in median survival time for mice 161 administered 5x10 6 HK C. neoformans Δsgl1, although all mice still succumbed to infection. 162

Remarkably, mice administered 5x10 7 HK C. neoformans Δsgl1 exhibited a 70% survival rate at 163 the endpoint of the experiment that was not statistically different from the complete protection 164 seen with live C. neoformans Δsgl1 (Fig. 1A) . 165

Surviving mice visually appeared healthy with normal weight gain, and endpoint organ 166 fungal burden analysis confirmed no extrapulmonary dissemination had occurred (Fig. 1B) . 167

Nevertheless, mice vaccinated with 5x10 7 HK C. neoformans Δsgl1 displayed a significantly 168 greater lung fungal burden compared to mice vaccinated with live C. neoformans Δsgl1 (Fig.  169   1B) . These data suggest that mice vaccinated with HK C. neoformans Δsgl1 exhibited 170 concentration-dependent partial protection with 5x10 7 HK C. neoformans Δsgl1 being the most 171 efficacious. 172 173 Two administrations of HK C. neoformans Δsgl1 confers complete host protection even during 174

We have unveiled that administration of a single dose of 5x10 7 HK C. neoformans Δsgl1 176 conferred similar host protection compared to vaccination with live C. neoformans Δsgl1, 177 although complete protection was not achieved (Fig. 1A) , and the endpoint lung fungal burden 178 was significantly greater than the live mutant vaccinated mice (Fig. 1B) protection (100% survival) at the experimental endpoint ( Fig. 2A) . Endpoint organ fungal burden 190 analysis showed that HK2d Δsgl1-vaccinated mice displayed no extrapulmonary dissemination 191 and a significantly lower lung fungal compared to live C. neoformans Δsgl1-vaccinated mice 192 (Fig. 2B) . In fact, 1 of the 7 HK2d Δsgl1-vaccinated mice fully cleared the WT yeast from the 193 lungs. These data suggest that vaccination with two subsequent doses of 5x10 7 HK C. 194 neoformans Δsgl1 confers complete host protection and aids in pulmonary clearance of the WT 195 fungal cells. 196 To assess if vaccination with HK C. neoformans Δsgl1 possessed clinical relevance, 197

CD4-deficient mice were also vaccinated with two subsequent doses of 5x10 7 HK C. neoformans 198

Δsgl1 and challenged mice with the WT strain. Interestingly, 100% host protection was achieved 199 in HK2d Δsgl1-vaccinated mice depleted of CD4 + T cells (Fig. 2C) , and endpoint organ fungal 200 burden analysis revealed no extrapulmonary dissemination of the WT strain (Fig. 2D) . There 201 was a significantly greater fungal burden in the lungs of HK2d Δsgl1-vaccinated mice depleted 202 of CD4 + T cells compared to HK2d Δsgl1-vaccinated immunocompetent mice. However, there 203 was a significantly lower lung fungal burden in HK2d Δsgl1-vaccinated mice depleted of CD4 + 204 T cells compared to live C. neoformans Δsgl1-vaccinated immunocompetent mice (Fig. 2D) . 205

Overall, these data indicate that vaccination with two subsequent administrations of 5x10 7 HK C. 206 neoformans Δsgl1 confers host protection from WT challenge in both immunocompetent and 207 CD4-deficient mice, and the HK vaccination strategy may provide a greater efficacy in host 208 clearance of the WT strain from the lungs compared to live vaccination strategy. 209 210 Vaccination with live or HK C. neoformans Δsgl1 confers long-lasting host immunity to lethal 211

Because administration of 2 subsequent doses of 5x10 7 HK C. neoformans Δsgl1 also 213 conferred complete host protection to the WT strain even during CD4-deficiency ( Fig. 2) , we 214 sought to investigate the efficacy of host protection after vaccination with either live or HK C. 215 neoformans Δsgl1 via alternations to our preventative vaccination model during 216 immunocompetency and CD4-deficiency. 217

