Comparison of Cryptococcus gattii/neoformans Species Complex to Related Genera (Papiliotrema and Naganishia) Reveal Variances in Virulence Associated Factors and Antifungal Susceptibility

Cryptococcosis is an infectious disease of worldwide distribution, caused by encapsulated yeasts belonging to the phylum Basidiomycota. The genus Cryptococcus includes several species distributed around the world. The C. gattii/neoformans species complex is largely responsible for most cases of cryptococcosis. However, clinical series have been published of infections caused by Papiliotrema (Cryptococcus) laurentii and Naganishia albida (Cryptococcus albidus), among other related genera. Here, we examined the pathogenic potential and antifungal susceptibility of C. gattii/neoformans species complex (clades I and II) and related genera (Papiliotrema and Naganishia) isolated from environmental and clinical samples. P. laurentii (clade III), N. liquefasciens/N. albidosimilis (clade IV); and N. adeliensis/N. albida (clade V) strains produced higher levels of phospholipase and hemolysins, whereas the C. gattii/neoformans species complex strains (clades I and II) had markedly thicker capsules, produced more biofilm biomass and melanin, which are known virulence attributes. Interestingly, 40% of C. neoformans strains (clade II) had MICs above the ECV established for this species to amphotericin B. Several non-C. gattii/neoformans species complex (clades III to V) had MICs equal to or above the ECVs established for C. deuterogattii and C. neoformans for all the three antifungal drugs tested. Finally, all the non-C. gattii/neoformans clinical isolates (clades III to V) produced more melanin than the environmental isolates might reflect their particularly enhanced need for melanin during in vivo protection. It is very clear that C. gattii/neoformans species complex (clades I and II) strains, in general, show more similar virulence phenotypes between each other when compared to non-C. gattii/neoformans species complex (clades III to V) isolates. These observations together with the fact that P. laurentii and Naganishia spp. (clades III to V) strains were collected from the outside of a University Hospital, identify features of these yeasts important for environmental and patient colonization and furthermore, define mechanisms for infections with these uncommon pathogens.


INTRODUCTION
Cryptococcosis is an infectious disease of worldwide distribution, caused by encapsulated yeasts belonging to the phylum Basidiomycota (Kwon-Chung and Bennett, 1992;Maziarz and Perfect, 2016). Cryptococcus spp., such as the Cryptococcus gattii/ neoformans species complex, are able to enter mammalian hosts and cause damage to the central nervous system (CNS) (Kwon-Chung et al., 2014;Lakoh et al., 2020) and respiratory tract (Lakoh et al., 2020). In the last few decades, cryptococcal meningoencephalitis has become an infection of global relevance, with up to 1 million new infections annually and significant attributable morbidity and mortality. Even with widespread effective antiretroviral therapy, there is still approximately 200,000 deaths per year especially among patients with human immunodeficiency virus (HIV) infection and AIDS (Park et al., 2009;Park et al., 2011). Besides HIV/AIDS infection, other risk factors for cryptococcosis include corticosteroids and other immunosuppressive therapies, solid organ transplantation, malignant and rheumatological diseases (Maziarz and Perfect, 2016).
Several virulence factors have classically been carefully characterized and validated for C. gattii/neoformans complex, such as the presence of a polysaccharide capsule that protects yeast cells against phagocytosis, the ability to grow at 37°C, biofilm formation, and the production of additional factors including melanin, phospholipase, metalloprotease, DNase, urease, superoxide dismutase and other antioxidant enzymes (Mancianti et al., 2002;Sanchez et al., 2008;Peng et al., 2018;Zaragoza, 2019). However, only a few studies have investigated the potential role of these and other virulence factors in the non-Cryptococcus gattii/neoformans species complex (Ikeda et al., 2002;Andrade-Silva et al., 2010;Araujo Gde et al., 2012;Ferreira-Paim et al., 2012;Araujo et al., 2017).
Because of the variations of virulence factors and clinical resistance to azole monotherapy, it is important to know and compare the relevant pathogenic potential and antifungal susceptibility profiling of Cryptococcus spp. collected from clinical and environmental sources. Therefore, the primary objectives of this study were to evaluate the pathogenic features and antifungal susceptibility of Cryptococcus-related genera isolated from pigeon droppings (environmental), compared to the Cryptococcus gattii/ neoformans species complex strains obtained from clinical samples of patients belonging to 3 different Brazilian states (Bahia, Pernambuco and Rio Grande do Norte) in the Northeast region. This is the first study to demonstrate that non-Cryptococcus gattii/ neoformans species complex may express a full range of known virulence factors in vitro and furthermore, some strains may exhibit increased minimal inhibitory concentrations (MICs) to antifungal drugs without previous known exposure to these compounds.

