Cryptococcus Gattii Isolates Chapter 7

PDF hosted at the Radboud Repository of the Radboud University Nijmegen

The following full text is a publisher's version.

For additional information about this publication click this link. http://hdl.handle.net/2066/128930

Please be advised that this information was generated on 2021-10-10 and may be subject to change. María Teresa Illnait Zaragozí Illnait Teresa María studies and in vitro Cryptococcus neoformans Cryptococcus complex: molecular, clinical clinical molecular, complex:

Cryptococcus neoformans complex: molecular, clinical and in vitro studies M.T. Illnait Zaragozí 2014 Cryptococcus neoformans complex: molecular, clinical and in vitro studies ISBN: 978-94-6203-627-7

Printed by CPI Wöhrmann Print Service, Zutphen, The Netherlands. Layout by Ferry Hagen.

The work described in this thesis was performed in several periods from 2007-2012 at the Canisius-Wilhelmina Hospital, Nijmegen, The Netherlands

Cover: Front: Almond tree, Havana, Cuba Back: Mango tree and pigeon, Havana, Cuba Pictures: María Teresa Illnait Zaragozí Cryptococcus neoformans complex: molecular, clinical and in vitro studies

Proefschrift

ter verkrijging van de graad van doctor aan de Radboud Universiteit Nijmegen op gezag van de rector magnificus prof. mr. S.C.J.J. Kortmann, volgens besluit van het college van decanen in het openbaar te verdedigen op woensdag 27 augustus 2014 om 10.30 uur precies

door

María Teresa Illnait Zaragozí geboren op 10 maart 1967 te Havana, Cuba Promotoren Prof. dr. A. Voss Prof. dr. P.E. Verweij

Copromotoren Dr. J.F.G.M. Meis (Canisius-Wilhelmina Ziekenhuis, Nijmegen) Dr. F. Hagen (Canisius-Wilhelmina Ziekenhuis, Nijmegen)

Manuscriptcommissie Prof. dr. M.G. Netea Prof. dr. B.E. de Pauw Prof. dr. G.S. de Hoog (CBS-KNAW Fungal Biodiversity Centre, Utrecht)

Paranimfen Dr. Ph.A. Van Damme N. Manraad Cryptococcus neoformans complex: molecular, clinical and in vitro studies

Doctoral Thesis

to obtain the degree of doctor from Radboud University Nijmegen on the authority of the Rector Magnificus prof. dr. S.C.J.J. Kortmann, according to the decision of the Council of Deans to be defended in public on Wednesday, August 27, 2014 at 10.30 AM

María Teresa Illnait Zaragozí Born on March 10, 1967 in Havana, Cuba Supervisors Prof. dr. A. Voss Prof. dr. P.E. Verweij

Co-supervisors Dr. J.F.G.M. Meis (Canisius-Wilhelmina Ziekenhuis, Nijmegen) Dr. F. Hagen (Canisius-Wilhelmina Ziekenhuis, Nijmegen)

Doctoral Thesis Committee Prof. dr. M.G. Netea Prof. dr. B.E. de Pauw Prof. dr. G.S. de Hoog (CBS-KNAW Fungal Biodiversity Centre, Utrecht)

Paranymphs Dr. Ph.A. Van Damme N. Manraad Sometimes we feel that what we do is just a drop in the sea, but the ocean would be less without a drop.

Mother Teresa of Kolkata

Table of contents

Chapter 1. General Introduction 11

Part I Molecular characterization of clinical and environmental C. neoformans species complex isolated from Cuba and The Netherlands

Chapter 2. Microsatellite typing of clinical and environmental 35 Cryptococcus neoformans var. grubii isolates from Cuba shows multiple genetic lineages

Chapter 3. Extensive genetic diversity within the Dutch clinical 43 Cryptococcus neoformans population

Chapter 4. Microsatellite typing and susceptibilities of serial 53 Cryptococcus neoformans isolates from Cuban patients with recurrent cryptococcal meningitis

Part II Antifungal susceptibility of Cuban and worldwide cryptococcal strains

Chapter 5. In vitro activity of the new azole isavuconazole (BAL 61 4815) compared with six other antifungal agents against 162 Cryptococcus neoformans isolates from Cuba

Chapter 6. In vitro antifungal susceptibilities and amplified 65 fragment length polymorphism genotyping of a worldwide collection of 350 clinical, veterinary, and environmental Cryptococcus gattii isolates Chapter 7. Cryptococcus neoformans - Cryptococcus gattii 73 species complex: an international study of wild- type susceptibility endpoint distributions and epidemiological cutoff values for fluconazole, itraconazole, posaconazole, and voriconazole

Part III Clinical and environmental studies

Chapter 8. Environmental isolation and characterization of 83 Cryptococcus species from living trees in Havana city, Cuba

Chapter 9. Reactivation of a Cryptococcus gattii infection in a 91 cheetah (Acinonyx jubatus) held in the National Zoo, Havana, Cuba

Chapter 10. Fatal Cryptococcus gattii genotype AFLP5 infection in 97 an immunocompetent Cuban patient

Chapter 11. Cryptococcus gattii induces a cytokine pattern that is 103 distinct from other cryptococcal species

Chapter 12. General discussion and future perspectives 115

Chapter 13. Summary - Samenvatting - Resumen 133

Chapter 14. Acknowledgments 145 Curriculum vitae List of publications Chapter General Introduction 1 Chapter 1

The incidence of cryptococcosis, initially considered unusual, began to increase with the arrival of acquired immunodeficiency syndrome (AIDS) in the 1980s to become one of the most important mycoses worldwide. After the advent of highly effective antiretroviral therapy in the mid-1990, its frequency started to decline in developed countries. However, cryptococcal meningitis remained the most common life- threatening opportunistic fungal infection with an estimated annual mortality of 625,000 persons, which ranks cryptococcosis as fourth among the most common causes of death due to infections in sub-Saharan Africa, higher than tuberculosis. (Armstrong et al. 2014; Park et al., 2009; La Hoz and Papas, 2013). Additionally, it has been reported that up to 6% of people with impaired cellular immunity due to other causes, are at risk of developing clinically apparent infections due to Cryptococcus (Bratton et al., 2012).

From discovery of the causative agent of cryptococcosis to its current taxonomy Probably the first case of cryptococcosis was descrited in 1861 by Zenker. However because of the lack of microbiological evidence, the first case description is usually attributed to doctors Otto Busse and Abraham Buschke from Germany. In 1894 they reported the isolation of a microorganism from an injury in a 31-year-old woman and named it Saccharomyces hominis. Almost simultaneously, Sanfelice described the presence of encapsulated yeasts obtained from fermented peach juice; he demonstrated its pathogenicity in laboratory animals and called it Saccharomyces neoformans. Sanfelice shortly afterwards recovered a similar agent from the lymph node in an ox and called it Saccharomyces lithogenes because there were particular differences between this organism and the first one (Casadevall and Perfect, 1998, Bovers et al., 2008a). A year after the first cryptococcosis case report by Busse and Buschke, Curtis Ferdinand in France, described a “plant parasite” yeast as an etiologic agent of an ulcerating abscess in a young Frenchman with an unremarkable medical history. Based on different morphological features in comparison with the previous described isolates (elongated or oval yeasts) and its affinity for subcutaneous tissue shown in animal models, Curtis labelled this agent Saccharomyces subcutaneous tumefaciens. Despite the absence of epidemiological data, it is currently considered the first report of an autochthonous infection by C. gattii in France (Hagen et al., 2012; Kwon-Chung et al., 2011).

These findings demonstrated three cardinal aspects: i) the ability to cause infection in humans and in animals, ii) this, added to the demonstration of Koch’s postulates by Sanfelice and Curtis, showed its pathogenic potential and iii) the fact that it can be recovered from natural sources revealed its saprophytic nature. However, the different denominations received yielded confusion about the true nature of this agent (Casadevall and Perfect, 1998; Bovers et al., 2008a). The genus Cryptococcus was first established in 1833 by Kützing to “accommodate” a new microorganism described as “globose hyaline microscopic organism with undetermined taxonomic position that produces whitish mucoid colonies”. Later, in 1901, Vuillemin argued that the microorganisms isolated by Busse and Sanfelice in 1894 actually belong to this genus. Thereafter, its definition underwent multiple

12 General Introduction proposed amendments until the gender name Cryptococcus Vuillemin and pointed C. neoformans as the type species. Nowadays, the genus includes about 100 pathogenic and non-pathogenic species (Kwon-Chung et al., 2011). 1 One of the main contributions to the current taxonomy of C. neoformans, was the recognition of its sexual stage. In 1975 the first teleomorph, Filobasidiella neoformans, was obtained by mating compatible isolates of serotypes A and D. The second species, Filobasidiella bacillispora was achieved by crosslinking compatible cells of serotypes B and C (Kwon-Chung, 1975). The correlation between the different serotypes and teleomorph states indicated that differences between isolates were not restricted to their antigenic properties. Subsequent studies showed that both species differ in their ecological niches, epidemiology, biochemistry, association with different host immune states, phylogeny and genomic arrangements (Kwon-Chung and Varma, 2006; Kwon- Chung et al., 2011).

Studies about this agent and the disease it causes, led to its current classification. What once was considered a “complex species”, is now recognized as a “species complex” (Fonseca et al., 2011). The F. bacillispora anamorph state which comprises serotypes B and C was raised to the upper level and named Cryptococcus bacillisporus; later on this denomination was replaced by C. neoformans var. gattii and it is currently recognized as the species C. gattii. On the other hand, the F. neoformans anamorph state (serotype A, D and AD), named as C. neoformans was reclassified asC. neoformans var. neoformans, however, based on it phylogenetic characteristics, this nomenclature changed as well. The latter variety today has been raised to species level that comprises C. neoformans var. grubii corresponding to serotype A, C. neoformans var. neoformans corresponding to serotype D and the intervarietal state C. neoformans corresponding to serotype AD (table 1) (Kwon-Chung and Varma, 2006; Ngamskulrungroj et al., 2009).

Table 1. Accepted nomenclature for the C. neoformans/C. gattii species complex Division Basidiomycota Class Hymenomycetes Order Tremellales Family Tremellaceae Genus Filobasidiella Teleomorph Anamorph F. bacillispora C. gattii (serotype B and C) F. neoformans C. neoformans Species, C. neoformans C. neoformans C. neoformans varieties var. grubii var. neoformans intervarietal state and (serotype A) (serotype D) (serotype AD) serotypes

13 Chapter 1

Cryptococcosis – What is it and what is the cause? Cryptococcosis, formerly known as Busse and Buschke disease, European blastomycosis or torulosis is the infection caused by the encapsulated yeast C. neoformans species complex which are ubiquitous microorganisms able to affect humans and animals. Although infection mostly occurs after inhalation of infective particles, the most frequent clinical presentation is not a pneumonia, but meningitis and disseminated disease (Lin and Heitman, 2006). Nearly 100 species have been described within the genus Cryptococcus, the C. neoformans/C. gattii species complex being primary responsible for cryptococcal infections in both humans and animals (Fonseca et al., 2011). The literature includes a budding number of case reports on infections due to Cryptococcus albidus and Cryptococcus laurentii, as the etiological cause of approximately 80% of infections due to other species different from C. neoformans species complex (Khawcharoenporn et al., 2007). Other members of this genus occur less frequently and are usually related to severe immunosuppression (Casadevall and Perfect, 1998; Khawcharoenporn et al. 2007; Kwon-Chung et al., 2011).

Correct identification of Cryptococcus different from the C. neoformans/C. gattii species complex has high clinical relevance, as they might have a different pattern of susceptibility to antifungal agents (Pfaller et al., 2009). However, for practical purposes, C. neoformans/C. gattii species complex is considered to be the only one that consistently produces disease (Kwon-Chung et al., 2011).

Yeasts belonging to the C. neoformans/C. gattii species complex are characterized by 4 to 6 μm rounded-oval cells, surrounded by a polysaccharide capsule of variable size: isolates recovered from environmental sources have a virtual or unapparent capsule (≤ 4 μm) while those from clinical sources used to show a wide capsule (≥ 30 μm). Based on antigenic differences in their structure, seven serotypes were identified: A, B, C, D, AD, AB and BD that, together with the biochemical, ecological, epidemiological and genetic divergences, permitted discrimination of the existing species from varieties that set up the complex (Casadevall and Perfect, 1998; Kronstad et al., 2011).

What are the sources of it causative agents? The C. neoformans/C. gattii species complex comprises of free-living saprophytic yeasts which can survive in diverse niches. Numerous studies have been carried out in order to achieve a better understanding of its life cycle in nature (Casadevall and Perfect, 1998; Randhawa et al., 2010; Kwon-Chung et al., 2011).

C. neoformans and C. gattii have been isolated ubiquitous worldwide from variable natural sources (soil, air, water, raw milk and its derivatives, citrus, vegetables such as green beans, beets, cabbage, radishes, lettuce , tomatoes and some tubers including potatoes). However, poultry manure, especially from domestic pigeons (Columba livia) has been identified as the most notorious source of infection-related isolates (Casadevall and Perfect, 1998; Cabral, 1999; Kwon-Chung et al., 2011). In the accumulated faeces protected from the sun’s ultraviolet rays up to 2-3 x 106 cells/g can be detected, which

14 General Introduction can survive for more than a year (Hamasha et al., 2004). The low concentration of C. neoformans in the peak, belly, legs and intestines of these birds suggests that pigeons are not its main reservoir in nature. These facts combined with the increased C. neoformans 1 reports from more than 22 tree species of different families and genera indicated that plants could be the primary ecological niche (Lazera et al., 2000; Randhawa et al., 2008; Randhawa et al., 2010). This hypothesis is supported by the potential of the yeast to degrade lignin which serves as growth substrate by using its phenoloxidase enzyme (Lazera et al., 2000).

On the other hand, C. gattii used to have a more restricted distribution to tropical and subtropical climates. However, more recently, the outbreak at the west coast of North America indicates a drastic change in the adaptation of this organism to other environments (Chowdhary et al., 2012a; Galanis and Macdougall, 2010; Hagen et al., 2013; Kidd et al., 2007; MacDougall et al., 2011). Until 1990 the ecological niche of this species was a mystery when the Australian researchers Ellis and Pfeiffer described its association with Eucalyptus camaldulensis. They confirmed the presence of basidiospores, suspended in the air near these trees and suggested that in this reservoir sexual reproduction with subsequent propagule dissemination to the environment could occur (Ellis and Pfeiffer, 1990). Subsequent reports have shown that this association is not limited to Eucalyptus trees or a particular geographic area contributing to a better understanding of the diversity of its natural sources that, interestingly, do not include poultry manure or contaminated soil (Chowdhary et al. 2012b; Ellis, 1987; Hamasha et al., 2004; Kidd et al., 2007; Lazera et al., 2000).

How do we become sick from these yeasts? Circumstantial evidence suggests that infection by the C. neoformans/C. gattii species complex usually develops after inhalation of infectious propagules (Velagapudi et al., 2009). This can also occur in various species of animals but proof of person to person and animal to person transmission or laboratory-acquired disease are scarce. Other routes are less common and include organ transplants from infected donors and traumatic cutaneous inoculation (Török et al., 2010). Serologic studies in the open population indicate a high prevalence of infections due to this yeast, which seems to be acquired already from childhood onwards. Despite this, whether a person/individual will became sick or not depends on three main factors: the state of the host defence mechanisms, the virulence of the infecting strain and the inoculum count (Casadevall, 2010).

Once inhaled, the infectious agent reaches the lungs where it meets the first line of defence that consists of nonspecific phagocytic effector cells: alveolar macrophages, neutrophils, natural killer cells and lymphocytes. Macrophages phagocytose the yeasts, express on their surface the processed antigen bound to the major histocompatibility complex class II and present it to CD4+ T cells responsible for triggering a specific immune response. In most immunocompetent subjects the response is able to eliminate the infection and the microorganism does not extend beyond the pulmonary

15 Chapter 1 focus with little or no symptoms. However, the ability of these cells to remove the yeast is limited. By consequence a large inoculum can overcome this defence mechanism allowing fungal multiplication and dissemination (Casadevall et al., 1993; Olszewski et al., 2010). Cryptococcosis is more common in individuals with impaired cell-mediated immunity, being HIV infection the most important predisposing condition (Byrnes et al., 2009; Galanis and MacDougall, 2010). Other well known risk factors are steroid treatment, solid organ transplantation, chronic obstructive pulmonary disease, diabetes, lymphoma, leukemia, sarcoidosis, cirrhosis and connective tissue diseases (e.g. lupus erythematosus and rheumatoid arthritis) (Baddley et al., 2011; Bratton et al., 2012; La Hoz and Pappas, 2013; Pappas, 2013). Besides, the hypothesis that an intact immune system is able to control the infection has been challenged. On one hand, documented cases prior to the AIDS epidemic, show that this agent is able to cause disease in individuals without apparent predisposing factors (Mitchell and Perfect, 1995) while, on the other hand, it has been widely demonstrated that C. gattii species is the main responsible agent of cryptococcosis in HIV-negative patients (Casadevall, 2010).

Unlike potent toxins released by pathogenic bacteria, the C. neoformans/C. gattii species complex does not seem to produce abundant products that directly contribute to the clinical manifestations of the infection. Cryptococcal pathogenesis largely depends on the integrity of the various host defence components in the compromised immune system (Casadevall et al., 2003; Idnurm et al., 2005; Kozubowski and Heitman, 2011).

The polysaccharide capsule and its constituents as well as melanin and extracellular enzymes (urease, superoxide dismutase, proteases, esterases and lipases) are among the most studied virulence factors of C. neoformans/C. gattii species complex that contribute to the seriousness of the disease (Rodriguez et al., 2009; Kronstad et al., 2011). More recently additional factors such as morphologic changes (development of filaments during mating as well as the expansion of the capsule and cell size in response to specific host factors during infection) and biofilm formation have been recognized that amplify the pathogenic power of these yeasts (Martínez et al. 2008; Lin, 2009; Kozubowski and Heitman, 2011; Kronstad et al., 2011).

Both, C. neoformans and C.gattii share the main adaptation and virulence mechanisms, which corroborates the classification of both species as true pathogens. Extrapulmonary dissemination is invariably associated with disease, meningoencephalitis being the most common clinical presentation. The basis of C. neoformans/C. gattii species complex neurotropism are not clear, but this propensity could be associated with the abundant presence of catecholamines (excellent substrates for the phenoloxidase enzyme) and the selective evasion of host defences. Protein molecules such as antibodies and complement factors, essential for efficient fungal opsonization, are effectively shut out by the blood-brain barrier. It indeed has been suggested that phagocytic cells resident in the central nervous system (CNS) are effective against the fungus, but their activity is critically depend on opsonins (Feldmesser et al., 2000; Casadevall, 2010; Charlier et

16 General Introduction al., 2010; Liu et al., 2012).

C. neoformans and C. gattii are able to survive in the hostile phagosome environment. 1 The development of these structures with large numbers of yeasts inside, suggests that early stages of the replication of the organism does occur within these vesicles subsequent to the escape through a non-destructive removal process (Johnston and May, 2010). In fact, such an expulsion is preceded by the phago-lysosome fusion, which seems to be coordinated by or dependent on the pathogen, since this phenomenon does not occur in the presence of dead yeasts (Alvarez and Casadevall, 2006). The development of giant phago-lysosomes led to speculate that C. neoformans/C. gattii species complex is able of altering the cytoplasmic membrane properties of these organelles, the pathogen to prevent death of the host cell and the activation of the inflammatory response. Furthermore, it has been demonstrated that yeasts can be transported between macrophages, a process that may contribute to the persistence or latency of the pathogen in the host (Ma et al., 2007; Johnston and May, 2010). Several investigations assessed the mechanism of persistence of Cryptococcus. Olszewski et al. provided a full summary of this complex phenomenon. According to these researchers, a small number of yeasts may persist for long periods or even forever within host tissues. In this case, the modulatory effects on the innate and adaptive immune response, the high adaptability to extreme conditions, and the development of tolerance to treatment, although insufficient to promote disease in most immunocompetent hosts, explains persistence and reactivation or relapse of subclinical infection (Olszewski et al., 2010).

What are the clinical manifestations of infection? The clinical picture is related to the affected organ. Although the load of yeasts at the affected sites is significantly higher in immunocompromised patients, the clinical manifestations are similar regardless of the host immune status. However, infections due to C. gattii can be distinguished from C. neoformans given a higher tendency to produce tumorous lesions and neurological involvement because of a slower response to treatment (Török et al., 2010; Pappas, 2013).

After more than 100 years of clinical experience, it has become clear that lungs are the first organ that the C. neoformans/C. gattii species complex face, but most cases stay asymptomatic. Clinical manifestations, if any, consist of coughing, fever, pleural pain, malaise and weight loss. Unlike infections with Histoplasma capsulatum, tissue calcification occurs exceptionally. This feature, combined with the unavailability of reliable intradermal tests, makes the diagnosis of infection due to this agent difficult (Harrison 2009; Török et al., 2010).

Most cases remain undiagnosed unless neurological signs and symptoms, usually quite insidiously appear. A wide range of CNS symptoms can be present, headache and fever (38-39 °C) being the most prominent. Since the clinical presentation is nonspecific, it is important to consider the diagnosis of cryptococcal meningitis in every HIV-positive

17 Chapter 1 patient with fever. Supplemented by monitoring for this possibility at regular intervals, even when the initial investigations are negative because brain and meningeal infection is the most frequent cause of death due to cryptococcal infection (Casadevall and Perfect, 1998; Harrison, 2009). C. neoformans can invade a single site or multiple organs simultaneously during dissemination. Other sites such as skin and prostate are often affected; the latter is considered an important yeast reservoir in males. (Casadevall and Perfect, 1998; Török et al., 2010).

How can the infection be established? Microbiological diagnosis of cryptococcosis is based on direct microscopic examination of a clinical sample as well as isolation and identification of the yeast using biochemical and serological tests. However, these methods have some disadvantages (Huston and Mody, 2009). Phenotypic methods for example, are laborious, time consuming and sometimes imprecise because of the subjectivity in result interpretation (Wengenack and Binniker, 2009). Automatic and semi-automatic methods such as API 20C AUX and Vitek respectively allow obtaining results within approximately 72 h. These tests, however, have their own limitations as they depend on complementary tests for final identification (Gündeş et al., 2001; Massonete et al., 2004).

Matrix-Assisted Laser Desorption Ionization-Time-of-Flight Mass Spectrometry (MALDI-TOF MS) is a rapid identification tool for the correct recognition of the two currently recognized human pathogenic Cryptococcus species and offers a simple method for the separation of the eight major molecular types and the detection of hybrid strains within this species complex in the clinical laboratory (Firacative et al., 2012; McTaggart et al., 2011; Stevenson et al., 2010).

For serological testing, false positive results due to cross-reactions with rheumatoid factor and Trichosporon spp. or false negative results because a pro-zone phenomenon have been described (Huston and Mody, 2009). In 2009, a lateral flow immunoassay (LFA) for the detection of cryptococcal antigen was developed by IMMY (ImmunoMycologics, U.S.A.) as a potential test for the diagnosis of cryptococcal infection (Jarvis et al., 2011). This system seems to be stable at room temperature, has a rapid turnaround time, requires very little technical skill, and can be performed with minimal laboratory infrastructure (McMullan et al., 2012). Beside these advantages, it has been shown that it is a sensitive test as compared with the gold standard (culture), and it has a high level of overlap with enzyme immunoassay (EIA) (Lindsley et al., 2011). However agreement between titres, which was more often observed in sera with low EIA optical density values was imperfect possibly reflecting the presence of low levels of circulating antigen (Lindsley et al., 2011; McMullan et al., 2012).

Although not suited for routine diagnosis, molecular methods are available for the detection of specific genetic sequences ofC. neoformans/C. gattii species complex from specimens and cultures (Bovers et al., 2007; Wengenack and Binnicker, 2009). These tools provide identification of C. neoformans/C. gattii species complex and other less

18 General Introduction frequently encountered species such as C. laurentii, and offer more complete results than conventional diagnostic procedures. They contribute to a better understanding of epidemiology and natural history of cryptococcosis. These tests can also be used to 1 determine the presence of certain genotypes in a given population, allowing the study of the species diversity in the environment (Leaw et al., 2006). Some of the most used techniques are described below:

DNA-DNA hybridization: Based on the intrinsic property of the complementarity of the nitrogenous bases. It is a powerful tool for the development of specific primers currently used in research centers to identify genetic variations among isolates from different geographical areas (Hu et al., 2008, Costa et al., 2010).

Electrokaryotyping: This was the first technique used for the chromosomal analysis of C. neoformans. It consists in the separation of large DNA fragments through their re-orientation on agarose gel under the action of electric alternating fields. Lebens and Polacheck were the first to visualize C. neoformans and C. gattii chromosomes by this system and revealed that these two species differ in the number of chromosomal bands. Due to high complexity its current use in C. neoformans studies is very limited (Polacheck and Lebens, 1989).

