|Home | About | Journals | Submit | Contact Us | Français|
Cryptococcus neoformans and Cryptococcus gattii are closely related pathogenic fungi that cause pneumonia and meningitis in both immunocompromised and immunocompetent hosts and are a significant global infectious disease risk. Both species are found in the environment and are acquired via inhalation, leading to an initial pulmonary infection. The infectious propagule is unknown but is hypothesized to be small desiccated yeast cells or spores produced by sexual reproduction (opposite- or same-sex mating). Here we characterize the morphology, germination properties, and virulence of spores. A comparative morphological analysis of hyphae and spores produced by opposite-sex mating, same-sex mating, and self-fertile diploid strains was conducted by scanning electron microscopy, yielding insight into hyphal/basidial morphology and spore size, structure, and surface properties. Spores isolated by microdissection were found to readily germinate even on water agarose medium. Thus, nutritional signals do not appear to be required to stimulate spore germination, and as-yet-unknown environmental factors may normally constrain germination in nature. As few as 500 CFU of a spore-enriched infectious inoculum (~95% spores) of serotype A C. neoformans var. grubii were fully virulent (100% lethal infection) in both a murine inhalation virulence model and the invertebrate model host Galleria mellonella. In contrast to a previous report on C. neoformans var. neoformans, spores of C. neoformans var. grubii were not more infectious than yeast cells. Molecular analysis of isolates recovered from tissues of infected mice (lung, spleen, and brain) provides evidence for infection and dissemination by recombinant spore products. These studies provide a detailed morphological and physiological analysis of the spore and document that spores can serve as infectious propagules.
Humans are exposed to infectious agents via inhalation and cutaneous exposure and from the microbiota. Transmission can involve direct human-human transmission, intermediate animal or insect vectors, or exposure to environmental sources. Animals are exposed to pathogenic fungi via direct/fomite transmission of dermatophytes that infect skin and/or nails, animal-animal transmission by the inhalation of the obligate pathogen Pneumocystis, and bloodstream penetration by Candida species from the gastrointestinal microbiota (46).
Many pathogenic fungi (dimorphic fungal pathogens, molds such as Aspergillus fumigatus, and the basidiomycete Cryptococcus) are environmental, and exposure occurs via the inhalation of conidia and/or spores, hyphal fragments, or yeast cells. Particles of >5 μm are subject to efficient mucociliary airway clearance, and smaller infectious propagules more readily deposit deep in the lungs and alveoli (18). As such, spores and conidia represent known or suspected infectious propagules for many pathogenic fungi. Moreover, as spores are often stress resistant, abundant, and readily aerially dispersed, animals may encounter spores more often than other infectious forms. Studies of Schizophyllum commune revealed abundant spores present in air above the ocean at distances of up to a mile from shore (24).
Cryptococcus neoformans is a common, opportunistic human fungal pathogen that causes meningoencephalitis in immunocompromised individuals and, if untreated, is uniformly fatal (6, 40). Cryptococcosis is caused by three varieties or sibling species that diverged 10 to 40 million years ago and that exhibit different environmental distributions and virulence properties (32, 43). C. neoformans var. grubii (serotype A) is the major cause (95%) of infections worldwide and >99% of infections in AIDS patients (6). C. neoformans var. neoformans (serotype D) strains account for <5% of infections worldwide, but serotype AD hybrids cause up to 20% of clinical infections in Europe (2, 9, 31). The serotype A and D lineages are globally distributed, associated with pigeon guano, and typically infect immunocompromised hosts.
The closely related species Cryptococcus gattii (serotypes B and C) is more geographically restricted to tropical and/or subtropical regions, associated with trees, and commonly infects immunocompetent hosts and less frequently AIDS patients and other immunocompromised hosts (6, 29). A C. gattii outbreak has been occurring on Vancouver Island since 1999 and recently expanded to the Canadian mainland and the United States (5, 12, 25, 38, 52).
The infectious propagules for many human fungal pathogens are thought to be spores. Humans are exposed to Cryptococcus by inhalation, leading to an initial pulmonary infection that can be asymptomatic or limited or can disseminate. For many individuals, the initial pulmonary infection is cleared; in others, the organism establishes a dormant latent granulomatous form in the hilar lymph nodes (10, 17). Individuals can die harboring granulomas without overt disease (1, 19, 48). In response to immunosuppression, either primary infection or recrudescence of latent infection results in dissemination via the bloodstream. The organism can infect many organs and tissues but exhibits a predilection to infect the central nervous system (45). The suspected infectious propagules for Cryptococcus are spores or small, desiccated, less encapsulated yeast cells that are an ideal size for alveolar deposition. C. neoformans var. neoformans serotype D spores have been reported to be up to 100 times more infectious than yeast cells in an immunocompromised murine inhalation model (50). Studies to purify spores and analyze their pathogenicity in animal models using different infection methods were previously reported, documenting that spores can serve as infectious propagules under certain conditions, including when directly inoculated intracerebrally, intravenously, or by inhalation (7, 50, 59).
Cryptococcus neoformans and C. gattii have defined sexual cycles that produce abundant basidiospores, which could represent infectious propagules (13, 23, 27, 28, 39, 42). However, if spores are the infectious propagule, and spores are produced by mating between a and α cells, how and where would this occur in the largely α unisexual population? A potential resolution was the discovery that α isolates can produce hyphae, basidia, and basidiospores via monokaryotic fruiting (11, 54), a modified sexual cycle that requires only one of the two mating types (4, 12, 21, 33-36, 49). This might be a route by which infectious propagules are produced in nature; however, serotype A C. neoformans var. grubii and C. gattii strains do not reproducibly undergo monokaryotic fruiting under laboratory conditions. Population genetic studies provided evidence that they do so in nature (4, 21, 36, 49), but this remains to be established under defined laboratory conditions.
