Search for Cryptococcus neoformans/gattii Complexes and Related Genera (Filobasidium, Holtermanniella, Naganishia, Papiliotrema, Solicoccozyma, Vishniacozyma) spp. Biotope: Two Years Surveillance of Wild Avian Fauna in Southern France

Fungi belonging to the Cryptococcus genus and related genera (Filobasidium, Holtermanniella, Naganishia, Papiliotrema, Solicoccozyma, Vishniacozyma) are encapsulated yeasts found in either the environment or animal sources. However, the precise biotopes of most species remain poorly defined. To assess whether wild birds from southern France can carry or spread the most pathogenic species (i.e., species belonging to the C. neoformans and C. gattii complexes), as well as lesser-studied species (non-neoformans/gattii Cryptococcus and former Cryptococcus spp.), 669 birds belonging to 89 species received for care over a two-year period at the Centre de Protection de la Faune Sauvage of Villeveyrac (Bird Protection League nongovernmental organization (NGO) care center) were sampled. Samples were cultured, and Cryptococcus and former Cryptococcus yeasts were identified by PCR sequencing. The purpose was to evaluate whether there was any health risk to local populations or care personnel in aviaries and gather new data on the ecological niches of lesser-known species. One hundred and seven birds (16%) were found to be positive for at least one Cryptococcus or former Cryptococcus species. No yeasts belonging to the highly pathogenic C. neoformans or C. gattii complexes were isolated. However, diversity was notable, with 20 different Cryptococcus or former Cryptococcus species identified. Furthermore, most bird–yeast species associations found in this study have never been described before.


Introduction
Cryptococcus spp. are yeasts with a natural biotope that varies from species to species. The genus has recently undergone major classification changes, with the separation of C. neoformans and C. gattii species into two complexes of two (C. neoformans/C. deneoformans) and five (C. gattii/C. bacilisporus/C. deuterogattii/C. tetragattii/C. decagattii) species, respectively [1,2]. Species other than C. neoformans/C. gattii have also undergone taxonomic reclassifications and have been redistributed between the Cryptococcus genus and other genera, including Cutaneotrichosporon, Cystofilobasidium, Filobasidium, Hanalea, Holtermanniella, Naganishia, Papiliotrema, Solicoccozyma, Vishniacozyma, and Vanrija. For instance, C. magnus, are performed in urban areas and in domestic animals or captured fauna (particularly in bird markets or zoos) because they search specifically the most clinically significant species; C. neoformans/C. gattii s.l. [46]. On the contrary, we decided to focus on wild fauna samples. This was done to avoid a possible bias related to sampling performed in animals that interact with humans (pets or captive animals) or in an anthropogenized environment (such as an urban area), in which the situation may be different from the wild.

Sampling Centre
Sampling was conducted at the Centre de Protection de la Faune Sauvage, Ligue de Protection des Oiseaux (LPO) Hérault (Centre for Wild Fauna Protection, Bird Protection Association, Hérault, France). This facility is located in Villeveyrac (34560 France, GPS: 43 • 28 47.6 N 3 • 36 49.3 E) and is the departmental center for the care of wild fauna in distress. It receives approximately 3000 animals per year on average from the Hérault, Aude, and Pyrénées Orientales areas (see Supplementary Data S1); the admitted animals are mostly wild birds (88% of animals), followed by mammals (11.7%) and a small number of reptiles (0.3%). Its primary purpose is to care for wild animals that are wounded or in distress and rehabilitate them for release back into the wild. It should be noted that the center does not care for pet animals or wild species that are considered domestic or invasive.

Animal Population Sampling
From May 2017 to July 2019, ten to twenty birds among the individuals who received daily care at the center were randomly chosen every 7 to 10 days for sampling. A total of 669 wild birds belonging to 71 genera and 89 species (+1 undetermined at the species level) were sampled.