Collection and Processing of Guano Samples From Pigeons
The study included samples of pigeon droppings collected outside the building of a tertiary care university hospital, located in Natal City, Rio Grande do Norte State, Brazil, from April 2012to March 2014and November 2017to February 2018 Approximately, 2g of dried and/or fresh pigeon guano was collected from several sites and diluted in saline solution (0.9% NaCl) in a ratio of 1:10, vortexed for 10 minutes and allowed to stand for 1h. Subsequently, 100 mL of the supernatant was inoculated on the surface of Sabouraud Dextrose Agar (SDA; Oxoid, Basingstoke, Hampshire, UK) and Rose Bengal (0.05 g/L) plates with the aid of a Drigalski loop. Each plate was incubated at 30°C for 48h. Yeast colonies with creamy to mucoid phenotypes were selected for subsequent identification.

Phenotypic Identification of Cryptoccocus spp. Related Genera Isolates
After colony growth, yeast cells were streaked onto the surface of CHROMagar Candida ® (CHROMagar Microbiology, Paris, France) to check for purity and screening for different color colonies (Baumgartner et al., 1996). The initial identification (screening for the Cryptococcus genus) was based on the characteristics of the yeast cells observed microscopically after cultivation on corn meal agar with Tween 80, glucose fermentation (non-fermentative), urease test, Niger seed agar (NSA) for melanin production and capsule visualization with an India ink stain (Yarrow, 1998). After identification, all cryptococcal yeast isolates were cultured in YPD broth (Yeast Peptone Dextrose, yeast extract 10 g/L, dextrose 20g/L, peptone 20g/L) overnight at 30°C, and then transferred to cryotubes containing 20% glycerol and stored at -80°C. The isolates now belong to the fungal culture collection of the Medical and Molecular Mycology Laboratory, Department of Clinical and Toxicological Analyses, Federal University of Rio Grande do Norte.

DNA Extraction
The isolates were grown on YPD (BD Difco, NJ, USA) plates for 72 h and then, a single colony was inoculated into 3.5 mL of 2xYPD broth in a 14 ml culture tube and incubated at 30°C with shaking (225 rpm) for 48 h. DNA was extracted using the MasterPure Yeast DNA Purification Kit (Epicentre Biotechnologies, Madison, WI, USA). DNA concentrations were measured using the Qubit dsDNA HS assay according to instructions (Life Technologies).

PCR Assay and ITS Region Sequencing
The PCR Hot Start Taq 2X Master Mix (New England Biolabs) was used to amplify all fragments generated in this work. Samples were amplified in a Bio-rad thermocycler (Thermo Cycler T100) using the following cycling parameters for both regions: one initial cycle of 95°C for 30 sec; followed by 30 cycles of 30 sec at 95°C, 1 min at 47°C and 1 min at 68°C and a final cycle of 5 min at 68°C. PCR products were treated with ExoSAP-IT (Thermo Fisher Scientific) according to the manufacturer's protocol. Purified PCR products were sequenced by the Sanger sequencing method on ABI 3730 xl DNA Sequencers by Eton Bioscience, Inc. For sequencing the ITS region, the internal primers ITS1 and ITS4 were used (White et al., 1990). Nucleotide sequences were submitted for BLAST analysis at the NCBI site (http://www.ncbi.nlm.nih.gov) for species identification. Only sequences deposited in GenBank showing high similarities with our query sequences and an E value of lower than 10 −5 were used in this study.

Genbank Accession Numbers
GenBank accession numbers may be seen in Table 1.

Phylogenetic Analysis
The ITS rDNA sequences generated from all the isolates were aligned with type strain sequences from Cryptococcus spp. and the other genera present as matches in the BLASTn searches (http:// blast.ncbi.nlm.nih.gov/Blast.cgi), according to taxonomic reclassification of members within the Tremellomycetes (Liu et al., 2015). All sequences were aligned using the Muscle algorithm implemented by MEGA X (Kumar et al., 2018). The phylogeny was inferred using the Neighbor-Joining method (Saitou and Nei, 1987), including gap positions, and the evolutionary distances were corrected by the Kimura 2-parameter method (Kimura, 1980). Bootstrap analysis (bt) (Felsenstein, 1985) was conducted by evaluating 1,000 pseudoreplicates of the alignment. All analyses were computed using MEGA X (Kumar et al., 2018).

Isolates Used in the Present Study
We have included 106 isolates of Cryptococcus spp. and related genera from our culture collection to perform the evaluation of these isolates virulence factor expressions and antifungal susceptibility testing in vitro. Thirty-eight of them were environmental: Papiliotrema laurentii (clade III; n=9), N. liquefasciens/N. albidosimilis (clade IV; n=19), N. adeliensis/N. albida (clade V; n=10). Sixty-eighth clinical isolates primarily from the Cryptococcus gattii/neoformans species complex were assessed including C. neoformans (clade II; n=45), C. gattii (clade I; n=20), P. laurenttii (clade III; n=1), N. liquefasciens/N. albidosimilis (clade IV; n=1) and N. adeliensis/N. albida (clade V; n=1). The clinical isolates were obtained from patients admitted to tertiary hospitals from three different Brazilian states (Rio Grande do Norte, Pernambuco and Bahia); all of them located in the Brazilian Northeast region ( Table 1).