As most previous studies of spore infectivity focused on serotype D, we focused our investigations on serotype A spores and virulence properties. Our analysis examined spore size, shape, and surface properties and conditions supporting spore germination. Importantly, we demonstrate that infectious inocula highly enriched for spores serve as efficient infectious propagules in both a murine inhalation model and an invertebrate host. Analyses of isolates recovered from infected animals provide evidence that recombinant spore progeny are infectious. Spores were not more virulent than yeast cells in either virulence assay. Taken together, our studies provide electron microscopic views of the spores and demonstrate that both spores and yeast cells can serve as infectious propagules.
Serotype A strains H99 (αA) (44), KN99a (aA) (42), Bt63 (aA) (37), diploid KN2B5#19 (aAAα) (42), serotype D strains JEC20 (aD) and JEC21 (αD) (20, 30), XL280 (αD) (34), and C. gattii strains WM276α (αB, VGI), R265α (αB, VGII), B4546a (aC, VGIII), NIH 312α (αC, VGIII), and MMRL2651α (αC, VGIV) (12) were used in mating assays and scanning electron microscopy (SEM) analyses. For matings, equal amounts of a and α cells were mixed together with a flat-end toothpick on MS medium (57) and incubated at room temperature in the dark for 2 to 4 weeks until spore chains were visible under the microscope at the culture edges.
Mating cultures were excised from the agar plate around the colony edge and fixed in 3% glutaraldehyde in 0.1 M Na cacodylate buffer (pH 6.8) for several days at 4°C. Samples were then rinsed in three 1-h changes of cold 0.1 M Na cacodylate buffer (pH 6.8) followed by a graded dehydration series of 1-h changes in cold 30% and 50% ethyl alcohol (EtOH) and held overnight in 70% EtOH. Dehydration was completed with 1-h changes of cold 95% and 100% EtOH at 4°C and warming to room temperature in 100% EtOH. Two additional 1-h changes of room temperature 100% EtOH completed the dehydration series. The samples were then critical-point dried in liquid CO2 (Samdri-795; Tousimis Research Corp., Rockville, MD) for 15 min at the critical point. The agar pieces were mounted onto stubs with double-stick tape, pressed down completely around the edge, and then sealed with silver paint to ensure good conductivity. The samples were sputter coated with 50 Å of Au/Pd (Hummer 6.2; Anatech U.S.A., Hayward, CA). Samples were held in the vacuum desiccator until viewed with a scanning electron microscope (Jeol JSM 5900LV) at 15 kV. Spore sizes were calculated by measuring at least 50 individual spores from each cross for all the samples. Statistical significance was determined by calculating the P value by using the t test provided by QuickCalcs (GrapPad Software).
To isolate basidiospores, strains from opposite mating types, H99α and KN99a (serotype A) or JEC20α and JEC21a (serotype D), were mixed in equal proportions on MS medium plates with flat-end toothpicks. Matings were conducted at room temperature in the dark for 2 to 4 weeks until robust sporulation was observed by microscopy. Basidiospores on the edges of a mating colony were removed with a glass Pasteur pipette and transferred onto a yeast extract-peptone-dextrose (YPD) plate, and individual basidiospores were transferred into fresh YPD medium by microdissection, as described previously (22). Isolated basidiospores were incubated at 30°C for 2 to 3 days to allow the spores to germinate, and the resulting yeast colonies were subcultured in fresh YPD medium. To determine the mating type, isolates were cocultured with reference tester strains JEC20a and JEC21α grown on V8 (pH 7) medium in the dark at room temperature (30). An isolate and tester strains were cultured alone on the same plate as controls. The mating reaction plates were examined for the formation of mating hyphae, which signaled the initiation of sexual reproduction. Genomic DNA extractions followed a procedure reported previously (51). Mating and mitochondrial DNA inheritance were determined by PCR analysis using specific primers for STE20a (5′-ATCAGAGACAGAGGAGCAAGAC and 5′-CTAACTCTACTACACCTCACGG) and STE20α (5′-AGCTGATGCTGTGGATTGAATAC and 5′-TGCAATCACAGCACCTTACATAG) (35), mitochondrial DNA inheritance was determined by using Da3/Da20 (5′-GCAATAGCATATACCATCCCG and 5′-GACACTACACAAGATGCCTC), and the PCR conditions were described previously (35, 51).