Polymerase chain reaction (PCR): Among the many tools to manipulate and gather data from DNA, this technique facilitates the production of multiple copies of small, specific segments of the genome. This technique accelerates the diagnosis of cryptococcosis with satisfactory specificity and sensitivity by detecting minimum amounts of specimens. The main advantages are i) rapidly and easily executable, ii) less expensive than hybridization, iii) can be fully automated, iv) applicable to contaminated samples or mixed cultures and v) can be used in combination with other techniques making it a useful tool for molecular epidemiology studies. Widely used variants for C. neoformans and C. gattii identification are conventional PCR and multiplex (real-time) PCR. The most commonly used target sequences are CAP59, IGS1, PLB1, URA5, M13 and ITS (18S, 5.8S and 28S). Particularly due to its high variability and the presence of about 100 copies per haploid genome, the latter are the most widely used for genomic sequence studies (Hsu et al., 2003; Pryce et al., 2006).

PCR-fingerprinting is based on the detection of hypervariable repetitive DNA sequences. It is considered to be one of the fastest molecular techniques for the detection of band patterns. It has been applied to confirm a particular molecular type in sporadic cases or for molecular epidemiological studies. By using short primers founded in arbitrary sequences it is not necessary to have prior knowledge of the sequence. As for many molecular methods, PCR fingerprinting has an enhanced discrimination power compared to biochemical and serological tests. Its high sensitivity and specificity for detecting differences allows rapid differentiation ofC. neoformans and C. gattii isolates. Within the field of fungal taxonomy the ribosomal DNA targets 18S, 5.8S and 28S are frequently used, but for the C. gattii/C. neoformans species complex several other

19 Chapter 1 genetic markers have been applied to study the genetic diversity and include CAP59, CAP64, IGS1, LAC1, PBL1, URA5. Studies based on genetic fingerprinting facilitate the classification of clinical and environmental isolates in eight major molecular types, but an exact correlation between serotype and molecular types of C. gattii does not exist (Meyer et al., 2003, 2009; Chen et al., 2008; Hagen et al., 2010; Hagen et al., 2012; Mueller and Wolfenbarger, 1999; Kidd et al., 2007).

Random amplification of polymorphic DNA (RAPD) is a technique that involves the amplification of genomic DNA by simple arbitrary primers. It detects polymorphisms without knowledge about the genome. This method allows the differentiation among varieties, serotypes and molecular types of major pathogenic species of the genus. It is used to determine cryptococcal genetic profiles from different sources of strains obtained from patients in a particular geographic region or from different anatomical sites. At present, it use has been replaced by newer methods that have a higher resolution (García-Hermoso et al., 1999; Hagen et al., 2010; Hagen et al., 2012; Litvintseva et al., 2005).

Restriction fragment length polymorphism (RFLP) or PCR-RFLP can be applied for the rapid determination of molecular types of C. gattii and C. neoformans. First, a specific region of interest is amplified, for C. gattii/C. neoformans the PLB1 and URA5-loci are widely used. The basic difference between both techniques is that PCR-RFLP needs specific primers that amplify the genetic region of interest followed by an enzymatic digestion. Molecular epidemiological studies have revealed that PCR-RFLP can be applied to determine the relationship between clinical and environmental cryptococcal isolates (Kidd et al., 2007; Meyer et al., 2003; Meyer et al., 2009).

Amplified Fragment Length Polymorphism (AFLP) is one of the most advanced techniques to efficiently obtain fingerprint data from large numbers of cryptococcal isolates. It is more efficient than PCR-RFLP, combining high resolution and high sampling power. Additionally, it is excellent in the detection of genetic variability as it explores the presence or absence of polymorphisms in restriction sites in greater detail. By this method it is possible to group C. neoformans and C. gattii species complex in thirteen molecular types and to compare the obtained results with other typing techniques (e.g. serotyping). Among its main limitations are the multiple stages involved in the process, expensive reagents and laboratory equipment and requirement of high quality DNA. Recently it was demonstrated that isolates with certain AFLP genotypes are associated with increased virulence (Mueller and Wolfenbarger, 1999; Boekhout et al., 2001; Barreto de Oliveira et al., 2004; Enache-Angoulvant et al., 2007; Bovers et al., 2008b; Leal et al., 2008; Litvintseva and Mitchell, 2009; Byrnes et al., 2009; Hagen et al., 2010; Hagen et al., 2012).

Multi-locus sequence typing (MLST) is currently one of the important tools for global epidemiological and population biological studies. It is based on the variation of the nucleotide sequence of multiple genes such as those encoding the capsule, urease,

20 General Introduction phospholipase, and laccase production. It is remarkably valuable for determining circulating genotypes in different parts of the world. Its advantages include a high reproducibility, and the possibility to perform analysis through specific software that 1 can be exchanged between laboratories. Moreover, the generated data can be used to differentiate isolates and to investigate evolutionary relationships. Recently the International Society for Human and Animal Mycology published a consensus MLST scheme for genotyping of C. neoformans and C. gattii (Litvintseva et al., 2006; Hagen et al., 2012; Meyer et al., 2009; Urwin and Maiden, 2003).

Short tandem repeat (STR) analysis was initially developed for forensic and paternity studies but the advantages favoured its expansion into other research fields. STR is actually the analysis of DNA fragment sizes of DNA-repeats that have a length of one to six nucleotides. The variation in the number of repeats creates different allele sizes. STRs are usually found in non-coding DNA regions and have a high mutation rate, making it highly polymorphic. As for MLST, STR has highly reproducible and readily comparable results intra-and inter-laboratory (Hanafy et al., 2008; de Valk et al., 2009).

Currently, molecular typing of microorganisms plays a much broader role than simply performing epidemiological studies. The development of these techniques introduced new possibilities to the fields of taxonomy, identification and diagnosis, as well as to that of clarification of the phylogeny and evolution. The accumulation of data generated by these methods is expected to have a positive impact on the control of resistant isolates and better management of the disease (Costa et al., 2010; Kwon-Chung et al., 2011).

What are the options for the management of the infection? Cryptococcal infection treatment has been one of the most studied among mycosis diseases, but there are still many unresolved aspects. Despite medical advances, the mortality rate during the first three months of treatment for neurocryptococcosis exceeds 20% and approaches 100% in patients who do not receive proper management (Perfect et al., 2010). Papers published on this subject agree that the application of an induction phase, among the evaluated schemes, is the most effective followed by consolidation and a maintenance phase (Perfect et al., 2010; Török et al., 2010).

The recommended treatment during the induction phase is based on a combination of two or more drugs. Amphotericin B (0.7 to 1 mg/kg/d) and 5-fluorocytosine (100 mg/kg/day) during 2 weeks are first choice. Alternatives are liposomal amphotericin B formulations and in cases of unavailability of, or intolerance, to any of the listed drugs, fluconazole (800-400 mg/d) or itraconazole (400 mg/d) (Perfect et al., 2010). It is recommended to use 8 weeks of fluconazole (400 mg/d) during the consolidation phase. In the next phase the dose can be halved for not less than a year or as long as patients exhibit CD4+ lymphocyte counts ≤ 200 cells/mL. Itraconazole (400 mg/d) and amphotericin B (1 mg/kg/week), as alternatives have also been evaluated (Perfect et al., 2010).

21 Chapter 1

Reduction or cessation of steroid treatment when possible, administration of interferon gamma and intrathecal or intraventricular amphotericin B, as well as surgical excision of the lesions, have been used for rescue of therapeutic failures (Perfect et al., 2010; Török et al., 2010). The optimal strategy for the management of intracranial pressure is not well established, but there is some evidence that the drainage of cerebrospinal fluid (CSF) by various methods produces the best results (Casadevall and Perfect, 1998; Török et al., 2010). Nevertheless, the high incidence of therapeutic failures and relapses hint at the lack of an optimum approach to control this infection. A wide array of therapeutic options, ranging from immunization to assessment of new antifungal compounds, have been tested (Mondon et al., 1999; Espinel-Ingroff, 2003; Thompson et al., 2008; Pfaller et al., 2009; Thompson et al., 2009; Saylo et al., 2010; Chow and Casadevall, 2011). Although some recently developed antifungals already have been licensed for use in humans, it remains problematic to treat some patients effectively. Therefore it is necessary to continue the search for alternative therapies. A new generation of azoles such as isavuconazole are currently in different phases of development (Guinea and Bouza, 2008; Fera et al., 2009). The antifungal spectrum of isavuconazol appears to be similar to that of voriconazole and posaconazole with potent in vitro activity against Candida spp., Aspergillus spp., Scedosporium spp., Blastomyces dermatitidis, Histoplasma capsulatum, zygomycetes as well as against less frequent, pigmented fungi (Exophiala spp., and Phialophora spp.). Moreover, cross-resistance with fluconazole has not been reported (Hagen et al., 2010; Thompson et al., 2009; Thompson and Wiederhold, 2010; Yamazaki et al., 2010). In addition, the pharmacokinetic profile of this drug has advantage over that of other antifungals allowing oral and parenteral administration (Thompson and Wiederhold, 2010). This drug is currently tested in clinical phase III trials of therapy for candidosis and aspergillosis (Guinea and Bouza, 2008; Guinea et al., 2010).

Testing the in vitro activity of antifungal compounds With the increase in number of available antifungals, a need for reliable methods for determining antifungal susceptibility testing has become apparent. Susceptibility testing became important in medical mycology when the AIDS epidemic began. The use of fluconazole, generally given at subtherapeutic doses for the treatment of oropharyngeal candidosis, led to the emergence of resistance (Tapia, 2009).

The first reports of resistance were questioned on the basis of the use of different methodologies with low intra- and interlaboratory reproducibility. Subsequent work showed the importance of adjusting the test conditions (inoculum, culture medium composition, pH, temperature and incubation time) because of their respective impact on inhibitory concentration values (MIC).

The first optimized and standardized method for the study of yeasts susceptibility was a macrodilution technique. This was developed by the National Committee for Clinical Laboratory Standards (NCCLS), now the Institute for Clinical Laboratory

22 General Introduction

Standards (CLSI). Afterward, this method was adapted to a microdilution plate format, which currently is the most widely accepted reference test (Johnson et al., 2008; CLSI, 2008a). The European Committee on Antimicrobial Susceptibility Testing (EUCAST) 1 has promoted modifications, which in practical terms made the broth microdilution test less laborious, easier to interpret (especially with isolates showing trailing) and faster, with definitive results at 24 h for Candida spp. and 48 hours for Cryptococcus spp. Although according to published studies both methods correlate well, it should be taken into account that they have their own individual standard cutoff (Espinel-Ingroff et al., 2005; Rodriguez-Tudela et al., 2007; Rodriguez-Tudela et al., 2010). Sensitire YeastOne® (TREK Diagnostic Systems Ltd., UK) is the closest approach to the reference methodology (Johnson et al., 2008).

Etest (Biodisk AB, Sweden) was recently developed and is one of the most attractive alternatives currently on the market. It is a simple method based on the application of a plastic strip with a gradient concentration of antifungal on a previously inoculated agar surface with the isolate for testing. After incubation the inhibition zone indicates the MIC as the point where the inhibition ellipse intersects the strip. Once the operator gains experience in reading and interpreting, the test is a reliable and reproducible method that correlates well with the reference method (Johnson et al., 2008; Tapia et al., 2009). Detection of resistant species usually should be checked by a second method, which constitutes a drawback of this method. This is particularly important for Candida tropicalis isolates and Cryptococcus spp. as a result of relatively poor growth on the recommended medium. The use of Etest also provides a much wider range of MIC values than those obtained by other methods. Hence, susceptible isolates tend to show lower MIC values while less susceptible isolates tend to have higher MIC values than those obtained by reference methods. The technique has been standardized for yeasts and filamentous fungi and is currently available for amphotericin B, 5-fluorocytosine, fluconazole, itraconazole, voriconazole, posaconazole and caspofungin (Johnson et al., 2008).

Disk diffusion is a simple and economical method to test the susceptibility of water- soluble antifungals such as 5-fluorocytosine, fluconazole and voriconazole. This is described by the CLSI standards for application and interpretation and is more useful for Candida (CLSI, 2008c; Johnson et al., 2008). The test provides an inhibition zone diameter that correlates with MIC values obtained by the reference method based on the evaluation of thousands of different isolates (Pfaller et al., 2004). Addition of methylene blue to the culture medium (0.5 mg/mL) gives a sharper demarcation of the inhibition zone that improves readability of the test (CLSI, 2008c). Susceptibility tests may predict the clinical response to treatment. However, within the framework of invasive fungal infections, there are still a large number of possibly ambiguous factors when one is trying to interpret the in vitro activity of antifungal agents. Pharmacodynamic and pharmacokinetic analyses in conjunction with the identification of resistance mechanisms have been useful in establishing cutoff values for at least some of the systemic antifungal agents (Tapia, 2009).

23 Chapter 1

Due to the relatively low frequency of resistant isolates and the multiple concurrent factors involved in invasive fungal infections, it is difficult to validate cutoff values. The interaction between antifungal drugs - host, fungus - host and drug - fungus is a complex process, so that the in vitro susceptibility tests constitutes only a small part of the information needed to predict treatment outcome.

Aim of this thesis The Cryptococcus neoformans/Cryptococus gattii species complex is a saprophytic encapsulated globally distributed yeast. Although the primary route of cryptococcal- host entry is through the inhalation of fungal conidia and/or desiccated yeast forms, meningitis is the commonest and more severe clinical form of disease. About one million new cases are reported annually and despite appropriate antifungal therapy, mortality remains high with death rates of up to 70% in HIV/AIDS patients resulting in at least 625 000 deaths per year. Understanding the ecology, epidemiology and pathogenesis of C. neoformans/C. gattii species complex as well as its susceptibility profile to old and new antifungal compounds may contribute to a better management and control of this important mycosis. With these aims we did the following investigations.

Part I includes three chapters (2-4) that deal with the application of molecular biological techniques to the study of clinical and environmental C. neoformans/C. gattii species complex isolates from Cuba and the Netherlands. Chapter 2 describes a novel approach for high resolution molecular subtyping of C. neoformans var. grubii based on nine microsatellite markers selected from the available genomic sequences of H99 strain. The panel of markers was applied to 190 Cuban Cryptococcus isolates, which included 122 clinical strains, all collected from different patients between 1987 and 2007 (mainly from cerebrospinal fluid representing almost every province of Cuba) and 68 environmental isolates from pigeon guano gathered in the period 1998 to 2007 from different locations across the island. AFLP was used to confirm the identification of the isolates as C. neoformans var. grubii, and to validate the typing result of the microsatellite analysis. Chapter 3 describes the study of 300 Dutch C. neoformans isolates, obtained from 237 patients during 1977 to 2007. For this purpose, relevant information about the strains was obtained from the database of the Netherlands Reference Laboratory for Bacterial Meningitis (Academic Medical Center, Amsterdam, The Netherlands). Four categories were determined based on the predisposing factors that may have influenced the onset of the Cryptococcus infection: HIV/AIDS-related immunodeficiency, non- HIV/AIDS-related immunodeficiency (e.g., immunosuppressive therapy and organ transplantation), immunocompetent individuals, and a category for which the presence or absence of predisposing factors was unknown. Isolates were molecularly characterized with PCR (mating- and serotype), AFLP and microsatellite. Correlation between genotype and susceptibility to seven antifungal compounds was investigated. In chapter 4, 19 C. neoformans var. grubii isolates from seven Cuban patients with recurrent infection were studied using the microsatellite set of nine markers previously mentioned. Isolates were recovered at different time intervals from cerebrospinal fluid (n = 16), blood, urine and semen (one each). Patients (6 HIV-positive and 1 HIV-

24 General Introduction negative) were admitted at Tropical Medicine Institute “Pedro Kourí” between 1995 and 2001, just before HAART was introduced in Cuba. Additionally, mating- and serotyping as well as in vitro susceptibility to amphotericin B, flucytosine, fluconazole, 1 itraconazole, voriconazole, posaconazole and isavuconazole by broth microdilution method were performed.

Part II encloses also three chapters (5-7) mainly devoted to antifungal susceptibility studies. Chapter 5 describes the largest study up to now on antifungal susceptibility of Cuban C. neoformans/C. gattii species complex isolates collected over 25 years. It was also the first study to report the in vitro susceptibility of C. neoformans var. grubii to isavuconazole. Chapter 6 is dedicated to the study of a worldwide collection of 350 Cryptococcus gattii isolates representing clinical (n = 215; 61.4%), veterinary (n = 57; 16.3%), environmental (n = 73; 20.9%) and unknown sources (n = 5; 1.4%). According to their geographical origin, the strains were grouped as Africa (n = 28; 8.0%), Asia (n = 22; 6.3%), Australia (n = 20; 5.7%), Europe (n = 101; 28.9%), North America (n = 95; 27.1%), South America (n = 81; 23.1%) and unknown origin (n = 3; 0.9%). The North American group contained 75 (24.1%) C. gattii isolates related to the Vancouver Island outbreak. Clusters obtained using AFLP fingerprinting, were correlated with the antifungal susceptibility profile for seven antifungal drugs including isavuconazole. Chapter 7 is based on an international study of wild-type (WT) susceptibility endpoint distributions and epidemiological cutoff values (ECVs) of fluconazole, itraconazole, posaconazole and voriconazole. The aggregated MIC distributions of the four triazoles (2,985 to 5,733 MICs for C. neoformans and 705 to 975 MICs for C. gattii gathered in 15 to 24 laboratories) were used. The ECV was based on MIC distributions (obtained in at least three laboratories) and defined as the highest WT cutoff susceptible value. In the absence of clinical breakpoints the obtained values may help in the detection of isolates with acquired resistance mechanisms.

Part III (chapers 8-11) involves the study of possible sources of infection from trees, the description of the firstC. gattii isolates in Cuba, as well as some of the possible immune mechanisms involved in host defense against C. neoformans and C. gattii infections. In chapter 8, different species of trees in Havana, Cuba were sampled to identify potential sources of C. neoformans/C. gattii species complex. A total of 662 samples were collected from 331 trees and cacti (65 Indian almond trees, 65 flamboyant trees, 65 Ficus spp. (mainly Indian laurel trees), 60 cacti of different species, 60 casuarins, five blue mahoe trees, three ceiba, two seagrape, and one each of copey, yellow oleander, yellow flamboyant, wallnut, caribbean pine and roble. Initial selection of the isolates was carried out by conventional techniques and further characterized using a combination of AFLP and DNA sequencing analysis. Chapters 9 and 10 describe the first veterinary and human cases of cryptococcosis due C. gattii in Cuba. In both settings the strains were identified using macro- and micro-morphological characteristics. The strains were subsequently studied by PCR (for mating- and serotype), AFLP as well as MLST using the standard loci (CAP59, GPD1, IGS1, LAC1, PLB1, SOD1), URA5, CAP10, MPD1 and TEF1 to compare the results with the data of other isolates with the same genotype.

25 Chapter 1

A bootstrap Maximum Likelihood phylogenetic analysis was performed using the settings as being used for a previously published C. gattii MLST study. Additionally, susceptibility testing using a broth microdilution method was performed in accordance with CLSI document M27-A3 guidelines. Chapter 11 highlights differences in the cytokine profile of human peripheral blood mononuclear cells of healthy individuals after in vitro stimulation with 40 different well-defined heat-killed isolates of C. gattii, C. neoformans and several hybrid isolates obtained from different sources. In addition, the involvement of TLR2, TLR4 and TLR9 in the pro-inflammatory cytokine response to C. gattii was investigated. Finally, the main results from this thesis are integrated and discussed in chapter 12 with suggestions for future research.

26 General Introduction References 1 Álvarez M, Casadevall A. Phagosome extrusion and host-cell survival after Cryptococcus neoformans phagocytosis by macrophages. Curr Biol 2006; 16: 2161-5.

Armstrong JD, Meintjes G, Brown GD. A neglected epidemic: fungal infections in HIV/AIDS. Trends Microbiol 2014; 22: 120-7.

Arikan S. Current status of antifungal susceptibility testing methods. Med Mycol 2007; 45: 569- 87.

Baddley JW, Schain DC, Gupte AA, Lodhi SA, Kayler LK, Frade JP, Lockhart SR, Chiller T, Bynon JS Jr, Bower WA. Transmission of Cryptococcus neoformans by organ transplantation. Clin Infect Dis 2011; 52: e94-8.

Barreto de Oliveira MT, Boekhout T, Theelen B, Hagen F, Baroni FA, Lazera MS, Lengeler KB, Heitman J, Rivera IN, Paula CR: Cryptococcus neoformans shows a remarkable genotypic diversity in Brazil. J Clin Microbiol 2004, 42:1356-9.

Boekhout T, Theelen B, Diaz M, Fell JW, Hop WC, Dromer F, Meyer W. Hybrid genotypes in the pathogenic yeast Cryptococcus neoformans. Microbiology 2001; 147: 891-907.

Bovers M, Diaz MR, Hagen F, Spanjaard L, Duim B, Visser CE, Hoogveld HL, Scharringa J, Hoepelman IM, Fell JW, Boekhout T. Identification of genotypically diverse Cryptococcus neoformans and Cryptococcus gattii isolates by Luminex xMAP technology. J Clin Microbiol. 45: 1874-83.

Bovers M, Hagen F, Boekhout T. Diversity of the Cryptococcus neoformans-Cryptococcus gattii species complex. Rev Iberoam Micol 2008(a); 25: S4-12.

Bovers M, Hagen F, Kuramae EE, Boekhout T. Six monophyletic lineages identified within Cryptococcus neoformans and Cryptococcus gattii by multi-locus sequence typing. Fungal Genet Biol 2008(b); 45: 400-21.

Bratton EW, El Husseini N, Chastain CA, Lee MS, Poole C, Stürmer T, Juliano JJ, Weber DJ, Perfect JR. Comparison and temporal trends of three groups with Cryptococcosis: HIV-infected, solid organ transplant, and HIV-negative/non-transplant. PLoS One 2012; 7: e43582.

Byrnes EJ III, Li W, Lewit Y, Perfect JR, Carter DA, Cox GM, Heitman J. First reported case of Cryptococcus gattii in the Southeastern USA: Implications for travel-associated acquisition of an emerging pathogen. PLoS ONE 2009; 4: 1-12.

Cabral LF. Wood, animals and human beings as reservoirs for human Cryptococcus neoformans infection. Rev Iberoam Micol 1999; 16: 40-2.

Casadevall A, Perfect JR. Cryptococcus neoformans. ASM Press, Washington DC 1998.

27 Chapter 1

Casadevall A, Spitzer ED, Webb D, Rinaldi MG. Susceptibilities of serial Cryptococcus neoformans isolates from patients with recurrent cryptococcal meningitis to amphotericin B and fluconazole. Antimicrob Agents Chemother 1993; 37: 1383-6.

Casadevall A, Steenbergen JN, Nosanchuk JD. ‘Ready made’ virulence and ‘dual use’ virulence factors in pathogenic environmental fungi-the Cryptococcus neoformans paradigm. Curr Opin Microbiol 2003; 6: 332-7.

Casadevall A. Cryptococci at the brain gate: break and enter or use a Trojan horse? J Clin Invest 2010; 120: 1389-92. Charlier C, Nielsen K, Daou S, Brigitte M, Chretien F, Dromer F. Evidence of a role for monocytes in dissemination and brain invasion by Cryptococcus neoformans. Infect Immun 2010; 77: 120-7.

Chen J, Varma A, Díaz MR, Litvintseva AP, Wollenberg KK, Kwon-Chung KJ. Cryptococcus neoformans strains and infection in apparently immunocompetent patients, China. Emerg Infect Dis 2008; 14: 755-62.

Chow SK, Casadevall A. Evaluation of Cryptococcus neoformans galactoxylomannan-protein conjugates as vaccine candidate against murine cryptococcosis. Vaccine 2011; 29: 1891-98.

Chowdhary A, Randhawa HS, Boekhout T, Hagen F, Klaassen CH, Meis JF. Temperate climate niche for Cryptococcus gattii in Northern Europe. Emerg Infect Dis. 2012; 18: 172-4 (a).

Chowdhary A, Randhawa HS, Prakash A, Meis JF. Environmental prevalence of Cryptococcus neoformans and Cryptococcus gattii in India: an update. Crit Rev Microbiol. 2012;38:1-16 (b).

Clinical and Laboratory Standards Institute (CLSI). Methods for antifungal disk diffusion susceptibility testing of yeasts: Approved standard M44-A. NCCLS, Wayne, PA, USA, 2008 (c).

Clinical and Laboratory Standards Institute (CLSI). Reference method for broth dilution antifungal susceptibility testing of yeasts. Approved standard, 3rd ed. M27–A3. Clinical and Laboratory Standards Institute, Wayne, PA, 2008 (a).

Clinical and Laboratory Standards Institute (CLSI). Reference method for broth dilution antifungal susceptibility testing of yeasts; 3th Informational Supplement documenent M27-S3. Clinical and Laboratory Standards Institute, Wayne, PA, 2008 (b).

Costa JJ, Cordeiro AK, Freire RA, Cordeiro RA, Nogueira RS, Araújo FE, Collares DS, de Araújo MP, Gadelha MF. Molecular methods for the diagnosis and characterization of Cryptococcus: a review. Can J Microbiol 2010; 56: 445-58. de Valk HA, Meis JF, Bretagne S, Costa JM, Lasker BA, Balajee SA, Pasqualotto AC, Anderson MJ, Alcázar-Fuoli L, Mellado E, Klaassen CH. Interlaboratory reproducibility of a microsatellite-based typing assay for Aspergillus fumigatus through the use of allelic ladders: proof of concept. Clin Microbiol Infect 2009; 15: 180-7. Ellis DH, Pfeiffer TJ. Natural habitat of Cryptococcus neoformans var. gattii. J Clin Microbiol 1990; 28: 1642-4.