Inoculum Preparation of Cryptococcus spp. and Related Genera Isolates for In Vitro Assessment of Virulence Factors
For the phenotypic characterizations of the different isolates, the yeasts were initially grown in NGY broth medium (Difco Neopeptone 1 g/L, Dextrose 4 g/L; Difco yeast extract 1 g/L) prior to the phospholipase and hemolysin enzymatic activity tests. Yeast cells were inoculated by "wet looping" in this medium with a ring loop loaded with a yeast suspension film, rapidly immersed in the medium and removed and incubated for 18-24h in a shaker incubator at 30°C at 200 rpm. An inoculum of   approximately 2x10 8 cells/mL is produced (Chaves et al., 2007). Cultures were spectrophotometrically measured at a wavelength of 600nm ranging from 0.8 and 1.2 (Biochrom Libra S32). Subsequently, yeast cells were diluted to obtain the same inoculum for each condition evaluated in vitro.

Cryptococcus spp. and Related Genera for Phospholipase Production
For detection of the phospholipase activity, the method of Price et al. (1982) (Price et al., 1982) adapted for Cryptococcus spp. was used (Pini et al., 2017). Overnight NGY cultures were diluted and standardized to a concentration of 2x10 8 yeast cells/mL and the suspension of yeast cells was inoculated in triplicate onto the surface of phospholipase agar (10 g peptone, 40 g dextrose, 16 g agar, 80 mL Egg Yolk Emulsion [Fluka] was added to 1000 mL of distilled water). The plates were incubated at 30°C for 10 days. After the incubation period, the diameters of the colonies and the halo formed around them were measured. The Pz (phospholipase zone) was determined by dividing the colony diameter by the precipitation zone plus colony diameter. The strains were classified as follows, according to tertiles distribution: Pz = 1 as negative phospholipase activity; 0.65 ≤ Pz ≤ 0.83 as weak; 0.46 ≤ Pz ≤ 0.64 as moderate; 0.27 ≤ Pz ≤ 0.45 as strong phospholipase producers.

Measurement of Hemolysin Production of Cryptococcus spp. and Related Genera
In order to evaluate hemolysin production, we followed the methodology proposed by Luo (Luo et al., 2001) with some adaptations. Yeast cells were initially cultured on SDA at 30°C for 24 h. Strains were next grown overnight in NGY broth. Ten microliters of the culture were streaked out in triplicate onto the surface of SDA containing 7% fresh sheep blood (Ebe-Farma) and 3% glucose, in Petri dishes of 155 mm of diameter. The plates were incubated for 48 h at 37°C in an atmosphere with 5% CO 2 . After the incubation period, the presence of a clear halo around the inoculum indicated positive hemolysis. The diameter of colonies and zones of hemolysis were measured in order to obtain the hemolysis index (HI) for each strain. HI was determined by dividing the colony diameter by the precipitation zone plus colony diameter, which allowed classification of isolates into strong, moderate and weak producers. The strains were classified as follows, according to tertiles distribution: HI= 1 as negative hemolytic index; 0.69 ≤ Pz ≤ 0.84 as weak; 0.53 ≤ HI ≤ 0.68 as moderate; 0.37 ≤ HI ≤ 0.52 as strong hemolysin producers.

Capsule Size Measurement in Cryptococcus spp. and Related Genera
To perform the capsule experiments, we used the protocol of Zaragoza et al., (Zaragoza et al., 2003) with adaptations. All the strains were grown in YPD broth for 24 h at 37°C with 200 rpm agitation. After washing with phosphate buffer saline (PBS), yeast cells were adjusted to 1×10 5 cells/mL using a hemocytometer to standardize the inoculum and inoculated in a solution containing 10% fetal bovine serum (FBS; Sigma-Aldrich; Missouri, Saint Louis, USA) diluted in PBS (5mL of final solution in conical tubes). The tubes were incubated for 24 h at 37°C in an atmosphere of 5% CO 2 . Subsequently, a drop of India ink stain was added to a small volume of yeast cells in suspension and placed on a microscope slide. The samples were imaged with a Nikon Eclipse E100 microscope. Pictures were taken using a Tsview 7 software. Total diameter (including capsule) (dt) and yeast cell diameter (dy) were measured for each strain. Capsule thickness (tc) was determined with the following equation: tc=1/2 (dt-dy) in micrometers (Fernandes et al., 2018).

Melanin Production in Cryptococcus spp. and Related Genera
To evaluate melanin production, yeast cells were cultured on SDA for 72 h at 30°C. Then, a single colony was streaked out thrice (parallel sides) on the surface of NSA medium at 30°C for 10 days for visualization of melanin production. Scores from 0 to 4 were attributed as follows: 0 (zero), no pigment; 1, light brown; 2,brown; 3, dark brown; and 4, for almost black colonies (Pedroso et al., 2009).