Spores were dissected by micromanipulation from mating cultures as described previously (22). To determine the signals that trigger spore germination, spores were dissected onto different media with limiting nutrients, including carbon and nitrogen sources. Plates were observed by microscopy at regular 6- to 8-h intervals for microcolonies to determine the germination efficiency. Ten to twenty individual spores from the cross between H99α and KN99a were dissected onto 2% water agar medium with no added nutrient, carbon, or nitrogen sources. Plates with dissected spores were incubated at 30°C and observed under the microscope at regular time intervals. In order to completely eliminate nutrients, 2% agarose water medium was prepared, and the germination efficiency of spores was observed. To determine the conditions favoring spore germination, agarose medium plates were incubated in an inverted position at 30°C, and control plates with dissected spores were incubated in the standard upright position. In another set of experiments, spores containing agarose plates were Parafilm wrapped to limit CO2 escape and to determine the effect of CO2 accumulation on germination, and control plates were incubated without Parafilm. To determine the effects of carbon and nitrogen sources on spore germination, spores were dissected onto 2% agarose medium supplemented with 2% glucose as well as 10 mM cyclic AMP separately in individual plates. Also, to analyze an effect between fungal and bacterial cultures, which might influence fungal development by releasing quorum-sensing molecules, we used bacterial cultures of Escherichia coli strain OP50 or Salmonella enterica strain SL1344 grown in filamentous broth (54) overnight. Agarose plates were prepared by mixing 2% agarose in culture supernatant, and spores were dissected onto these plates directly by micromanipulation. Plates were incubated at 30°C, and microcolonies were observed at the place of spore deposition. We further compared the germination rates between the serotype A and serotype D spores by dissecting the spores onto 2% water agarose medium from the mating crosses between H99α and KN99a (serotype A) and JEC21α and JEC20a (serotype D), incubating cultures at 30°C until the colonies were formed, and observing them by microscopy at regular intervals.
Basidiospores were isolated carefully from the edges of mating colonies between H99α and KN99a on MS medium with a glass Pasteur pipette tip without touching the surrounding yeast colonies and suspended into Eppendorf tubes containing phosphate-buffered saline (PBS). Yeast cells (H99α and KN99a) were grown overnight on MS medium for the same time as mating cultures, harvested, and washed with PBS. The inoculum size was confirmed by counting cells in a hemocytometer and by CFU determinations.
Animals were infected as previously described (8). Groups of 4- to 6-week-old female A/JCr mice were anesthetized by intraperitoneal injection of phenobarbital (~0.035 mg/g of body weight). Five or ten animals for each sample of spores and H99α and H99α/KN99a coinfections (1:1 ratio, from both medium plates and cultures grown overnight) were infected by carefully instilling an inoculum of 500 CFU each in 50 μl of PBS by directly dropping into the nares using a Pipetteman to imitate an inhalation mode of infection (42). After the inoculation of fungal cells, animals were monitored twice daily, and the animals showing signs of severe morbidity (weight loss, extension of the cerebral portion of the cranium, abnormal gait, paralysis, seizures, convulsions, or coma) were sacrificed by CO2 inhalation. Survival was plotted against time, and P values to determine the significance of survival curves were calculated by plotting Kaplan-Meier survival curves and performing the log-rank test using Prism software (version 4.0a; Prism Computational Sciences, Inc., Madison, WI).
To determine the number of recombinants inhaled by the animals and their dissemination to cause infection, the lungs, spleen, and brain from two animals from each group were harvested, weighed, and homogenized in 5 ml of sterile PBS. Serial dilutions of the organ samples were plated onto Sabouraud dextrose agar (1% peptone, 2% dextrose, 1% agar) plates containing 100 μg/ml chloramphenicol and incubated at 37°C overnight. Colonies were tested for mating types (a and α) both by PCR using STE20a- and STE20α-specific primers (51) and by mating the isolates with tester strains (JEC20a and JEC21α) on V8 (pH 7) medium incubated in dark at room temperature. Mitochondrial DNA inheritance was determined by PCR with primer pair Da3/Da20 (51).
Greater wax moth virulence assays were performed according to a protocol reported previously (41). Greater wax moth larvae were purchased from Vanderhorst, Inc. (St. Marys, OH). Larvae from a single shipment were used for each experiment. An inoculum of 500 yeast cells (1:1, a and α) or a spore-enriched preparation (500 CFU; ~93% spores) were injected into the larvae in the posterior pseudopod, and the larvae were incubated at 37°C and at room temperature in the dark postinoculation. Larvae were examined daily, and those not responding to touch were scored as being inviable.
J774A.1 cells were removed from T75 flasks by scraping, centrifuged at 1,000 × g for 8 min, and resuspended in complete medium. Live cells were counted with a hemocytometer using trypan blue exclusion to identify dead cells. Cells were then adjusted to a concentration of 1 × 105 cells/ml, and 300 μl of this suspension was plated into eight-well coverslip chambers at a concentration of 1 × 105 cells/ml, allowed to adhere, and activated with 300 μl fresh complete medium containing 100 ng/ml lipopolysaccharide (serotype O111:B4) and 200 U/ml mouse gamma interferon (R&D Systems) overnight. C. neoformans cells of strain H99α or spores (H99α and KN99a mating) were washed three times and resuspended in sterile PBS at a concentration of 2 × 106 yeast cells/ml. Yeast cells were then incubated with 2 μg/ml of 18B7 anticapsular antibody or alone for 1 h at 37°C with shaking in an Eppendorf (Westbury, NY) 5436 thermomixer. For macrophage infection, 10 μl of pretreated yeast cells (20,000 yeast cells) was inoculated into wells containing activated macrophages at time zero. The number of yeast cells or spores was chosen to represent a multiplicity of infection of 1:1, with the assumption that the number of macrophages doubled overnight (53). Data represent data from experiments performed at least three times with similar results.
Time lapse microscopy was performed using a Zeiss Axiovert time lapse station (Carl Zeiss Microimaging, Inc., Thornwood, NY) equipped with a Hamamatsu (Bridgewater, NJ) Orca AG camera. An exposure of 240 min was used along with an LD Acroplan 40× objective for optimal bright-field cell imaging. Images were obtained from five independent fields per well at 1-min time intervals for a total of 120 min. Movies were compiled using Metamorph software.