28 General Introduction

Ellis DH. Cryptococcus neoformans var. gattii in Australia. J Clin Microbiol 1987; 25: 430-1.

Enache-Angoulvant A, Chandenier J, Symoens F, Lacube P, Bolognini J, Douchet C, Piorot JL, 1 Hennequin C. Molecular identification of Cryptococcus neoformans serotypes. J Clin Microbiol 2007; 45: 1261-5.

Espinel-Ingroff A, Barchiesi F, Cuenca-Estrella M, Pfaller MA, Rinaldi M, Rodríguez-Tudela JL, Verweij PE. International and multicenter comparison of EUCAST and CLSI M27-A2 broth microdilution methods for testing susceptibilities Candida spp. to fluconazole, itraconazole, posaconazole, and voriconazole. J Clin Microbiol 2005; 43: 3884-9.

Espinel-Ingroff A. In vitro antifungal activities of anidulafungin and micafungin, licensed agents and the investigational triazole posaconazole as determined by NCCLS methods for 12,052 fungal isolates: review of the literature. Rev Iberoam Micol 2003; 20: 121-36.

Feldmesser M, Kress Y, Novikoff P, Casadevall A. Cryptococcus neoformans is a facultative intracellular pathogen in murine pulmonary infection. Infect Immun 2000; 68: 4225-37.

Fera MT, La Camera E, De Sarro A. New triazoles and echinocandins: Mode of action, in vitro activity and mechanisms of resistance Expert Rev Anti Infect Ther 2009; 7: 981-98.

Firacative C, Trilles L, Meyer W. MALDI-TOF MS Enables the rapid identification of the major molecular types within the Cryptococcus neoformans/C. gattii species complex. PLoS ONE 2012; 7: e37566.

Fonseca A, Boekhout T, Fell JW. Cryptococcus Vuillemin (1901). In: Kurtzman CP, Fell JW, Boekhout T. (Eds.) The yeasts: a taxonomic study. Elsevier Science & Technology, Amsterdam, 2011; p: 1665-741.

Galanis E, Macdougall L. Epidemiology of Cryptococcus gattii, British Columbia, Canada, 1999- 2007. Emerg Infect Dis 2010; 16: 251-7.

García-Hermoso D, Janbon G, Dromer F. Epidemiological evidence for dormant Cryptococcus neoformans infection. J Clin Microbiol 1999; 37: 3204-9.

Guinea J, Bouza E. Isavuconazole: a new and promising antifungal triazole for the treatment of invasive fungal infections. Future Microbiol 2008; 3: 603-15.

Guinea J, Hagen F, Pelaez T, Boekhout T, Tahoune H, Torres-Narbona M, Bouza E. Antifungal susceptibility, serotyping, and genotyping of clinical Cryptococcus neoformans isolates collected during 18 years in a single institution in Madrid, Spain Med Mycol 2010; 48: 942-8.

Gündeş SG, Gulenc S, Bingol R. Comparative performance of Fungichrom I, Candifast and API 20C AUX systems in the identification of clinically significant yeasts. J Med Microbiol 2001; 50: 1105-10.

Hagen F, Ceresini PC, Polacheck I, Ma H, van Nieuwerburgh F, Gabaldón T, Kagan S, Pursall ER, Hoogveld HL, van Iersel LJ, Klau GW, Kelk SM, Stougie L, Bartlett KH, Voelz K, Pryszcz LP, Castañeda E, Lazera M, Meyer W, Deforce D, Meis JF, May RC, Klaassen CH, Boekhout

29 Chapter 1

T. Ancient dispersal of the human fungal pathogen Cryptococcus gattii from the Amazon rainforest. PLoS ONE. 2013; 8: e71148.

Hagen F, Colom MF, Swinne D, Tintelnot K, Iatta R, Montagna MT, Torres-Rodriguez JM, Cogliati M, Velegraki A, Burggraaf A, Kamermans A, Sweere JM, Meis JF, Klaassen CH, Boekhout T. Autochthonous and dormant Cryptococcus gattii infections in Europe. Emerg Infect Dis. 2012; 18: 1618-24.

Hagen F, Illnait MT, Bartlett KH, Swinne D, Geertsen E, Klaassen CH, Boekhout T, Meis JF. In vitro antifungal susceptibilities and AFLP genotyping of a worldwide collection of 350 clinical, veterinary and environmental Cryptococcus gattii isolates. Antimicrob Agents Chemother 2010; 54: 5139-45.

Hamasha AM, Yildiran ST, Gonlum A, Saracli MA, Doganci L. Cryptococcus neoformans varieties from material under the canopies of Eucalyptus trees and pigeon dropping samples from four major cities in Jordan. Mycopathologia 2004; 158: 195-9.

Hanafy A, Kaocharoen S, Jover-Botella A, Katsu M, Iida S, Kogure T, Gonoi T, Mikami Y, Meyer W. Multilocus microsatellite typing for Cryptococcus neoformans var. grubii. Med Mycol 2008; 46: 685-96.

Harrison TS. The burden of HIV-associated cryptococcal disease. AIDS 2009; 23: 531-2.

Hsu MC, Chen KW, Lo HJ, Chen YC, Liao MH, Lin YH, Li SY. Species identification of medically important fungi by use of real-time LightCycler PCR. J Med Microbiol 2003; 52: 1071-6.

Hu G, Liu I, Sham A, Stajich JE, Dietrich FS, Kronstad JW. Comparative hybridization reveals extensive genome variation in the AIDS-associated pathogen Cryptococcus neoformans. Genome Biol 2008; 9(2): R41.

Huston SM, Mody CH. Cryptococcosis: an emerging respiratory mycosis. Clin Chest Med 2009; 30: 253-64.

Idnurm A, Bahn YS, Nielsen K, Lin X, Fraser JA, Heitman J. Deciphering the model pathogenic fungus Cryptococcus neoformans. Nat Rev Microbiol 2005; 3: 753-64.

Jarvis JN, Percival A, Bauman S, Pelfrey J, Meintjes G, Williams GN, Longley N, Harrison TS, Kozel TR. Evaluation of a novel point-of-care cryptococcal antigen test on serum, plasma, and urine from patients with HIV-associated cryptococcal meningitis. Clin Infect Dis 2011; 53: 1019-23.

Johnson E, Espinel-Ingroff A, Szekely A, Hockey H, Troke P. Activity of voriconazole, itraconazole, fluconazole and amphotericin B in vitro against 1763 yeasts from 472 patients in the voriconazole phase III clinical studies Int J Antimicrob Agent 2008; 32: 511-4.

Johnston SA, May RC. The human fungal pathogen Cryptococcus neoformans escapes macrophages by a phagosome emptying mechanism that is inhibited by Arp2/3 complex- mediated actin polymerisation. PLoS Pathog 2010; 6, e1001041.

Khawcharoenporn T, Apisarnthanarak A, Mundy LM. Non-neoformans cryptococcal infections:

30 General Introduction

A systematic review. Infection 2007; 35: 51-8.

Kidd SE, Chow Y, Mak S, Bach PJ, Chen H, Hingston AO, Kronstad JW, Bartlett KH. 1 Characterization of environmental sources of the human and animal pathogen Cryptococcus gattii in British Columbia, Canada, and the pacific northwest of the United States. Appl Environ Microbiol 2007; 73: 1433-43.

Kozubowski L, Heitman J. Profiting a killer, the development of Cryptococcus neoformans. FEMS Microbiol Rev 2011; Doi:10.1111/j.1574-6976.2011.

Kronstad JW, Attarian R, Cadieux B, Choi J, D’Souza CA, Griffiths EJ, Geddes JM, Hu G, Jung WH, Kretschmer M, Saikia S, Wang J. Expanding fungal pathogenesis: Cryptococcus breaks out of the opportunistic box. Nat Rev Microbiol 2011; 9: 193-203.

Kwon-Chung KJ, Boekhout T, Wickes BL, Fell JW. Systematics of the genus Cryptococcus and its type species C. neoformans. In: Cryptococcus: From human pathogen to model yeast. Heitman J. (ed) ASM Press, Washington, DC. 2011; p: 3-15.

Kwon-Chung KJ, Varma A. Do major species concepts support one, two or more species within Cryptococcus neoformans? FEMS Yeast Res 2006; 6: 574-87.

Kwon-Chung KJ. A new genus, Filobasidiella, the perfect state of Cryptococcus neoformans. Mycologia 1975; 67: 1197-200.

La Hoz RM, Pappas PG. Cryptococcal infections: changing epidemiology and implications for therapy. Drugs 2013; 73:495-504.

Lazera MS, Salmito MA, Londero AT, Trilles L, Nishikawa MM, Wanke B. Possible primary ecological niche of Cryptococcus neoformans. Med Mycol 2000; 38: 379-83.

Leal AL, Faganello J, Bassanesi MC, Vainstein MH. Cryptococcus species identification by multyplex PCR. Med Mycol 2008; 46: 377-83.

Leaw SN, Chang HC, Sun HF, Barton R, Bouchara JP, Chang TC. Identification of medically important yeast species by sequence analysis of the internal transcribed spacer regions. J Clin Microbiol 2006; 44: 693-9.

Lin X, Heitman J. The biology of the Cryptococcus neoformans species complex. Annu Rev Microbiol 2006; 60: 69-105.

Lin X. Cryptococcus neoformans: Morphogenesis, infection, and evolution. Infect Genet Evol 2009; 9: 401-16.

Lindsley MD, Mekha N, Baggett HC, Surinthong Y, Autthateinchai R, Sawatwong P, Harris JR, Park BJ, Chiller T, Balajee SA, Poonwan N. Evaluation of a newly developed lateral flow immunoassay for the diagnosis of cryptococcosis. Clin Infect Dis 2011; 53: 321-5.

Litvintseva AP, Kestenbaum L, Vilgalys R, Mitchell TG. Comparative analysis of environmental and clinical populations of Cryptococcus neoformans. J Clin Microbiol 2005; 43: 556-64.

31 Chapter 1

Litvintseva AP, Mitchell TG. Most environmental isolates of Cryptococcus neoformans var. grubii (serotype A) are not lethal for mice. Infect Immun 2009; 77: 3188-95.

Litvintseva AP, Thakur R, Vilgalys R, Mitchell TG. Multilocus sequence typing reveals three genetic subpopulations of Cryptococcus neoformans var. grubii (serotype A), including a unique population in Botswana. Genetics 2006; 172: 2223-38.

Liu TB, Perlin DS, Xue C. Molecular mechanisms of cryptococcal meningitis. Virulence 2012; 3: 173-81.

Ma H, Croudace JE, Lammas DA, May RC. Direct cell-to-cell spread of pathogenic yeast. BMC Immunol 2007; 8: 15.

Macdougall L, Fyfe M, Romney M, Starr M, Galanis E. Risk factors for Cryptococcus gattii infection, British Columbia, Canada. Emerg Infect Dis 2011; 17: 193-9.

Martínez LR, Ibom DC, Casadevall A, Fries BC. Characterization of phenotypic switching in Cryptococcus neoformans biofilms. Mycopathologia. 2008; 166: 175-80.

Massonet C, van Eldere J, Vaneechoutte M, de Baere T, Verhaegen J, Lagrou K. Comparison of VITEK2 with ITS2-fragment length polymorphism analysis for identification of yeast species. J Clin Microbiol 2004; 42: 2209-11.

McMullan BJ, Halliday C, Sorrell TC, Judd D, Sleiman S, Marriott D, Olma T, Chen SC. Clinical utility of the cryptococcal antigen lateral flow assay in a diagnostic mycology laboratory. PLoS ONE 2012; 7: e49541. Doi: 10.1371/journal.pone.0049541.

McTaggart L, Lei E, Richardson SE, Hoang L, Fothergill A. Rapid identification of Cryptococcus neoformans and Cryptococcus gattii by matrix-assisted laser desorption ionization-time of flight mass spectrometry. J Clin Microbiol 2011; 49: 3050-3.

Meyer W, Aanensen DM, Boekhout T, Cogliati M, Diaz MR, Esposto MC, Fisher M, Gilgado F, Hagen F, Kaocharoen S, Litvintseva AP, Mitchell TG, Simwami SP, Trilles L, Viviani MA, Kwon- Chung J. Consensus multi-locus sequence typing scheme for Cryptococcus neoformans and Cryptococcus gattii. Med Mycol 2009; 47: 561-70.

Meyer W, Castaneda A, Jackson S, Huynh M, Castaneda E. Molecular typing of Iberoamerican Cryptococcus neoformans isolates. Emerg Infect Dis 2003; 9: 189-195.

Mitchell TG, Perfect JR. Cryptococcosis in the era of AIDS -100 years after the discovery of Cryptococcus neoformans. Clin Microbiol Rev 1995; 8: 515-48.

Mondon P, Petter R, Amalfitano G, Luzzati R, Concia E, Polacheck I, Kwon-Chung KJ. Heteroresistance to fluconazole and voriconazole in Cryptococcus neoformans. Antimicrob Agents Chemother 1999; 43: 856-61.

Mueller UG, Wolfenbarger LL. AFLP genotyping and fingerprinting. Trends Ecol Evol 1999; 14: 389-94.

32 General Introduction

Olszewski MA, Hang Y, Huffnagle GB. Mechanisms of cryptococcal virulence and persistence. Future Microbiol 2010; 5: 1269-88.

Pappas PG. Cryptococcal infections in non-HIV-infected patients. Trans Am Clin Climatol Assoc. 2013; 124: 61-79.

Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, Chiller TM. Estimation of the current global burden of cryptococcal meningitis among persons living with HIV/AIDS. AIDS 2009; 23: 525-30.

Perfect JR, Dismukes WE, Dromer F, Goldman DL, Graybill JR, Hamill RJ, Harrison TS, Larsen RA, Lortholary O, Nguyen MH, Pappas PG, Powderly WG, Singh N, Sobel JD, Sorrell T. Clinical practice guidelines for the management of cryptococcal disease: 2010 update by the Infectious Diseases Society of America. Clin Infect Dis 2010; 50: 291-322.

Pfaller MA, Diekema DJ, Gibbs DL, Newell VA, Bijie H, Dzierzanowska D, Klimko NN, Letscher- Bru V, Lisalova M, Muehlethaler K, Rennison C, Zaidi M. Results from the ARTEMIS DISK Global Antifungal Surveillance Study, 1997 to 2007: 10.5-year analysis of susceptibilities of non-Candidal yeast species to fluconazole and voriconazole determined by CLSI standardized disk diffusion testing. J Clin Microbiol 2009; 47: 111-23.

Pfaller MA, Messer SA, Boyken L, Hollis RJ, Rice C, Tendolkar S, Diekema DJ. In vitro activities of voriconazole, posaconazole, and fluconazole against 4,169 clinical isolates of Candida spp. and Cryptococcus neoformans collected during 2001 and 2002 in the ARTEMIS global antifungal surveillance program. Diagn Microbiol Infect Dis 2004; 48: 201-5.

Polacheck I, Lebens GA. Electrophoretic karyotype of the pathogenic yeast Cryptococcus neoformans. J Gen Microbiol 1989; 135: 65-71.

Pryce TM, Palladino S, Price DM, Gardam DJ, Campbell PB, Christiansen KJ, Murray RJ. Rapid identification of fungal pathogens in BacT/ALERT, BACTEC, and BBL MGIT media using polymerase chain reaction and DNA sequencing of the internal transcribed spacer regions. Diagn Microbiol Infect Dis 2006; 54: 289-97.

Rambali B, Fernandez JA, Van Nuffel L, Woestenborghs F, Baert L, Massart DL, Odds FC. Susceptibility testing of pathogenic fungi with itraconazole: a process analysis of test variables. J Antimicrob Chemother 2001; 48: 163-77.

Randhawa HS, Kowshik T, Chowdhary A, Prakash A, Khan ZU, Xu J. Seasonal variations in the prevalence of Cryptococcus neoformans var. grubii and Cryptococcus gattii in decayed wood inside trunk hollows of diverse tree species in north-western India: A retrospective study. Med Mycol 2010. Doi: 10.3109.

Randhawa HS, Kowshik T, Chowdhary A, Preeti Sinha K, Khan ZU, Sun S, Xu J. The expanding host tree species spectrum of Cryptococcus gattii and Cryptococcus neoformans and their

33 Chapter 1 isolations from surrounding soil in India. Med Mycol 2008; 46: 823-33.

Rodríguez ML, Fonseca FL, Frases S, Casadevall A, Nimrichter L. The still obscure attributes of cryptococcal glucuronoxylomannan Med Mycol 2009; 3: 1-7.

Rodríguez-Tudela JL, Arendrup MC, Cuenca-Estrella M, Donnelly JP, Lass-Flörl C. EUCAST breakpoints for antifungals. Drug News Perspect 2010; 23: 93-7.

Rodríguez-Tudela JL, Donnelly JP, Pfaller MA, Chryssantou E, Warn P, Denning DW, Espinel- Ingroff A, Barchiesi F, Cuenca-Estrella M. Statistical analysis of correlation between fluconazole MICs for Candida spp. assessed by standard methods set forth by European Committee on Antimicrobial Susceptibility Testing. J Clin Microb 2007; 45: 109-11.

Saylo CA, Dadachova E, Casadevall A. Murine IgG1 and IgG3 isotype switch variants promote phagocytosis of Cryptococcus neoformans through different receptors. J Immunol 2010; 184: 336-43.

Stevenson LG, Drake SK, Shea YR, Zelazny Am, Murray PR. Evaluation of matrix-assisted laser desorption ionization-time of flight mass spectrometry for identification of clinically important yeast species. J Clin Microbiol 2010; 48: 3482-6.

Tapia CV. Actualización en pruebas de susceptibilidad antifúngica. Rev Chil Infect 2009; 26: 144-50.

Thompson GR III, Wiederhold NP, Fothergill AW, Vallor AC, Wickes BL, Patterson TF. Antifungal susceptibilities among different serotypes of Cryptococcus gattii and Cryptococcus neoformans. Antimicrob Agents Chemother 2009; 53: 309-11.

Thompson GR III, Wiederhold NP. Isavuconazole: a comprehensive review of spectrum of activity of a new triazole. Mycopathologia 2010; 170: 291-313.

Török E, Moran E, Cooke F. Cryptococcus neoformans In Oxoford handbook of infectious diseases and microbiology. Oxford University Press, New York 2010 p. 527-9.

Urwin R, Maiden MC. Multi-locus sequence typing: a tool for global epidemiology. Trends Microbiol 2003; 11: 479-87.

Velagapudi R, Hsueh YP, Geunes-Boyer S, Wright JR, Heitman J. Spores as infectious propagules of Cryptococcus neoformans. Infect Immun. 2009; 77: 4345-55.

Wengenack NL, Binnicker MJ. Fungal molecular diagnostics. Clin Chest Med 2009; 30: 391-408.

Yamazaki T, Inagaki Y, Fujii T, Ohwada J, Tsukazaki M, Umeda I, Kobayashi K, Shimma N, Page MG, Arisawa M. In vitro activity of isavuconazole against 140 reference fungal strains and 165 clinically isolated yeasts from Japan Int J Antimicrob Agents 2010; 36: 324-31.

34 Chapter Microsatellite typing of clinical2 and environmental Cryptococcus neoformans var. grubii isolates from Cuba shows multiple genetic lineages

Published in PLoS ONE 2010;5:e9124. Chapter 2

Microsatellite Typing of Clinical and Environmental Cryptococcus neoformans var. grubii Isolates from Cuba Shows Multiple Genetic Lineages

Maria-Teresa Illnait-Zaragozi1, Gerardo F. Martıńez-Machıń1, Carlos M. Fernańdez-Andreu1, Teun Boekhout2, Jacques F. Meis3, Corne´ H. W. Klaassen3* 1 Instituto Pedro Kouri, Havana, Cuba, 2 Centraalbureau voor Schimmelcultures, Utrecht, The Netherlands, 3 Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, Nijmegen, The Netherlands

Abstract

Background: Human cryptococcal infections have been associated with bird droppings as a likely source of infection. Studies toward the local and global epidemiology of Cryptococcus spp. have been hampered by the lack of rapid, discriminatory, and exchangeable molecular typing methods.

Methodology/Principal Findings: We selected nine microsatellite markers for high-resolution fingerprinting from the genome of C. neoformans var. grubii. This panel of markers was applied to a collection of clinical (n = 122) and environmental (n = 68; from pigeon guano) C. neoformans var. grubii isolates from Cuba. All markers proved to be polymorphic. The average number of alleles per marker was 9 (range 5–51). A total of 104 genotypes could be distinguished. The discriminatory power of this panel of markers was 0.993. Multiple clusters of related genotypes could be discriminated that differed in only one or two microsatellite markers. These clusters were assigned as microsatellite complexes. The majority of environmental isolates (.70%) fell into 1 microsatellite complex containing only few clinical isolates (49 environmental versus 2 clinical). Clinical isolates were segregated over multiple microsatellite complexes.

Conclusions/Significance: A large genotypic variation exists in C. neoformans var. grubii. The genotypic segregation between clinical and environmental isolates from pigeon guano suggests additional source(s) of human cryptococcal infections. The selected panel of microsatellite markers is an excellent tool to study the epidemiology of C. neoformans var. grubii.

Citation: Illnait-Zaragozi M-T, Martıńez-Machıń GF, Fernańdez-Andreu CM, Boekhout T, Meis JF, et al. (2010) Microsatellite Typing of Clinical and Environmental Cryptococcus neoformans var. grubii Isolates from Cuba Shows Multiple Genetic Lineages. PLoS ONE 5(2): e9124. doi:10.1371/journal.pone.0009124 Editor: Francoise Dromer, Pasteur Institute, France Received November 19, 2009; Accepted January 13, 2010; Published February 9, 2010 Copyright: � 2010 Illnait-Zaragozi et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This study was supported by a research and travel grant from the International Society for Human and Animal Mycology to MTIZ. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction associated with alcoholism, organ transplants and immunological disorders. Since the first cases of AIDS in Cuba in 1986, the Cryptococcosis ranks as one of the three common life- number of patients infected by this fungus has increased over the threatening opportunistic infections in persons with AIDS [1]. years. The annual number of infected individuals ranged from 8 to Global estimates indicate more than 900,000 annual cases, causing 15 cases per year [14] while a study of 211 serial autopsies of an estimated 624,700 deaths [2]. Other patient groups with patients with HIV/AIDS infection in Cuba over a period of 10 impaired T-cell function have an up to 6% lifetime risk of years, showed that systemic or central nervous system cryptococ- developing clinically manifest cryptococcosis [3]. Seventy species cosis was a serious and common disorder in 29% of cases [15]. Up belonging to the genus Cryptococcus have been described, but only to now all clinical isolates from patients in Cuba have been members of the C. neoformans complex are mostly associated with identified as C. neoformans var. grubii [16]. human infections. This species complex has been considered to Multiple molecular typing methods have been described to contain two pathogenic species: C. neoformans involving the varieties study the epidemiology of C. neoformans complex. The most neoformans (serotype D) and grubii (serotype A), and C. gattii commonly used approaches to date involve AFLP, PCR (serotypes B and C) [4–7]. Six monophyletic lineages have been fingerprinting and or PCR-RFLP approaches as well as mating- identified that also may represent species [4,5,8] as well as some and/or serotype-specific PCRs [17–23]. These techniques have hybrids [9–12]. proven useful to discriminate between the different sero- and In the Caribbean the disease appears to be not very common mating types but have not been shown to be very useful for with an estimated 7800 patients and 4300 casualties reported discrimination within specific C. neoformans complex members and annually [2]. In Cuba the disease was first reported in the early varieties. Multi-locus sequence typing (MLST) has been applied to 1950s [13]. Since then sporadic cases of cryptococcosis were collections of C. neoformans and C. gattii from various origins