Biofilm Formation of Cryptococcus spp. and Related Genera
Biofilm formation assays were performed according to Palanco et al. (2017)  were used for C. deuterogattii (clade I) and C. neoformans (clade II). The ECV determines the upper limit of wild-type distribution (without intrinsic or acquired resistance mechanisms), distinguishing from non-wild-type isolates harboring intrinsic or acquired resistance mechanisms (CLSI, 2020).

Statistical Analysis
Data were analyzed using the statistical software "Graph Pad version 8.0". Results were presented as mean ± standard deviation, and differences were analyzed by the Mann-Whitney test. For all the analyses, P was considered a default value of 0.05 and the confidence interval of 95%. In addition, the values obtained for some of the virulence attribute tests in vitro were divided onto tertile categories as weak, moderate or strong producers.

Phenotypic Preliminary Identification of the Strains Used in the Present Study
All the isolates produced encapsulated blastoconidia, did not ferment glucose and were able to hydrolyze urea. Some isolates, but not all of them, produced melanin, when grown on NSA. Therefore, they were presumptively identified as Cryptococcus spp. and related genera.

Molecular Identification of the Strains Used in the Present Study
For molecular species identification, DNA fragments of the ITS region were amplified and DNA sequences obtained were sent to the GenBank genome database at the NCBI website (http://ncbi. nlm.nih.gov) using BLAST to compare gene sequences. All the Cryptococcus gattii/neoformans strains were unambiguously identified by sequencing, as follows: 45 as C. neoformans (formerly C. neoformans var. grubii) and 20 as C. deuterogattii (formerly C. gattii VGII genotype). It is important to recognize that the identification of these cryptococcal strains was further confirmed with the Multi-Locus Sequence Typing (MLST) consensus scheme of seven genetic loci (unpublished data). The same trend of accurate identification was also observed for P. laurentii (10 isolates). Nevertheless, for the other non-Cryptococcus gattii/neoformans species complex strains, a less precise identification was observed. For some of the strains (n=20), BLAST comparisons revealed 100% identity (E-value of 0) with three different Naganishia spp. (N. liquefasciens/N. albidosimilis/N. diffluens), while 10 of them were 99.8% identical to N. albida and N. adeliensis).

Phylogenetic Analysis
Phylogenetic analysis confirmed the species identification of 20 isolates as C. deuterogattii with bt of 54 (Clade I). The other isolates were confirmed as C. neoformans (45) and P. laurentii (10), respectively (Clades II and III), with high bt values (99). All the other yeast isolates were identified as Naganishia spp. due to low confidence values in the analysis to confirm BLASTn species identification. Twenty isolates belong to the N. liquefasciens/N. albidosimilis clade (bt=99), excluding N. diffluens from the most probable species (clade IV), while 11 isolates were identified within the N. adeliensis/N. albida clade, grouping closer to the first species (bt=62; clade V; Figure 1).

Colony Phenotypes and Melanin Production of the Strains Used in the Present Study
Cryptococcus gattii/neoformans species complex (clades I and II and N. adeliensis/N. albida (clade V) Chromagar Candida colony colors ranged from pale pink to purple (mainly pink for C. neoformans isolates), while all the isolates belonging to P. laurentii (clade II) and N. liquefasciens/N. albidosimilis (clade IV) were dark purple. All non-Cryptococcus gattii/neoformans species complex isolates (clades III to V) grew very poorly on Chromagar Candida (incubation at 37°C), whereas the opposite growth happened to clade I and II isolates (Cryptococcus gattii/ neoformans species complex) All of the strains showed creamy to mucoid colony growth phenotypes when grown on YPD at 30°C, with all C. neoformans isolates being highly mucoid. Of note, strains belonging to clades III to V (non-Cryptococcus gattii/ neoformans species complex) produced occasionally a pink pigment when grown on this culture medium ( Figure 2). Approximately,67% of the isolates belonging to clades I and II produced dark colonies (scores 3 to 4) on NSA, while the other isolates (clades III to V) produced either light brown colonies (80.5%; score 1) or did not produce melanin at all (14.6%; score 0), except for isolates LMMM1398 (clade III), LMMM1460 (clade IV) and LMMM221 (clade V) colonies which were classified as score 3. Interestingly, these strains were non-Cryptococcus gattii/ neoformans species complex strains obtained from patients (CSF samples) and not the environment (Table 1).

Cryptococcus spp. and Related Genera Phospholipase Production
Phospholipase production was detected in all of the isolates (100%) evaluated. However, the levels of enzyme production varied among strains of each species. Phospholipase activity is inversely proportional to Pz, where values equal to 1 mean negative activity. The enzyme production ranged from 0.27 ± 0.02 (LMMM460; P. laurentii; clade III) to 0.83 ± 0.04 (LMMM644; C. neoformans clade II). Most of Cryptococcus gattii/neoformans species complex (clades I and II strains) were considered weak (60%) to moderate phospholipase producers (35.4%; Table 1 and Figure 3), while 51.2% of non-Cryptococcus gattii/neoformans species complex (clades III to V strains) were considered strong phospholipase producers and 41.5% showed moderate production of this enzyme. These data were reinforced with average value comparisons for the production of this enzyme, which showed significantly different results between C. gattii/neoformans species complex strains (clades I and II) and the other strains belonging to non-Cryptococcus gattii/ neoformans species complex (clades III to V; Figure 3). Of note, the average value of P. laurentii (clade III) strains for phospholipase production was also considered lower than the results found for N. liquefasciens/N. albidosimilis (clade IV) and N. adeliensis/N. albida (clade V)strains (P<0.05; Figure 3). Our three clinical non-C. gattii/neoformans species complex isolates were considered strong (P. laurentii; clade III; LMMM1398; 0.45 ± 0.05 and N. adeliensis/N. albida;clade V; LMMM1429 0.45 ± 0.015) to moderate producers (N. liquefasciens/N. albidosimilis; clade IV; LMMM1460; 0.60 ± 0.09; Table 1 and Figure 3).