Method of Sampling
During the study, each bird to be tested was kept in a separate cage during the care period. Cages were cleaned with sodium hypochlorite between each individual to avoid cross-contamination. The age, species, reason for admission, treatment, and swab method were documented for each individual. Sampling was performed by direct cloacal swabbing when possible. Alternatively, cage swabbing was performed when concerns about the wellbeing and compliance of animals or excessive risks to care personnel made direct cloacal swabbing impossible. Samples were stored in a 0.9% sodium saline solution containing 10 mg/L chloramphenicol at 4 • C. Once 10 to 20 samples (15 on average) were gathered over one week, they were sent the next Monday by postal mail to the Laboratoire de Parasitologie et Mycologie Médicale for processing. The transport delay ranged from 2 to 4 days, with a mean time of 3 days.

Treatment of Samples
Upon reception, samples were treated as per the following protocol adapted from the standardized protocol used by the International Society for Human and Animal Mycology (ISHAM) Cryptococcal Working Group [7]; in brief, they were seeded on 2 Niger Seed agar plates (supplemented with 20 U/mL penicillin, 40 U/mL gentamycin, 0.25 mg/mL chloramphenicol, and 5 mg/mL benomyl) and left to grow separately at 25 • C and 35 • C for 7 days. Each possible Cryptococcus colony, selected by examination based on the phenotypic characteristics and color of the colony, was checked for the presence of a capsule by India ink staining and tested for its urease capability (Biomerieux, Marcy l'Etoile, France). Colonies with a capsule and positivity on the urease test were characterized biochemically with an API ID 32C strip.

DNA Extraction
Genomic DNA was extracted for each strain using a NucleoSpin blood quick extraction kit (Macherey-Nagel Gmb and Co. KG, Duren, Germany) with modifications as previously described [47].

Sequence Analyses
Sequencing of the amplicons was performed by Genewiz (Azenta Life Sciences, Leipzig, Germany) on both strands (forward and reverse) using IT1 and ITS4 primers for sequencing. Sequences were compared to the global NCBI standard nucleotide collection database (nr/nt) by BLAST. Identification was considered acceptable when a hit was obtained with an E-value of 0.0, a query coverage of at least 99%, homology of 100% for one strand, and at least ≥99.45% for the other strand (only if the mismatches occurred in or next to the sequence priming or termination areas). When species discrimination could not be achieved with ITS1-ITS4 fragment sequencing, further amplification with the 28 S large nuclear subunit primers NL1 (5 GCATATCAATAAGCGGAGGAAAAG-3 ) and NL4 (5 -GGTCCGTGTTTCAAGACGG-3 ) [49] was performed and processed as described above.

Statistical Analyses
Fisher's exact test for count data with simulated p-values (based on 2000 replicates) was used to assess whether there were associations of bird species, order, age, clinical condition, and treatment with positivity for RCFC.
The chi 2 test was used to assess whether there was an association between the type of sampling and positivity for RCFC.
The chi 2 test was used to determine whether there was an association between antibiotic treatment, use of a nonsteroidal anti-inflammatory drug (NSAID, meloxicam) or steroidal anti-inflammatory drug (rapidexon), or any combination of these drugs and positivity for RCFC.
The chi 2 test was used to determine if birds' diet had an impact on positivity for RCFC and RCFC species distribution. For the analysis, birds were classified into seven groups according to their main diet at the adult stage: grain eaters, fruit eaters, small vertebrate eaters, insectivores, omnivorous birds, opportunistic feeders, and, finally, others. Grain eaters group corresponds to birds feeding mainly on grain and/or seeds; fruit eaters group corresponds to birds feeding mainly on fruit and/or berries. The small vertebrate eaters group includes birds feeding on a variety of small terrestrial vertebrates, such as mice, voles, small reptiles, and small passerine birds. Insectivore group contains birds feeding on a wide variety of insects and other terrestrial arthropods such as spiders. The omnivorous feeders group contains birds with no preferential diet between plants/grains and animals. Opportunistic feeders group corresponds to birds able to feed on a wide variety of food sources including human wastes. Finally, the other group includes all birds whose diet does not fit into any of the previous groups; specialist feeders, birds feeding on invertebrates other than insects such as annelids, molluscs, or crustaceans, birds feeding on aquatic food sources such as fishes or amphibians, carrion eaters, etc.
Birds with two different eating habits (for example, grain eater but also fruit eater) were counted once in each group for the statistical analysis.