36 Microsatellite typing of Cryptococcus neoformans var. grubii isolates from Cuba

[5,24,25] but this technique is laborious, has a long turn-around repeat markers and 3 tetranucleotide repeat markers was selected time and is associated with significant costs. Microsatellites are from the candidate markers using previously described criteria increasingly popular molecular typing targets since they provide [30]. PCR amplification primers for each of the markers are cost-effective genotyping with fast turn-around times. On a according to Table 1. Three subpanels (CNA2, CNA3 and CNA4 theoretical basis, and, as has been shown for other fungi, respectively) of 3 markers each were amplified using a multicolor molecular typing by using microsatellites is more discriminatory multiplex PCR approach. Within each panel, one of the than by using MLST [26]. Like MLST data, microsatellite typing amplification primers carried a fluorescent label consisting of data is transportable and exchangeable [27]. Here we describe the either FAM (6-carboxyfluorescein), HEX (hexachlorofluorescein) 2 use of a 9-marker microsatellite panel consisting of 3 dinucleotide or TET (tetrachlorofluorescein). In addition to the amplification repeat markers, 3 trinucleotide repeat markers and 3 tetranucle- primers, each 50 ml amplification reaction contained approxi- otide repeat markers. Each panel of 3 markers was amplified using mately 1 ng of genomic DNA, 1 U FastStart Taq DNA a multiplex multicolor PCR approach. Amplified products were polymerase (Roche diagnostics), 2 mM MgCl2 and 0.2 mM analyzed on a high resolution capillary electrophoresis platform dNTP’s in 1x reaction buffer (Roche diagnostics). The amplifica- allowing precise determination of repeat numbers in each marker. tion profile consisted of a 10 min denaturation/activation step We applied this panel to a collection of clinical and environmental followed by 35 cycles of 94uC for 30 s, 60uC for 30 s and 72uC for C. neoformans var. grubii isolates from Cuba. Part of this work was 1 min. After an additional 10 min incubation at 72uC, the presented at the 46th Interscience Conference on Antimicrobial reactions were cooled to room temperature. Agents and Chemotherapy (ICAAC), San Francisco, 2006, Abstr. M904. AFLP Analysis Approximately 50 ng of genomic DNA was subjected to a Materials and Methods combined restriction-ligation procedure containing 5 pmol of EcoR I adapter, 50 pmol Mse I adapter, 2 U of EcoR I (New Ethics Statement England Biolabs, Beverly, MA, USA), 2 U of Mse I (New England This research was approved by the Institutional Scientific and Biolabs) and 1 U of T4 DNA ligase (Promega, Leiden, The Ethical Committee of the Instituto Pedro Kouri, Havana, Cuba. Netherlands) in a total volume of 20 ml of 1x reaction buffer for All data were analyzed anonymously. 1 hour at 20uC. Next, the mixture was diluted five times with 10 mM Tris/HCl pH 8.3 buffer. Adapters were made by mixing Isolates equimolar amounts of complementary oligonucleotides (59- A total of 190 clinical and environmental Cryptococcus isolates CTCGTAGACTGCGTACC-39 and 59-AATTGGTACGCAG- from the collection of the mycology laboratory at the Tropical TC-39 for EcoR I; 59-GACGATGAGTCCTGAC-39 and 59- Medicine Institute ‘‘Pedro Kourı´’’, were included in the study. TAGTCAGGACTCAT–39 for Mse I) and heating to 95uC, Clinical strains (n = 122) were collected between 1987 and 2007, subsequently followed by cooling slowly to ambient temperature. the large majority (91%) of isolates were from cerebrospinal fluid. One microliter of the diluted restriction-ligation mixture was used The remaining isolates were from urine, blood, tissue biopsy and for amplification in a volume of 25 ml under the following bronchoalveolar lavage samples. For ,70% of all clinical isolates, conditions: 1 mM EcoR I primer with two selective residues (59- information about the origin of the patients was available. Most Flu-GTAGACTGCGTACCCGTAC-39), 1 mM MseI primer patients (,65%) were from Havana City. The remaining patients with one selective residues (59-GATGAGTCCTGACTAAG-39), inhabited almost every province of Cuba (Figure 1). Approxi- 0.2 mM each dNTP and 1 U of Taq DNA polymerase (Roche mately 77% of all patients were HIV positive. All clinical isolates Diagnostics) in 1x reaction buffer containing 1.5 mM MgCl2. were from different patients. Environmental isolates (n = 68) were Amplification was done as follows. After an initial denaturation isolated from pigeon guano collected in the period 1998 to 2007 step for 4 min at 94uC in the first 20 cycles a touch down from well distributed locations across Cuba (Figure 1). In case an procedure was applied: 15 s denaturation at 94uC; 15 s annealing environmental sample yielded multiple different colony morphol- at 66uC with the temperature for each successive cycle lowered by ogies suspected of being Cryptococcus, these were analyzed 0.5uC and 1 min of extension at 72uC. Cycling was then separately. continued for further 30 cycles with an annealing temperature of Species identification was initially performed by standard 56uC. After completion of the cycles an additional incubation at mycological methods [28] and confirmed with a commercial 72uC for 10 min was performed before the reactions were cooled identification system (Auxacolor 2; Bio-Rad, Marnes-la-Coquette, to room temperature. The amplicons were then combined with France) as well as by using AFLP analysis. the ET400-R size standard (GE Healthcare, Diegem, Belgium) and analyzed on a MegaBACE 500 automated DNA platform DNA Isolation (GE Healthcare), according to the manufacturer’s instructions. A suspension of freshly grown cells was prepared in lysis buffer (Roche Diagnostics, Almere, The Netherlands) and subjected to Capillary Electrophoresis mechanical lysis in a MagNA Lyser for 30 s at 6500 rpm (Roche Following amplification, the reaction products were diluted 10- Diagnostics). Next, DNA was purified using a MagNAPure LC fold with distilled water. One ml of diluted products was combined instrument in combination with a MagNAPure LC DNA Isolation with 0.25 ml of ET-ROX 550 size marker and 8.75 ml of distilled Kit III as recommended (Roche Diagnostics). DNA yield and water. After a 1 min denaturation step at 94uC, the samples were purity were estimated by UV absorbance measurements. quickly cooled to room temperature and injected onto a MegaBACE 500 automated DNA analysis platform equipped Microsatellite Analysis with a 48 capillary array as recommended by the manufacturer Candidate short tandem repeat markers were identified in the (GE Healthcare). Electropherograms were analyzed using Frag- available genomic sequences from the H99 strain using the ment Profiler 1.2 software (GE Healthcare). Assignment of repeat Tandem Repeats Finder software [29]. A 9 marker microsatellite numbers was relative to the results obtained using the H99 strain, panel consisting of 3 dinucleotide repeat markers, 3 trinucleotide which was used as a control strain in all experiments. According to

37 Chapter 2

Figure 1. Origin of samples and relationships between genotypes. A: Map of Cuba. Colored dots indicate provinces from which clinical and/ or environmental samples were available. B: Minimum spanning tree (MST) based on a multistate categorical analysis representing the genotypes of 190 C. neoformans var. grubii isolates from Cuba. Each circle represents a unique genotype. The size of the circle corresponds to the number of isolates with that genotype. Genotypes are linked to their closest relative. Numbers and connecting lines correspond to the number of different markers between genotypes. Genotypes with identical colors and connected by a shaded background are part of a microsatellite complex (MC). In yellow are unique genotypes that are not part of a MC. C: Same as B, but now showing cross-links between all genotypes that differ in no more than 2 markers. D: Same MST as in B, but now showing genotypes obtained from clinical and environmental samples. doi:10.1371/journal.pone.0009124.g001

38 Microsatellite typing of Cryptococcus neoformans var. grubii isolates from Cuba

Table 1. Basic characteristics of the selected microsatellite markers.

Repeat Labeled primer Unlabeled primer Conc. No. alleles Panel Marker Chr: position unit sequence (59-39) sequence (59-39)* (mM) (range)

CNA2 CNA2A 11: 389201–389350 CT FAM- CGAGGTCATGTTGTGAGTCC GTGACCGTCTCGTTCTTCTCA 0.3 15 (10–60) CNA2B 9: 191563–191711 TG HEX- TCGTCAACGATGCAAGTCTC GGGCCTGGGAAATAGGTAGA 0.3 6 (8–21) 2 CNA2C 10: 307773–306912 TA TET- AGAAGCACATGGGGAAAGG GCGCAGTTTGAAGATGAGAA 1.0 16 (8–44) CNA3 CNA3A 11: 1281429–12814716 CTA FAM- ACCCCCTGCCCATCATA GCACAGGCATAAAGCTAAGTGTGA 0.3 9 (19–69) CNA3B 4: 339525–339664 TCT HEX- TGGGGATATCGATTCCTTCTC GATTGGTATGGGAAGCGTTG 0.3 5 (5–17) CNA3C 7: 285123–285270 CCA TET- TGGAAGAGGATGGAGCGTAT GCATAGTTTATCGTTTTCTCTTTTC 0.3 10 (8–38) CNA4 CNA4A 5: 233120–233445 TTAT FAM- CGTCGAAGACTGCACAAAAA GTTCTGTATGACAGGTCGCAAA 1.0 51 (15–119) CNA4B 4: 1021855–1022020 ATCC HEX- CGGATGAGATGGAAAGAAGG GTGCGTCTGTCAAAAGATTGC 0.3 10 (5–25) CNA4C 14: 131866–132031 TATT TET- AGATGTCCTGGCGATGTTG GAGGAGCAAGCAATCAAACC 0.3 11 (1–18)

*The underlined residue(s) are not a match to the genomic sequence. These were introduced to minimize the formation of minus A peaks, a well known PCR artifact that may complicate interpretation of the results [34]. doi:10.1371/journal.pone.0009124.t001 the genomic sequence, the genotype of the H99 strain was 27-20- genotypes (from 18 isolates) were observed that did not belong 20-57-17-14-55-19-16 for markers 2A-2B-2C-3A-3B-3C-4A-4B- to a microsatellite complex, bringing the total number of different 4C respectively. genogroups to 20. Four MC’s (MC1-MC4) were the most prevalent and contain more than 70% of all isolates. Data Analysis Not all markers yielded a PCR product with all of the isolates, Typing data was imported into BioNumerics v5.0 software especially markers CNA3a scored negative on a substantial part (Applied Maths, Sint-Martens-Latem, Belgium). Microsatellite (85%) of the isolates. When a negative result was obtained, the data was analyzed using the multistate categorical similarity particular marker was reamplified in a monoplex PCR reaction. coefficient. Microsatellite complexes (MC’s) were defined as When still negative, the marker was scored as ‘‘0’’. Exclusion of groups of 2 or more genotypes differing by a maximum of 2 marker CNA3a from the microsatellite panel did not influence the markers. AFLP data was analyzed by UPGMA clustering using clustering of the isolates over the different MC’s (not shown). the Pearson correlation coefficient. The distribution of the clinical and environmental isolates over the different MC’s is shown in Table 4. Very interestingly, MC1 Results contained a large majority of isolates from environmental origin (.96% from pigeon guano) and only few human clinical isolates. A selection of 9 microsatellite markers was made from genomic This difference was highly significant (p,0,001). Though small in sequences from Cryptococcus neoformans var. grubii strain H99. All size, MC9 also exclusively contained isolates from environmental markers proved to be polymorphic displaying a minimum of 5 and origin . The non environmental MC’s were all found in HIV up to 51 different alleles per marker (Table 1). The specificity of positive patients with one exception: MC6 contained 3 isolates and the markers was tested by including C. neoformans var. neoformans those were obtained from HIV negative patients. isolates (serotype D) as well as C. gattii isolates (serotype B). None of The temporal distribution of the largest MC’s with clinical these isolates yielded amplification products confirming the isolates showed their presence over prolonged periods of time since specificity of these primers for C. neoformans var. grubii. In Table 2, the discriminatory power for each of the individual markers, panels and entire set of markers were calculated. When Table 2. Overview of the discriminatory power of the all markers are combined, the nine marker microsatellite panel individual markers, panels of markers and the entire set of yielded a discriminatory power of greater than 0.993. With this markers. Calculated values are based on the Simpson’s index collection of 190 isolates, 104 different genotypes could be of diversity and are expressed in a value of ‘D’ [35]. discriminated. The AFLP analysis confirmed that all isolates were C. neoformans var. grubii (not shown) in line with previous observations [16]. This panel of markers thus provides a highly Marker D Panel D Set D discriminatory typing assay for C. neoformans var. grubii. The relationship between the different genotypes is illustrated in CNA2a 0.842 CNA2 0.906 CNA 0.993 Figure 1. Within the large diversity of genotypes, complexes of CNA2b 0.789 closely related genotypes are recognized and indicated as CNA2c 0.828 microsatellite complexes (MC’s). Within the MC’s most of the CNA3a 0.282 CNA3 0.868 genotypic variation is the result from variations in few microsatellite markers (Table 3). The most discriminatory marker was CNA3b 0.618 marker CNA4a. Elimination of this marker from the dataset CNA3c 0.819 resulted in a reduction of the number of different genotypes by CNA4a 0.972 CNA4 0.992 approximately 50%, but did not affect the distribution of the CNA4b 0.712 isolates over the different MC’s nor did it affect the segregation CNA4c 0.688 between the different MC’s (results not shown). Eleven MC’s are recognized containing up to 51 isolates each. Nine further doi:10.1371/journal.pone.0009124.t002

39 Chapter 2

Table 3. Signature profiles of the 4 most prevalent microsatellite complexes and the reference isolate H99 upon whose genome the selection of markers was made.

CNA2a CNA2b CNA2c CNA3a CNA3b CNA3c CNA4a CNA4b CNA4c

MC1 53–60 11 9 0 13 35–38 85–108 8 5 MC2 11 8 22–25 0 14 16 30–37 8 5 MC3 10 12 10 0 14 14 66–78 6 9–11 MC4 12 10 8 0 5 8 103–119 5 1 H99 27 20 20 57 17 14 55 19 16

doi:10.1371/journal.pone.0009124.t003

they were found repeatedly since the early years of the study and microsatellite markers, there appears to be no close relationship thus have been present for almost 2 decades. between these MC’s (Figure 1C; Table 3) and these were therefore considered to reflect different unrelated MC’s. Discussion One marker that was selected from the H99 genome (CNA3a) did not yield an amplification product in 85% of all tested isolates. A 9-marker microsatellite panel is described for high resolution This could be the result of an actual deletion of this locus in the sub typing of C. neoformans var. grubii isolates, one of the members genome from these isolates or it could be the result of one or more of the C. neoformans complex, in particular serotype A isolates. This sequence polymorphisms underneath either of the two amplifica- panel of markers provides a highly discriminatory panel allowing tion primers. Alternatively, the size of the amplified fragment excellent discrimination between isolates from various origins. The could be beyond the reach of the size marker on the capillary numerical typing result allows easy storage into global databases as electrophoresis runs. This was not further investigated. Whether or well as portability of the results. not this observation is specific for the Cuban population of C. From the results, it is obvious that certain genotypes appear to neoformans var. grubii, which may very well be explained by the be more closely related to each other than to other genotypes. geopolitical isolation of this country, remains to be established. MC’s were arbitrarily defined by genotypes differing in up to 2 In our study, the large majority of environmental isolates from microsatellite markers from each other. Within each MC, the pigeon droppings co-clustered in one MC (MC1) containing only amount of variation is attributable to only one or two very few isolates from human clinical samples. This MC was microsatellite markers as the likely result of instability of these widespread in multiple environmental locations across Cuba. Vice specific markers. This mostly involved the markers CNA2a, versa, several of the other MC’s predominantly contained isolates CNA2c and CNA4a. Not surprisingly, these are among the most from human clinical samples and only few from environmental discriminatory markers from the entire set (Table 2). Likewise, origin. Possibly, isolates from MC1 are more adapted to the within an MC, there was very limited to no variation in the less pigeon host and may be less pathogenic for humans. Several MC’s discriminatory markers (Table 3). The difference between the were identified that contained only clinical isolates and were never MC’s is attributable to multiple microsatellite markers ($3 found in environmental samples which suggests the presence of markers difference). Although MC1 and MC2 could be connected additional niches of C. neoformans var. grubii that may cause human to each other by sequential genotypes differing in only 2 infections. One MC was identified (MC6) containing only isolates

Table 4. Distribution of isolates from clinical (including HIV status) or environmental origin over the 11 microsatellite complexes.

Complex Total number of isolates Env. Clin. HIV pos. HIV neg. HIV status unknown

MC1 51 49 2 2 0 0 MC2 32 2 30 16 6 8 MC3 39 2 37 27 9 1 MC4 14 0 14 14 0 0 MC5 8 2 6 5 0 1 MC6 3 0 3 0 3 0 MC7 9 5 4 4 0 0 MC8 4 1 3 3 0 0 MC9 3 3 0 0 0 0 MC10 2 0 2 1 1 0 MC11 7 0 7 4 1 2 other 18 4 14 9 5 0 Total 190 68 122 85 25 12

Env.: Environmental isolates; Clin.: Clinical isolates; pos.: positive; neg.: negative. doi:10.1371/journal.pone.0009124.t004

40 Microsatellite typing of Cryptococcus neoformans var. grubii isolates from Cuba

from humans without HIV infection. This may indicate that this cultures showed that the genotypes were not stable over time, particular genotype may be more virulent to humans but this limiting the usefulness of this approach. In addition, the authors clearly needs more confirmation. However, despite careful did not report specific genotypes being associated with either selection of multiple colonies from environmental samples, we clinical or environmental isolates. Our results confirm the cannot rule out the possibility that there could have been a superiority of this panel of microsatellite markers as molecular sampling bias towards genotypes that are more abundantly present typing targets for C. neoformans var. grubii and simultaneously allow in pigeon guano. distinguishing between isolates from clinical and environmental In this study, we used AFLP analysis for a dual purpose. Firstly origin. The lack of temporal and spatial (east to west Cuba is 2 to confirm the identification of the isolates as C. neoformans var. 1000 km) variability suggest a clonal relationship between the grubii, and secondly to validate the typing result of the majority (.70%) of isolates from pigeon guano in Cuba. This too microsatellite analysis. The specific combination of restriction may be explained by the isolated location of the country. In enzymes and selective residues used here was reported before [17]. contrast, most clinical cases were from Havana City and these Interestingly, we found only limited genetic variation in this involved multiple different MC’s additionally suggesting the collection of isolates. AFLP showed no segregation between the presence of alternative sources for human infections. isolates in MC1 and the majority of the other isolates (results not The use of microsatellite markers in typing studies with C. shown). On the other hand, a relatively small number of clinical neoformans var. grubii was reported before. Hanafy et al. reported a isolates (11%) did indeed segregate into a separate AFLP cluster. selection of 15 microsatellite markers for C. neoformans var. grubii of This subdivision is fully supported by the microsatellite data since which only 3 proved to be polymorphic [33]. The PCR products this AFLP cluster is recognized as a separate MC (MC4, Figure 1), were analyzed by agarose gel electrophoresis. This technique well separated from the other MC’s. Since AFLP analysis amplifies suffers from insufficient resolution and does not allow use of the fragments from multiple random locations in the genome, this may full potential of microsatellite markers as molecular typing targets. point to relatively few or small genomic differences between MC1 In summary, we have developed a novel approach for high isolates and the majority of the other isolates versus substantial resolution molecular sub typing of C. neoformans var. grubii. This will genomic differences between the majority of the isolates and those help the study of the global epidemiology of this opportunistic from MC4. Probably, the discriminatory power of the AFLP pathogenic yeast. We also show that microsatellites are excellent analysis can be increased by testing other combinations of multilevel genotyping targets allowing recognition of individual restriction enzymes and/or selective residues but this was not genotypes as well as clusters of related genotypes. The selected attempted. C. neoformans var. grubii has since long been associated with bird microsatellite markers are sufficiently stable for use in long-term droppings: according to the Centers for Disease Control and longitudinal studies. Finally, our results point to additional sources Prevention, people with weakened immune systems should avoid other than bird droppings as origin of human infections. areas contaminated with bird droppings and contact with birds [31]. However, this assumed link has received little attention using Acknowledgments molecular typing studies. Our results suggest that certain clinical We thank I. Curfs-Breuker, J. Jansen, A. Stevens-Krebbers, and K. isolates may originate from additional ecological niche(s). This is Driessen for expert technical assistance and F. Hagen for providing supported by recent evidence from Litvintseva et al. that reference isolates of the different C. neoformans complex members. environmental isolates from pigeon excreta were less pathogenic in a mouse model than isolates from human clinical samples [32]. Author Contributions In this prior work, neither AFLP nor MLST proved useful to distinguish between these clinical and environmental isolates. Conceived and designed the experiments: MTIZ CHWK. Performed the experiments: MTIZ CHWK. Analyzed the data: MTIZ TB JFM CHWK. Instead, a retrotransposon based Southern blotting procedure Contributed reagents/materials/analysis tools: MTIZ GFMM CMFA enabled them to distinguish between individual isolates with JFM. Wrote the paper: MTIZ TB JFM CHWK. identical AFLP/MLST genotypes. However, analysis of serial

References 1. Bicanic T, Harrison TS (2004) Cryptococcal Meningitis. Br Med Bull 72: 10. Bovers M, Hagen F, Kuramae EE, Hoogveld HL, Dromer F, et al. (2008) AIDS 99–118. Patient Death Caused By Novel Cryptococcus Neoformans X C. Gattii Hybrid. Emerg 2. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, et al. (2009) Infect Dis 14: 1105–1108. Estimation Of The Current Global Burden Of Cryptococcal Meningitis Among 11. Cogliati M, Esposto MC, Tortorano AM, Viviani MA (2006) Cryptococcus Persons Living With HIV/AIDS. AIDS 23: 525–530. Neoformans Population Includes Hybrid Strains Homozygous At Mating-Type 3. Levitz SM, Boekhout T (2006) Cryptococcus: The Once-Sleeping Giant Is Fully Locus. FEMS Yeast Res 6: 608–613. Awake. FEMS Yeast Res 6: 461–462. 12. Lengeler KB, Cox GM, Heitman J (2001) Serotype AD Strains Of Cryptococcus 4. Bovers M, Hagen F, Boekhout T (2008) Diversity Of The Cryptococcus Neoformans- Neoformans Are Diploid Or Aneuploid And Are Heterozygous At The Mating- Cryptococcus Gattii Species Complex. Rev Iberoam Micol 25: S4–12. Type Locus. Infect Immun 69: 115–122. 5. Bovers M, Hagen F, Kuramae EE, Boekhout T (2008) Six Monophyletic 13. Curbelo A (1951) Septicemia Y Meningitis A Cryptococcus Neoformans. Arch Hosp Lineages Identified Within Cryptococcus Neoformans And Cryptococcus Gattii By Univ (Cuba) 9: 324–330. Multi-Locus Sequence Typing. Fungal Genet Biol 45: 400–421. 14. Illnait-Zaragozı´ MT, Valde´z-Hernańdez IC, Fernańdez-Andreu CM, Mendoza- 6. Casadevall A, Perfect JR (1998) Cryptococcus Neoformans. Washington DC: ASM Llanez D, Perurena-Lancha MR, et al. (2001) Criptococosis: Una Alerta Press. Necesaria. Rev Cub Med Trop 53: 167–169. 7. Kwon-Chung KJ, Boekhout T, Fell JW, Dıáz M (2002) Proposal To Conserve 15. Arteaga-Hernańdez E, Capo´ De Paz V, Pe´rez-Fernańdez-Terań ML (1998) The Name Cryptococcus Gattii Against C. Hondurianus And C. Bacillisporus Micosis Oportunistas Invasivas En El Sida. Un Estudio De 211 Autopsias. Rev (Basidiomycota, Hymenomycetes, Tremellomycetidae). Taxon 51: 804– Iberoam Micol 15: 33–35. 896. 16. Fernańdez-Andreu CM, Martıńez-Machıń GF, Illnait-Zaragozı´ MT, Perurena- 8. Ngamskulrungroj P, Gilgado F, Faganello J, Litvintseva AP, Leal AL, et al. Lancha MR, Gonza´lez-Miranda M (1998) Identificacioń De Cryptococcus (2009) Genetic Diversity Of The Cryptococcus Species Complex Suggests That Neoformans Var. Neoformans En Aislamientos Clıńicos Cubanos. Rev Cub Med Cryptococcus Gattii Deserves To Have Varieties. Plos One 4: E5862. Trop 50: 167–169. 9. Bovers M, Hagen F, Kuramae EE, Diaz MR, Spanjaard L, et al. (2006) Unique 17. Boekhout T, Theelen B, Diaz M, Fell JW, Hop WC, et al. (2001) Hybrid Hybrids Between The Fungal Pathogens Cryptococcus Neoformans And Cryptococcus Genotypes In The Pathogenic Yeast Cryptococcus Neoformans. Microbiology 147: Gattii. FEMS Yeast Res 6: 599–607. 891–907.

41 Chapter 2

18. Meyer W, Castaneda A, Jackson S, Huynh M, Castaneda E (2003) Molecular 26. Klaassen CH (2009) MLST Versus Microsatellites For Typing Aspergillus Typing Of Iberoamerican Cryptococcus Neoformans Isolates. Emerg Infect Dis 9: Fumigatus Isolates. Med Mycol 47: S27–S33. 189–195. 27. De Valk HA, Meis JF, Bretagne S, Costa JM, Lasker BA, et al. (2009) 19. Trilles L, Lazera M, Wanke B, Theelen B, Boekhout T (2003) Genetic Interlaboratory Reproducibility Of A Microsatellite-Based Typing Assay For Characterization Of Environmental Isolates Of The Cryptococcus Neoformans Aspergillus Fumigatus Through The Use Of Allelic Ladders: Proof Of Concept. Species Complex From Brazil. Med Mycol 41: 383–390. Clin Microbiol Infect 15: 180–187. 20. Raimondi A, Ticozzi R, Sala G, Bellotti MG (2007) Genotype-Based 28. Hazen KC, Howell SA (2003) Candida, Cryptococcus And Other Yeasts Of Medical Differentiation Of The Cryptococcus Neoformans Serotypes By Combined PCR- Importance. In: Murray PR, Baron EJ, Yolken RH, eds. Manual Of Clinical RFLP Analysis Of The Capsule-Associated Genes CAP10 And CAP59. Med Microbiology, 8th Ed Vol 2 ASM Press, Wshington, DC. Mycol 45: 491–501. 29. Benson G (1999) Tandem Repeats Finder: A Program To Analyze DNA 21. Viviani MA, Cogliati M, Esposto MC, Lemmer K, Tintelnot K, et al. (2006) Sequences. Nucl Acids Res 27: 573–580. Molecular Analysis Of 311 Cryptococcus Neoformans Isolates From A 30-Month 30. De Valk HA, Meis JFGM, Curfs IM, Muehlethaler K, Mouton JW, et al. (2005) ECMM Survey Of Cryptococcosis In Europe. FEMS Yeast Res 6: 614–619. Use Of A Novel Panel Of Nine Short Tandem Repeats For Exact And High- 22. Feng X, Yao Z, Ren D, Liao W, Wu J (2008) Genotype And Mating Type Resolution Fingerprinting Of Aspergillus Fumigatus Isolates. J Clin Microbiol 43: Analysis Of Cryptococcus Neoformans And Cryptococcus Gattii Isolates From China 4112–4120. That Mainly Originated From Non-HIV-Infected Patients. FEMS Yeast Res 8: 31. Disease Listing: Cryptococcus General Information. Available: Http://Www. 930–938. Cdc.Gov/Nczved/Dfbmd/Disease_Listing/Cryptococcus_Gi.Html. Accessed 23. Escandon P, Sanchez A, Martinez M, Meyer W, Castaneda E (2006) Molecular Epidemiology Of Clinical And Environmental Isolates Of The Cryptococcus 2009 Nov 19. Neoformans Species Complex Reveals A High Genetic Diversity And The 32. Litvintseva AP, Mitchell TG (2009) Most Environmental Isolates Of Cryptococcus Presence Of The Molecular Type VGII Mating Type A In Colombia. FEMS Neoformans Var. Grubii (Serotype A) Are Not Lethal For Mice. Infect Immun 77: Yeast Res 6: 625–635. 3188–3195. 24. Litvintseva AP, Thakur R, Vilgalys R, Mitchell TG (2006) Multilocus Sequence 33. Hanafy A, Kaocharoen S, Jover-Botella A, Katsu M, Iida S, et al. (2008) Typing Reveals Three Genetic Subpopulations Of Cryptococcus Neoformans Var. Multilocus Microsatellite Typing For Cryptococcus Neoformans Var. Grubii. Med Grubii (Serotype A), Including A Unique Population In Botswana. Genetics 172: Mycol 46: 685–696. 2223–2238. 34. De Valk HA, Meis JF, Klaassen CH (2007) Microsatellite Based Typing Of 25. Meyer W, Aanensen DM, Boekhout T, Cogliati M, Diaz MR, et al. (2009) Aspergillus Fumigatus: Strengths, Pitfalls And Solutions. J Microbiol Methods 69: Consensus Multi-Locus Sequence Typing Scheme For Cryptococcus Neoformans 268–272. And Cryptococcus Gattii. Med Mycol 47: 561–570. 35. Simpson EH (1949) Measurement Of Diversity. Nature 163: 688.

