Plants promote mating and dispersal of the human pathogenic fungus Cryptococcus

Infections due to Cryptococcus are a leading cause of fungal infections worldwide and are acquired as a result of environmental exposure to desiccated yeast or spores. The ability of Cryptococcus to grow, mate, and produce infectious propagules in association with plants is important for the maintenance of the genetic diversity and virulence factors important for infection of animals and humans. In the Western United States and Canada, Cryptococcus has been associated with conifers and tree species other than Eucalyptus; however, to date Cryptococcus has only been studied on live Arabidopsis thaliana, Eucalyptus sp., and Terminalia catappa (almond) seedlings. Previous research has demonstrated the ability of Cryptococcus to colonize live plants, leaves, and vasculature. We investigated the ability of Cryptococcus to grow on live seedlings of the angiosperms, A. thaliana, Eucalyptus camaldulensis, Colophospermum mopane, and the gymnosperms, Pseudotsuga menziesii (Douglas fir), and Tsuga heterophylla (Western hemlock). We observed a broad-range ability of Cryptococcus to colonize both traditional infection models as well as newly tested conifer species. Furthermore, C. neoformans, C. deneoformans, C. gattii (VGI), C. deuterogattii (VGII) and C. bacillisporus (VGIII) were able to colonize live plant leaves and needles but also undergo filamentation and mating on agar seeded with plant materials or in saprobic association with dead plant materials. The ability of Cryptococcus to grow and undergo filamentation and reproduction in saprobic association with both angiosperms and gymnosperms highlights an important role of plant debris in the sexual cycle and exposure to infectious propagules. This study highlights the broad importance of plants (and plant debris) as the ecological niche and reservoirs of infectious propagules of Cryptococcus in the environment.

Introduction exposure to these reservoirs can greatly influence the risk of cryptococcal disease in immunocompromised individuals.
The environment plays an essential role as the exposure reservoir and breeding ground for the propagation and dispersal of infectious propagules (desiccated yeast and spores). Historically, C. deneoformans/C. neoformans was associated with pigeon and bird guano. Mating was first observed and described as the result of incubation on malt extract [27], sporulation agar [27], minimal medium minus thiamine agar [28], V8 agar [28], and more recently described on pigeon guano [29], in association with live plants [30], and on plant debris agars [31]. Cryptococcus gattii was distinguished from C. neoformans in the 1970's and matings were observed and described on sporulation agar [32], V8 agar [33] and A. thaliana plants [30]. Laboratory matings of C. gattii (VGI, VGII, VGIII) are slower to progress and are less frequently observed than those involving C. neoformans isolates. Hybrid interspecies and intermolecular type matings and have been demonstrated in the laboratory but were associated with reduced viability [34,35]. Less frequently, productive hybrid-matings can result in hybrid vigor, enhanced virulence, and increased antifungal resistance in hosts [34][35][36]. Opposite (MATα + MATa) and same-sex (MATα + MATα) matings were first described for C. neoformans/C. deneoformans and later identified in C. gattii (VGI, VGII, VGIII) [27,37]. The accumulation of 40 years of research now illustrates that both C. neoformans, C. deneoformans, C. deuterogattii, C. bacillisporus, and C. gattii can complete their lifecycle in the environment. Whereas C. deneoformans is more frequently isolated from bird guano, C. gattii (VGI, VGII, VGIII, and VGIV) is more frequently isolate from trees, and C. neoformans is associated with both bird guano and trees. Filamentation, mating, and the production of spores in association with plants may enhance nutrient accumulation and dispersal in the environmental niche [38]. Therefore, the genetic drivers of pathogenicity and virulence are a direct result of interactions and selective forces Cryptococcus encounters in the environment. In Western Canada and the Western United States, Cryptococcus has been associated with many tree species, but to date Cryptococcus has only been studied on live A. thaliana, Eucalyptus sp., and Terminalia catappa (Almond) seedlings. However C. deuterogattii (VGII) is frequently associated with conifers in the Western USA and information pertaining to the growth of Cryptococcus in these naturally associated plant species is lacking [30,31,[39][40][41][42]. We therefore sought to extend the known interaction of Cryptococcus with plant host plants encountered within the United States, including live conifers and saprobic matter to further understand the role of plants in the infectious lifecycle and virulence of Cryptococcus.