A detailed analysis of hyphal morphology and spores produced during C. neoformans and C. gattii sexual reproduction was conducted by SEM (Fig. (Fig.1;1; see Fig. S1A and B in the supplemental material). Hyphae, basidia, and spores produced by serotype A × A, D × D, B × C, and C × C strain crosses were analyzed and compared. These matings involve an initial cell-cell fusion, and the resulting zygote then forms a filamentous dikaryon that produces hyphae decorated with fused clamp connections, basidial fruiting structures, and long chains of basidiospores emanating from the basidial surface.
Hyphae terminated in basidia that were globular at the apex and cylindrical at the base. Four independent basidiospore chains were seen arising from the basidial surfaces, which were otherwise smooth (Fig. (Fig.1).1). Sterigmata on the basidial surface were not observed for any of the crosses examined. The spores have rough surfaces, are elongated, and form four long, intact chains unless disrupted. Serotype A and D spores were 1 μm wide but differed modestly in length. Spores from serotype A strains (H99α × KN99a [VNI × VNI] or H99α and Bt63a [VNI × VNB]) were rough and 2 to 3 μm long (Fig. (Fig.11 and and2).2). Spores produced by a self-fertile aAAα diploid strain (KN2B5#19) were smoother, slightly curved at the tip, and shorter (2 to 2.5 μm long). Spores from crosses of congenic serotype D strains (JEC21α × JEC20a) were 2 to 2.5 μm long (Fig. (Fig.11 and and2).2). Approximately 50 individual spores from each cross were measured to determine the spore size for all the samples. Statistical significance was determined by calculating P values where the two-tailed P value equals 0.0012. Both serotype A and D spores produced by mating had a roughened, prominent surface, with numerous spike-like surface protrusions, which may represent novel cell surface proteins or carbohydrates (cell wall/capsule). The spores produced by monokaryotic fruiting of serotype D strain XL280α were roughened but modestly longer than those produced by serotype D a-α matings, ranging from 2 to 3 μm in length and 1 to 1.25 μm in width. As shown in Fig. Fig.22 and Fig. S1A in the supplemental material, the spore shape was asymmetric, rounded on the basidial distal end and flattened on the basidial proximal end. Spores in chains were always arranged in a head-to-tail fashion. These differences in the spore ends likely reflect the process of budding and the action of specific hydrolytic enzymes that lead to spore-spore, and spore-basidial, separation. In summary, spores produced by mating and monokaryotic fruiting of serotype A and D strains are morphologically similar, with a rough surface, and have similar albeit not identical sizes and shapes.
Interestingly, no capsular staining was observed in our studies using anti-GXM antibody in spores from serotype A mating cultures (see Fig. S3 in the supplemental material), whereas control yeast cells showed affinity toward anti-GXM antibody, suggesting the presence of polysaccharide surrounding the yeast cell surface (see Fig. S3 in the supplemental material).
Interestingly, spores from Cryptococcus gattii matings were considerably more elongated than spores obtained from C. neoformans var. grubii (serotype A) and C. neoformans var. neoformans (serotype D) (Fig. (Fig.11 and and3).3). Spores from C. gattii matings were dramatically elongated, in some cases slightly curved, with a smooth surface (Fig. (Fig.3).3). In C. gattii mating cultures, four basidiospores arose from the basidia and elongated during their development. Long spore chains were rarely seen in this species, possibly because the elongated spores are more readily discharged from the chains. No sterigmata on the basidia were observed in these crosses, similar to the findings for C. neoformans. Different C. gattii molecular types were compared to spores produced by serotype A and D strains (see Fig. S1B in the supplemental material). In all cases, C. gattii spores were much longer, ranging from 3 to 5 μm in length and 1 μm in width. Control yeast cells differed in size and structure compared to spores produced by both mating and fruiting. Yeast cells in all of the backgrounds measured ~5 μm or larger and were round and encapsulated. Taken together, our findings are in accord with data from previous and recent studies (3, 47) and reveal additional insights into spore morphology.
We next addressed signals that trigger spore germination. Typically, microdissected spores are germinated on rich YPD medium for genotypic and phenotypic characterization. To determine the signals influencing spore germination, different media were tested with limited nutrients, such as reduced carbon and nitrogen sources. Germination efficiency was analyzed using spores isolated from H99α and KN99a matings by micromanipulation. On distilled water-2% agar medium plates with no added nutrients, spores were isolated by microdissection and incubated at 30°C. Spores were observed by microscopy at 6- to 8-h intervals, and the production of 2 to 3 cell microcolonies was scored as being indicative of spore germination and yeast cell division. Under these conditions, spore germination was modestly delayed (<48 h) compared to that when YPD medium was used (<24 h), but there was no difference in the frequencies of spore germination (~70 to 80%), and within 1 week of incubation, microcolonies that were visible to the naked eye were produced (see Fig. S2 in the supplemental material). To test whether spore germination might result from trace contaminating nutrients in the agar, similar studies were conducted using distilled water medium prepared with 2% agarose. Spores microdissected onto 2% agarose medium with no added nutrients or carbon and nitrogen sources still germinated (within 48 to 72 h) to form microcolonies with a similar efficiency. The addition of 10 mM cyclic AMP to this medium had no effect. Thus, it appears that there is little to no obligatory nutritional signal required to trigger spore germination under these experimental conditions.
We next tested if conditions in nature or the laboratory normally constrain spore germination. We tested spores on agarose medium plates incubated under different conditions, including Parafilm sealing plates to restrict airflow and increase CO2 levels, incubation upright versus inverted to alter humidification and desiccation, and using conditioned medium from bacterial cultures. None of the conditions increased or decreased spore germination levels. We conclude that spores are self-sufficient for nutrients required for germination and germinate under a variety of environmental conditions. On the other hand, intact spore chains produced from mating remain dormant on mating medium for months. Thus, other signals, such as contact sensing, may inhibit germination in spore chains or stimulate germination when an aerial spore lands on a solid surface in nature.