42 Chapter 3 Extensive genetic diversity within the Dutch clinical Cryptococcus neoformans population

Published in Journal of Clinical Microbiology 2012;50: 1918-26. Chapter 3

Extensive Genetic Diversity within the Dutch Clinical Cryptococcus neoformans Population

Ferry Hagen,a,b María-Teresa Illnait-Zaragozí,c,d Jacques F. Meis,c,e William H. M. Chew,a Ilse Curfs-Breuker,c Johan W. Mouton,c,e Andy I. M. Hoepelman,b Lodewijk Spanjaard,f Paul E. Verweij,e Greetje A. Kampinga,g Ed J. Kuijper,h Teun Boekhout,a,b and Corné H. W. Klaassenc Department of Yeast and Basidiomycete Research, CBS-KNAW Fungal Biodiversity Centre, Utrecht, The Netherlandsa; Department of Internal Medicine and Infectious Diseases, University Medical Centre Utrecht, Utrecht, The Netherlandsb; Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, Nijmegen, The Netherlandsc; Tropical Medicine Institute Pedro Kouri, Havana, Cubad; Department of Medical Microbiology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlandse; Department of Medical Microbiology, Netherlands Reference Laboratory for Bacterial Meningitis, Academic Medical Center Amsterdam, Amsterdam, The Netherlandsf; Department of Medical Microbiology, University Medical Center Groningen, Groningen, The Netherlandsg; and Department of Medical Microbiology, Center of Infectious Diseases, Leiden University Medical Center, Leiden, The Netherlandsh

A set of 300 Dutch Cryptococcus neoformans isolates, obtained from 237 patients during 1977 to 2007, was investigated by determining the mating type, serotype, and AFLP and microsatellite genotype and susceptibility to seven antifungal compounds. Al- most half of the studied cases were from HIV-infected patients, followed by a patient group of individuals with other underlying diseases and immunocompetent individuals. The majority of the isolates were mating type � and serotype A, followed by �D isolates and other minor categories. The most frequently observed genotype was AFLP1, distantly followed by AFLP2 and AFLP3. Microsatellite typing revealed a high genetic diversity among serotype A isolates but a lower diversity within the serotype D set of isolates. One patient was infected by multiple AFLP genotypes. Fluconazole and flucytosine had the highest geometric mean MICs of 2.9 and 3.5 �g/ml, respectively, while amphotericin B (0.24 �g/ml), itraconazole (0.08 �g/ml), voriconazole (0.07 �g/ml), posaconazole (0.06 �g/ml), and isavuconazole (0.03 �g/ml) had much lower geometric mean MICs. One isolate had a high flucytosine MIC (>64 �g/ml), while decreased susceptibility (>16 �g/ml) for flucytosine and fluconazole was found in 9 and 10 C. neoformans isolates, respectively.

he encapsulated basidiomycetous yeast Cryptococcus neofor- VNII), C. neoformans variety neoformans has one genotype Tmans, the causative agent of cryptococcosis, can cause life- (AFLP2/VNIV), and the C. neoformans intervariety serotype AD threatening infections, such as severe lung infections and menin- hybrid has one (AFLP3/VNIII). Cryptococcus gattii can be divided goencephalitis. This fungal disease was rarely seen until the early into five genotypes: AFLP4/VGI, AFLP6/VGII, and AFLP10/ 1980s, but this changed due to the HIV pandemic (5, 42, 44). VGIV with serotype B isolates, and AFLP5/VGIII and AFLP7/ Annually nearly one million HIV-infected subjects develop cryp- VGIV are represented by serotype C isolates (21, 30). Several in- tococcosis, mainly in sub-Saharan Africa, where the majority of terspecies hybrids have been described. C. neoformans variety patients die due to ineffective treatment or a lack of treatment neoformans AFLP2/VNIV � C. gattii AFLP4/VGI is known as options (35, 44). Besides a high incidence of cryptococcosis AFLP8 and C. neoformans variety grubii AFLP1/VNI � C. gattii among HIV-infected patients, individuals with other underlying AFLP4/VGI has been designated AFLP9 (7, 8). A third interspecies diseases might be at risk due to immune disorders or immuno- variety has recently been described and was found to be a hybrid suppressive therapy (e.g., organ transplant recipients and individ- between C. neoformans var. grubii AFLP1/VNI and C. gattii uals receiving anti-cancer therapy) (5, 35, 44). AFLP6/VGII (1). According to the current taxonomic classification, the genus A retrospective survey of HIV-infected patients in the Nether- Cryptococcus contains more than 100 species (19). However, only lands identified 268 patients with cryptococcosis during 1986 to C. neoformans and C. gattii are frequent causes of cryptococcal 1999 (42). A major increase in the incidence of the disease was disease in humans and animals. Cryptococcus gattii has been rec- observed since 1987, which correlated with the increasing number ognized as a primary pathogen, and C. neoformans is mainly an of HIV/AIDS patients. The introduction of highly active antiret- opportunistic infection. The latter is currently divided into two roviral therapy (HAART) in 1996 resulted in a significant decrease varieties, C. neoformans variety grubii (serotype A) and C. neofor- of cryptococcosis patients in the Netherlands (42). A clinical epi- mans variety neoformans (serotype D) (5). Cryptococcus neofor- demiological analysis was made of the patient characteristics, but mans variety gattii (serotype B and C), was raised to a separate species named C. gattii (5, 28). PCR fingerprinting, restriction fragment length polymor- Received 20 December 2011 Returned for modification 18 January 2012 phism analysis of the PLB1 and URA5 loci, amplified fragment Accepted 14 March 2012 length polymorphism (AFLP) fingerprint analysis, and several Published ahead of print 21 March 2012 multilocus sequence typing (MLST) studies divided C. neofor- Address correspondence to Teun Boekhout, [email protected]. mans and C. gattii into 10 distinct genotypes (4, 6, 26, 30, 33). Copyright © 2012, American Society for Microbiology. All Rights Reserved. Within C. neoformans variety grubii, three genotypes can be dis- doi:10.1128/JCM.06750-11 cerned (named AFLP1/VNI, AFLP1A/VNB/VNII, and AFLP1B/

44 Extensive genetic diversity withing the Dutch clinical Cryptococcus neoformans population

a genotypic analysis of the involved C. neoformans isolates has not TABLE 1 Primer sequences for microsatellite typing of C. neoformans been performed (42). In this study, we investigated Dutch C. neo- var. neoformansa formans isolates that were collected during a 30-year period (1977 Name Sequence Target to 2007) to investigate the genetic diversity. AFLP fingerprinting, CND2-1 Fwd FAM-AATCCTGAGACGGTTTGTGG CND2A microsatellite typing, and mating type and serotype analyses were CND2-1 Rvd GAACATTGTGCCCGACTTTT performed by using PCRs to assess the genetic diversity among CND2-4 Fwd TET-ATCGAAGGTCTCGTCGTCTG CND2B Dutch clinical Cryptococcus strains. Antifungal susceptibility pro- CND2-4 Rvd GAATGCTCTGGATGGAAGGA files were determined for the antifungal compounds amphotericin CND3-1 Fwd FAM-TGCTGAGCTTGTTGAGGTTG CND3A B, flucytosine, fluconazole, itraconazole, and the three novel tria- CND3-1 Rvd GAATGGCGCACTAATCTACCC zoles, posaconazole, voriconazole, and isavuconazole. CND3-2 Fwd HEX-AGGCAAGACTGACGTTGAGC CND3B CND3-2 Rvd GCAACTTCCTCCCCTCTTTCC MATERIALS AND METHODS CND3-3 Fwd GTTGGTCTGTCTCCCTCGAT CND3C 3 Patient-related data. Clinically relevant information was extracted from CND3-3 Rvd TET-TTTCAAAGATGCTTGATGAGGA the database of the Netherlands Reference Laboratory for Bacterial Men- CND3-4 Fwd GTCGGTAAGGTTTTGGGTTGA CND3D ingitis (Academic Medical Center, Amsterdam, The Netherlands) and CND3-4 Rvd HEX-AGCACCAAAAGATGGGTACAA included the source and date of Cryptococcus isolation, date of birth, gen- CND4-3 Fwd GCGCCGCTCTTAGAATGAAGA CND4A der, and place of residence of the patient. Patient data were requested from CND4-3 Rvd TET-TCTCCGTCTGTCGTCTTTTG the contributing clinicians or hospitals in the case that no or limited data a The primer sequences for the seven microsatellite serotype D-specific primer pairs are were available. Relapses of cryptococcal infections were defined by a pe- given. Primer sequences for the nine microsatellite serotype A-specific primer pairs are riod between two positive cultures of more than 90 days. Four patient from Illnait-Zaragozi et al. (25). categories were determined based on the predisposing factors that may have influenced the onset of the Cryptococcus infection, namely, HIV/ To characterize the genetic structure of C. neoformans variety neofor- AIDS-related immunodeficiency, non-HIV/AIDS-related immunodefi- mans (serotype D) isolates, a microsatellite panel was developed based on ciency (e.g., immunosuppressive therapy and organ transplantation), im- the genome sequence of reference strain JEC21 (CBS10513) using the munocompetent individuals, and a patient category for which the Tandem Repeats Finder software package (3). This panel consists of two presence or absence of predisposing factors was unknown. dinucleotide repeat loci, four trinucleotide repeat loci, and one tet- Cryptococcal isolates and media. Cryptococcal isolates (n � 328; 300 ranucleotide repeat locus (Table 1) that are all specific for isolates harbor- were viable) were revived from the collection of the Netherlands Refer- ing serotype D; no cross-specificity was observed when tested with C. ence Laboratory for Bacterial Meningitis (held at �80°C; Academic Med- neoformans var. grubii or C. gattii isolates (data not provided). For each ical Center, Amsterdam, The Netherlands) or were requested from par- locus, one of the amplification primers was labeled with a fluorescent dye ticipating Dutch hospitals. Isolates were cultured on 1% yeast extract, 1% (FAM, 6-carboxyfluorescein; HEX, hexachlorofluorescein; or TET, tetra- peptone, 2% D-glucose agar (YPGA) medium and incubated at 25°C for 2 chlorofluorescein). The C. neoformans var. grubii genome-sequenced ref- days. erence strain JEC21 was used as a control. Amplification conditions, the Genomic DNA extraction. Isolates were subcultured for 2 days at processing of amplicons, and data analysis were identical to the methods 30°C on YPGA media supplemented with 0.5 M NaCl to prevent the and conditions used for the analysis of the C. neoformans variety grubii formation of polysaccharide capsules. Genomic DNA was extracted as microsatellite panel. described previously by Hagen et al. (21). The purity and quantity of the Antifungal susceptibility testing. The MICs of amphotericin B (Bristol genomic DNA solution was measured using a Nanodrop ND-1000 spec- Myers Squibb, Woerden, The Netherlands), flucytosine (Valeant Pharma- trophotometer (Nanodrop, Wilmington, DE), and final work solutions ceuticals, Zoetermeer, Netherlands), fluconazole and voriconazole (Pfizer with a concentration of 100 ng/�l were prepared. Central Research, Sandwich, Kent, United Kingdom), itraconazole (Janssen Determining mating type and serotype by PCR and AFLP genotyp- Cilag, Tilburg, The Netherlands), posaconazole (Schering-Plough Corp. ing. The mating type and serotype determination of the isolates was per- [now Merck], Kenilworth, NJ), and isavuconazole (Basilea Pharmaceutica formed as described previously (2). Reference strains 125.91 (CBS10512; [now Astellas], Basel, Switzerland) were determined using a broth microdi- mating type aA; AFLP1), H99 (CBS8710; �A; AFLP1), JEC20 (CBS10511; lution method in accordance with CLSI document M27-A3 (13). The proce- aD; AFLP2), and JEC21 (CBS10513; �D; AFLP2) were included as con- dure was performed as described by Hagen et al. (21). trols. Isolates that failed to yield an amplicon subsequently were subjected Statistical analysis. The 50 and 90% minimum inhibitory concentra- to a set of two PCRs to amplify the C. gattii mating type STE12a or STE12� tions (MIC50 and MIC90, respectively) were determined by ordering the allele (7, 21). For these PCRs, the reference strains CBS1930 (aB; AFLP6) MIC values for each antifungal in ascending order and selecting the me- and WM276 (�B; AFLP4) were included as controls. dian and 90th quantile, respectively, of the MIC distribution. Geometric The genotype was determined using AFLP fingerprint analysis as de- mean MICs were calculated using Office Excel 2007 (Microsoft, Red- scribed previously by Boekhout et al. (4). Reference strains for each of the mond, WA), for which purpose values “�X” were set equal to “0.5X” and genotypes were included according to Hagen et al. (21). “�Y” values were set equal to “2Y”. A two-tailed Mann-Whitney-Wil- Microsatellite amplification and analysis. The genetic relatedness of coxon test was performed to compare the MIC values between different C. neoformans variety grubii (serotype A) isolates was further investigated groups (e.g., patient categories versus AFLP genotypes or microsatellite using nine short tandem repeat markers (STRs) as described recently (25). cluster), and a P value of �0.05 was regarded as statistically significant. The C. neoformans var. grubii genome-sequenced reference strain H99 Significant correlations between Cryptococcus characteristics versus any of (CBS8710) was used as a control. Fragment analysis on a MegaBACE 500 the other values were calculated using a chi-square test. Statistical analyses capillary electrophoresis platform (GE Healthcare Life Sciences, Diegem, were performed using the SAS software package (SAS, Cary, NC). Belgium) was carried out as described by Illnait-Zaragozi et al. (25). The discriminatory power of the microsatellite typing experiments Bionumerics version 6.1 (Applied Maths, Sint-Martens-Latem, Belgium) was calculated according to Simpson (37) using the software package was used to calculate a minimum spanning tree. A multistate categorical Haplotype Analysis v1.05 (18). similarity coefficient was used, and microsatellite complexes (MCs) were defined by a minimum group of two genotypes that contain at least five RESULTS isolates and with a maximum neighbor distance of a two-locus difference Cryptococcal isolates, patients, and patient categories. Detailed among the nine loci studied (25). data of the total number of Cryptococcus isolates and the number

45 Chapter 3

TABLE 2 Cryptococcosis patient dataa Cryptococcus isolate and Other predisposing patient parameter Total HIV/AIDS positive factors Immunocompetent Unknown immune status Isolates No. (%) of strainsb 300 (100) 137 (45.7) 68 (22.7) 24 (8) 71 (23.7) No. (%) of CSF culturesc 201 (67) 105 (76.6) 37 (54.4) 13 (54.2) 46 (64.8) No. (%) of blood culturesc 44 (14.7) 20 (14.6) 7 (10.3) 4 (16.7) 13 (18.3) No. (%) of other sourcesc 55 (18.3) 12 (8.8) 24 (35.3) 7 (29.2) 12 (16.9)

Patients All No. (%)b 237 (100) 115 (48.5)d,e 50 (21.1)d 18 (7.6) 54 (22.8)f Male/female ratio 4.2:1 8.4:1 1.8:1 3.5:1 3.5: 1 Age distribution (yr) 14–83 20–63d 17–78d 15–83 14 �69 Avg age (yr) 43.2 38.2d 53.9d 50.8 41.2 Median age (yr) 41 37d 60d 48 38

Male No. (%)b 188 (100) 101 (53.7)e 31 (16.5) 14 (7.4) 42 (22.3)f Age distribution (yr) 17–83 20–63 17–75 27–83 19–69 Avg age (yr) 42.7 39.0 52.7 53.1 40.6 Median age (yr) 41 38 59 48 35

Female No. (%)b 45 (100) 12 (26.7) 17 (37.8) 4 (8.9) 12 (26.7) Age distribution (yr) 14–78 23–43 30–78 15–73 14–69 Avg age (yr) 45.4 31.0 53.8 43.0 42.9 Median age (yr) 42 32.8 60 42 49 a The number of included C. neoformans isolates is listed for the categories HIV/AIDS patients, patients with other predisposing factors, immunocompetent patients, and patients with unknown immune status. Data are divided into four sections, providing the source of isolation of the Cryptococcus isolates, data related to the patient, and patient-related data for male and female patients. b Numbers and percentages indicated in these patient categories are the cumulative numbers and percentages indicated in the column Total. c Numbers and percentages of isolates per source (CSF, blood, or other source) are the cumulative numbers and percentages of isolates applied to that patient category. d Including two patients without known gender. e Including one male patient without known age. f Including three male patients without known age. of patients included, as well as number of isolates per defined isolates were �A(n � 219; 73.0%). Thirty-eight isolates (12.7%) patient category, are listed in Table 2. The isolates were isolated were serotype D, of which 32 were mating type � and 6 mating from cerebrospinal fluid (CSF) (n � 201; 67%), blood (n � 45; type a. The remaining isolates belonged to minor mating type and 15%), lung (n � 28; 9.3%), bone (n � 5; 1.7%), spleen, urine serotype categories as listed in Table 3. (each n � 2; 0.67%), pus, neck gland (each n � 1; 0.33%), and The mating type and serotype of 11 isolates (3.7%) could not unknown sources (n � 15; 5%). be determined, and these were subsequently subjected to a C. gat- The 300 isolates originated from 237 patients, of whom 45 were tii-specific PCR. This resulted in two (0.7%) isolates that were C. female (19.0%), 188 were male (79.3%), and 4 were of unknown gattii mating type �, while the remaining nine (3%) isolates re- gender (1.7%). The male-to-female ratio was 4.2 to 1 with an mained untypeable. overall age distribution of 14 to 83 years and male and female age AFLP fingerprinting divided the isolates into six AFLP distributions of 17 to 83 and 14 to 78 years, respectively. The groups. The majority of isolates (n � 245; 81.7%) clustered average age at which the 237 patients developed cryptococcosis together with reference strains for genotype AFLP1, which rep- was 43.2 years with a median of 41 years. resents C. neoformans var. grubii (serotype A) and includes Almost half the studied cases of cryptococcosis occurred in seven (2.3%) isolates that belong to the minor genotype cluster HIV-infected patients (n � 115; 48.5%), followed by the group AFLP1B. Thirty-six isolates clustered together with the refer- with other underlying diseases (n � 50; 21.1%) and immunocom- ence strain for genotype AFLP2 (n � 36; 12%), representing C. petent individuals (n � 18; 7.6%). Fifty-four (22.8%) of the epi- neoformans var. neoformans (serotype D), and 14 clustered to- sodes occurred in patients for which no clinical data were avail- gether in genotype AFLP3 (n � 14; 4.7%), which represents the able. Two or more isolates were available from 42 patients serotype AD hybrids. The remaining isolates belonged to mi- (17.7%), and 2 of these patients had a relapse of cryptococcosis. nor AFLP genotypes, as listed in Table 3. Mating type, serotype, and AFLP genotype determinations. Some discrepancies between the conventional methods for de- Detailed data on the mating type, serotype, and AFLP genotype termining mating type and serotype versus AFLP fingerprinting distribution among investigated Cryptococcus spp. isolates per pa- were observed. Four isolates (1.2%) were determined by PCRs for tient category are listed in Table 3. The PCR-based determination mating type and serotype as �A or aD, but AFLP genotyping re- of mating type and serotype revealed that the majority of the 300 vealed that they were actually hybrid serotype AD isolates (geno-

46 Extensive genetic diversity withing the Dutch clinical Cryptococcus neoformans population

TABLE 3 Genetic background of Cryptococcus neoformans isolates per patient category based on mating type, serotype, and genotypea Other predisposing Typing method and strain Total HIV/AIDS factors Immunocompetent Unknown immune status PCR-based mating type and serotype determination No. (%) of �A strains 219 (73) 95 (69.3) 47 (69.1) 20 (83.3) 57 (80.3) No. (%) of aD strains 6 (2) 3 (2.2) 2 (2.9) 1 (1.4) No. (%) of �D strains 32 (10.7) 9 (6.6) 13 (19.1) 1 (4.2) 9 (12.7) No. (%) of �A-aA strains 21 (7) 14 (10.2) 2 (2.9) 2 (8.3) 3 (4.2) No. (%) of �A-�D strains 1 (0.3) 1 (0.7) No. (%) of �A-aD strains 6 (2) 5 (3.6) 1 (1.5) No. (%) of �D-aA strains 4 (1.3) 4 (2.9)e 3 No. (%) of C. gattii (�) isolates 2 (0.7) 1 (0.7) 1 (1.5) No. (%) of unknown isolates 9 (3) 5 (3.6) 2 (2.9) 1 (4.2) 1 (1.4)

AFLP genotyping (no. [%] positive) AFLP1 245 (81.7) 111 (81)d 51 (75)c 21 (87.5)b 62 (87.3) AFLP2 36 (12) 13 (9.5)e 13 (19.1) 2 (8.3) 8 (11.3) AFLP3 14 (4.7) 12 (8.8) 1 (1.5) 1 (4.2) AFLP4 1 (0.3) 1 (0.7) AFLP6 1 (0.3) 1 (1.5) AFLP8 3 (1) 2 (2.9) 1 (1.4) a The numbers and percentages of C. neoformans isolates per patient category are given for the mating type and/or serotyping results and AFLP genotyping. Percentages are cumulative for the PCR-based mating type and serotype determination group and the AFLP genotyping group. b Includes one isolate with genotype AFLP1B. c Includes two isolates with genotype AFLP1B. d Includes four isolates with genotype AFLP1B. e One isolate was found to be �D-aA by conventional PCR mating type and serotype determination, while AFLP genotyping revealed that this isolate was AFLP2. type AFLP3) that lack either the serotype A or D genetic back- microsatellite cluster analysis was found to be represented by the ground. Three isolates that were determined by mating type and microsatellite loci CNA3B, CNA3C, and CNA4C, which all had a serotype PCR as aD were identified by AFLP genotyping as inter- low genotypic diversity and low discriminatory power. Statistical species hybrids between C. neoformans var. neoformans and C. analysis revealed no significant differences between the MCs com- gattii that were assigned to AFLP8 (7). A significant correlation pared to the clinical origin, date of isolation, AFLP genotype, or between the genotypic AFLP1 and AFLP2 groups versus the im- geographic locality (data not shown). mune status of the patients was observed (P � 0.029). Microsatellite (STR) genotyping of serotype D isolates. A set Microsatellite (STR) genotyping of serotype A isolates. Based of 53 C. neoformans isolates that included 36 serotype D (genotype on serotype and AFLP fingerprint analysis, a set of 259 C. neofor- AFLP2), 14 serotype AD (genotype AFLP3), and 3 serotype BD mans isolates was further investigated using a serotype A-specific (genotype AFLP8) isolates (Table 3) was studied by serotype D- microsatellite panel. This included 245 serotype A isolates belong- specific microsatellite analysis. When all seven microsatellite CND ing to genotype AFLP1 and 14 serotype AD hybrid isolates belong- markers were combined, 32 different microsatellite profiles could ing to genotype AFLP3 (Table 3). be distinguished among 53 isolates (Fig. 2). No microsatellite clus- When all nine serotype A-specific microsatellite markers were ters could be determined due to the number of isolates used for combined, 196 different microsatellite genotypes and 11 MCs analysis with this novel microsatellite typing panel. The discrimi- could be distinguished among 259 isolates (Fig. 1). Using the natory power for the complete set of seven microsatellite markers Simpson’s diversity index (D), the discriminatory power for the was found to be 0.966. Among the seven studied microsatellite complete set of nine microsatellite markers was 0.994, and that of loci, CND2A and CND2B had the highest genotypic diversity, the three separate panels was 0.929, 0.953, and 0.985 for CNA2, CNA3, and CNA4, respectively. Among the nine studied micro- with nine alleles (D � 0.813 and 0.73, respectively), while locus satellite loci, CNA4A had the highest genotypic diversity with 59 CND3C had the lowest genotypic diversity and the lowest resolu- types (D � 0.967), while locus CNA3B had the lowest genotypic tion, with three alleles (D � 0.466). There were no cross-reactions diversity with nine types (D � 0.649). However, locus CNA3C observed when the serotype D microsatellite typing was per- showed the lowest resolution, with a D value of 0.578. formed on a set of non-serotype D C. neoformans and C. gattii Microsatellite loci CNA3A and CNA4A showed a relatively isolates (data not shown). Statistical analysis revealed no signifi- high degree of negative results for 35 and 6% of the investigated cant differences between the MCs compared to the clinical origin, isolates, respectively, whereas the other microsatellite markers date of isolation, AFLP genotype, or geographic locality. showed only the occasional absence of a microsatellite locus. Mixed C. neoformans infections determined by AFLP and However, the assignment of MCs remained intact when microsat- microsatellite typing. Multiple isolates were available from 38 ellite loci CNA3A and CNA4A were discarded from the analysis patients. The majority of these isolates were sampled on the same (data not shown). A step-by-step removal of the most discrimina- day but from different clinical specimens. While microsatellite tory microsatellite loci showed that the backbone structure of the profiles of most of these isolates were found to be identical, a few