Production of Hemolysins by Cryptococcus spp. and Related Genera
Almost 91% of the isolates were able to produce hemolysis. HI is also inversely proportional to the diameter of the halo, where values equal to 1 mean negative activity. Most of the isolates in the present study were able to show medium (41.5%) to large halos (38.7%), meaning low to moderate HI. Interestingly, 7 C. deuterogattii (clade II) isolates did not produce halos suggesting that the enzyme(s) was not produced. The average values for non-C. gattii/neoformans species complex strains (clades III to V) of hemolysin production was significantly higher than those of Cryptococcus gattii/neoformans species complex (clades I and II) isolates (P<0.05; Figure 3). Most clade I and II strains were considered moderate hemolysin producers (53.8%) while larger halos (strong hemolysin production) were observed for 65.9% of clades III to V strains, being considered strong hemolysin producers, including all three clinical isolates within these clades. The average value for P. laurentii (clade III) strains hemolysin production was lower than the results found for   Figure 3). In addition, there was no statistically significant difference between P. laurentii (clade III) expression of this virulence factor in vitro and C. gattii/neoformans species complex (clades I and II)isolates. The isolates with the highest hemolytic activity was LMMM383, belonging to N. liquefasciens/N. albidosimilis (clade IV) (0.37 ± 0.06; Table 1).

Determination of Capsule Thickness in Cryptococcus spp. and Related Genera
There was a huge interspecies variation in average values of capsule thickness among certain isolates of different clades. It ranged from 4.1 ± 1.27 µm (LMMM1397) to 36.07 ± 9.33 µm (LMMM787). Interestingly, both of these isolates belonged to C. deuterogatti (clade I) showing wide strain variation. When this species was analyzed separately, it was observed that this species contained the isolates with greatest capsule thickness (LMMM1241; 35.87 ± 7.33 µm and LMMM782; 30.61 ± 4.55 µm), and several others with sizes above 20 µm were observed in this species. Furthermore, there was not a significant difference when the average values for the two clades (I and II) were compared, due to the enormous variation in measurements found within C. deuterogattii (clade I) strains, despite the observation that the greatest capsule thickness observed for C. neoformans (clade II) strains was 20.95 ± 5.28 µm (LMMM620; Table 1). All the other isolates belonging to the non-Cryptococcus gattii/neoformans species complex (clades III to V) had significantly smaller capsules when compared to Cryptococcus gattii/neoformans species complex strains (clades I and II) and no differences in average sizes were found among them (Figure 3).

Evaluation of Biofilm Formation in
Cryptococcus spp. and Related Genera All the isolates were able to form biofilms on polystyrene microtiter plates. However, there was a large variation in biofilm biomass among the isolates. The OD570nm readings ranged from 0.21 ± 0.03 (N. adeliensis/N. albida; clade V; LMMM406) to OD570nm of 1.14 ± 0.06, a C. deuterogattii (clade I) isolates (clade I; LMMM700 (P<0.05; Table 1 and Figure 3). In fact, this was the unique virulence factor expressed in vitro that showed a significant difference between C. deuterogatti (clade I) and C. neoformans (clade II) isolates. For instance, 90% of C. deuterogattii (clade I) isolates were considered strong biofilm producers, while 75.6% of C. neoformans (clade II) isolates were included in this category. Taken together, 80% of Cryptococcus gattii/neoformans species complex isolates (clades I and II) were considered strong biofilm producers. The average values for clades I and II OD570nm readings were greater than the results found for the within the non-Cryptococcus gattii/ neoformans species complex (clades III to V; P<0.05; Figure 3). Furthermore, the vast majority of isolates belonging to clades I and II were not considered weak biofilm producers (except for LMMM644, a C. neoformans (clade II) strain; OD570nm of 0.45 ± Hemolytic index after incubation on SDA supplemented with sheep's blood and glucose at 37°C for 48 h, 5% CO 2 . (C) Capsule thickness determined after cells incubation in 10% fetal bovine serum in PBS, incubated for 24 h at 37°C in an atmosphere of 5% CO 2 and (D) Biofilm formation induced after cells incubation in 96-wells microtiter plates containing BHI medium at 37°C, for 48 h. Each bar represents mean ± standard deviation of the triplicate obtained for each isolate. *P < 0.05; ***P < 0.001.