Total Cryptococcal Species Distribution
One hundred and seven of the 669 birds (16%) belonging to 15 different orders and 37 different species were positive for at least one RCFC species. No statistical correlation was found between bird orders or species and positivity for yeasts (p = 0.64 and p = 0.9, respectively) ( Table 1 and Supplementary Data S2).

Assessment of Highly Pathogenic Cryptococcus Species
No species belonging to the highly pathogenic C. neoformans or C. gattii species complexes were found in the bird population sampled.
The distribution of positive bird species can be found in Table 1 and Supplementary Data S2.

Distribution by Age
Ninety-four of the birds positive for RCFC were adults (total adults 486, 19.3%), eight were juveniles (total juveniles 132, 6%) and five were immatures (total immatures 41, 12.2%). A significant (p = 0.0005, Fisher's exact test) association was observed between age class and RCFC positivity in birds. The most significant difference was between adults and juveniles. Indeed, under the null hypothesis of independence, the expected number of positive samples was 77.7 and 21.1 for these respective age classes, whereas the observed number of positive samples was 94 and 8. Therefore, an over-representation of adults and an under-representation of juveniles were observed among the positive birds (Table 3 and Supplementary Data S3).

Distribution by Health Issue
Eleven of 107 birds (11.3%) had a reported health issue, including the presence of lead shot (5/11), an infectious disease (5/11), or electrocution (1/11). The other 96 birds had no reported diseases or wounds. According to Fisher's exact test, there was no significant (p = 0.48) association between any documented health condition and positivity for RCFC. The power of the test was weak, so rare conditions (<5%) were grouped together, but Fisher's exact test remained nonsignificant (p = 0.52). Health conditions were grouped further (physical/trauma ailments or infectious diseases), but there was still no significant correlation with bird positivity (p = 1) ( Table 3 and Supplementary Data S3).
Notably, no bird had documented symptoms of fungal infection, including symptoms of cryptococcal infection (Table 3).

Distribution by Medical Treatment
Fourteen birds positive to RCFCs, including twelve for which no wounds or diseases were reported, had a history of drug administration, mostly a combination of antibiotics (amoxicillin or amoxicillin + clavulanate) and anti-inflammatory drugs (meloxicam or rapidexon). According to Fisher's exact test, there was a nonsignificant (p = 0.89) association between any treatment and positivity for RCFC. However, the power of the test was weak. To increase the power of the test, some rare treatments (<5%) were grouped. However, Fisher's exact test remained nonsignificant (p = 0.36). We thus focused on the treatments that increased susceptibility to yeast carriage (antibiotics, NSAIDs, steroidal anti-inflammatory drugs, and any combination of these three drugs), but again, no statistically significant correlation was found (p = 0.61, p = 0.33, p = 1 and p = 1, respectively) ( Table 3 and Supplementary Data S3).
Chi 2 analysis of positivity for RCFC with regard to feeding habits showed that there was no correlation between the two elements (p = 0.26).
No correlation was found between the feeding habit and the RCFC species distribution among the birds (p = 0.78).   Table 4. Distribution of birds according their main diet at adult stage. Grain eaters indicates birds feeding mainly on grain and/or seeds at adult stage; Fruit eaters indicates birds feeding mainly on fruit and/or berries at adult stage; Small vertebrate eaters indicates birds feeding on a variety of small terrestrial vertebrate preys such as mice, voles, small reptiles and small birds at adult stage; Insectivore indicates birds feeding on a wide variety of insects and other terrestrial arthropods such as spiders at adult stage; Omnivorous feeders indicates birds with no preferential diet between plants/grain and animals; Opportunistic feeders indicates birds with a wide variety of food sources that also scavenge on human wastes; Other group indicates birds with a diet that does not fit into any of the previous groups. This includes specialist feeders, birds feeding on invertebrates other than insects such as annelids, molluscs or crustaceans, birds feeding on aquatic food sources such as fishes or amphibians. Birds with several main diets (for example grain eater and fruit eaters) were counted once in each category.