47 Chapter 3

FIG 1 Genotypic diversity of Cryptococcus neoformans variety grubii isolates based on serotype A-speciﬁc microsatellite typing. Minimum spanning tree showing 259 C. neoformans var. grubii isolates based on a nine-locus microsatellite typing panel. Each circle corresponds to a unique genotype, and connected shaded circles belong to a speciﬁc microsatellite cluster. The size of the circles corresponds to the number of isolates of the same genotype. Connecting lines correspond to the number of differences between genotypes, with a solid thick line connecting genotypes that differ in one locus, a solid thin line connecting genotypes that differ in up to three loci, a dashed line connecting genotypes that differ in four loci, and a dotted line connecting microsatellite genotypes that differ in more than four loci. (A) Microsatellite clusters (MCs) that are numbered according to Illnait-Zaragozi et al. (25). MC13 and MC14 are novel MCs not observed in previous microsatellite typing studies, and yellow-colored circles represent genotypes that do not belong to a known microsatellite cluster. (B) Identical to panel A, except colors represent the AFLP genotype of the isolates: green, AFLP1; dark blue, AFLP1B; red, AFLP3. (C) Identical to panel B, except colors represent the source of isolation: green, CSF; red, blood; dark blue, source other than blood or CSF.

showed minor differences of one to three repeat units between one these groups, the MIC50 and the geometric mean MIC values dif- and two of the nine microsatellite loci studied. fered by less than 1 log2 dilution step, except for fluconazole and Some patients were found to be infected with multiple C. neo- itraconazole. formans strains. One immunosuppressed patient yielded nine se- The overall MIC ranges for each of the seven antifungal com- rotype A isolates during a period of 2 weeks, and two showed pounds were 0.063 to 1 �g/ml for amphotericin B, 0.125 to �64 major differences in three out of the nine microsatellite loci stud- �g/ml for flucytosine, 0.25 to 64 �g/ml for fluconazole, 0.016 to ied. A three-locus difference was also observed between isolates 0.5 �g/ml for itraconazole and posaconazole, 0.016 to 1 �g/ml for sampled on the same day from CSF and blood of one patient. The voriconazole, and �0.016 to 0.5 �g/ml for isavuconazole. Flu- two isolates obtained from a patient that experienced a cryptococ- conazole and flucytosine had the highest geometric mean MICs of cosis relapse were found to be different for six loci and were a mix 2.86 and 3.5 �g/ml, respectively, while amphotericin B (0.24 �g/ of minor (1 or 2) and major (�10) changes in repeat numbers. ml), itraconazole (0.08 �g/ml), voriconazole (0.07 �g/ml), po- This pattern of minor and major differences in repeat numbers saconazole (0.06 �g/ml), and isavuconazole (0.03 �g/ml) had was also observed between two isolates sampled on the same day much lower geometric mean MIC values. Isolates with MICs out- from a patient that was also found to be infected with C. neofor- side the normal range were tested on two different occasions and mans serotype D, which was isolated 21 days prior and 24 days showed the same results. The Cryptococcus isolate with a high flu- after the isolation of the two serotype A isolates. These two sero- cytosine MIC (�64 �g/ml) was cultured from blood obtained type D isolates were found to belong to two different genotypes of C. neoformans var. neoformans based on microsatellite typing, from an HIV-positive patient. However, a second CSF isolate since they differed for six of the seven microsatellite loci. Two from this patient from the same day had a low MIC (8 �g/ml), interspecies hybrid isolates of C. neoformans var. neoformans and while microsatellite typing revealed that both isolates were nearly C. gattii (genotype AFLP8) were cultured with an 88-day interval identical and should be considered microvariants of the same ge- from an immunocompromised patient and were found to differ notype. Nine and 10 C. neoformans isolates were less susceptible for five of the seven serotype D microsatellite loci. (�16 �g/ml) for flucytosine and fluconazole, respectively. Patients from whom multiple C. neoformans isolates were cul- Antifungal susceptibility testing. MIC ranges, MIC50s, MIC90s, and geometric mean MICs of seven antifungal com- tured, either at the same time from different body sites or at dif- pounds are presented in Table 4. Susceptibility data for each of the ferent time points, generally showed the same pattern of MIC seven antifungal compounds are presented for all 300 Cryptococ- values for the seven tested antifungal compounds. However, two cus spp. isolates as well as for the three major genotypes AFLP1 HIV-positive patients were infected with serotype A isolates that (n � 245), AFLP2 (n � 36), and AFLP3 (n � 14). For all four of showed large differences in fluconazole susceptibility (MICs of 4

48 Extensive genetic diversity withing the Dutch clinical Cryptococcus neoformans population

FIG 2 Genotypic diversity of Cryptococcus neoformans variety neoformans isolates based on serotype D-specific microsatellite typing. Minimum spanning tree showing 53 C. neoformans var. neoformans isolates based on a novel seven-locus microsatellite typing panel. Each circle corresponds to a unique genotype. The size of the circles corresponds to the number of isolates of the same genotype. Connecting lines correspond to the number of differences between genotypes, with a solid thick line connecting genotypes that differ in one locus, a solid thin line connecting genotypes that differ in up to three loci, a dashed line connecting genotypes that differ in four loci, and a dotted line connecting microsatellite genotypes that differ in more than four loci. (A) The microsatellite genotypes. (B) Identical to panel A, except colors represent the AFLP genotypes of the isolates: green, AFLP2; dark blue, AFLP3; red, AFLP8. (C) Identical to panel B, except colors represent the source of isolation: green, CSF; red, blood; dark blue, source other than blood or CSF. and 64 �g/ml), while the isolates were genetically nearly all iden- this opportunistic pathogenic yeast in the Netherlands, a set of 300 tical, with a minor difference for one microsatellite locus. C. neoformans isolates from 237 patients obtained during 1977 to Statistical analysis showed that genotype AFLP1 (serotype A) 2007 was investigated. isolates were significantly less susceptible to amphotericin B (0.26 Nearly half of the cryptococcosis patients were found to be HIV �g/ml) than isolates with genotypes AFLP2 (serotype D) and positive (n � 115; 48.5%). When patients with a known immune AFLP3 (serotype AD) (0.18 and 0.16 �g/ml, respectively) (P � status were considered, the percentage of HIV-positive patients 0.0001). AFLP2 isolates were found to be significantly (P � 0.003) was 62.8%, which is similar to data from other European studies. less susceptible to flucytosine (5.24 �g/ml) than genotype AFLP1 A French study showed that 77% (n � 177) of the patients with and AFLP3 isolates (3.30 and 3.81 �g/ml, respectively). Genotype cryptococcosis were HIV infected (17); in a cohort of 77 Austrian, AFLP2 isolates were significantly more susceptible to fluconazole, German, and Swiss patients this was 68% (n � 52) (40), a Spanish itraconazole, voriconazole, posaconazole (P � 0.0001 for each), study observed 87% (n � 48) (20), and a 30-month European and isavuconazole (P � 0.0022) than AFLP1 and AFLP3 isolates, survey reported 77% (n � 435) of HIV-positive patients (43). In as shown by their geometric mean MIC values (Table 4). How- the current study, 18 (7.6%) patients were found to have no un- ever, based on the geometric mean MIC values, only those of derlying disease or any other predisposing factors, such as corti- genotype AFLP2 isolates were found to differ by more than 1 log2 costeroid treatment. When corrected for patients with a known dilution step for fluconazole compared to AFLP1 and AFLP3 iso- immune status, the percentage of immunocompetent patients was lates, whereas genotype AFLP2 isolates also differed a log2 dilution found to be 9.9%. A French study observed nine (4%) immuno- step for itraconazole compared to genotype AFLP1 isolates. No competent patients within the studied cohort (17). Remarkably, significant differences in antifungal susceptibility patterns were several recent reports from far-east Asia noted that large numbers observed between the four patient categories. of cryptococcosis patients are immunocompetent, with values ranging from 13 to 92% of the investigated cohort of patients from DISCUSSION China, South Korea, Taiwan, and Vietnam (9–11, 14, 29, 34, 45). A cohort of Dutch HIV-positive cryptococcosis patients has been Interestingly, these immunocompetent patients were found to be investigated previously from a clinical epidemiological point of infected with C. neoformans var. grubii, which was also the case for view, but no data were presented on the Cryptococcus neoformans the majority (n � 20; 83.3%) of Dutch immunocompetent cryp- isolates (42). To obtain detailed insights into the epidemiology of tococcosis patients. Recent studies from South Korea and Viet-

49 Chapter 3

TABLE 4 Antifungal susceptibility per C. neoformans AFLP genotypea MIC (�g/ml)

Strain and antifungal agent Range MIC50 Geometric mean MIC90 C. neoformans variety grubii serotype A, AFLP1 (n � 245) Amphotericin B 0.125–1 0.25 0.26 0.5 Flucytosine 0.125–�64 4 3.3 8 Fluconazole 0.25–64 4 3.1 8 Itraconazole 0.016–0.5 0.125 0.09 0.25 Voriconazole 0.016–1 0.063 0.08 0.125 Posaconazole 0.016–0.5 0.063 0.06 0.125 Isavuconazole �0.016–0.5 0.031 0.04 0.125

C. neoformans variety neoformans serotype D, AFLP2 (n � 36) Amphotericin B 0.063–1 0.125 0.18 0.25 Flucytosine 0.5–64 4 5.24 16 Fluconazole 0.25–16 1 1.39 8 Itraconazole 0.016–0.5 0.031 0.04 0.25 Voriconazole 0.016–0.5 0.031 0.04 0.125 Posaconazole 0.016–0.25 0.031 0.04 0.125 Isavuconazole �0.016–0.25 0.031 0.02 0.125

C. neoformans variety neoformans serotype AD, AFLP3 (n � 14) Amphotericin B 0.063–1 0.125 0.16 0.25 Flucytosine 2–8 4 3.81 8 Fluconazole 1–32 4 4.2 8 Itraconazole 0.031–0.5 0.063 0.07 0.125 Voriconazole 0.031–0.25 0.063 0.08 0.125 Posaconazole 0.016–0.125 0.063 0.05 0.063 Isavuconazole 0.016–0.25 0.031 0.04 0.063

C. neoformans and C. gattii, all genotypes (n � 300) Amphotericin B 0.063–1 0.25 0.24 0.5 Flucytosine 0.125–�64 4 3.51 8 Fluconazole 0.25–64 4 2.87 8 Itraconazole 0.016–0.5 0.125 0.08 0.25 Voriconazole 0.016–1 0.063 0.07 0.125 Posaconazole 0.016–0.5 0.063 0.06 0.125 Isavuconazole �0.016–0.5 0.031 0.03 0.125 a The range, MIC50, geometric mean MIC, and MIC90 are listed for the seven tested antifungal compounds for each of the four genotypic C. neoformans groups. nam have observed a correlation between the immune status of genotype AFLP2 (serotype D) (20). Other epidemiological sur- the patient and the genotype of the isolated C. neoformans var. veys from Brazil, China, South Korea, Taiwan, and Vietnam grubii strain (10, 14). revealed that serotype D isolates are rarely found or are absent In contrast to C. neoformans var. grubii, the epidemiology of C. (2, 9–11, 29). neoformans var. neoformans has been poorly studied. From the few Until recently, it was believed that C. neoformans had a predi- available studies it is known that serotype D isolates are more lection to cause disease in immunocompromised patients, while prevalent in Europe (27, 43). In the current study, only 12% of the C. gattii was found to preferentially cause disease in immunocom- clinical isolates were serotype D (Table 3). A higher percentage of petent individuals (5). However, several studies observed that a isolates (20.5%) was found among 410 French C. neoformans iso- substantial number of cryptococcal infections in African HIV- lates (16). However, in Italy, 71% of the C. neoformans isolates infected patients are caused by C. gattii, thus indicating that the were found to be serotype D (41). A European survey observed presumed preference for a certain patient category is not valid (32, that a large proportion of the 311 C. neoformans isolates were 38). In the Netherlands, two patients were found to be infected by serotype D (30%) and serotype AD (19%) (43). A higher propor- C. gattii. One case was caused by a travel-related C. gattii AFLP6 tion of infections caused by C. neoformans var. neoformans was infection in an immunocompromised Dutch patient who devel- observed in several Mediterranean countries than in the north- oped cryptococcosis after corticosteroid treatment (22). The sec- western European area (20, 43). Furthermore, genotyping studies ond case was caused by a C. gattii AFLP4 isolate that was cultured revealed that 33.9% of the isolates from Madrid, Spain, fell into from a 54-year-old Dutch male who became HIV infected during the hybrid genotype cluster AFLP3 (serotype AD) and 19.6% in his stay in Africa. MLST analysis revealed that the latter isolate fell

50 Extensive genetic diversity withing the Dutch clinical Cryptococcus neoformans population

into a clade with African C. gattii AFLP4 isolates. This particular gal compounds inhibited the growth of C. neoformans. A global strain (N114) was further investigated by Ferry Hagen and Teun antifungal susceptibility study revealed that the observed resis- Boekhout (unpublished data). These cases suggest that C. gattii tance against amphotericin B, flucytosine, and fluconazole was can be dormant in the human host, and that it can be activated less than 1% of the tested isolates (36). Interestingly, these authors when the immune status is attenuated due to disease or treatment found that, during the time span of their study, C. neoformans with corticosteroids. Although C. gattii AFLP4 has been isolated became more susceptible to flucytosine and fluconazole. This phe- from environmental sources in the Netherlands, an autochtho- nomenon, however, has not been observed in the current study. It nous infection has not yet been observed in the current study (12). seems that the number of C. neoformans isolates resistant against Three Cryptococcus isolates from the current study were found to the conventional antifungal compounds remains limited. Several be interspecies hybrids between C. gattii AFLP4 and C. neoformans recently published antifungal susceptibility studies, including the AFLP2 and caused disease in an apparently healthy and an immu- current one, showed that posaconazole, voriconazole, and isavu- nocompromised patient, respectively (7). conazole are promising candidates to replace the conventional 3 Microsatellite typing has recently been applied to C. neofor- drugs, since the latter have potential toxic side effects (20, 21, 23, mans var. grubii and was shown to be an excellent tool to discrim- 36, 39, and this study). The present study shows that genotype inate isolates as well as to study their epidemiology (24, 25). This AFLP1 and AFLP3 isolates are less susceptible to fluconazole than molecular typing tool was first applied using a set of 122 clinical AFLP2 isolates. This genotypic difference was only recently ob- and 68 environmental C. neoformans var. grubii isolates from served in a Croatian study that found a similar result for flucona-

Cuba and distinguished 104 different genotypes (DCuba � 0.993) zole, although this difference was not signiﬁcant between AFLP2 and 11 MCs. The genetic diversity of the Dutch clinical C. neofor- and AFLP3 isolates (31). The current study shows that there is a mans var. grubii population is slightly higher than that observed in trend toward lower MIC values of amphotericin B for genotype

Cuba (DNetherlands � 0.994) with 196 genotypes among 259 stud- AFLP2 isolates, similarly to that observed in the Croatian and ied isolates, including the two novel clusters MC13 and MC14 Spanish studies (20, 31). The Croatian and Spanish studies did not (Fig. 1). This slightly higher genetic diversity of the Dutch C. neo- observe a significant difference between any of the tested antifun- formans var. grubii isolates may be due to one or more of the gal compounds versus the AFLP genotypes, except for Spanish following. First, the time span of this retrospectively study is 30 AFLP1 isolates that were less susceptible to amphotericin B than years, which is longer than the 20-year time frame of the Cuban Spanish AFLP3 isolates (20, 31). Similarly to our study, the Span- study. Second, the composition of the Dutch population is char- ish study did not observe a significant difference between patient acterized by extensive immigration, which includes people from categories versus antifungal susceptibility. the former overseas colonies of Surinam, Netherlands Antilles, In conclusion, C. neoformans var. grubii is the major cause of and Indonesia, as well as immigrants from northern Africa and cryptococcosis among immunocompromised and immunocom- eastern Europe (CBS Statistics Netherlands; www.cbs.nl; popula- petent individuals in the Netherlands. AFLP genotyping and mi- tion data for 1899 to 2010). It is likely that a proportion of the crosatellite typing showed that the Dutch clinical C. neoformans Dutch cryptococcosis patients are immigrants that acquired a var. grubii population is genetically more diverse than other re- subclinical infection with C. neoformans prior to immigration to cently studied populations. In vitro antifungal susceptibility test- the Netherlands, and this may contribute to the observed higher ing showed that resistance and decreased susceptibility are not genetic diversity in the studied cohort. Third, it has recently been major issues in the Netherlands, and the novel triazoles are prom- shown that C. neoformans can produce in vivo diploid offspring ising candidates for treatment strategies against cryptococcosis. from haploid parental isolates, via either cell fusion or endorepli- cation (15). Thus, a further possibility is that dormant C. neofor- ACKNOWLEDGMENTS mans isolates become activated upon immune suppression and We thank Collin Gerritzen and Bart Theelen (CBS-KNAW), Wendy that during the course of infection and/or antifungal therapy, C. Keijzers, Agaath Arends, and Virma Godfried (Netherlands Reference neoformans cells fuse to form hybrids that are more resistant to the Laboratory for Bacterial Meningitis, Amsterdam, The Netherlands), and antifungal compounds (14). The recent observation of clinical Maaike de Ruiter (CWZ) for excellent technical assistance. J.F.M. has been a consultant to Astellas, Basilea, Merck, and Schering- interspecies hybrids between C. neoformans and C. gattii from Plough and received speaker fees from Gilead, Janssen Pharmaceutica, clinical sources suggests that these hybrids can be formed in vivo Merck, Pfizer, and Schering-Plough. P.E.V. has received research grants (7, 8). from Pfizer, Gilead, Basilea, Merck, Bio-Rad, and Schering-Plough. Antifungal susceptibility testing has frequently been applied to J.W.M. has been a consultant to Astellas, Basilea, Merck, Pfizer, and C. neoformans. The development of new antifungal drugs, such as Wyeth and received speaker fees from Merck, Pfizer, and Wyeth. A.I.M.H. the novel triazoles and the reported increase in antifungal resis- is a member of the advisory board of MSD, VIIV, Janssen Pharmaceutica, tance, highlights the importance of monitoring the susceptibility and Gilead and received grants from MSD and Pfizer which are unrelated profiles of clinical C. neoformans isolates. The number of resistant to this study. C.H.K. received a research grant from Pfizer. All other au- and less susceptible isolates observed in the current study re- thors report no potential conflicts of interest. mained low. One isolate was found to be flucytosine resistant, and REFERENCES nine were less susceptible for this antifungal compound. High 1. Aminnejad M, et al. 2012. Identification of novel hybrids between Cryp- MICs remain exceptional; a few were found, including for the two tococcus neoformans var. grubii VNI and Cryptococcus gattii VGII. Myco- isolates from CSF and blood from the same patient, with at least a pathologia doi:10.1007/s11046-011-9491-x. 4 log dilution difference in MIC of flucytosine (8 and �64 �g/ 2. Barreto de Oliveira MT, et al. 2004. Cryptococcus neoformans shows a 2 remarkable genotypic diversity in Brazil. J. Clin. Microbiol. 42:1356– ml). In our collection, flucytosine resistance was rare, but there are 1359. countries where it is emerging, such as Indonesia (34). Nine iso- 3. Benson G. 1999. Tandem Repeats Finder: a program to analyze DNA lates were less susceptible for fluconazole, while all other antifun- sequences. Nucleic Acids Res. 27:573–580.