Oliveira et al.
Cryptococcus spp. Reveal Phenotypic Differences 0.09). In contrast, the non-C. gattii/neoformans species complex strains (clades III to V) had no isolates identified as high biofilm producers. All N. liquefasciens/N. albidosimilis (clade IV) and 95% of N. adeliensis/N. albida (clade V strains were actually considered weak biofilm producers. In addition, 60% of P. laurentii strains (clade III) showed low biofilm formation, and the other 40% were considered moderate biofilm producers (Table 1). If we compare the average values of OD570nm readings among all three clades, P. laurentii (clade III) showed a trend for higher biofilm production compared to the other 2 clades (IV and V) and this difference was statistically significant between clades III and IV ( Figure 3). Interestingly, the clinical isolates belonging to the non-Cryptococcus gattii/neoformans species complex (clades III to V) were also considered weak to moderate biofilm producers ( Table 1 and Figure 3).

Antifungal Susceptibility Testing for Cryptococcus spp. and Related Genera
All MIC values obtained by the reference strains were compatible with the values expected by the CLSI methodology, assuring the reliability of the results obtained for the isolates tested. MIC ranges and geometric means (GMs) for all the antifungal drugs tested are displayed in Table 2. Because there are no breakpoints established to Cryptococcus spp. and related genera, we have included ECVs, according to the CLSI M59-Ed3 document for C. deuterogattii (clade I) and C. neoformans (clade II) strains (species that ECVs are available). Most strains had low MICs to fluconazole, while a few of them showed increased MICs. Two out of 45 (4.4%) C. neoformans strains (LMMM629 and LMMM631) where considered non-wild-type. All P. laurentii isolates had low MICs to fluconazole. Four out of 20 (20%) clade IV strains and a single N. albida isolates (LMMM1430; clade V; 9.1%), had MICs ≥16 µg/mL ( Table 2 and Supplementary Table 1).
All the strains tested had MICs below the ECVs established for C. deuterogattii (clade I) and C. neoformans (clade II) to itraconazole. Some strains belonging to the non-Cryptococcus gattii/neoformans species complex (clades III to V) had higher MICs, when compared to clade I and II strains MICs ≥ 0.25 µg/ mL ( Table 2 and Supplementary Table 1).