To assess the longevity of the vaccine-induced memory T cells, we increased the time 218 between the administration of the vaccine and WT challenge. Immunocompetent or CD4-219 deficient mice were administered either live C. neoformans Δsgl1 or PBS and challenged with 220 the WT strain 90 days later (a 3-fold increase between vaccination and WT challenge). 221

Interestingly, all vaccinated mice survived the WT challenge, while all unvaccinated mice 222 succumbed to the WT infection (Fig. 3A) . Endpoint organ fungal burden in surviving mice 223 showed that no extrapulmonary dissemination was observed (Fig. 3B) . Similar to our previous 224 studies, the lung fungal burden in CD4-deficient vaccinated mice was significantly greater than 225 in immunocompetent vaccinated mice. These data show that vaccination with live C. neoformans 226 Δsgl1 confers long term host immunity to lethal WT challenge, which strongly suggests long-227 lived memory T cells even during CD4-deficiency. 228

Because vaccination with live C. neoformans Δsgl1 promoted long term immunity 229 resulting in complete host protection to the WT strain, we then asked if HK C. neoformans Δsgl1 230 provided the same protection. Immunocompetent or CD4-deficient mice were administered 2 231 subsequent doses of either HK C. neoformans Δsgl1 or PBS on days -90 and -75 and challenged 232 with the WT strain on day 0. We observed a 90% and 70% survival rate in immunocompetent 233 and CD4-deficient mice, respectively (Fig. 3C) . Nonetheless, the difference between the median 234 survival time for immunocompetent mice and CD4-deficient mice was not statistically different, 235 the endpoint lung fungal burdens were nearly identical, and no extrapulmonary dissemination of 236 the WT yeast was observed in either group (Fig. 3D) . Altogether, these data suggest that 237 vaccination with live or HK C. neoformans Δsgl1 provides long-lived memory T cells conferring 238 robust host protection and lung containment even after a 3-fold increase of the vaccination 239 window. To test for induced T cell anergy, immunocompetent or CD4-deficient C. neoformans 250 Δsgl1-vaccinated mice underwent multiple WT challenges, monitored for survival, and lung fungal burden was assessed at the end of each WT challenge period (experimental design 252 schematic: Fig. 4A ). Very interestingly, C. neoformans Δsgl1-vaccinated mice completely 253 survived for a total of 105 days after three subsequent lethal WT challenges on days 0, 45, and 254 75 (Fig. 4B) . Endpoint lung fungal burden analysis showed that there was a significant decrease 255 of persistent WT yeast in the lungs of mice that were WT challenged a second time (Fig. 4C) . 256

This decrease in lung burden did not further decrease after a third challenge. In addition, the 257 decrease in the persistent lung fungal burden from the subsequent WT challenge resulted in no 258 statistical difference between the lung burden in isotype-treated and CD4-deficient mice ( Fig.  259 4C). Overall, these data indicate that immunocompetent and CD4-deficient C. neoformans 260

Δsgl1-vaccinated mice are protected from at least three subsequent lethal WT challenges 261 resulting in fewer persisting WT cells in the lungs strongly suggesting that vaccination produces 262 memory T cells that do not undergo anergy due to chronic infection. nearly identical at all timepoints, and histopathology at these timepoints displayed a decreased 272 percentage of inflamed lung tissue and increased formation isolated nodules of contained yeast 273 cells [52] . In this study, we found for the first time that vaccination with two subsequent doses of 274 HK C. neoformans Δsgl1 results in a significant decrease of lung fungal burden compared to live 275 C. neoformans Δsgl1-vaccination for both immunocompetent and CD4-deficient mice (Fig. 2D) . 276

Moreover, vaccinated mice that received more than one WT challenge displayed a significant 277 reduction in the lung fungal burden compared to mice that were received only one WT challenge 278 ( Fig. 4C) . Because subsequent WT challenges decreased the lung fungal burden, we asked if 279 administration of our vaccine could be used as a therapeutic strategy and administered after the 280 WT challenge. 281