Arabidopsis thaliana infection model
Cryptococcus cells were subcultured twice in yeast peptone dextrose (YPD) broth at 25˚C shaking at 180 rpm. Cells were collected by centrifugation, washed in ddH 2 O, and resuspended to OD 600 = 0.1 (~10 7 cells/mL) in autoclaved ddH 2 O. Four whole leaves of three-to four-week old soil-grown A. thaliana plants (Col-0 wildtype, jar1-1, or npr1-1) were infiltrated with a suspension of Cryptococcus cells. Plants were incubated in 16 hours light at 24˚C and 8 hours of dark at 20˚C, or 12 hours of light and 12 hours of darkness.
One leaf per plant was harvested, homogenized in ddH 2 O, and plated onto YPD and Niger seed agar to analyze colony forming units (CFUs). Plates were incubated at 30˚C for two to five days and CFUs were counted.
Individual and mating mixtures of C. deneoformans, C. neoformans or C. gattii, C. deuterogattii, or C. bacillisporus were inoculated on three different branches or leaves of the same individual plant and replicated on three individual plants. For Mopane, Eucalyptus, Douglas fir, and Eastern hemlock experiments multiple leaves/branches of the same plant were used for different treatments, starting from top, ddH 2 O control, 3 branches for Cryptococcus inoculation (Strain 1, Strain 2, Strain 1 + Strain 2), and remaining branches served as inoculated controls. 20 μL of a OD 600 = 0.1 Cryptococcus cell suspension (~2.0 x 10 5 cells) were drop inoculated onto the adaxial surface of two to three Mopane or Eucalyptus leaves. Three individual branch tips of Douglas fir or Western hemlock seedlings (composed of multiple needles and emerging buds) were dipped into the OD 600 = 0.1 cell suspension, excesses drips were blotted off with sterile Kimtech Science Ã KIMWIPES Ã , and allowed to air dry in a BSLIII hood. Additionally, one branch per tree was dip-inoculated with autoclaved ddH 2 O as experimental controls. Plants were incubated in an enclosed incubation chamber in 12 hours of light and 12 hours of dark that temperatures ranged from 30˚C to 25˚C respectively. Plant leaves or needles were harvested three weeks post-infection, homogenized in 1 mL autoclaved ddH 2 O and plated on YPD, YPD + NAT (100 mg/L), YPD + NEO (100 mg/L), and YPD + NAT (100 mg/L) + NEO (100 mg/L).

Growth and mating on plant materials and agar
Freshly grown Cryptococcus cells were harvested from YPD agar plates and suspended in 1 mL autoclaved ddH 2 O and OD 600 was measured. Cells were diluted to OD 600 = 1.0. Then, 10 μL of each strain were spotted individually or in combination with the opposite mating partner on agar or directly on autoclaved plant materials placed on top of 2% water agar (20 g Difco Bacto 1 agar/L) in petri-plates. Plant-based agars contained 30 g Difco Bacto 1 agar + 20 g ground plant materials (fresh A. thaliana plants, Black cherry (Prunus serotina) wood chips, Almond, Coco, Long leaf pine (Pinus palustris) needles, pine wood chips, Sugar maple (Acer saccharum) wood chips, Western red cedar (Thuja plicata), Douglas fir (Pseudotsuga menziesii), Western hemlock (Tsuga heterophylla) wood chips, or Mopane wood chips + 1 mL 20% glucose per L and pH was not adjusted. Plant materials were ground uniformly with a coffee bean grinder that was cleaned thoroughly with water and ethanol between each preparation. Black cherry and sugar maple wood chips (obtained from Dr. Paul Manion, Cazenovia, New York); A. thaliana plants (Dong lab, Duke University); Almonds (Whole, dry, and non-salted, Kroger brand); pine shavings (Petco, bedding); Douglas fir, Western red cedar, and Hemlock chips (obtained from Jamie, Murry, GEM Shavings, Auburn, Washington) were homogenized in a spice grinder. Coca agar was composed of 20 g Hershey's powder cocoa + 20 g Difco Bacto 1 agar and pH was not adjusted. V8 agar pH 5, V8 agar pH 7, MS agar, Niger seed agar, and Filament agar were made following standard methodology. V8 fusion agar was made substituting V8 Strawberry Banana fusion juice (50 mL) for standard V8 juice + 0.5 g KH 2 PO 4 + 30 g Difco Bacto 1 agar, and pH was adjusted to pH 5 or pH 7. Fusion agar consisted of 100 mL V8 Strawberry Banana fusion juice + 30 g Difco Bacto 1 agar and pH was not adjusted.
Tree needles (Hemlock, Long leaf pine, Eastern red cedar), leaves (Sugar maple), wood chips (Black cherry), Pine button plug (General Unfinished Flat Head no.315038 pine, Home Depot), or Oak button plugs (General Unfinished Flat Head no.313038 oak, Home Depot) were autoclaved, allowed to cool overnight, and aseptically placed in petri plates on top of 2% water agar before being inoculated with 15 μL of H99α, KN99a, or H99α + KN99a mating mixture; JEC21α, JEC20a, or JEC21α + JEC20a mating mixture; or NIH444α, NIH194a, or NIH444α + NIH194a mating mixture. Plates were incubated in a dark drawer at room temperature and observed weekly for the production of hyphae or basidia.