51 Chapter 3

4. Boekhout T, et al. 2001. Hybrid genotypes in the pathogenic yeast Cryp- of serial Cryptococcus neoformans isolates from Cuban patients with recur- tococcus neoformans. Microbiology 147:891–907. rent cryptococcal meningitis. BMC Infect. Dis. 10:e289. 5. Bovers M, Hagen F, Boekhout T. 2008. Diversity of the Cryptococcus 25. Illnait-Zaragozí MT, et al. 2010. Microsatellite typing of clinical and neoformans-Cryptococcus gattii species complex. Rev. Iberoam. Micol. 25: environmental Cryptococcus neoformans var. grubii isolates from Cuba S4–S12. shows multiple genetic lineages. PLoS One 5:e9124. 6. Bovers M, Hagen F, Kuramae EE, Boekhout T. 2008. Six monophyletic 26. Kidd SE, et al. 2004. A rare genotype of Cryptococcus gattii caused the lineages identified within Cryptococcus neoformans and Cryptococcus gattii cryptococcosis outbreak on Vancouver Island (British Columbia, Can- by multi-locus sequence typing. Fungal Genet. Biol. 45:400–421. ada). Proc. Natl. Acad. Sci. U. S. A. 101:17258–17263. 7. Bovers M, et al. 2006. Unique hybrids between the fungal pathogens 27. Kwon-Chung KJ, Bennett JE. 1984. Epidemiologic differences between Cryptococcus neoformans and Cryptococcus gattii. FEMS Yeast Res. 6:599– the two varieties of Cryptococcus neoformans. Am. J. Epidemiol. 120:123– 607. 130. 8. Bovers M, et al. 2008. AIDS patient death caused by novel Cryptococcus 28. Kwon-Chung KJ, Boekhout T, Fell JW, Diaz M. 2002. Proposal to neoformans � C. gattii hybrid. Emerg. Infect. Dis. 14:1105–1108. conserve the name Cryptococcus gattii against C. hondurianus and C. ba- 9. Chau TT, et al. 2010. A prospective descriptive study of cryptococcal cillisporus (Basidiomycota, Hymenomycetes, Tremellomycetidae). Taxon meningitis in HIV uninfected patients in Vietnam-high prevalence of 51:804–806. Cryptococcus neoformans var. grubii in the absence of underlying disease. 29. Liaw SJ, Wu HC, Hsueh PR. 2010. Microbiological characteristics of BMC Infect. Dis. 10:e199. clinical isolates of Cryptococcus neoformans in Taiwan: serotypes, mating 10. Choi YH, et al. 2010. Prevalence of the VNIc genotype of Cryptococcus types, molecular types, virulence factors, and antifungal susceptibility. neoformans in non-HIV-associated cryptococcosis in the Republic of Ko- Clin. Microbiol. Infect. 16:696–703. rea. FEMS Yeast Res. 10:769–778. 30. Meyer W, et al. 2009. Consensus multi-locus sequence typing scheme for 11. Chowdhary A, et al. 2011. In vitro antifungal susceptibility profiles and Cryptococcus neoformans and Cryptococcus gattii. Med. Mycol. 47:561– genotypes of 308 clinical and environmental isolates of Cryptococcus neo- 570. formans var. grubii and Cryptococcus gattii serotype B from north-western 31. Mlinaric´-Missoni E, et al. 2011. In vitro antifungal susceptibilities and India. J. Med. Microbiol. 60:961–967. molecular typing of sequentially isolated clinical Cryptococcus neoformans 12. Chowdhary A, et al. 2012. Temperate climate niche for Cryptococcus gattii strains from Croatia. J. Med. Microbiol. 60:1487–1495. in northern Europe. Emerg. Infect. Dis. 18:172–174. 32. Morgan J, et al. 2006. Cryptococcus gattii infection: characteristics and 13. Clinical Laboratory and Standards Institute. 2008. Reference method for epidemiology of cases identified in a South African province with high broth dilution antifungal susceptibility testing of yeasts. Approved stan- HIV seroprevalence, 2002–2004. Clin. Infect. Dis. 43:1077–1080. dard M27-A3. Clinical Laboratory and Standards Institute, Wayne, PA. 33. Ngamskulrungroj P, et al. 2009. Genetic diversity of the Cryptococcus 14. Day JN, et al. 2011. Most cases of cryptococcal meningitis in HIV- species complex suggests that Cryptococcus gattii deserves to have varieties. uninfected patients in Vietnam are due to a distinct amplified fragment PLoS One 4:e5862. length polymorphism-defined cluster of Cryptococcus neoformans var. 34. Pan W, et al. 2012. Resistance of Asian Cryptococcus neoformans serotype grubii VNI. J. Clin. Microbiol. 49:658–664. A is confined to few microsatellite genotypes. PLoS One 7:e32868. 15. Desnos-Ollivier M, et al. 2010. Mixed infections and in vivo evolution in 35. Park BJ, et al. 2009. Estimation of the current global burden of crypto- the human fungal pathogen Cryptococcus neoformans. mBio 1:e00091–10. coccal meningitis among persons living with HIV/AIDS. AIDS 23:525– 16. Dromer F, Mathoulin S, Dupont B, Letenneur L, Ronin O. 1996. Individual and environmental factors associated with infection due to 530. Cryptococcus neoformans serotype D. Clin. Infect. Dis. 23:91–96. 36. Pfaller MA, et al. 2005. Global trends in the antifungal susceptibility of 17. Dromer F, Mathoulin-Pélissier S, Launay O, Lortholary, and French Cryptococcus neoformans (1990 to 2004). J. Clin. Microbiol. 43:2163–2167. Cryptococcosis Study Group O. 2007. Determinants of disease presen- 37. Simpson EH. 1949. Measurement of diversity. Nature 163:688. tation and outcome during cryptococcosis: the CryptoA/D study. PLoS 38. Steele KT, Thakur R, Nthobatsang R, Steenhoff AP, Bisson GP. 2010. Med. 4:e21. In-hospital mortality of HIV-infected cryptococcal meningitis patients 18. Eliades NG, Eliades DG. 2009. Haplotype Analysis: software for analysis with C. gattii and C. neoformans infection in Gaborone, Botswana. Med. of haplotype data. Forest Genetics and Forest Tree Breeding, Georg- Mycol. 48:1112–1115. August University, Goettingen, Germany. http://www.uni-goettingen.de 39. Thompson GR, et al. 2009. Antifungal susceptibilities among different /en/134935.html. serotypes of Cryptococcus gattii and Cryptococcus neoformans. Antimicrob. 19. Fonseca Á, Boekhout T, Fell JW. 2011. Cryptococcus Vuillemin, p 1661– Agents Chemother. 53:309–311. 1738. In Kurtzman CP, Fell JW, Boekhout T (ed), The yeasts, a taxonomic 40. Tintelnot K, Lemmer K, Losert H, Schär G, Polak A. 2004. Follow-up of study, 5th ed. Elsevier, Amsterdam, The Netherlands. epidemiological data of cryptococcosis in Austria, Germany and Switzer- 20. Guinea J, et al. 2010. Antifungal susceptibility, serotyping, and genotyp- land with special focus on the characterization of clinical isolates. Mycoses ing of clinical Cryptococcus neoformans isolates collected during 18 years in 47:455–464. a single institution in Madrid, Spain. Med. Mycol. 48:942–948. 41. Tortorano AM, et al. 1997. Prevalence of serotype D in Cryptococcus 21. Hagen F, et al. 2010. In vitro antifungal susceptibilities and amplified neoformans isolates from HIV positive and HIV negative patients in Italy. fragment length polymorphism genotyping of a worldwide collection of Mycoses 40:297–302. 350 clinical, veterinary, and environmental Cryptococcus gattii isolates. 42. Van Elden LJ, et al. 2000. Declining number of patients with cryptococ- Antimicrob. Agents Chemother. 54:5139–5145. cosis in the Netherlands in the era of highly active antiretroviral therapy. 22. Hagen F, van Assen S, Luijckx GJ, Boekhout T, Kampinga GA. 2010. AIDS 14:2787–2788. Activated dormant Cryptococcus gattii infection in a Dutch tourist who 43. Viviani MA, et al. 2006. Molecular analysis of 311 Cryptococcus neofor- visited Vancouver Island (Canada): a molecular epidemiological ap- mans isolates from a 30-month ECMM survey of cryptococcosis in Eu- proach. Med. Mycol. 48:528–531. rope. FEMS Yeast Res. 6:614–619. 23. Illnait-Zaragozí MT, et al. 2008. In vitro activity of the new azole isavu- 44. Warkentien T, Crum-Cianflone NF. 2010. An update on Cryptococcus conazole (BAL4815) compared with six other antifungal agents against among HIV-infected patients. Int. J. Sex. Transm. Dis. AIDS 21:679–684. 162 Cryptococcus neoformans isolates from Cuba. Antimicrob. Agents 45. Zhu LP, et al. 2010. Cryptococcal meningitis in non-HIV-infected pa- Chemother. 52:1580–1582. tients in a Chinese tertiary care hospital, 1997–2007. Med. Mycol. 48:570– 24. Illnait-Zaragozí MT, et al. 2010. Microsatellite typing and susceptibilities 579.

52 Chapter 4 Microsatellite typing and susceptibilities of serial Cryptococcus neoformans isolates from Cuban patients with recurrent cryptococcal meningitis

Published in BMC Infectious Diseases 2010;10:289. Chapter 4

Microsatellite typing and susceptibilities of serial Cryptococcus neoformans isolates from Cuban patients with recurrent cryptococcal meningitis María T Illnait-Zaragozí1, Gerardo F Martínez-Machín1, Carlos M Fernández-Andreu1, Ferry Hagen2, Teun Boekhout2, Corné HW Klaassen3, Jacques F Meis3*

Abstract Background: Cryptococcus neoformans is commonly associated with meningoencephalitis in immunocompromised patients and occasionally in apparently healthy individuals. Recurrence of infection after initial treatment is not uncommon. We studied C. neoformans isolates from 7 Cuban patients with recurrent cryptococcal meningitis. Antifungal susceptibility and genotyping with microsatellite molecular typing were carried out. Methods: Isolates (n = 19) were recovered from cerebrospinal fluid, blood, urine and semen. Antifungal susceptibilities for amphotericin B, fluconazole, flucytosine, itraconazole, voriconazole, posaconazole and isavuconazole were tested by CLSI M27A3 broth microdilution method. Genotyping was done using a panel of 9 microsatellite (STR) markers: (CT)n, (TG)n, (TA)n, (CTA)n, (TCT)n, (CCA)n, (TTAT)n, (ATCC)n and (TATT)n. Results: The average number of isolates/patient was 2.71. The mean time interval between the collection of any two isolates was 52.5 days. All strains were identified as C. neoformans var. grubii (serotype Aa). Although none of the strains were resistant to the studied drugs, in serial isolates from two patients, MICs values of triazoles increased 4-5 log2 dilutions over time. STR patterns showed 14 distinctive profiles. In three patients the recurrent infection was associated with genotypically identical isolates. The four other patients had relapse isolates which were genotypically different from the initial infecting strain. Conclusion: Recurrences of cryptococcal meningitis in our series of patients was not associated with development of drug resistance of the original strain but by an initial infection with different strains or a reinfection with a new strain.

Background relapse despite effective antifungal therapy. Early studies The incidence of cryptococcosis started to increase with performed with karyotyping and restriction fragment the beginning of the acquired immune deficiency syn- length polymorphism (RFLP) analysis of serial isolates of drome (AIDS) epidemic in the early 1980 s. Its fre- AIDS patients in New York concluded that recurrent quency declined in the Western world since the mid infection was caused through persistence of the original 1990 s due to the use of highly active antiretroviral ther- strain and not from infection with new strains [3,4]. A apy (HAART). However, cryptococcal meningitis is still later study using the same typing technique but with one of the most common life-threatening opportunistic isolates from AIDS patients from Uganda found that fungal infections in immunocompromised patients, par- among 17 patients with more than 1 cerebrospinal fluid ticularly among those with AIDS in Sub-Saharan Africa (CSF) isolate of Cryptococcus neoformans, sequential iso- [1] and Asia [2]. These patients show a high tendency to lates were identical or highly related in 12 patients [5]. Several treatment strategies have been used in patients with cryptococcal meningitis but the optimum regimen * Correspondence: [email protected] is still not clear. Amphotericin B with or without flucy- 3Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, Nijmegen, The Netherlands tosine remain the agents of choice for induction therapy Full list of author information is available at the end of the article

© 2010 Illnait-Zaragozí et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

54 Microsatellite typing and susceptibilities of serial Cuban Cryptococcus neoformans isolates

while fluconazole has proven to be superior for long- flucytosine (Valeant Pharmaceuticals, The Netherlands), term maintenance therapy [6,7]. High dose fluconazole fluconazole (Pfizer Central Research, U.K.), itraconazole with flucytosine is an oral treatment alternative although (Janssen-Cilag, The Netherlands), voriconazole (Pfizer this regimen is not as effective as amphotericin B and Central Research, U.K.), posaconazole (Schering Plough, flucytosine [8]. Different molecular typing methods have USA), and isavuconazole (Basilea Pharmaceutica, been used in the epidemiological analyses of clinical Switzerland) were used. The stock solutions of the drugs and/or environmental isolates of C. neoformans, includ- were prepared in the appropriate solvent. The final con- ing electrophoretic karyotyping, PCR fingerprinting, ran- centrations of the antifungal agents were 0.016 to 8 μg/ dom amplified polymorphic DNA analysis, RFLP mL for amphotericin B, itraconazole, voriconazole, and analysis, MLST and amplified fragment length poly- posaconazole; 0.063 to 32 μg/mL for flucytosine and flu- morphism analysis with divergent results [9-11]. To conazole; and 0.004 to 4.00 μg/mL for isavuconazole. determine whether recurrences of cryptococcal meningi- After 72 h incubation at 35°C the minimum inhibitory tis in seven Cuban patients were due to development of concentration (MIC) was defined as the lowest concen- drug resistance or to infection by multiple strains, we tration of drug showing absence of growth for amphoter- 4 tested the in vitro antifungal susceptibility and deter- icin B and a prominent reduction of growth (≥50%) for mined the genotypes of the sequential isolates using a the other antifungal agents compared to the drug-free recently described microsatellite based assay [12]. growth control. The MICs were read optically and spectrophotometrically at 420 nm after agitation. Candida Methods parapsilosis ATCC 22019 and C. krusei ATCC 6258 were Fungal isolates used as quality control [13]. From the stock collection of the mycology laboratory at the Instituto de Medicina Tropical “Pedro Kourí” (IPK) Mating-, sero- and genotyping in Havana, Cuba, 19 clinical isolates of C. neoformans C. neoformans isolates were grown on Sabouraud’s dex- from patients with recurrent infection were selected for trose agar at 30°C for 48 h and DNA was obtained from this study. The isolates were recovered at different time freshly grown cells using a MagNA lyser/MagNA Pure intervals from cerebrospinal fluid (n = 16), blood, urine protocol (Roche Diagnostics, Almere, the Netherlands). and semen (one each) from 7 patients (6 HIV positive The mating- and serotype was determined using four and 1 HIV negative) admitted at the institute between different PCRs that specifically amplify the mating-type 1995 and 2001, just before HAART was introduced in a or a allele of the STE20 locus for either serotype A or Cuba (Table 1). All patients were admitted to the clini- D isolates [14]. The reference isolates CBS9172 (aA), cal AIDS care division at the IPK. Initial isolates were CBS8710 (aA), CBS10511 (aD) and CBS10513 (aD) obtained at diagnosis and before any antifungal therapy were included as positive control for each of the four was used and follow up isolates were from patients dur- PCRs. ing or after treatment with antifungal drugs. Repeat STR analysis was performed in two steps as described lumbar puncture was only performed in those cases previously [12]: i) Amplification of STR loci by PCR: which did not react within two weeks of therapy or three separate multiplex PCRs were used (CNA2, those who showed a clinical deteriorisation during or CNA3, and CNA4, respectively), each amplifying three after treatment. From each culture positive sample a sin- different STRs. For every multiplex PCR, one of the gle isolate from morphological similar colonies was amplification primers was labelled with carboxyfluores- archived. Species identification was initially done by cein (FAM), hexachlorofluorescein (HEX), or tetrachlor- growth on canavanine-glycine-bromothymol blue (CGB) ofluorescein (TET) at the 5’ end, respectively. In agar and strains were stored in sterile water at room addition to the amplification primers (concentrations temperature until the study was carried out. From initial according to reference [11]), each PCR mixture con- isolation until the study the isolates were typically sub- tained 0.2 mM deoxynucleoside triphosphates, 1 U of cultured between 4 and 6 times. Before use, the identifi- FastStart Taq DNA polymerase (Roche Diagnostics), 2 cation of all isolates was repeated with a commercial mM MgCl2 and 1 ng of genomic DNA in 1 × reaction identification system (Auxacolor 2; Bio-Rad, Marnes-la- buffer. Thermocycling was performed in a T1 thermocy- Coquette, France). cler (Biometra, Göttingen, Germany) by using the following thermal protocol: 10 min of denaturation at 95° Antifungal agents and susceptibility testing C, followed by 35 cycles of 30 s of denaturation at 95°C, Broth microdilution testing was performed in accordance 30 s of annealing at 60°C, and 1 min of extension at 72° with Clinical and Laboratory Standards Institute docu- C. Before the reaction mixtures were cooled to room ment M27-A3 guidelines [13]. Standard antifungal pow- temperature, an additional incubation for 10 min at 72° ders of amphotericin B (Sigma, The Netherlands), C was performed. All temperature transitions were

55 Chapter 4

Table 1 Clinical data, origin and date of isolation of the studied strains Patient Sex HIV/ Sample Isolation Interval between each Total of days Antifungal Nr. Year of diagnosis date isolate in days treatment (number) 1 Male +/1991 Blood 14/04/95 0 0 AmB+Flu (08-36-09-92) Flu (maintenance) CSF 14/04/95 0 0 AmB+Flu (08-36-09-75) Flu (maintenance) CSF 18/07/95 96 96 AmB+Flu (08-36-10-01) 2 Male +/1990 CSF 05/09/95 0 0 Flu+Itr (08-36-09-97) Urine 08/12/95 95 95 Flu (08-36-09-70) CSF 10/01/96 34 129 Flu (08-36-09-91) 3 Female +/1994 CSF 18/01/96 0 0 AmB+Flu (08-36-09-71) Flu (maintenance) CSF 29/02/96 34 34 AmB+Flu (08-36-09-72) 4 Male +/1995 CSF 23/04/97 0 0 AmB+FC+Flu (08-36-09-82) CSF 06/05/97 14 14 AmB+FC+Flu (08-36-09-89) 5 Male +/1998 CSF 15/04/00 0 0 AmB (08-36-09-24) Flu (maintenance) CSF 26/10/00 194 194 AmB+FC (08-36-09-25) 6 Female +/1996 CSF 12/06/00 0 0 AmB+FC (08-36-10-56) Flu (maintenance) CSF 12/10/00 123 123 AmB (08-36-10-52) Flu (maintenance) CSF 30/10/00 19 142 AmB+FC (08-36-10-53) 7 Male - CSF 09/01/01 0 0 AmB+Itr (08-36-09-26) Flu (maintenance) CSF 17/04/01 99 99 AmB+Flu (08-36-09-30) Flu (maintenance) CSF 07/06/01 52 151 Flu+Itr (08-36-09-78) Semen 19/06/01 13 164 Liposomal AmB (08-36-09-42)

performed with maximal heating and cooling settings (5° differed in up to two loci were considered to be geneti- C/s). ii) Detection and sizing of amplification products cally related [12]. Genotypes differing in more than two with subsequent assignment of repeat numbers: the loci were considered to be unrelated. The study was fragments obtained were combined with the ET550-R approved by the Scientific Council and Ethics Commit- size standard (GE Healthcare, Diegem, Belgium) and tee of the Instituto Pedro Kouri, Havana, Cuba. analyzed on a MegaBACE 500 automated DNA platform (GE Healthcare), according to the instructions of the Results manufacturer. Electropherograms were analyzed using The majority of strains were obtained from HIV positive Fragment Profiler 1.2 software (GE Healthcare). Identi- patients, except those recovered from patient 7 for whom cal isolates were those that possessed alleles with the no underlying disease could be demonstrated (Table 1). same number of repeat units in all nine loci. Consistent The mean number of isolates/patient was 2.71 (range 2-4 with the previous separation of genotypes into microsa- isolates/patients). The mean time between collection of tellites complexes (MC’s), isolates with genotypes that any two isolates was 52.5 days (range 13-123 days). All

56 Microsatellite typing and susceptibilities of serial Cuban Cryptococcus neoformans isolates

patients received antifungal treatment but only the STR patterns of the 19 isolates showed 14 distinct HIV negative patient (nr. 7) survived. All clinical profiles. Three patients (patients 2, 3 and 6) had genoty- strains were identified as C. neoformans var. grubii ser- pically identical isolates over the course of time. The otype A and mating-type a. serial isolates from patient 6 were genotypically identical For the antifungal susceptibility testing, no differences but also showed the largest difference in MIC for the between visual and spectrophotometric readings were triazoles including drugs which have not been used observed, and the MICs for the quality control strains (itraconazole, voriconazole, isavuconazole). The other were all within the suggested reference ranges (data not four patients were probably infected by more than one shown). Table 2 summarizes the in vitro susceptibilities genotype (Figure 1) although this might be biased of the isolates according to the origin of the strains. because from each positive CSF culture only one colony When all data were considered together, the widest was archived for future study. ranges and highest MICs were for fluconazole (0.25 to 8 μg/mL) and flucytosine (0.5 to 8 μg/mL) and the lowest Discussion were for isavuconazole, voriconazole, and posaconazole. No previous studies of the antifungal susceptibility and 4 Amphotericin B, posaconazole and isavuconazole exhib- genetic diversity of sequential isolates of C. neoformans ited similar MICs patterns among all the studied iso- obtained from individual patients have been done in lates. Isolates from patients 6 and 7 exhibited a stepwise Cuba. Here we studied 19 serial clinical strains of C. increase among serial isolates for some of the drugs. neoformans from 7 patients with recurrent cryptococcal Increased MICs values of at least 4 log2 dilutions over meningitis during the pre-HAART period by analyzing the time for fluconazole, itraconazole, voriconazole and their antifungal susceptibility and molecular profiles. isavuconazole were observed in patient 6. In patient 7 Our results on the in vitro activities of the main anti- higher MICs were found between initial and last isolate fungal drugs are similar to those published previously for fluconazole (4 log2), itraconazole (6 log2) and vorico- [15-17]. Amphotericin B has long successfully been used nazole (5 log2). Amphotericin B, flucytosine and posaco- to treat various yeasts and mould infections. In this ser- nazole demonstrated maximally only a 1-3 log2 difference ies there was only one log2 dilution difference between among serial isolates. the sequential isolates with a highest observed MIC of

Table 2 Minimum inhibitory concentration of all C. neoformans var. grubii isolates for seven antifungal drugs Patient Nr. Strain Minimum Inhibitory Concentration (μg/mL) reading at 72 h Nr. AmB FC Flu Itr Vor Pos Isa 1 08-36-09-92 0.125 8 4 0.125 0.016 0.125 0.016 08-36-09-75 0.25 2 4 0.063 < 0.016 0.063 < 0.004 08-36-10-01 0.25 4 4 0.25 0.125 0.125 < 0.004 2 08-36-09-97 0.125 2 1 0.031 0.063 0.125 < 0.004 08-36-09-70 0.25 2 2 0.063 0.063 0.125 0.016 08-36-09-91 0.125 1 1 0.031 < 0.016 0.063 0.008 3 08-36-09-71 0.25 2 2 0.063 0.063 0.016 0.016 08-36-09-72 0.125 8 2 0.125 0.125 0.031 < 0.004 4 08-36-09-82 0.25 2 2 0.031 0.031 0.031 0.008 08-36-09-89 0.25 1 4 0.031 0.125 0.063 0.008 5 08-36-09-24 0.25 2 1 < 0.016 < 0.016 0.016 0.004 08-36-09-25 0.25 1 0.25 0.031 < 0.016 0.016 < 0.004 6 08-36-10-56 0.125 0.5 0.25 < 0.016 < 0.016 0.063 < 0.004 08-36-10-52 0.125 0.5 4 0.125 0.125 0.016 0.031 08-36-10-53 0.25 1 8 0.125 0.25 0.016 0.063 7 08-36-09-26 0.25 0.5 0.5 < 0.016 < 0.016 0.031 0.016 08-36-09-30 0.25 0.5 0.5 0.031 0.031 0.016 0.004 08-36-09-78 0.125 1 1 0.031 0.063 0.016 < 0.004 08-36-09-42 0.25 4 8 0.5 0.25 0.031 < 0.004 Abreviations: AmB: amphotericin B, FC; flucytosine, Flu; fluconazole, Itr; itraconazole, Vor; voriconazole, Pos; posaconazole, Isa; isavuconazole.

57 Chapter 4

Figure 1 Details of the 19 C. neoformans isolates from 7 patients and relationship between the obtained genotypes. The numbers below the genotype correspond to the number of repetitions observed in markers CNA2a, CNA2b, CNA2c, CNA3a, CNA3b, CNA3c, CNA4a, CNA4b and CNA4c, respectively. A hyphen indicates that no result was obtained. The dendrogram is based on a categorical analysis using UPGMA clustering. The scale bar indicates the percentage similarity.

0.25 μg/mL way below the suggested breakpoint for susceptibility profiles. Because the short time in between resistance of MIC ≥2 μg/mL, which was found to be each isolation it is suggestive that this was not a recur- associated with therapeutic failure [18]. All flucytosine rent infection, it is possible that the patient was infected MICs were < 16 μg/mL which are regarded as suscepti- simultaneously with both strains. Patients 2 and 5 had ble [7,14]. Among the azoles, fluconazole showed the three and two isolates respectively with similar suscept- lowest in vitro activity. In fact, previous reports have ibility patterns. In the first case only the CSF isolates already demonstrated the low activity of this drug were genotypically similar. This might imply that this against C. neoformans isolates, even though it has pro- patient was also infected simultaneously with more than ven to be more active in vivo [19]. According to these one strain however there might be a bias because mixed authors, the good therapeutic results obtained are lar- infections in one sample of morphologically similar gely attributable to its high concentrations in cerebrosp- cryptococci might have been missed. In the second case, inal fluid. Although no resistance has been found in the isolates differed in one marker and were thus considered present collection of isolates (MIC < 16 μg/mL) [20], a to be genetically related. These observations suggest stepwise increase of 4 and 5 dilutions was found in 2 microevolution of C. neoformans during human infec- patients suggesting development of reduced levels of tion. This process may allow the fungal population to antifungal susceptibility [7]. In one patient (nr. 6) the change and escape eradication by the immune system, sequential isolates were genotypically identical. Of inter- and thus cause chronic infections as suggested by Jain et est is the finding that the increase of fluconazole MICs al. [24]. Patient 1 had two baseline isolates, from blood over time in patients 6 and 7 parallels data with itraco- and CSF, that had similar susceptibility but were genoty- nazole, voriconazole and isavuconazole but not with pically different. Since both genotypes were isolated posaconazole. These observations support previous work simultaneously this could mean that the patient was that suggests the development of cross-resistance of flu- infected with both strains at time of first sampling. As conazole with other triazoles [20]. Isavuconazole is an has been stated before we might have missed mixed experimental broad-spectrum antifungal triazole active infection at baseline because not all (morphologically) against clinically relevant yeasts and moulds [21]. Our similar colonies were studied. After more than three results are in agreement with previously published stu- months of fluconazole maintenance therapy, the patient dies which demonstrated a high in vitro activity of this was re-admitted because signs and symptoms of relapse. drug for C. neoformans [17,22]. It has been demon- The CSF isolate obtained at that time had higher MIC strated that there are no trends towards higher MICs values for voriconazole and a different genotype com- for strains isolated from patients who failed to respond pared with the previous isolates. These findings suggest to a given therapy compared to isolates from patients two possibilities: i) the patient was re-infected with a who did not [23]. On the other hand, patient 4 was new strain during the maintenance therapy or ii) the infected with two different genotypes with similar initial strains underwent genetic microevolution.