To investigate the therapeutic potential of HK C. neoformans Δsgl1 administration in C. 282 neoformans Δsgl1-vaccinated mice, immunocompetent and CD4-deficient mice were challenged 283 with the WT strain first and then received either 1 or 2 subsequent administrations of HK C. 284 neoformans Δsgl1 (see experimental design schematic: Fig. 5A ). We found a significant decrease 285 in the lung fungal burden after therapeutic administration of HK C. neoformans Δsgl1 in both 286 immunocompetent and CD4-deficient mice (Fig. 5B) . From the baseline lung fungal burden on 287 day 30 post challenge, there was a significantly greater lung burden in CD4-deficient mice 288 compared to the isotype-treated as we have seen previously. In addition, the lung fungal burden 289 in mice that were treated with either 1 or 2 administrations of PBS (control groups) on days 30 290 and 45, respectively, was nearly identical to the baseline lung fungal burden (Fig. 5B) . 291

Interestingly, there was a significant decrease in the lung burden in mice that received 1 or 2 292

administrations of HK C. neoformans Δsgl1 post WT challenge compared to the PBS-treated 293 groups at those timepoints as well as in the lung burdens between mice that received 1 or 2 294

administrations of HK C. neoformans Δsgl1 (Fig. 5B) . Together, these data suggest that We tested this hypothesis by challenging naive mice with the WT strain and then 305 administered HK C. neoformans Δsgl1, live C. neoformans Δsgl1, or PBS on either day 3 or day 306 7 and monitored for survival (experimental design schematic: Fig. 6A ). While all mice 307 administered PBS fatally succumbed to infection, all mice administered HK C. neoformans 308

Δsgl1 or live C. neoformans Δsgl1 on day 3 post WT challenge survived to the experimental 309 endpoint (Fig. 6B ). In addition, mice administered HK C. neoformans Δsgl1 or live C. 310 neoformans Δsgl1 on day 7 post WT challenge exhibited a 70% and 60% survival rate at the 311 experimental endpoint, respectively. Nevertheless, there were no differences in the lung fungal 312 burden between any of the surviving groups (Fig. 6B) . Of note, all surviving mice displayed 313 extrapulmonary dissemination of the WT strain to the brain. Interestingly, mice that received 314 therapeutic administration of live or HK C. neoformans Δsgl1 on day 3 had fewer brain CFU 315 compared to mice administered on day 7 (Fig. 6C) . Overall, these data suggest that live or HK C. 316 neoformans Δsgl1 aids to significantly prolong the survival of mice from fatal WT infection. We have now shown that live or HK C. neoformans Δsgl1 can be effectively used both 321 preventatively (Figs. 2-4) and therapeutically (Figs. 5-6) to elicit robust host protection. 322

However, C. neoformans is not only a primary pathogen since fungal cells can be contained 323 within lung granulomas in immunocompromised hosts for extensive periods of time but 324 immunosuppressive conditions, such as CD4-lymphopenic HIV/AIDS patients, can cause 325 granuloma breakdown, latent cell proliferation, and brain dissemination potentially resulting in 326 fatal meningoencephalitis [11, 31] . Because of this, we investigated the ability of C. neoformans 327

Δsgl1 to protect mice from cryptococcal reactivation from a lung granuloma. 328

To test this, mice were intranasally inoculated with C. neoformans Δgcs1, a mutant strain 329 lacking glucosylceramide synthase, that has been previously reported to induce pulmonary 330 granuloma formation in mice over 30 days. At 30 days post Δgcs1 administration, we 331 administered live C. neoformans Δsgl1, HK C. neoformans Δsgl1, or PBS. After another 30 days, 332 all groups of mice underwent either corticosteroid-induced immunosuppression to induce 333 leukopenia or CD4 + T cell depletion to induce CD4 lymphopenia, and mice were monitored for 334 survival (simplified experimental design schematic: Fig. 7A ; detailed experimental design 335 schematic: Supplementary Fig. S1 ). Extraordinarily, we observed that mice administered live 336 C. neoformans Δsgl1 or HK C. neoformans Δsgl1 exhibited a 75% and 62.5% survival rate, 337 respectively, at the experimental endpoint post corticosteroid-induced immunosuppression, while 338 all PBS-treated mice fully succumbed to fatal reactivation (Fig. 7B) . Similarly, mice 339 administered live C. neoformans Δsgl1 or HK C. neoformans Δsgl1 exhibited a 100% and 87.5% 340 survival rate, respectively, at the experimental endpoint post CD4 + T cell depletion, which were 341 significantly greater than the PBS-treated mice that displayed a 37.5% survival rate (Fig. 7C) . 342