In considering if spores serve as infectious propagules, a key question is whether spores become airborne. Previous studies provided evidence that particles small enough to be spores are present in air, but whether these are spores, small desiccated yeast cells, or both is unknown (14, 25). Serotype A mating mixtures and control yeast cells on the surface of solid MS medium were exposed to different conditions to test if spores are released from mating reactions. No CFU were observed when mating reaction mixtures or control yeast cells were gently agitated overnight or vigorously vortexed for 2 min above fresh YPD medium, indicating that spores are not readily released.
We next tested if exposing plates to an airstream mimicking wind would trigger spore release. Mating and yeast cultures on MS medium were exposed to a standard laboratory airstream, and the resulting cells were trapped onto a closely attached fresh YPD agar plate. Both mating and yeast culture plates exposed to an airstream led to the release and capture of CFU. Seven hundred forty-two CFU were captured from the mating plate, while only 24 CFU were captured from the control yeast plate. Thirty-two CFU derived from the mating plate were randomly selected and analyzed for genomic and mitochondrial markers. Among the 32 isolates analyzed, cells that were MATα (12 isolates) all exhibited MATa-type mitochondria. These results suggest that although both spores and yeast cells are readily aerosolized by wind, spores can be discharged at a higher efficiency (~20-fold). These findings provide evidence as to how spores might become airborne infectious propagules in nature.
For further studies, we investigated a variety of different approaches to separate spores and yeast cells of serotype A. We focused on serotype A spores, as these are the cause of >95% of cryptococcal infections globally. We first tested filtration approaches based on size. Mating mixtures with spores, hyphae, and yeast cells were passed through a filter with a 3-μm exclusion size. We hypothesized that the larger yeast cells would be retained on the filter because of their larger size (>5 μm) and that spores would pass through the filter, yielding a spore-enriched preparation. Spores were detected as recombinant products of mating using serotype A mating parents with dominant drug resistance markers such that spores could be distinguished from parental yeast cell genotypes. Using this approach, we observed a high proportion of contamination attributable to yeast cells of the parental genotypes (data not shown), and we attribute this to the presence in mating cultures of a high proportion of smaller yeast cells of a size similar to that of spores, which may result from desiccation-promoted mating.
Second, in attempt to separate serotype A spores and yeast cells based on size and buoyancy, step and/or discontinuous gradients were used under a variety of conditions. Commercially available gradients (Percoll and Renografin) and laboratory-prepared sucrose gradients were used at concentrations from 40 to 90% for gradients. None yielded sufficiently enriched spores to eliminate contamination by yeast cells (not shown).
Third, the selective killing of yeast cells at different temperatures for various incubation periods was tested. The treatment of a mixture of spores and yeast cells at a higher temperature (i.e., 55°C for 10 min) resulted in ~80% spore preparations at best, but this level of enrichment varied (data not shown).
As a final attempt to obtain purified spores, we used a novel approach. A standard micromanipulation procedure, similar to that used to microdissect individual spores from mating cultures, was adapted to obtain sufficient quantities of spores to analyze virulence. Spore chains from mating culture edges were carefully dissected using the tip of a capillary glass tube and then resuspended in water or PBS for further analysis. By microscopic inspection, these samples contained a large proportion of spores (data not shown).
The proportion of spores in these preparations was estimated by germinating CFU, isolating genomic DNA from isolates, and subjecting this to PCR analyses for mitochondrial DNA inheritance (Da3/Da20), mating type (STE20a and STE20α), and genotyping using sequence-specific primers. Parental strains H99α and KN99a are congenic but differ at the mating-type locus, harbor unique mitochondrial genomes (51), and differ in one other genomic region. KN99a (and KN99α) harbor a recombinant mitochondrial genome that was generated during the backcrosses used to derive this congenic strain pair from strains H99α and 125.91a as their progenitors (51). Previous studies documented that the mitochondria are uniparentally inherited from the a parent following mating (55, 58).
By scoring the segregation of the mitochondrial genome and mating type, we can ascertain that the α progeny that inherit the a parental mitochondrial genome are products of mating, and we can infer that an equivalent number of a progeny are similarly derived. In other words, H99α crossed with KN99a yields F1 progeny with four theoretically possible genotypes: two parental (α mito 1 and a mito 2) and two recombinant (α mito 2 and a mito 1). However, because of uniparental mitochondrial inheritance, the vast majority of progeny inherit mitochondrial genotype 2, and there are, in practice, two types of progeny: α mito 2 and a mito 2. The frequency of recombinant F1 progeny can therefore be calculated as twice the number of the α mito 2 recombinants. This method of analysis might underestimate the number of spores analyzed given that mitochondrial recombination and rare mitochondrial transmission from the α parent also occur. We also analyzed the segregation of one genomic region that perdured in the KN99a/α genome from the original parental strain 8-1, which differs from H99α. This genomic region segregated in a Mendelian fashion (1:1) and independently of the MAT locus based on restriction fragment length polymorphism (RFLP) analysis, providing direct genomic evidence that these represent products of meiosis. The purity of the spore preparations and/or inocula were calculated to be at least ~93% spores-7% yeast cells by molecular and mating-type analyses, and these spore preparations were used for further analyses (see Table S1 in the supplemental material).