In vivo murine model
Six-week-old female A/JCr mice (Cat. No. 01A24, NCI-Frederick) or male BALB/c mice (Cat. No. 01B05, NCI-Frederick) were used. All animal studies were conducted in the Division of Laboratory Animal Resources (DLAR) facilities at Duke University Medical Center (DUMC) and animals were handled according to the guidelines defined by the United States Animal Welfare Act and in full compliance with the DUMC Institutional Animal Care Use Committee (IACUC). Animal models were reviewed and approved by DUMC IACUC under IACUC protocol # A217-11-08. Mice were acclimated in the facility for one week prior to infection and were housed in cages at 21˚C and 50% humidity with a 12 hrs. light/12 hrs. dark cycle and given ample food and water daily. Cryptococcus cells were grown on Arabidopsis agar (20 g/L A. thaliana plants + 20 g/L agar) at 30˚C for six days. Cells were harvested by scraping and washing from the agarose plates with autoclaved ddH 2 O and resuspended in 10 mL of autoclaved ddH 2 O. Cell were counted with a hemocytometer and resuspended to 2 x 10 6 cells/mL. All mice were sedated with Nembutal (sodium pentobarbital) prior to intranasal inoculation with 10 6 Cryptococcus cells in 40 μl. We expect the mice will become ill from inoculation with Cryptococcus which may exhibit as social isolation, lack of grooming, weight loss, loss of balance, inability to feed, loss of sternal recumbency, limb paralysis, seizures, convulsion, and coma. Animals were monitored daily by the primary author and Duke University DLAR staff for signs of disease development, distress, or suffering as described previously and were euthanized utilizing CO 2 when weight loss ! 15% of original body weight or they exhibited neurological symptoms or cranial swelling. Kaplan-Meier survival curves were constructed by GraphPad Prism version 6.03 (Windows, GraphPad Software, La Jolla California USA, www. graphpad.com).

Electron microscopy
To examine mating reactions for morphological features associated with mating, scanning electron microscopy studies were conducted. First, one centimeter square blocks of agar, whole needles, or sectioned leaf tissues were collected and fixed in 2% glutaraldehyde (Electron Microscopy Sciences, EMS, Hatfield, PA, USA) with 0.05% malachite green oxalate (EMS) in 0.1 M sodium cacodylate buffer and incubated at 4˚C until further processing. The fixation buffer was removed, and blocks were dehydrated by ethanol series, critical point dried (Pelco CPD2, Ted Pella, Inc., Redding, California, USA), sputter coated, and imaged with the FEI XL30 SEM-FEG (FEI Company, Hillsboro, Oregon, USA) at the electron microscopy facility at North Carolina State University.