These data suggest that vaccination with live or HK C. neoformans Δsgl1 can be used to protect the host from cryptococcal reactivation from a lung granuloma in the event that they become 344

To examine the efficacy of our vaccine strategy in the experimental reactivation model, 346

the endpoint lung fungal burden in mice pre-immunosuppression (day 0) was compared to the 347 fungal burden in the lungs of mice that survived until the experimental endpoint post-348 immunosuppression (day 30). We first observed there were no differences in the lung fungal 349 burdens between any of the groups pre-immunosuppression (Fig. 7D) . Interestingly, there were 350 no statistical differences between the endpoint lung fungal burdens post-immunosuppression in 351 C. neoformans Δsgl1-vaccinated mice or between the endpoint lung fungal burdens in mice pre-352 immunosuppression compared to in mice post-immunosuppression. In fact, the only observed 353 statistically significant difference was between the surviving PBS-treated CD4-deficient mice 354 and live C. neoformans Δsgl1-vaccinated mice, which further supports that vaccination with C. 355 neoformans Δsgl1 protects mice from lethal reactivation upon immunosuppression (Fig. 7D) . 356

Comparably, all surviving PBS-treated mice displayed significantly greater brain dissemination 357 compared to vaccinated mice, which were almost fully absent of any brain fungal burden 358 (Supplementary Fig. S2) . Overall, these data suggest that administration of live or HK C. vaccine-inducing mutants, including this current work, report 100% protection to the lethal WT 380 challenge. However, the true standout characteristic for a clinically relevant vaccine formulation 381 is the ability to induce protection in a model most associated with a disease, which is CD4-382 deficiency for cryptococcosis [36, 40] . 383

In comparison to our present findings where we report 100% protection in CD4-384 deficiency with HK C. neoformans Δsgl1 vaccination, the only other HK vaccine-inducing 385 mutant to demonstrate protection during CD4-deficiency was from Wang and colleagues using a 386 HK F-box protein (Δfbp1) mutant strain [55] . Similarly, both HK C. neoformans Δsgl1 and the 387 HK Δfbp1 mutants demonstrated complete protection in both immunocompetent and CD4-388 deficient CBA/J mice, although differences between the two are noteworthy. First, our present 389 work with HK C. neoformans Δsgl1 resulted in a ~1 log lower endpoint lung fungal burden for 390 isotype-treated mice compared to isotype-treated mice in the report by Wang and colleagues 391 (although the endpoint lung fungal burden for CD4-deficient mice was nearly identical) [55] 392 (Fig. 2) . Second, we observed a complete lack of any extrapulmonary dissemination in mice 393 vaccinated with HK C. neoformans Δsgl1, while several mice displayed fungal CFU in the brain 394 and spleen in the study by Wang and colleagues [55] . Finally, the WT challenge dose used in our 395 work was 15x greater than used by Wang and colleagues. 396 With regards to our vaccine, we aimed to test the rigor and robustness of C. neoformans 397

Δsgl1 in the preventative model of vaccination via functional alterations to our experimental 398 design. Since T cell mediated immunity is a well-established keystone of anti-cryptococcal 399 immunity [60, 61] as well as the need for either CD4 + or CD8 + T cells for C. neoformans Δsgl1 400

fold increase in the time between vaccination and WT challenge, where vaccination began 90 402 days prior to WT challenge for both live and HK C. neoformans Δsgl1. All immunocompetent 403 and CD4-deficient mice vaccinated with live C. neoformans Δsgl1 survived the lethal WT 404 challenge, and a respective 90% and 70% survival was observed in mice vaccinated with HK C. 405 neoformans Δsgl1 (Fig. 3A, C) . The protection observed in the extended rest period suggests the 406 induction of long-lived memory T cells post vaccination with C. neoformans Δsgl1. Future 407 immunophenotyping assays will be aimed to define the type of circulating memory T cells, such 408 as central memory, tissue-resident memory, or effector memory. 409