Data from previous studies indicated that spores from serotype D are up to 100 times more infectious than yeast cells in a murine inhalation model (50). Here, we addressed the pathogenicity of serotype A spores. Spores produced by mating of serotype A strains H99α and KN99a were isolated by micromanipulation, and 500 CFU (~93% spores) were resuspended in saline as the infectious inoculum. The spore-enriched inoculum was compared with three different types of yeast cell inocula: (i) H99α alone, (ii) a 1:1 mixture of H99α and KN99a yeast cells grown in liquid culture, and (iii) a 1:1 mixture of H99α and KN99a yeast cells (grown individually on V8 mating medium and mixed just prior to infection) to control for desiccation. Two experiments were conducted, in which groups of 5 (experiment 1) or 10 (experiment 2) mice per sample were infected via intranasal inoculation and monitored daily. Because spores germinate within 48 to 72 h, even on water agarose medium, spore preparations were prepared in PBS and used for animal infections within 4 h of preparation.
The spore-enriched inoculum used for these infections was analyzed for mitochondrial DNA inheritance and mating types by PCR using gene-specific primers and by mating on V8 (pH 7) plates with tester strains JEC20a and JEC21α (Fig. (Fig.4A;4A; see Table S1 in the supplemental material). Based on this analysis, we calculate that the spore preparation used for infections contained ~93% spores and ~7% contaminating parental yeast cells. For the infectious inoculum of 500 CFU, which was validated by counting spores and/or cells with a hemocytometer and plating assays, this corresponds to 465 spores and 35 yeast cells.
All infectious inocula, including the spore-enriched preparation and the yeast cell controls, were found to be highly infectious in the murine inhalation model. In both independent experiments, all inocula resulted in 100% lethal infection by 34 days postinfection (Fig. (Fig.4B4B and C). While the spore-enriched preparation was highly infectious, progression to lethal infection was modestly delayed compared to that of yeast cells in both independent experiments, with 100% mortality by day 25 for H99 yeast cells and day 30 for the spore-enriched inoculum in experiment 1 and by day 23 for H99 yeast cells and day 34 for the spore inoculum in experiment 2. This difference in the virulence of yeast cells and spores was statistically significant (P < 0.0027 for spores versus H99 yeast cells in experiment 1 and P < 0.0018 for experiment 2). Notably, there was no statistically significant difference in the relative virulences among the three different yeast inocula (P < 0.2799 for H99 versus H99/KN99 liquid culture and P < 0.1606 for H99 versus H99/KN99 plate culture for experiment 1; P < 0.9237 for H99 versus H99/KN99 liquid culture and P < 0.0816 for H99 versus H99/KN99 plate culture for experiment 2). These findings support the conclusion that both yeast cells and a highly spore-enriched preparation are infectious in animals but that the virulence of spores is modestly delayed compared to that of yeast cells.
Given that the spore-enriched infectious inoculum contained a majority of spores but also a small fraction of contaminating yeast, further studies were conducted to document that animals had been infected with spores. Lung, spleen, and brain tissues were recovered from sacrificed animals that had been infected with the spore-enriched preparation. CFU were recovered and analyzed for mating type, mitochondrial markers, and the genomic marker by PCR using gene-specific primers, for mating on plates with tester strains (JEC21α and JEC20a), and for RFLP analysis (Fig. (Fig.5;5; see Tables S2 to S5 in the supplemental material). As shown in Fig. 5A to C, of the isolates recovered from lung, spleen, and brain, the vast majority had inherited the mitochondrial genome from the a mating parent, whereas roughly equal proportions of a and α isolates were recovered. As discussed above, the recovery of α isolates with the a mitochondrial genome provides evidence that these isolates derived from spores produced by sexual reproduction. This conclusion is supported by segregation analysis of the genomic region that differs between H99α and KN99a (see Table S5 in the supplemental material). Based on this analysis, the vast majority of the recovered isolates were recombinant. Thus, following inhalation, spores are deposited in the lungs, germinate, and disseminate to both the spleen and the brain of infected animals.
To determine the possibility of murine infection caused by contaminating yeast cells (~7%) in pure spore preparations, we further conducted an experiment with low infection inocula (5, 25, 100, and 300 CFU per mice) of H99α yeast cells (see Fig. S4 in the supplemental material). Mice infected with 300 CFU succumbed to lethal infection by 31 to 34 days postinfection, and with 5 CFU, the survival of mice was prolonged until day 70 postinfection. Compared to mice infected with the spore-enriched preparation (465 spores/35 yeast cells), which showed severe infection by days 33 to 35 postinfection, with the lower yeast cell inocula alone (25 yeast cells), survival plots demonstrate the initiation of infection symptoms 50 days postinfection, with 100% mortality being delayed until day 61 (see Fig. S4 in the supplemental material). Thus, the potent virulence of the spore-enriched inoculum cannot be ascribed to the small percentage of contaminating yeast cells.
C. neoformans spores were also virulent in the invertebrate host model Galleria mellonella. Similar to murine virulence studies, larvae were infected with 500 CFU of spores produced by H99α and KN99a mating, H99α yeast cells, or a 1:1 mix of H99α and KN99a cells either from a liquid culture grown overnight or under plate conditions. Infections were performed by injecting 5 μl of the infectious inoculum in PBS into the terminal pseudopod, and infected larvae were incubated at room temperature (24°C) or at 37°C. Saline-injected larvae served as controls and in all cases survived for the duration of the experiment (8 to 14 days).