Cryptococcus colonizes live Arabidopsis thaliana plants
Cryptococcus can colonize a broad variety of live plants. We demonstrate that C. neoformans (VNIV, Fig 1) C. deuterogattii (VGII), and C. gattii (VGI, S1 Fig) can colonize mature soil grown A. thaliana plants as previously described [30,39]. Mature soil grown A. thaliana plants were colonized more readily by H99α in comparison to KN99a (p < 0.0001). Survival and colonization of A. thaliana by Cryptococcus is strain-dependent (S1 Fig). Colonization of A. thaliana by C. neoformans is not dependent on laccase, capsule, or calcineurin (p > 0.100). However, we observed reduced colonization of C. deuterogattii (VGII) calcineurin (cna1Δ) mutant strains on A. thaliana plants (p < 0.001) S1 Fig). Chlorotic symptoms were only observed in leaves inoculated with mated pairs (MATa x MATα) and were not correlated with increased fungal colonization (Fig 1, KN99a versus H99α x KN99a, p > 0.100). A. thaliana mutants jar1-1 (p < 0.001) and npr1-1 (12 dpi and H99α x KN99a p < 0.001) were more permissible to C. neoformans colonization in comparison to Wild-type (Col-0) (Fig 2). Scanning electron microscopy indicates that C. neoformans colonizes live soil-grown wild type and Cryptococcus colonizes live Douglas fir, Western hemlock, Mopane, and Eucalyptus seedlings C. deneoformans, C. neoformans, C. deuterogattii, C. bacillisporus, and C. gattii can colonize Douglas fir and Western hemlock seedlings. Electron microscopy demonstrates that C. deneoformans (VNIV), C. neoformans (VNI), C. deuterogattii (VGII), and C. gattii (VGI) can colonize live Douglas fir and Western hemlock needles in mixed communities with other naturally acquired microorganisms (Fig 3). Douglas fir seedlings appeared healthy at one week post inoculation (Fig 4), but developed browning needles and progressive disease symptoms on young buds by three weeks (Fig 5). Disease symptoms progressively developed with prolonged incubation. Obvious disease symptoms were not observed on Western hemlock seedlings even after prolonged incubation (Fig 5). Douglas fir seedlings inoculated with C. neoformans, C. bacillisporus (VGIII), and C. gattii (VGI) displayed more severe needle browning and bud symptoms than those inoculated with C. deneoformans. Disease symptoms were associated with both mated mixtures and individually inoculated strains. Disease symptoms were not observed on Eucalyptus or Mopane seedlings inoculated with C. deneoformans, C. neoformans, C. bacillisporus, or C. gattii (S4 Fig). Cryptococcus recovery from infected tree seedlings appears to be plant-and strain-dependent, but C. deneoformans, C. neoformans, C. deuterogattii C. bacillisporus, and C. gattii were able to colonize tree seedlings (Fig 6). C. neoformans were more consistently recovered from inoculated Douglas fir and Western hemlock seedlings compared with C. deneoformans and C. deuterogattii C. bacillisporus, or C. gattii (Fig 6a). Cryptococcus cell recovery from Eucalyptus was inconsistent between strains (Fig 6b). Cryptococcus recovery was enhanced by utilizing Cryptococcus strains containing a drug-resistance marker (Fig 6b). No double drug resistant colonies were isolated from trees inoculated with mixed mating strains. Furthermore, we observed by light microscopy (S5 Fig) and scanning electron microscopy (Fig 7) that Cryptococcus can grow saprobically, filament (Oak, Hemlock, Maple, and Long leaf pine), and mate on dead plant materials including Sugar maple leaves and needles of Cedar and Long leaf pine (Fig 7).
Cryptococcus can mate on agar containing plant materials C. neoformans is highly fertile and prolifically mates on plant agars. Mating of C. deneoformans and C. neoformans was observed on Black cherry chip, Arabidopsis, Niger seed, Coca, Douglas fir, Sugar maple, Longleaf pine, and Almond agars, as well as on classic V8 and MS agar and newly concocted V8-fusion and Fusion agars (Fig 8 and S6 Fig). Strains of C. bacillisporus (VGIII) x C. gattii (VGI) had fewer observed matings and were slower to mate on many of the tested media (Fig 8 and S6 Fig). Mating of C. bacillisporus (VGIII) x C. gattii (VGI) was observed on Douglas fir, Cocoa, V8-fusion, Fusion, and classic V8 (pH 5) agar (Fig 8).   basidia on standard mating media in contrast to mating on newly described plant-based agars displayed in S8 Fig (colony morphology) and S10 Fig (filamentation and basidia).

Cryptococcus gattii can grow in extract broth
C. deuterogattii and C. bacillisporus, can grow and proliferate in A. thaliana and Pigeon guano extract broth. The OD of C. deuterogattii and C. bacillisporus cells followed normal growth dynamics over 72 hours, logarithmically increasing over 24 hours before reaching stationary phase (S11a Fig). Viable cells were obtained at the conclusion of the growth curve (S11b Fig). A. thaliana and pigeon guano extract broth have no additional nutrients, and growth was similar to YPD broth over 8 hours but then rapidly reached stationary phase around 12 hours; viable cells were recovered at the termination of the growth curve at 72 hours.
The impact on virulence due to growth on Arabidopsis agar is strain dependent C. deneoformans, C. neoformans and C. gattii, C. deuterogattii, C. bacillisporus, and C. tetragattii strains were grown on Arabidopsis agar for one week, harvested from plates, and tested for virulence in the intranasal murine model (Fig 9). Hypervirulence as a result of growth on plant materials was only observed for Cryptococcus neoformans isolate A1-22 (Fig 9). Hypovirulence as a result of growth on Arabidopsis agar was observed for C. bacillisporus VGIII isolate NIH312 (Fig 9). No other statistically significant differences (P > 0.05) were observed for growth on YPD agar or Arabidopsis agar were observed for NIH444, WM779, EJB18, WM276, H99, A7-35-33 or C45 (Fig 9).