Complete host protection was not observed in 100% of the CD4 + deficient mice when 410 they were vaccinated with HK C. neoformans Δsgl1 90 days prior to WT challenge. This 411 suggests that the immunological memory induced was either less robust or shorter-lived 412 compared to vaccination with the live cell strain. It is noteworthy to mention that the WT 413 challenge dose was doubled in this experimental design due to the increased age of the mice at 414 the time of challenge. However, the decreased length of antigen encounter using HK mutant 415 strains may have potentially resulted in less robust naïve T cell stimulation and fewer memory T 416 cells following the contraction phase [54, 62] . Optimization of the dosing regimen will be 417 required in future studies. Potential adjustments could include increasing the number of doses, 418

increasing the time between the first and second dose, or altering the concentrations to induce 419 more robust immunity with a lower first dose and a greater second dose. 420

The second functional alteration to the preventative vaccination model experimental 421 design was to increase the number of WT challenges administered to vaccinated mice. During 422 chronic infections, such as when fungal cells are persisting in the lungs, T cells may become 423 tolerized to antigens remaining alive for extended periods of time in a hyporesponsive state 424 known as T cell anergy [57, 63] . Mice vaccinated with C. neoformans Δsgl1 exhibited the 425 opposite, however. First, all mice that received two or three subsequent WT challenges exhibited 426 100% survival even during CD4 + T cell deficiency (Fig. 4B) . Second, the endpoint lung fungal 427 burden in mice that received at least 2 WT challenges displayed a ~2 log decrease compared to 428 mice that received only 1 WT challenge (Fig. 4C) mice has been reported in other cryptococcal vaccine studies as well [55, 56, 58, 59, 64, 65] . 438

Because we observed a decrease in the lung fungal burden after a second WT challenge (Fig.  439   4C) , we investigated the immunotherapeutic ability of HK C. neoformans Δsgl1 to decrease 440 further WT fungal cells remaining in the lungs. The first administration of HK C. neoformans 441

Δsgl1 significantly decreased the persistent fungal burden to a similar degree as mice that 442 received a second WT challenge ( Fig. 5B and Fig. 4C) . Interestingly, mice that received a 443 second administration of HK C. neoformans Δsgl1 significantly decreased the remaining fungal 444 cells to an even further extent compared to mice that were administered PBS or mice that 445 received only 1 therapeutic dose of HK C. neoformans Δsgl1 (Fig. 5B) . In fact, 2 of the 3 mice 446 fully cleared the WT fungal cells from the lungs. Thus, HK C. neoformans Δsgl1 exhibits robust 447 immunotherapeutic potential in previously vaccinated immunocompetent mice. 448

Collectively, the therapeutic potential of HK C. neoformans Δsgl1 administration has 449 demonstrated highly efficacious host protection in both previously vaccinated (Fig. 5) and 450 unvaccinated mice (Fig. 6) . While this adds an entirely new dimension to our vaccine, (Fig. 7B, C) . To our knowledge, this is the first time a vaccine against the reactivation 468 infection has been reported in the literature. Previous work in our lab had shown that mice 469 treated with FTY720, a prescribed treatment for relapsing remitting multiple sclerosis, was 470 linked to granuloma breakdown with a disorganization of the peripheral macrophages with a 471 shift towards an M2 polarized state [11] . In addition, our findings also validate the reactivation 472 model, as it showed that the C. neoformans Δgcs1-induced granuloma in mice can lose integrity 473 upon immunosuppression resulting in fungal proliferation in the lung, brain dissemination, and 474 ultimately death. 475

In fact, clinical cases can occur due to the reactivation of granuloma-contained fungal 476 cells from either immunosuppression or comorbidities (HIV/AIDS progression) [27, 69] . 477