As shown in Fig. Fig.6,6, all four infectious inocula were highly virulent in G. mellonella larvae, resulting in 100% lethal infection by day 7 or day 13 at 37°C or 24°C, respectively. In the invertebrate host at 24°C, spores and yeast cells were of equivalent virulences at 24°C (P < 0.4161), and there was little or no difference in virulence among the three infectious yeast inocula (P < 0.2751 for H99 versus H99/KN99 liquid culture and P < 0.0417 for H99 versus H99/KN99 plate culture). This experiment was repeated, with similar results (see Fig. S5 in the supplemental material). Similar to the murine model, spores were modestly less virulent than yeast cells in the invertebrate host at 37°C in one virulence test (P < 0.0027 for H99 versus spores) (Fig. (Fig.6A),6A), whereas spores and yeast were equally virulent in the second independent experiment (P < 0.1149 for H99 versus spores) (see Fig. S5 in the supplemental material). These findings for an invertebrate host provide additional evidence that both yeast cells and spores are infectious.
The capacity of murine macrophages to phagocytose C. neoformans spores and yeast cells was evaluated by time lapse microscopy. As positive and negative controls, unopsonized yeast cells and yeast cells opsonized with monoclonal antibody 18B7, respectively, were assessed in parallel. Giles et al. recently found that spores do not require opsonization for phagocytosis (15). Our findings confirm their conclusions, and phagocytosis of unopsonized C. neoformans spores was clearly observed under our experimental conditions (see Movie S1 in the supplemental material), with the first phagocytosis events being detected by 55 min postincubation. Efficient phagocytosis of antibody-opsonized yeast cells by macrophages was evident by 30 min postincubation (see Movie S3 in the supplemental material). As expected, no phagocytosis of unopsonized control yeast cells was observed (see Movie S2 in the supplemental material). Control acapsular mutant strain cap59 was also tested and was found to be readily phagocytosed by macrophages, in accordance with data from previous studies showing that the capsule can serve to inhibit phagocytosis (see Movie S4 in the supplemental material) (26). In summary, the phagocytosis of yeast cells required opsonization, as expected, whereas the phagocytosis of spores did not.
As described above, our spore preparations are highly enriched for spores (~93%), but some yeast cells are still present in the spore sample, serving to mimic what might normally transpire during infections acquired in nature. Intriguingly, some yeast cells present in the spore sample were observed to be phagocytosed by macrophages (see Movie S1 in the supplemental material); this result was unexpected since no phagocytosis of unopsonized encapsulated yeast cells alone was observed. These results suggest that an unknown factor(s) present in the spore sample may potentiate the capacity of mouse macrophages to phagocytose encapsulated yeast cells, a phenomenon that typically requires the presence of an opsonin such as complement or antibody. Alternatively, the activation of phagocytosis by a spore may lead to the concomitant uptake of a yeast cell by macrophages. Taken together, our results confirm recent findings of others (15) showing that macrophages can readily phagocytose C. neoformans spores in the absence of opsonin, an event that may be critical for further disease progression or the establishment of latency.
The infectious propagules in several predominant human fungal pathogens are thought to be spores. Fungal spores are readily aerosolized and function as effective dispersal agents because of their shape, size, and ability to survive under harsh environmental conditions. Like many other fungal infections, such as those caused by Aspergillus, Histoplasma, and Coccidioides, cryptococcal infection is initiated when airborne infectious particles are inhaled into the human respiratory tract. Infectious particles larger than 5 μm are cleared by mucociliary action, whereas particles smaller than 2 μm penetrate and lodge in the alveoli, subsequently disseminate, and cross the blood-brain barrier to penetrate the central nervous system (18, 50).
In previous studies, attempts were made to isolate pure spores from mating cultures, and studies were performed using murine inhalation models with spores to analyze their pathogenicity (50, 59). Sukroongreung et al. previously performed studies using spores from C. neoformans var. neoformans (serotype D) and reported that the spores from this lineage are up to 100 times more infectious than yeast cells (50). Spores from Cryptococcus measure ~3 μm and readily lodge in the alveoli, compared to yeast cells, which are typically >5 μm. Desiccated yeast cells, which are similar in size to spores, may also efficiently enter the alveoli and disseminate. Our studies focused on spores to understand their morphology, germination signals, and host responses with controlled infectious inocula. We focused on the clinically most significant serotype, C. neoformans var. grubii (serotype A), which is the major cause (95%) of infections worldwide and more than 99% of infections in AIDS patients (6).
Comparative studies of spore morphologies from different C. neoformans strains and varieties provide a detailed view of surface properties and size and shape differences between spores from different serotypes and control yeast cells. Spores from serotype A measured between 2 and 3 μm, compared to serotype D spores, measuring 2 to 2.5 μm. Spore sizes were determined by measuring approximately 50 individual spores from each cross and by statistical analysis using t test results in which the two-tailed P value equals 0.0012, which is considered to be statistically significant. Spike-like protrusions on the surfaces of spores from serotypes A and D were commonly observed. We hypothesize that the rough surfaces of the spores may represent novel cell surface proteins or carbohydrates. In a self-fertile diploid strain (aAAα), the spores were smooth and measured 2 to 2.5 μm. Control yeast cells are round, >5 μm, and highly encapsulated. No visible capsule was observed on the spore surface (Fig. (Fig.11 and and2).2). Furthermore, in studies with anti-GXM antibody, no signal for polysaccharide capsule was observed on the spore surface, resulting in the exposure of glucans (α-1,3 and β-1,3) on the cell surface, as detected by staining with specific antibodies (see Fig. S3 in the supplemental material). Control yeast cell capsule was stained with anti-GXM antibody, whereas no staining was seen with antiglucan antibodies, indicating that the capsule overlays glucan structures (see Fig. S3 in the supplemental material). Interestingly, spores from C. gattii are more elongated than spores C. neoformans var. grubii and C. neoformans var. neoformans. C. gattii spores are elongated, with smooth surfaces, measuring 3 to 5 μm in length (Fig. (Fig.3).3). Sterigmata were not observed on the basidial surfaces for any of the crosses studied. This study of comparative morphology between different serotypes of Cryptococcus revealed distinguishable features between spores of C. neoformans, C. grubii, and C. gattii and both confirms and extends previously reported findings (47).