RCFC Diversity
Studies of the carriage of Cryptococcus and species formerly belonging to the Cryptococcus genus in free-ranging wild birds are rare because they are technically challenging as capture needs to be done without harming the animal. Additionally, a significant number of wild birds are considered protected species in Europe. To circumvent these challenges and to assess whether wild birds can act as carriers or spreaders of the most pathogenic Cryptococcus species (Cryptococcus neoformans/gattii s.l.) in southern France, we sampled animals admitted for care at a care center managed by the French Bird Protection League. No yeasts belonging to the highly pathogenic C. neoformans/C. gattii s.l. complexes were found during this study. This seems contradictory to various studies suggesting that southern France, where we performed the sampling, should harbor these two species [50,51] but is in accordance with literature reporting that isolation of C. neoformans in wild birds is rare [52]. On the other hand, the RCFC diversity in the bird population we sampled was high, with twenty different species isolated. Among them, one species (S. aeria) has never been isolated from any animal source. Five others, namely, Cryptococcus ovalis, N. chernovii, H. festucosa, F. oeirense, and F. wieringae, were previously isolated from various infected or asymptomatic mammals, such as cats, dogs, or horses, but have not been isolated from any birds [40,53,54]. The fourteen remaining species had already been isolated from bird sources but mostly from pigeon droppings [55][56][57][58] and captive or pet birds [59][60][61][62]. In fact, very few wild birds have been described as possible sources for RCFC. N. adeliensis and F. magnum were previously isolated from a cormorant [63]. N. albida was isolated from a number of wild ducks and teals [63]. P. laurentii was also isolated from a very limited number of migratory birds, Canadian goose feces, a bird of prey, and a house sparrow [64][65][66]. N. globosa was isolated from a bird belonging to the Charadriiforme order [67]. By contrast, we found that RCFC species were distributed in fifteen bird orders, thirty-four bird genera, and thirty-seven bird species. Thus, most of the RCFC species identified in this study have not been previously associated with the source birds described herein.
This lack of information about RCFC carriage in birds could be explained in two ways. First, a significant number of RCFC species are considered to be rare. Furthermore, most studies on Cryptococcus focus on the most medically important species; C. neoformans and C. gattii s.l., and thus limit themselves to the known reservoirs of these species i.e., pigeon droppings in urban areas and tree hollows [46]. Very few studies expand sampling to a wider variety of birds, and when doing so, it is mainly limited to captive or pet birds [59]. However, urban environments and captivity constitute specific situations that can influence the microbial composition of the animals tested and thus lead to possible bias. Indeed, population density, environmental stress, and food access in cities or in captivity are very different from those in the wild. It has been shown that proximity to humans and human activities influence the number and variety of yeasts present in bird feces [68]. It has also been described that nutritional status influences bird susceptibility to various pathogens [69]. Finally, it has been shown that bird density, as well as the cage and environmental hygiene, play significant roles in the presence of microbial pathogens in bird feces and/or in the shedding of pathogens by animals [70,71]. Therefore, focusing on pigeons in urban areas (or on captive or pet birds) may lead to a lack of data on the non-neoformans/non gattii RCFC possible reservoirs. We thus suggest that any study on RCFC yeasts should, when possible, expand sampling beyond usual urban Columbiformes (pigeon droppings and nests) sources.