Because of this, we tested our vaccination strategy in this mouse model during prolonged 478 corticosteroid-induced immunosuppression as well as CD4-deficiency. We observed a 70% and 479 60% survival rate in mice vaccinated with live or HK C. neoformans Δsgl1, respectively, at the endpoint after corticosteroid-induced immunosuppression with cortisone acetate (Fig. 7B) , and a 481 100% and 90% survival rate in mice vaccinated with live or HK C. neoformans Δsgl1 at the 482 endpoint after depletion of CD4 + T cells (Fig. 7C) . Interestingly, the corticosteroid-induced 483 immunosuppression was more lethal than the depletion of CD4 + T cells, which may be attributed 484 to the mechanism of immunosuppression. Corticosteroid-induced immunosuppression induces 485 leukopenia, inhibits phagocytosis, and decreases antigen presentation capabilities [70, 71] , while 486 depletion of CD4 + T cells ablates circulating CD4 + lymphocytes. So, we speculate that the 487 difference in lethality of the infection may be the speed at which the immunosuppression took 488

Although there was an observed difference in survival between the two modes of 490 immunosuppression, the endpoint lung fungal burden between the two modes of 491 immunosuppression were nearly identical (Fig. 7D) . In fact, there were also no differences 492 between the endpoint lung fungal burden of mice pre-immunosuppression and C. neoformans 493 Δsgl1-vaccinated mice post-immunosuppression. This suggests that vaccination with either live 494 or HK C. neoformans Δsgl1 controls the proliferation of the latent fungal cells in the lungs even 495 after the immunosuppressive regime. This is further supported from the endpoint lung fungal 496 burden in unvaccinated CD4-deficient mice being significantly greater than the lung fungal 497 burden in the vaccinated mice, which indicate that fungal cells extensively proliferate in 498 unvaccinated mice upon immunosuppression. The same was observed for extrapulmonary 499 dissemination to the brain (Supplementary Fig. S2) . While there were only 1-2 C. neoformans 500

Δsgl1-vaccinated mice that displayed fungal dissemination, all the surviving unvaccinated mice 501 had significant fungal burden in the brain. Overall, vaccination with either live or HK C. 502 neoformans Δsgl1 demonstrated remarkable efficacy in this cryptococcal model of reactivation.

In conclusion, we have shown here that HK C. neoformans Δsgl1 demonstrates a highly 504 efficacious vaccine candidate that goes beyond the canonical preventative model of primary 505 disease prevention. We have expanded not only to a more clinically relevant HK formulation but 506 also to additional models of vaccine strategies to protect against cryptococcosis during CD4-507 deficiency, including using our vaccine as a therapeutic mean and using our vaccine to prevent 508 reactivation of a latent infection upon immunodepression. Here forth, the tools for investigation 509 into the protective immunity against fungal reactivation from pulmonary granulomas in mice are 510 now available, which greatly opens future possibilities to significantly add to this completely 511 absent portion of the literature. 512

The fungal strains used in this study were wild-type (WT) C. neoformans var. grubii 516 strain H99, C. neoformans Δsgl1, a mutant strain accumulating sterylglucosides developed by 517 our group [45] , and C. neoformans Δgcs1, a mutant strain lacking glucosylceramide synthase 518

[72]. For all experiments, fungal strains were freshly recovered from a -80ºC freezer stock on 519 YPD agar at 30ºC for 3-4 days. An isolated colony was added to 10ml of YPD broth and grown 520 for 16-18hr with shaking, washed three times in sterile PBS, counted with a hemocytometer, and 521 resuspended in sterile PBS at the desired concentration. For HK strains, the desired concentration 522 of live yeast was resuspended in PBS and added to an 80ºC heat block for 1hr. All HK strains 523 were confirmed to be fully dead by plating the mixture on YPD plates at 30ºC for 4 days and 524 observing no growth. 525

Both male and female CBA/J mice were purchased from Envigo. All animals were 528 housed under specific pathogen free conditions and had access to food and water ad libitum. 529

Mice were allowed one week to acclimate upon arrival before any procedures began. All animal 530 procedures were approved by the Stony Brook University Institutional Animal Care and Use 531

Committee (protocol no. 341588) and followed the guidelines of American Veterinary Medical 532