To examine the role of nutrients in triggering germination, we tested different media with reduced or no nutrients. When spores from serotype A (H99α and KN99a) matings were microdissected on 2% water agarose medium plates, microcolonies were observed (see Fig. S2 in the supplemental material). Our findings indicate that spores germinate readily, even under severe nutrient-limiting conditions. We also tested the ability of spores and yeast cells to aerosolize under laboratory conditions by exposing mating plates to an airstream mimicking natural wind currents. These studies indicate that spores are aerosolized and disperse more rapidly than yeast cells.
To test spores as infectious propagules, we enriched spores (~93%) from mating cultures (H99α and KN99a) and used these as infectious inocula in both murine and invertebrate host models. All of the infection experiments were conducted using spores from serotype A (H99α and KN99a) matings. In these studies, C. neoformans var. grubii spores were fully virulent (100% lethal infection) both in the murine inhalation virulence model and in the invertebrate host model Galleria mellonella. In two independent murine experiments, 500 CFU were used as the infectious inocula. Infectious inocula used for infection studies were typed for molecular and mating markers, confirming that ~ 93% of CFU were recombinant. The dissemination of spores in infected mice was analyzed by typing CFU isolated from lung, spleen, and brain of infected animals. These CFU were also typed for molecular and mating-type markers, confirming that the infections were caused by recombinant mating products. Our results support the conclusion that spores are able to enter the lungs, germinate, and cause infection in the murine model.
Animals or larvae infected with yeast (mono- or coinoculated) progressed more rapidly to lethal infection than did animals infected spores, and this result is statistically significant (Fig. (Fig.44 and and6).6). In contrast, previous studies showed that serotype D spores were up to 100 times more infectious than yeast cells (50). We hypothesize that there could be serotype differences in the virulence patterns between spores from serotypes A and D. Sukroongreung et al. previously conducted experiments with serotype D spores, whereas all of our virulence studies were conducted with serotype A spores. These serotypes are now recognized as distinct species that diverged 18.5 million years ago (32, 56), so they may differ in virulence. There could also be host species-specific differences such that spores are more infectious than yeast cells in humans but not in mice. This might be influenced by species-specific details of the airway anatomy or innate or adaptive immune differences. Spores might remain in a dormant phase before causing infection, which could explain the delayed mean time to mortality that we observed for mice instilled with spores compared to those instilled with yeast cells. Spores may evade host immune cell recognition and subsequent responses via efficient uptake by macrophages and delayed germination. To test this and mimic human immune responses during latency, a rat model (16) may be more appropriate than mice, for which disease progresses more rapidly.
In nature, infectious spores are likely produced by α-α same-sex mating in the largely unisexual population. Spores produced by unisexual mating could differ from those produced by opposite-sex mating. It could also be that the course of infection differs when animals are infected with a mixture of mating types compared to spores from unisexual mating of only one mating type. Previous studies revealed the potential for cell-cell interactions during coinfection with a and α yeast cells (42); similar interactions might occur or be magnified during infection with a mixture of spores of opposite mating types. Given that spores are readily phagocytosed without opsonin, alveolar macrophages may contain both a and α spores, and pheromone-mediated cell-cell communication in this setting could be more robust and influence the course of infection. This would not occur during infection with just α yeast cells or spores. Finally, our studies employed a nasal instillation model, and a frank inhalation model might be required to appreciate differences between spores and yeast cells.
In our in vitro analysis with alveolar mouse macrophages, spores were readily phagocytosed without opsonin, whereas control encapsulated yeast cells required opsonization to be phagocytosed by macrophages. We observed that yeast cells mixed with spores could be phagocytosed by macrophages when spores were also present. These observations confirm and extend the previously reported findings of Giles et al., who discovered that spores do not require opsonin to be phagocytosed by macrophages (15). We postulate that β-1,3-glucans on spores and acapsular yeast cells are readily exposed to dectin 1, leading to rapid phagocytosis by macrophages. Whether the phagocytosis of encapsulated yeast cells in the presence of spores involves factors released by spores, or concomitant uptake triggered by the spore surface, remains to be investigated.
We thank Valerie Knowlton at NC State University for assistance with scanning electron microscopy images, Chaoyang Xue and Anna Floyd for assistance with virulence experiments, Yonathan Lewit for outstanding technical assistance with RFLP analysis, John Perfect and Andrew Alspaugh for critical reading of the manuscript, and Edmond Byrnes for providing information for genotyping and for critical reading of the manuscript.
These studies were supported by NIH/NIAID R01 grant AI39115 to Joseph Heitman and NIH/HHLBI grant RO1 HL-30923 to Scarlett Geunes-Boyer and Jo Rae Wright.
Editor: A. Casadevall
Published ahead of print on 20 July 2009.
†Supplemental material for this article may be found at http://iai.asm.org/.