Distribution of RCFC According to Bird Orders, Species, Diet, Clinical Data, Age and Type of Sampling
No significant correlations between order, species, or feeding habit/diet and RCFC carriage were found in this study. RCFC species distribution shows that the same yeast species can be shared by different carrier birds with different ecological niches and feeding habits. For example, N. albida was present in birds of prey, carrion eaters, opportunistic feeders, grain eaters, and insectivorous birds. N. diffluens was present in small vertebrate eaters, fish eaters, insectivores, and omnivorous and opportunistic birds, such as magpies and jackdaws. F. magnum was found in birds of prey, grain eaters, berry eaters and fish eaters. This result suggests that RCFC species may be present through the whole trophic chain from plants to invertebrates [72] to free-ranging small vertebrates [73] that serve as bird food. Alternatively, it is possible that bird carriage of RCFC species is not linked to their feeding regimen but to contact with an external source of contamination, such as decomposing plant material, where some RCFC can be found [74].
There was no significant correlation between recorded health issues or medical treatment and RCFC carriage. In particular, the intake of drugs, including antibiotics, NSAIDs and steroid anti-inflammatory drugs, which are known to facilitate fungal colonization, had no influence on sample positivity in our study.
On the other hand, a correlation between yeast carriage and bird age was found. Adult birds were over-represented in the RCFC-positive animals, while juveniles were under-represented. Studies on the association between Cryptococcus sp. (or species formerly belonging to the Cryptococcus genus) carriage and animal age are rare. When performed in cats and koalas, age was not a determinant of carriage [40,75]. It has been, however, shown that colonization with other fungi (mainly Candida krusei and Candida albicans) increases with age in house sparrows [66], suggesting a cumulative risk of carriage with age. Our data seem to be in accordance with the latter result. However, this needs confirmation because our cohort of aged birds was small, with only ten individuals labeled as such and none of them positive for RCFC.
A correlation between the method of sampling and positivity to RCFC was also observed in this study. Indeed, dry cage swabs were over-represented in comparison with wet cage and direct cloacal swabs among positive birds. This is comparable with what is known about the growth of the most studied Cryptococcus species; Cryptococcus neoformans s.l.: (i) the growth of these yeasts are inhibited in a medium undergoing alkalinization due to bacterial growth, such as wet droppings [76]; and (ii) dry droppings have a higher concentration of key nutrients, such as urea and creatinine, and are thus supposedly a more favorable growth medium for the yeasts [77]. On the other hand, RCFC isolates were also found in wet droppings, while C. neoformans/C. gattii s.l. species have almost never been isolated from such media [78]. The fact that some RCFC species, particularly Naganishia spp., seem to be able to survive in environments poor in carbon and nitrogen and with an alkaline or acidic pH, situations that would be hostile to C. neoformans/C. gattii s.l., may explain this last observation [79].