DISCUSSION
Our molecular identification and phylogenetic analysis revealed five different Cryptococcus spp. and related genera clades. Regarding the C. gattii/neoformans species complex isolates (clades I and II), which comprised the vast majority of the clinical isolates, BLAST searches showed that our C. gattii isolates (clade I) belonged to the VGII genotype, named as C. deuterogatti (Hagen et al., 2015). This is the main species recovered in Brazil from both environmental and clinical samples (Dos Freire et al., 2012;Maruyama et al., 2019;Santos Bentes et al., 2019). Following the same trend of other studies performed in Brazil (Cogliati, 2013;Alves et al., 2016;Ferreira-Paim et al., 2017), our C. neoformans strains were all considered the former C. neoformans var. grubii, but because of the recent recognition as a separate species in the C.gattii/neoformans species complex (Hagen et al., 2015), we will use the nomenclature C. neoformans, as recommended by Hagen et al., (Hagen et al., 2017).
Regarding the non-C. gattii/C. neoformans species complex (clades III to V) isolates identification, we faced a limitation using ITS region sequencing, particularly among Naganishia spp.
(clades IV and V) isolates. When our sequences were submitted to GenBank, the final identification was not precise, because they were considered identical to three different species as follows: N. liquefaciens, N.albidosimilis and N. diffluens (clade IV). However, our phylogenetic analysis excluded N. diffluens. The same level of indistinguishability was found for N. albida/adeliensis (clade V). Furthermore, all of them were previously considered Cryptococcus spp., but Liu et al. (Liu et al., 2015) reclassified several Tremellomycetes isolates, based on the results of phylogenetic analyses from a seven-genes dataset, including the large subunit ribosomal DNA (LSU rDNA locus). Because we have only used a single locus in our study (ITS region), we labeled our isolates as Naganishia spp. only and have submitted the sequences to the GenBank database as such. Further typing using DNA sequence of different loci will be needed for accurate species identification. Similar to our findings, Brito et al., (Brito et al., 2019) reported the recovery of P. laurentii, N. albida and N. liquefaciens among other former Cryptococcus spp. from bird excreta, collected outside a Brazilian University hospital, but no C. neoformans strains were found. P. laurentii was also recovered from bird droppings samples in Brazil, together with C. neoformans in the same samples (Andrade-Silva et al., 2010;Ferreira-Paim et al., 2011). N. albida, N. uzbekistanensis, N. adeliensis and C. neoformans were isolated from pigeon nests. N. adeliensis was the only species isolated from Eucalyptus trees in Ilam, Iran (Kamari et al., 2017).
Although most human infections and specifically, those related to the CNS are caused by C. gattii/neoformans species complex isolates, less common former Cryptococcus spp. have been reported to cause invasive disease due to P. laurentii and N. albida (Cheng et al., 2001;Averbuch et al., 2002;McCurdy and Morrow, 2003;Shankar et al., 2006;Khawcharoenporn et al., 2007;Furman-Kuklinska et al., 2009;Gullo et al., 2013;Neves et al., 2015). A systematic review performed by Londero et al., (Londero et al., 2019) reported 35 cases of deep-seated infections due to P. laurentii. In this study, 16.1% of cases were catheterrelated, 67.7% had positive blood (54.8%) or CSF (12.9%) cultures. Most patients were immunosuppressed and 48.4% of these patients had neoplasia as an underlying disease. Choe et al. (Choe et al., 2020) reported a case series of 20 patients with N. albida (or the former C. albidus infections), and these patients had high mortality rates (40%) despite antifungal treatment. The isolates came from the bloodstream in 9 cases, CNS in 6 cases, lower respiratory tract in 5 cases, and a single sample from peritoneal fluid. Seven cases of non-invasive infections were also reported, including colonization of the skin and fungal keratitis. Furthermore, a case of catheter-related fungemia due to N. liquefaciens has been described in a 71-year-old man with Bcell lymphoma (Takemura et al., 2015). In addition, a CSF mixed infection caused by N. liquequefaciens and Mycobacterium tuberculosis species complex has been reported in an HIVpositive female patient who died despite antifungal and antibacterial therapy. The final identification of this isolate was only possible with the sequencing of the D1/D2 domains of the large subunit of the rRNA region, since ITS rDNA sequencing was not able to differentiate between N. liquefaciens and N. albidosimilis (Conde-Pereira et al., 2015) and similar to our findings.
All C. gattii/C. neoformans species complex strains (clades I and II) isolates were able to grow at 37°C, but the high temperature growth phenotype was not observed for non-C. gattii/C. neoformans species complex strains (clades III to V), even if they were clinical isolates (LMMM1398, 1460 and 221). It is widely documented that the ability to grow at human physiological temperature is an important virulence factor for Cryptococcus spp. (Zaragoza, 2019). Nevertheless, other opportunistic fungi such as Rhodotorula spp. grow poorly in vitro in temperatures above 33°C, and yet, can cause human infections such as fungemia (Ioannou et al., 2019) and may also belong to the human gut microbiota (Borges et al., 2018).
It is very clear that C. gattii/neoformans species complex isolates (clades I and II), in general, show more similar virulence phenotypes among each other when compared to non-C. gattii/ C. neoformans species complex isolates (clades III to V) and vice versa. However, occasionally, P. laurentii (clade III) strains do behave closer to C. gattii/neoformans species complex isolates (clades I and II) such as with phospholipase and hemolysins production. While clades I and II strains have thicker capsules, produced more melanin and formed greater biofilms, clades III to V strains were able to produce the enzymatic activities of phospholipase and hemolysins. However, it is important to emphasize that because we do not have a balanced number of clinical and environmental strains for each clade, comparisons were performed independently of the source of the strain isolation We recognize that if more clinical strains of P. laurentii and Naganishia spp. (clades III to V) or environmental strains of C. gattii/neoformans species complex isolates (clades I and II) were included, our results could have been different because of strain adaptations to different ecological niches. Therefore, we recognize this fact as a limitation of our study. Ferreira-Paim et al. (2012) were the first to describe hemolysin production by the former C. laurentii isolates obtained from pigeon droppings. The fact that non-C. gattii/ neoformans species complex strains are able to efficiently produce hemolysins may partially explain the reason why these species are more frequently described in cases of fungemia, rather than meningeal or deep-seated tissue infections (Londero et al., 2019). Conversely, phospholipase production has been largely reported for C. gattii/C. neoformans species complex strains (Pini et al., 2017). This enzyme in C. neoformans is thought to trigger capsule enlargement, reduce phagocytosis by macrophages (Chrisman et al., 2011) and secreted phospholipase B (PLB) activity favors the survival and replication of C. neoformans in macrophages in vitro, but it seems that this enzyme is not directly involved in the establishment of CNS infections (Santangelo et al., 2004). Interestingly, Pedroso et al. (2009) reported greater phospholipase activity in C. albidus than in C. laurenttii isolates recovered from several organs in a murine model of infection, consistent with our findings. However, it appears that these extracellular enzymes may be critical in breaking down the environmental milieu rather than participating in mammalian host tissue invasion. Most of the clinical isolates when grown on NSA had colonies that were darkly pigmented (scores 3 to 4), including the non-C. gattii/C. neoformans species complex clinical isolates (LMMM1398, LMMM1460 and LMM221) and supports the importance of this as a mammalian virulence factor. Furthermore, melanin production is crucial for Cryptococcus spp. virulence, since mutants unable to produce this pigment have a significant attenuation in virulence (Salas et al., 1996). Melanin confers resistance to oxidative stress, heat and ionizing radiation (Zaragoza, 2019), besides impairing proper antifungal drug actions (Nosanchuk and Casadevall, 2006).
In our biofilm formation experiments, C. deuterogatti (clade I) was more efficiently able to form biofilm than C. neoformans (clade II) strains. Tavares et al. (2019) recently investigated C. gattii (VGI) kinetics of biofilm formation and described its structure as a dense network of yeast cells deeply encased in an extracellular polymeric matrix (Tavares et al., 2019). In terms of clinical significance, C. gattii species are frequently described in the immunocompetent host, cause more damage to the lungs and frequently produce cryptococcomas, which are considered biofilm-like structures (Chen et al., 2014;Aslanyan et al., 2017;Rodriguez-Goncer et al., 2018). In our study, there was also trend of greater biofilm formation in P. laurentii (clade III) than in Naganishia spp. (clades IV and V), but the three clades together were considered significantly less effective in biofilm formation than C. gattii/C. neoformans species complex strains (clades I and II). However, P. laurentii biofilm was recently described to have dense growth and colonization with extensive polymeric substances around the yeast cells (Ajesh and Sreejith, 2012).
C. deuterogatti (clade I) strains showed a large variation in capsule sizes, with some of the isolates possessing the largest capsules in the present study. Giant yeast cells (capsule thickness ≥ 15 µm) and variation in capsule thickness are more common in C. gattii complex than in the C. neoformans complex. However, both "large" and "small" capsular phenotypes are important in different stages of infection (early and late, respectively), reinforcing the advantage of strains with plasticity in capsule and cell sizes to cause cryptococcosis (Fernandes et al., 2018). The non-Cryptococcus gattii/neoformans species complex strains (clades III to V) had markedly thinner capsules. However, Araujo et al., (Araujo et al., 2017) described that polysaccharides of C. liquefaciens have strikingly similar ultrastructural and biological properties to those of C. neoformans, and found that this species led to mortality rates similar to C. neoformans in a virulence model with the larvae of Galleria melonella, reinforcing its pathogenic potential.
There are no antifungal breakpoints yet established for Cryptococcal species so discussions about drug susceptible and resistant isolates need to wait until breakpoints are determined. However, it is clear from our results most strains fit under the ECVs established for C. deuterogattii (clade I) and C. neoformans (clade II) and surely do not possess high-level resistance since azole MICs are within achievable drug levels in the blood. On the other hand, the isolates have some variability in their azole susceptibility. For the polyene in vitro activity, some of the MICs are relatively high and this may be more in vitro media importance rather than clinical importance. In a Chinese multicenter study almost 24% of 312 C. neoformans strains had non-wild-type phenotypes to fluconazole with 5 isolates possessing had MICs≥ 32 µg/mL (Ajesh and Sreejith, 2012) but Pfaller et al. with over 1,811 clinical isolates reported only 1% with very high MICs. Ferreira and Pahim (Ferreira-Paim et al., 2012) found that 71% of 38 environmental C. laurentii isolates had elevated azole MICs and Kantarcioglu et al. (2009) showed decreased susceptibility of 7/ 11 C. diffluens to both azoles and polyenes. Both our data and others suggest that azoles and polyenes will likely have clinical activity against both species complexes, the C. gattii/neoformans and non-C. gattii/neoformans species complex. However, some isolates push toward the higher antifungal MICs especially with the non-C. gattii/neoformans species complex strains which also possess less clinical treatment experience. Therefore, assessment of in vitro susceptibility for these isolates may be helpful to the clinician during clinical management.