Association. 533 534

All primary infections and immunizations were carried out in both male and female 536 CBA/J mice 4-6 weeks old. Mice were first intraperitoneally (IP) anesthetized with a 537 ketamine/xylazine solution (95mg of ketamine and 5mg of xylazine per kg of animal body 538 weight). Anesthetized mice were then intranasally (IN) injected with the desired concentration of 539 the specified yeast cells in 20μl of PBS. For fungal burden analysis, mice were euthanized via 540 CO 2 inhalation on pre-determined timepoints. The lungs, brain, spleen, kidneys, and liver were 541 removed, homogenized in 10ml of sterile PBS using a Stomacher 80 blender (Seward, UK), and 542 serial dilutions were grown on YPD plates at 30ºC for 3-4 days before being counted and total 543 organ burden calculated. 544 545

Cortisone 21-acetate (CA) (Sigma; cat # C3130) was used to induce leukopenia. Mice 547 were sub-cutaneously administered 250mg/kg/mouse CA in PBS every other day for a set 548 timeline. IP administration of anti-CD4 monoclonal antibody (clone: GK1.5; BioXCell) was used to deplete mice of CD4 + T cells. Antibody dilutions were prepared from the stock solution 550 in PBS each time. Mice were administered 400μg/100μl every 4 days for the duration of the 551 experiment to maintain cell depletion. Control group mice were administered isotype-matched 552 antibody at the same concentration and administration timeline. 553 554

Three different vaccination models were used in this study. For survival studies, any 556 animal that appeared to be moribund, exhibited labored breathing or neurological infection, or 557 had lost more than 20% body weight was euthanized via CO 2 . 

All analyses were performed using GraphPad Prism 9 software. The sample size, 587 statistical analysis, and statistical significance is described in each figure caption. The Mantel-588

Cox log-rank test was used to calculate significance for survival studies. A two-tailed unpaired t 589 test was used to calculate statistical significance between two samples, and either an ordinary 590 one-way ANOVA using Tukey's multiple comparisons test for P value adjustment or a two-way 591 ANOVA using Šídák's multiple comparisons test for P value adjustment was used to calculate 592 statistical significance between more than two samples. Live Δsgl1, 5x10 5 , 5x10 6 , or 5x10 7 HK Δsgl1, or PBS. After 30 days, mice were challenged with 772

Invasive endemic fungi of the Western Hemisphere

Invasive fungal disease in humans: are we aware of the real impact? 611 Memórias do Instituto Oswaldo Cruz

A neglected epidemic: fungal 614 infections in HIV/AIDS

Colliding Epidemics and the Rise of Cryptococcosis

Developing a Rational Approach to Patient Therapy

Hidden killers: human fungal infections

Cryptococcosis in solid organ transplant 622 recipients

Serologic evidence for reactivation of cryptococcosis in solid-organ 624 transplant recipients. Clin Vaccine Immunol

A titanic drug resistance threat in Cryptococcus neoformans

FTY720 reactivates cryptococcal granulomas in mice through S1P 628 receptor 3 on macrophages

Immunologic mechanisms of 630 fingolimod and the role of immunosenescence in the risk of cryptococcal infection: A case report 631 and review of literature

Cryptococcal meningitis after fingolimod 633 discontinuation in a patient with multiple sclerosis

Cryptococcal meningitis reported with Fingolimod treatment: case 635 series

Immunity to Invasive Fungal Diseases

Cryptococcal Meningitis and Anti-virulence

Global warming impact on the expansion of fundamental niche of 641

Implications of climate change (global 643 warming) for the healthcare system

Threatening and Debilitating Fungal Diseases in Humans

Pulmonary aspergillosis and cryptococcosis as a 647 complication of COVID-19

High mortality co-infections of COVID-19 patients: mucormycosis and 649 other fungal infections. Discoveries (Craiova)

Overview on the Prevalence of Fungal Infections

Microbiome Role in COVID-19 Patients

Opportunistic Fungal Infection Associated With COVID-19

Root Causes of Fungal Coinfections in COVID-19 Infected Patients

The hidden pathogenic potential of 659 environmental fungi

Mechanisms of Pulmonary Escape and Dissemination 665 by Cryptococcus neoformans

Granuloma and cryptococcosis

The granuloma in cryptococcal disease

The Formation and Function of Granulomas

Reactivation of dormant/latent fungal infection

Animal Models of Cryptococcus 696 neoformans in Identifying Immune Parameters Associated With Primary Infection and 697 Reactivation of Latent Infection