Discussion
The human pathogen, Cryptococcus, can utilize plant hosts as reservoir for mating and dispersal. It is presently understood that C. deneoformans, C. neoformans, and VGI, VGII, and VGIII crosses can colonize and mate on A. thaliana in the laboratory setting, but little is  the filamentation, mating, and the generation of double drug resistant progeny by Cryptococcus was only observed under those circumstances [30]. However, we observed filamentation, production of clamp connections, and basidia by Cryptococcus in saprobic association with sterile plant matter and agar supplemented with plant matter indicating the sexual lifecycle can also be completed in saprobic association with plant matter. It is possible that the pathogenic Cryptococcus species (VN and VG inclusive) could form epiphytic and endophytic associations with plants because C. laurentii and C. albidus, C. flavus, C. podzolicus, C. hungaricus, and other Cryptococcus sp. have been reported to form non-pathogenic associations with a variety of host plants [43][44][45][46]. Consistent with previous reports we observed A. thaliana is readily colonized by C. deneoformans, C. neoformans, C. gattii, C. deuterogattii, and C. bacillisporus. Evidence has been previously reported that Cryptococcus mating triggers a jasmonic acid (JA)-dependent defense response in Arabidopsis and that Salicylic acid (SA) signaling pathway involving NPR1 is suppressed [30]. Here, we show that both JA and SA signaling pathways reduce the colonization of Cryptococcus in Arabidopsis as the JA and SA signaling mutants, jar1-1 and npr1-1, both show increased fungal colonization. This is consistent with previous reports that Arabidopsis SA signaling or synthesis mutants eds1 (enhanced disease susceptibility 1; lipase/signal transducer/triacylglycerol lipase), nahG (transgenic line degrading salicylic acid; SA), sid2 (SAinduction deficient), and npr1 (nonexpressor of PR genes 1; pathogenesis-related 1) are all more susceptible to Cryptococcus infection [39,41]. Other mutant A. thaliana ecotypes rpm1 (resistance to Pseudomonas syringae pv maculicola 1), pad4 (phytoalexin deficient 4), and Atprn1 (PRN1; one of four members of an iron-containing subgroup of the cupin superfamily, accumulates flavonoids) that are components of SA-mediated pathways, were also observed to be more susceptible to Cryptococcus infection [39,41]. The data suggest that both the SA and JA pathways are required to hinder the colonization and growth of Cryptococcus in Arabidopsis and are not necessarily antagonistic to each other consistent with resistance to the pathogenic fungus Fusarium graminearum [47].
Furthermore, in laboratory trials, Cryptococcus mutants (ste12α, cap59, lac1) show reduced ability to colonize A. thaliana plants and could be differentially regulated between species and molecular types as in mating and virulence [41,48]. Laccase was not essential for the colonization of A. thaliana by C. deuterogattii [42] but was important for virulence in C. neoformans, suggesting that different Cryptococcus species may have adapted different strategies to cope with the evolutionary pressures encountered in the plant ecological niche [41].
The interactions between Cryptococcus and plants are likely dependent on both the environmental conditions, the genetics of Cryptococcus, and the genetics of the host, A. thaliana. Conserved factors that are utilized for both plant colonization and virulence within animal host could be maintained in the natural ecological niche providing for the natural reservoirs of infectious genotypes. Furthermore, both yeast and mated co-cultures were observed to colonize live plants and plant tissues that were associated with the appearance of disease symptoms [30,39,42].
In this study, we demonstrate the ability of C. deneoformans, C. neoformans and C. gattii, C. bacillisporus, C. deuterogattii, and C. tetragattii to colonize live non-sterile Douglas fir and Hemlock trees and to undergo saprobic filamentation, mating, and the production of spores on dead plant material, implicating the potential for long-term association of Cryptococcus with plants. Therefore, plants may serve an important role in maintaining, optimizing, and enhancing virulence factors during the environmental lifecycle of the opportunistically acquired human pathogen, Cryptococcus. Supporting information S1 Fig. C. deneoformans, C. neoformans, C. gattii, C. deuterogattii, and C.   Robust filamentation of C. deneoformans is observed in association with black cherry chips and oak but limited filamentation is also observed in association with Sugar maple leaf, Hemlock needle, and long leaf pine needle. Robust filamentation of C. neoformans was observed in association with oak and limited filamentation with Sugar maple and Long leaf pine.