CONCLUSIONS
In conclusion, we have shown different phenotypic traits between C. gattii/neoformans species complex (clades I and II) and other related genera such as Papiliotrema (clade III) and Naganishia (clades IV and V). In addition, P. laurentii (clade III) shows phenotypic properties more similar to C. gattii/C. neoformans species complex isolates (clades I and II), when compared to Naganishia spp (clades IV and V). But Naganishia spp. produced higher enzymatic levels of phospholipase and hemolysins. Furthermore, C. gattii/C. neoformans species complex (clades I and II) isolates had markedly larger capsules, produced more biofilm biomass and melanin. In addition, more isolates belonging to non-Cryptococcus gattii/neoformans species complex (clades III to V) possessed higher MICs to fluconazole, itraconazole and amphotericin B than those belonging to the C. gattii/neoformans species complex (clades I and II). However, 40% of C. neoformans strains had MICs equal to or above the ECV to amphotericin B. C. deuterogattii and C. neoformans showed very similar phenotypes among themselves. However, C. deuterogatti strains (clade I) produced statistically significant more biofilm biomass than C. neoformans (clade II). Another interesting observation is that all the non-C. gattii/neoformans clinical isolates (clades III to V) produced darker colonies, suggesting higher melanin production, when compared to their environmental counterparts. More isolates are needed to determine if there is an association of melanin production with non-C. gattii/C. neoformans clinical isolates. These findings, together with the fact that P. laurentii and Naganishia spp. strains were collected outside of a University Hospital, could be important for environmental to patient transmission that produces nosocomial infections especially in immunocompromised patients.