Role of Sterylglucosidase 1 (Sgl1) on the pathogenicity of Cryptococcus 699 neoformans: potential applications for vaccine development. Front Microbiol

Beta-sitosterol and beta-sitosterolglucoside stimulate human peripheral 703 blood lymphocyte proliferation: Implications for their use as an immunomodulatory vitamin 704 combination

A randomised placebo-controlled trial of the efficacy of beta-sitosterol 706 and its glucoside as adjuvants in the treatment of pulmonary tuberculosis. The International 707 Journal of Tuberculosis and Lung Disease

Ginsenoside Rg1 helps mice resist to disseminated candidiasis by 709

Th1 type differentiation of CD4+ T cell

Immunoregulatory activity by daucosterol, a beta-sitosterol glycoside, 711 induces protective Th1 immune response against disseminated Candidiasis in mice. Vaccine

Structure and inhibition of Cryptococcus neoformans 714 sterylglucosidase to develop antifungal agents

Cryptococcus neoformans Δsgl1 Vaccination

Requires Either CD4+ or CD8+ T Cells for Complete Host Protection. Frontiers in Cellular and 717 Infection Microbiology

Sterylglucoside Are Required for Host Protection in an Animal Vaccination Model. mBio

Development, trafficking, and function of memory T-cell subsets

A Heat-Killed Cryptococcus Mutant Strain Induces Host Protection 724 against Multiple Invasive Mycoses in a Murine Vaccine Model. mBio

Development of protective inflammation and cell-mediated immunity 726 against Cryptococcus neoformans after exposure to hyphal mutants. mBio

Cryptococcus neoformans Cda1 and Cda2 coordinate deacetylation 730 of chitin during infection to control fungal virulence

Induction of Protective Immunity to Cryptococcal Infection in Mice 732 by a Heat-Killed, Chitosan-Deficient Strain of Cryptococcus neoformans. mBio

Cryptococcus antigens and immune responses: 734 implications for a vaccine

Immunity to Cryptococcus neoformans and C. gattii 736 during cryptococcosis

Resident memory T cells show that it is never too late to 738 change your ways

Tolerance and exhaustion: defining mechanisms of T 740 cell dysfunction

Vaccine-mediated immune responses to experimental pulmonary 742

Cryptococcus gattii infection in mice

Vaccination with Glucan Particles Containing Cryptococcus Alkaline Extracts. mBio

A cryptococcal capsular polysaccharide 747 mimotope prolongs the survival of mice with Cryptococcus neoformans infection

Therapeutic efficacy of a conjugate vaccine 750 containing a peptide mimotope of cryptococcal capsular polysaccharide glucuronoxylomannan

Modulation of the Pulmonary Type 2 T-Cell Response to 753 Cryptococcus neoformans

by Intratracheal Delivery of a Tumor Necrosis Factor Alpha-Expressing Adenoviral 755

Infection and Immunity

Dynamic Immune Cell Recruitment After Murine Pulmonary 759

Aspergillus fumigatus Infection under Different Immunosuppressive Regimens. Front Microbiol

Therapeutic and toxicologic studies in 762 a murine model of invasive pulmonary aspergillosis

Glucosylceramide synthase is an essential regulator of 764 pathogenicity of Cryptococcus neoformans

WT) (day 0) and monitored for survival. B. Endpoint organ 773 fungal burden was quantified in the lungs, brain, spleen, kidney, and liver in surviving mice

Significance was determined using a two-tailed unpaired t-test (B) and 776 significance is denoted as ****, P < 0.001

Cox log-rank test (A) and denoted on graph A: ns, not significant (P > 0.05) for Live

WT vs. 5x10 7 HK Δsgl1  WT

001 for 5x10 7 HK Δsgl1  WT vs. 5x10 6 HK Δsgl1

001 for 5x10 6 HK Δsgl1  WT vs. 5x10 5 HK Δsgl1  WT