The Particular Case of Direct Cloacal Swabs
In our study, fourteen different RCFC species were isolated from eighteen birds sampled by direct cloacal swabs. This excludes the possibility of ground or cage contamination of the samples and suggests that the yeasts may come from the digestive tract of the animals. Notably, two species, C. ovalis, and S. aeria, were found by such method of sampling, suggesting that they can pass through birds' digestive system, while they were never isolated from any bird source before. However, none of C. ovalis or S. aeria isolates found in this study grew at a temperature of 35 • C after seeding on Niger seed agar plates, but only at 25 • C. One possible explanation is that the yeasts originated in the environment and were accidentally present on the feathers around the cloaca of the birds. According to this hypothesis, detection by cloacal swabbing may be due to contamination of the swabs by contact with colonized feathers. To test this hypothesis, attempts to grow these two species at 30 • C, a temperature that is below the bird's skin temperature, were made. S. aeria, but not C. ovalis, showed (poor) growth at 30 • C, suggesting a possible bird feathers origin for this species. On the other hand, no S. aeria was found by sampling methods inducing a risk of contamination from the environment (cage swabs). The two different samples positive for S. aeria obtained during this study were found only by cloacal swabs, raising questions about the true colonization site in birds: feathers or digestive tract, for this species. The twelve other species isolated from bird cloaca (F. chernovii, F. floriforme, F. magnum, F. wieringae, N. adeliensis, N. albida, N. liquefasciens, N. randhawae, N. uzbekistanensis, P. flavescens, V. carnescens) have all been previously isolated from pigeon sources or from other pet/captive birds but only from cages, nests, or dry ground droppings. The fact that these species were found in cloacal samples suggests an ability to survive in the birds' digestive tract too. However, as previously observed for C. ovalis, none of the isolates showed thermotolerance in culture. Thus far, we have no explanation for the discrepancy between species retrieved from a source with a high-temperature body and the lack of thermotolerance exhibited in culture unless the aforementioned species share the same escape mechanisms described for C. neoformans. Indeed, it was shown that C. neoformans can pass through the complete intestinal system of birds [80][81][82]. While this passage requires adaptation to the high body temperature of birds, it is possible for a minority of C. neoformans cells by vomocytosis or cell enlargement [83]. However, to our knowledge, no such mechanisms were yet described for any RCFC. Nevertheless, the proportion of animals positive for RCFC by cloacal swab appears high, particularly in comparison with what can be achieved for C. neoformans/C. gattii s.l. For these species' complexes, positivity by cloacal swab is rare in literature and ranges from 0% to 4.8% at most [52], while we observed 16.8% RCFC positivity in the bird population we studied. Furthermore, some cloacal swab samples were positive for two different RCFC, suggesting that different species can survive simultaneously in the same bird digestive system. Infection by up to three different Cryptococcus neoformans/gattii s.l. species in the same patient has already been described in humans [47]; but the carriage of multiple RCFC species in the digestive tract of the same carrier (or reservoir) animal has never been described so far.

Risk Assessment for Humans
Concerning the risk assessment for humans, including caregivers in protection and sheltering centers, no highly pathogenic species, such as C. neoformans s.l., were found in this study, even in Columbiformes (pigeons and doves) that are known reservoirs for this pathogen. Isolation of C. neoformans has been documented in the literature in a limited number of other birds than pigeons, such as Anas crecca, A. platyrhinchos [52,63], Falco tinnunculus [64], and various passerine birds [66,84]. However, the positive samples obtained in these studies came in fact from dead birds, birds in an aviary, or birds captured on buildings in an urban area or in villages. Although the same bird species are present in our bird cohort, they were negative for C. neoformans in our study. Very few other studies were performed in exclusively wild populations, and they did not detect C. neoformans either [85]. These observations may cast doubt on the presence of C. neoformans s.l. in wild fauna, where stress, population concentration, range, and access to food are different from those in captured birds or flocks of birds living in urbanized or anthropogenized areas. Nevertheless, this study shows that the risk of exposure to highly pathogenic Cryptococcus species in nonurban human populations and bird caretakers is limited. For the less pathogenic RCFC species, particularly those already known to infect humans [5], the risk may be elevated in care personnel working in aviaries or wild fauna protection centers. The risk may also be present for peri/suburban human populations because residential gardens are attractive to some carrier birds, particularly Passeriformes. However, given that RCFC rarely causes clinically significant infections in humans [86], the overall probability of infection seems very low but not impossible.
Conclusion: In this study, owing to the diversity of our study population, we elucidated the relationship between various bird species and the carriage of RCFC species. A large number of the reported relationships of RCFC species and the birds they were isolated from have not been previously described in the literature. Thus, this study may help to clarify the role of wild birds as ecological niches for rare Cryptococcus and species formerly belonging to the Cryptococcus genus. No sequences or strains were submitted to any database or library. However, they are accessible upon request.