58 Microsatellite typing and susceptibilities of serial Cuban Cryptococcus neoformans isolates

Cerebrospinal fluid from patient 7 remained microscopi- Authors’ contributions cally (Indian ink) and culture positive during more than MTIZ and FH carried out the susceptibility testing and molecular typing, and drafted the manuscript. MTIZ GFMM and CMFA participated in the design of 5 months, despite antifungal therapy. Only the second the study, contributed materials and cared for the clinical management of and third isolate showed the same genotype which was the patients. TB CHWK and JFM conceived of the study, and participated in different from the first and the fourth isolate. This could its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript. be explained by simultaneous infection with more than one strain or by re-infection with a new strain during Competing interests treatment which developed genetic changes over time. JFM received grants form Astellas, Basilea, Merck, and Schering-Plough. He has been a consultant to Basilea, Merck and Schering-Plough and received This is corroborated by the increased MICs values for speakers fees from Gilead, Janssen Pharmaceutica, Merck, Pfizer, and fluconazole, itraconazole and voriconazole of the last Schering-Plough. CHWK received a grant from Pfizer. All other authors: no isolate. potential conflicts of interest. Other authors have studied the molecular relationship Received: 20 May 2010 Accepted: 4 October 2010 of C. neoformans isolates obtained from the same epi- Published: 4 October 2010 sode of infection or during a recurrent infection with References 4 several different techniques [10,11,24-26]. Although the 1. Bicanic T, Harrison TS: Cryptococcal meningitis. Br Med Bull 2005, obtained results are variable, most authors suggest that 72:99-118. persistence or recurrence of the infection is caused by 2. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, Chiller TM: relapse rather than re-infection and/or microevolution Estimation of the current global burden of cryptococcal meningitis among persons living with HIV/AIDS. AIDS 2009, 23:525-30. of the original isolate [20,25-28]. Multiple strain infec- 3. Spitzer ED, Spitzer SG, Freundlich LF, Casadevall A: Persistence of initial tion was rarely considered until a recent study reported infection in recurrent Cryptococcus neoformans meningitis. Lancet 1993, a high frequency of mixed infections in 20% of patients 341:595-596. 4. Casadevall A, Spitzer ED, Webb D, Rinaldi MG: Susceptibilities of serial with cryptococcal disease and speculated that multiple Cryptococcus neoformans isolates from patients with recurrent strains could be exogenously acquired from the environ- cryptococcal meningitis to amphotericin B and fluconazole. Antimicrob ment, either simultaneous or sequentially [29]. A bias Agents Chemother 1993, 37:1383-1386. 5. Pfaller M, Zhang J, Messer S, Tumberland M, Mbidde E, Jessup C, was entered in our study because only one colony was Ghannoum M: Molecular epidemiology and antifungal susceptibility of selected for archiving from each sampling point Cryptococcus neoformans isolates from Ugandan AIDS patients. Diagn (because all growing colonies were morphologically Microbiol Infect Dis 1998, 32:191-199. 6. Dromer F, Bernede-Bauduin C, Guillemot D, Lortholary O, French similar). Morphological similar colonies might have dif- Cryptococcosis Study Group: Major role for amphotericin B-flucytosine ferent genotypic backgrounds as has been shown combination in severe cryptococcosis. PLoS One 2008, 3(8):e2870. recently [29]. 7. Perfect JR, Dismukes WE, Dromer F, Goldman DL, Graybill JR, Hamill RJ, Harrison TS, Larsen RA, Lortholary O, Nguyen MH, Pappas PG, Powderly WG, Singh N, Sobel JD, Sorrell TC: Clinical practice guidelines for the Conclusions management of cryptococcal disease: 2010 update by the Infectious STR typing as presented in this study has not been lar- Diseases Society of America. Clin Infect Dis 2010, 50:291-322. 8. Nussbaum JC, Jackson A, Namarika D, Phulusa J, Kenala J, Kanyemba C, gely used for Cryptococcus molecular characterization. Jarvis JN, Jaffar S, Hosseinipour MC, Kamwendo D, van der Horst CM, This new typing technique allowed the observation of Harrison TS: Combination flucytosine and high-dose fluconazole high genetic variability among the studied clinical iso- compared with fluconazole monotherapy for the treatment of cryptococcal meningitis: a randomized trial in Malawi. Clin Infect Dis 2010, lates keeping in mind the previous mentioned limitation 50:338-344. of this study and we confirm the observation of within- 9. Fusco-Almeida AM, Teruyuki-Matsumoto M, Baeza LC, Bellan de Oliveira R, host strain diversity [29]. Recurrence of infection in the Pizzirani-Kleiner AA, de Souza Carvalho-Melhem M, Mendes-Giannini MJS: Molecular typing and antifungal susceptibility of clinical sequential majority of these patients were not associated with drug isolates of Cryptococcus neoformans from Sao Paulo State, Brazil. FEMS resistance but probably by co-infection with different Yeast Res 2007, 7:152-164. strains or strains genetically modified during the long 10. Haynes KA, Sullivan DJ, Coleman DC, Clarke JC, Emilianus R, Atkinson C, Cann KJ: Involvement of multiple Cryptococcus neoformans strains in a maintenance therapy. single episode of cryptococcosis and reinfection with novel strains in recurrent infection demonstrated by random amplification of polymorphic DNA and DNA fingerprinting. J Clin Microbiol 1995, Acknowledgements 33:99-102. This study was supported by a research and travel grant from the 11. Sullivan D, Haynes K, Moran G, Shanley D, Coleman D: Persistence, International Society for Human and Animal Mycology to MTIZ. We thank M replacement, and microevolution of Cryptococcus neoformans strains in Perurena-Lancha for her assistance. The funding agency had no role in study recurrent meningitis in AIDS patients. J Clin Microbiol 1996, 34:1739-1744. design, data collection and analysis, decision to publish, or preparation of 12. Illnait-Zaragozi MT, Martínez-Machín GF, Fernández-Andreu CM, Boekhout T, the manuscript. Meis JF, Klaassen CH: Microsatellite typing of clinical and environmental Cryptococcus neoformans var. grubii isolates from Cuba shows multiple Author details genetic lineages. PLoS One 2010, 9;5(2):e9124. 1Department of Bacteriology and Mycology, Instituto Pedro Kourí, Havana, 13. Clinical and Laboratory Standards Institute: Reference method for Cuba. 2CBS-Fungal Biodiversity Centre, Utrecht, The Netherlands. broth dilution antifungal susceptibility testing of yeasts. Approved 3Department of Medical Microbiology and Infectious Diseases, Canisius standard. M27-A3. Clinical and Laboratory Standards Institute, Wayne, PA ,3 Wilhelmina Hospital, Nijmegen, The Netherlands. 2008.

59 Chapter 4

14. Barreto de Oliveira MT, Boekhout T, Theelen B, Hagen F, Baroni FA, Lazera MS, Lengeler KB, Heitman J, Rivera IN, Paula CR: Cryptococcus doi:10.1186/1471-2334-10-289 neoformans shows a remarkable genotypic diversity in Brazil. J Clin Cite this article as: Illnait-Zaragozí et al.: Microsatellite typing and Microbiol 2004, 42:1356-1359. susceptibilities of serial Cryptococcus neoformans isolates from Cuban 15. Pfaller MA, Messer SA, Boyken L, Hollis RJ, Rice C, Tendolkar S, Diekema DJ: patients with recurrent cryptococcal meningitis. BMC Infectious Diseases In vitro activities of voriconazole, posaconazole, and fluconazole against 2010 10:289. 4,169 clinical isolates of Candida spp. and Cryptococcus neoformans collected during 2001 and 2002 in the ARTEMIS global antifungal surveillance program. Diagn Microbiol Infect Dis 2004, 48:201-205. 16. Messer SA, Moet GJ, Kirby J, Jones RN: Activity of contemporary antifungal agents, including the novel echinocandin anidulafungin, tested against Candida spp., Cryptococcus spp., and Aspergillus spp.: Report from the SENTRY Antimicrobial Surveillance Program (2006 to 2007). J Clin Microbiol 2009, 47:1942-1946. 17. Illnait-Zaragozi MT, Martínez FG, Curfs-Breuker I, Fernández CM, Boekhout T, Meis JF: In vitro activity of the new azole isavuconazole (BAL4815) compared with six other antifungal agents against 162 Cryptococcus neoformans isolates from Cuba. Antimicrob Agents Chemother 2008, 52:1580-1582. 18. Lozano-Chiu M, Paetznick VL, Ghannoum MA, Rex JH: Detection of resistance to amphotericin B among Cryptococcus neoformans clinical isolates: performances of three different media assessed by using E-test and National Committee for Clinical Laboratory Standards M27-A methodologies. J Clin Microbiol 1998, 36:2817-2822. 19. Aller AI, Martin-Mazuelos E, Lozano F, Gomez-Mateos J, Steele-Moore L, Holloway WJ, Gutierrez MJ, Recio FJ, Espinel-Ingroff A: Correlation of fluconazole MICs with clinical outcome in cryptococcal infection. Antimicrob Agents Chemother 2000, 44:1544-1548. 20. Mondon P, Petter R, Amalfitano G, Luzzati R, Concia E, Polacheck I, Kwon- Chung KJ: Heteroresistance to fluconazole and voriconazole in Cryptococcus neoformans. Antimicrob Agents Chemother 1999, 43:856-861. 21. Guinea J, Bouza E: Isavuconazole: a new and promising antifungal triazole for the treatment of invasive fungal infections. Future Microbiol 2008, 3:603-615. 22. Thompson GR III, Wiederhold NP, Fothergill AW, Vallor AC, Wickes BL, Patterson TF: Antifungal susceptibilities among different serotypes of Cryptococcus gattii and Cryptococcus neoformans. Antimicrob Agents Chemother 2009, 53:309-311. 23. Dannaoui E, Abdul M, Arpin M, Nguyen AM, Piens MA, Favel A, Lortholary O, Dromer F: Results obtained with various antifungal susceptibility testing methods do not predict early clinical outcome in patients with cryptococcosis. Antimicrob Agents Chemother 2006, 50:2464-2470. 24. Jain N, Wickes BL, Keller SM, Fu J, Casadevall A, Jain P, Ragan MA, Banerjee U, Fries BC: Molecular epidemiology of clinical Cryptococcus neoformans strains from India. J Clin Microbiol 2005, 43:5733-5742. 25. Litvintseva A P, Kestenbaum L, Vilgalys R, Mitchell TG: Comparative analysis of environmental and clinical populations of Cryptococcus neoformans. J Clin Microbiol 2005, 43:556-564. 26. Igreja RP, Dos Santos-Lazera M, Wanke B, Gutierrez-Galhardo MC, Kidd SE, Meyer W: Molecular epidemiology of Cryptococcus neoformans isolates from AIDS patients of the Brazilian City, Rio de Janeiro. Med Mycol 2004, 42:229-238. 27. Fries BC, Casadevall A: Serial isolates of Cryptococcus neoformans from patients with AIDS differ in virulence for mice. J Infect Dis 1998, 178:1761-1766. 28. Friese G, Discher T, Füssle R, Schmalreck A, Lohmeyer J: Development of azole resistance during fluconazole maintenance therapy for AIDS- associated cryptococcal disease. AIDS 2001, 15:2344-2345. Submit your next manuscript to BioMed Central 29. Desnos-Ollivier M, Patel S, Spaulding AR, Charlier C, Garcia-Hermoso D, and take full advantage of: Nielsen K, Dromer F: Mixed infections and in vivo evolution in the human fungal pathogen Cryptococcus neoformans. mBio 1(1), pii: e00091- 10. • Convenient online submission • Thorough peer review Pre-publication history • No space constraints or color ﬁgure charges The pre-publication history for this paper can be accessed here: http://www.biomedcentral.com/1471-2334/10/289/prepub • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

60 Chapter 5 In vitro activity of the new azole isavuconazole (BAL4815) compared with six other antifungal agents against 162 Cryptococcus neoformans isolates from Cuba

Published in Antimicrobial Agents and Chemotherapy 2008;52:1580-2. Chapter 5

In Vitro Activity of the New Azole Isavuconazole (BAL4815) Compared with Six Other Antifungal Agents against 162 Cryptococcus neoformans Isolates from Cuba� Maria-Teresa Illnait-Zaragozi,1,2 Gerardo F. Martı´nez,2 Ilse Curfs-Breuker,1 Carlos M. Ferna´ndez,2 Teun Boekhout,3 and Jacques F. Meis1* Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, Nijmegen, The Netherlands1; Department of Medical Microbiology, Laboratory of Mycology, Institute of Tropical Medicine Pedro Kouri, Havana, Cuba2; and Centraalbureau voor Schimmelcultures, Utrecht, The Netherlands3

Received 26 October 2007/Returned for modiﬁcation 19 November 2007/Accepted 11 January 2008

Cuban Cryptococcus isolates (n � 165) were tested in vitro against amphotericin B, ﬂucytosine, ﬂuconazole,

itraconazole, voriconazole, posaconazole, and isavuconazole, giving MIC90 values of 0.25, 8, 4, 0.25, 0.125, 0.016, and 0.016 �g/ml, respectively. Isavuconazole and posaconazole seem to be potentially active drugs for treating cryptococcal infections.

Antifungals available for therapy of cryptococcosis are lim- were stored on sterile water at room temperature until the ited to amphotericin B, flucytosine, and fluconazole; however, study was carried out. Before use, the identities of all isolates the side effects associated with administration of amphotericin were confirmed with a commercial identification system B and flucytosine may restrict their use (17, 25). A new gen- (Auxacolor 2; Bio-Rad, Marnes-la-Coquette, France). Can- eration of triazoles, including posaconazole, voriconazole, dida krusei ATCC 6258 and Candida parapsilosis ATCC 22019 ravuconazole, and isavuconazole, has been developed. These strains were used for quality control. agents possess potent broad-spectrum activity and favorable Standard antifungal powders of amphotericin B, flucytosine, pharmacokinetic profiles (14, 18, 23). Isavuconazole is a water- fluconazole, itraconazole, voriconazole, posaconazole, and isa- soluble triazole that is suitable for oral and intravenous admin- vuconazole were used and were obtained from Sigma (The istration; its active moiety, BAL4815, is a potent inhibitor of Netherlands), Valeant Pharmaceuticals (The Netherlands), ergosterol biosynthesis. In vitro, the active drug shows activity Pfizer Central Research (United Kingdom), Janssen-Cilag against all major opportunistic and pathogenic fungi (14, 20, (The Netherlands), Pfizer Central Research (United States), 21, 23). Schering Plough (United States), and Basilea Pharmaceutica Previous studies have been done regarding the activity of the (Switzerland), respectively. The stock solutions of the drugs new azoles against a wide variety of fungi. This report summa- were prepared in the appropriate solvent (12). rizes the in vitro activities of isavuconazole and six other anti- Broth microdilution testing was performed in accordance fungal drugs on a large number of Cuban Cryptococcus isolates with the guidelines in CLSI document M27-A2 (12). The final from clinical and environmental origins by a broth microdilu- concentrations of the antifungal agents were 0.016 to 8 �g/ml tion method in accordance with the CLSI (formerly NCCLS) M27-A2 guidelines (12). To our knowledge, this is the first for amphotericin B, itraconazole, voriconazole, and posacon- report evaluating the susceptibility of this yeast to this new azole; 0.063 to 32 �g/ml for flucytosine and fluconazole; and triazole and the largest study ever done with Cuban Cryptococ- 0.004 to 4.00 �g/ml for isavuconazole. Drug-free and yeast-free cus strains. controls were included. The MICs at 48 and 72 h were read (Part of this work was presented at the 46th Interscience optically and spectrophotometrically at 420 nm after agitation. Conference on Antimicrobial Agents and Chemotherapy, San The MIC was defined as the lowest concentration of drug Francisco, CA, 27 to 30 September 2006 [10a].) showing no growth for amphotericin B and a prominent re- A total of 165 Cryptococcus strains were included: 117 were duction of growth (�50%) for the other antifungals compared obtained from clinical samples, including one isolated from a to the drug-free growth control. cheetah imported from South Africa. Forty-five strains were All strains were identified as Cryptococcus neoformans var. isolated from pigeon guano in three different geographical grubii, except one obtained from an animal (Cryptococcus gat- locations of Cuba: Havana City (n � 8), Cienfuegos (n � 6), tii) and two obtained from environmental samples from Pinar and Pinar del Rıó ( n � 31). Species identification was initially del Rio (Cryptococcus albidus and Cryptococcus albidisimilus). performed by standard mycological methods (9), and samples No differences between visual and spectrophotometric readings were observed, and the MICs for the quality control strains were all within the reference ranges (data not shown). * Corresponding author. Mailing address: Department of Medical Table 1 summarizes the in vitro susceptibilities of all the iso- Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, lates according to the origin of the strain. In this study, most of P.O. Box 9015, 6500 GS Nijmegen, The Netherlands. Phone: 31243657514. Fax: 31243657516. E-mail: [email protected]. the C. neoformans isolates showed quite uniform patterns of � Published ahead of print on 22 January 2008. susceptibility to the antifungal agents tested. All of the strains

62 In vitro activity of isavuconazole (BAL4815) compared with six other antifungal compounds

TABLE 1. In vitro susceptibilities of Cryptococcus isolates to by CLSI for C. neoformans. For this reason, breakpoints were amphotericin B, flucytosine, fluconazole, itraconazole, used according to Nguyen and Yu (13) and Lozano et al. (11). voriconazole, posaconazole, and isavuconazole according They define an isolate as resistant with a MIC of �2 �g/ml, to origin and underlying disease of patients which was found to be associated with therapeutic failure. MIC (�g/ml) Although isolates from non-HIV patients showed higher am- Origin (no. of strains) Antifungal agent Range 50% 90% photericin B MICs, the studied strains appear to be susceptible to this drug. Environmental (45) Amphotericin B 0.016–0.5 0.25 0.5 Flucytosine 0.5–64 4 8 Flucytosine is a drug with a limited spectrum of action that Fluconazole 0.25–64 4 8 includes Candida spp. and C. neoformans (5). The CLSI Itraconazole 0.016–0.5 0.063 0.25 M27-A2 document recommends that isolates for which MICs Voriconazole 0.016–0.25 0.063 0.25 are 32 �g/ml be regarded as resistant to this drug (12). By this Posaconazole 0.016–0.125 0.31 0.63 definition, none of the studied clinical isolates was resistant; Isavuconazole 0.002–0.031 0.004 0.016 nevertheless, a justified fear of the emergence of resistance has Clinical (117) Amphotericin B 0.031–1 0.25 0.25 led to its use in combination with amphotericin B in vitro and Flucytosine 0.063–8 4 8 in patients with cryptococcosis (4, 7, 22, 25). Fluconazole 0.25–8 2 4 Among the azoles, fluconazole showed the lowest activity. In Itraconazole 0.016–1 0.031 0.25 Voriconazole 0.016–0.25 0.031 0.125 fact, previous reports have already demonstrated the low ac- Posaconazole 0.016–0.5 0.004 0.016 tivity of this drug against C. neoformans isolates, even though Isavuconazole 0.002–0.063 0.004 0.016 it has proven to be more active in vivo (1). According to these authors, the good therapeutic results obtained are largely at- AIDS patients (85) Amphotericin B 0.031–1 0.25 0.25 5 tributable to its high concentrations in cerebrospinal fluid. Flucytosine 0.125–8 4 8 Fluconazole 0.25–8 2 2 Although no resistance has been found in the present collec- Itraconazole 0.016–0.5 0.031 0.125 tion of isolates, we continue paying attention to the emerging Voriconazole 0.016–0.25 0.031 0.125 resistance to this antifungal agent due to its widespread use as Posaconazole 0.016–0.125 0.031 0.125 primary prophylaxis in AIDS patients. Isavuconazole 0.002–0.063 0.008 0.016 Fortunately, the antifungal armamentarium has increased Non-AIDS patients Amphotericin B 0.125–1 0.25 1 during the past two decades with the addition of several agents (32) Flucytosine 0.063–8 2 4 and, although the echinocandins seem to be inactive, other Fluconazole 0.25–8 2 8 new triazoles are expected to be licensed shortly (5, 14). Vori- Itraconazole 0.016–1 0.031 0.063 conazole has in vitro activity against cryptococcal yeasts (15) Voriconazole 0.016–0.25 0.063 0.063 Posaconazole 0.016–0.5 0.063 0.125 including those that are resistant in vitro to fluconazole (13). Isavuconazole 0.004–0.063 0.008 0.031 Our results show that voriconazole was more potent than itraconazole against Cuban Cryptococcus isolates. However, these Total Amphotericin B 0.016–1 0.25 0.25 promising in vitro results should be complemented by clinical Flucytosine 0.063–64 4 8 Fluconazole 0.25–64 2 4 confirmation. Itraconazole 0.016–1 0.031 0.25 Posaconazole and isavuconazole belong to the latest gener- Voriconazole 0.016–0.25 0.031 0.125 ation of azole antifungal agents that are being investigated for Posaconazole 0.002–0.125 0.004 0.016 their role in treating serious infections due to yeasts and molds. Isavuconazole 0.002–0.063 0.004 0.016 In addition to potent activity, they are well tolerated and offer a 50% and 90%, MIC50 and MIC90, respectively. a diminished toxicity profile compared with other currently marketed systemic antimycotics and their pharmacokinetics are characterized by slow elimination, low plasma clearance, were susceptible to amphotericin B, voriconazole, posacon- and extensive tissue distribution (14, 26, 27). azole, and isavuconazole. Posaconazole is the broadest-spectrum azole licensed to When all of the strains were considered together, the widest date. It exhibits linear pharmacokinetics in volunteers (6) and ranges and highest MICs were for flucytosine (0.063 to 64 has demonstrated safety and efficacy comparable to those of �g/ml) and fluconazole (0.25 to 64 �g/ml). The lowest MICs fluconazole in human and animal studies, although there is no were for isavuconazole, voriconazole, and posaconazole. intravenous formulation available (26). In agreement with our The environmental isolates seem to be less susceptible to results with Cryptococcus, other in vitro studies have docu- fluconazole than the clinical ones. mented a potency and spectrum of activity against clinically Strains obtained from human immunodeficiency virus-neg- important yeasts and Aspergillus spp. similar to those of itra- ative patients showed lower MIC90s, especially for amphoter- conazole and superior to those of fluconazole. Consistent with icin B and fluconazole, compared with those from AIDS pa- these results, in vivo studies have demonstrated efficacy in tients. treating infections due to Candida spp. and C. neoformans with Our results on the in vitro activities of the main antifungal this antifungal agent (2, 10). drugs against C. neoformans are similar to those published The prodrug isavuconazole is a water-soluble triazole pre- previously (3, 7, 8, 16, 19, 24). Amphotericin B has long suc- cursor that is suitable for oral and intravenous administration. cessfully been used to treat various yeast and mold infections. Its active moiety, BAL4815, is a potent inhibitor of ergosterol Unfortunately, its clinical use is hindered by intrinsic nephro- biosynthesis (14, 20, 21). In vitro, the drug shows broad-spec- toxicity (5). There are no defined amphotericin B breakpoints trum activity against all major opportunistic and true patho-

63 Chapter 5

genic fungi (14, 23). In animals as well as in humans, the hout, and J. Meis. Abstr. 46th Intersci. Conf. Antimicrob. Agents Che- pharmacokinetics of this drug are characterized by slow elim- mother., abstr. M-1587, p. 420. 11. Lozano-Chiu, M., V. L. Paetznick, M. A. Ghannoum, and J. H. Rex. 1998. ination, low plasma clearance (approximately 10% of liver Detection of resistance to amphotericin B among Cryptococcus neoformans blood flow), and extensive tissue distribution (20, 21). To our clinical isolates: performances of three different media assessed by using E-test and National Committee for Clinical Laboratory Standards M27-A knowledge, this is the first report evaluating the action against methodologies. J. Clin. Microbiol. 36:2817–2822. Cryptococcus spp. of isavuconazole, which seems to be the most 12. National Committee for Clinical Laboratory Standards. 2002. Reference effective antifungal drug in terms of in vitro activity. In con- method for broth dilution antifungal susceptibility testing of yeasts. Approved standard M27-A2. National Committee for Clinical Laboratory clusion, our results indicate that there has been no significant Standards, Wayne, PA. shift in the MICs of amphotericin B and fluconazole for Cuban 13. Nguyen, M. H., and C. Y. Yu. 1998. In vitro comparative efficacy of voricon- C. neoformans (8), despite the widespread use of these agents azole and itraconazole against fluconazole-susceptible and -resistant Crypto- coccus neoformans isolates. Antimicrob. Agents Chemother. 42:471–472. in persons with AIDS. Given the high oral bioavailability and 14. Odds, F. C. 2006. Drug evaluation: BAL-8557, a novel broad-spectrum the well-tolerated nature of the new azoles, they might become triazole antifungal. Curr. Opin. Investig. Drugs 7:766–772. 15. Perfect, J. R., K. A. Marr, T. J. Walsh, R. N. Greenberg, B. Dupont, J. an important addition to the armamentarium of antifungal Torres-Cisneros, G. Just-Nubling, H. T. Schlamm, I. Lutsar, A. Espinel- agents: both voriconazole and posaconazole show strong anti- Ingroff, and E. Johnson. 2003. Voriconazole treatment for less-common, fungal activity against Cuban Cryptococcus strains in vitro, emerging, or refractory fungal infections. Clin. Infect. Dis. 36:1122–1131. 16. Pfaller, M. A., S. A. Messer, L. Boyken, R. J. Hollis, C. Rice, S. Tendolkar, while isavuconazole showed even lower MICs. These promis- and D. J. Diekema. 2004. In vitro activities of voriconazole, posaconazole, ing results still need to be correlated with clinical outcome. and fluconazole against 4,169 clinical isolates of Candida spp. and Crypto- coccus neoformans collected during 2001 and 2002 in the ARTEMIS global antifungal surveillance program. Diagn. Microbiol. Infect. Dis. 48:201–205. This study was made possible by an unrestricted research grant from 17. Saag, M. S., J. R. Graybill, R. A. Larsen, P. G. Pappas, J. R. Perfect, W. G. the International Society of Human and Animal Mycology (ISHAM) Powderky, J. D. Sobel, and W. E. Dismukes. 2000. Practice guidelines for the to M.T.I.-Z. management of cryptococcus disease. Clin. Infect. Dis. 30:710–718. 18. Sabatelli, F., R. Patel, P. A. Mann, C. A. Mendrick, C. C. Norris, R. Hare, D. Loebenberg, T. A. Black, and P. M. McNicholas. 2006. In vitro activities of REFERENCES posaconazole, fluconazole, itraconazole, voriconazole, and amphotericin B 1. Aller, A. I., E. Martin-Mazuelos, F. Lozano, J. Gomez-Mateos, L. Steele- against a large collection of clinically important molds and yeasts. Antimi- Moore, W. J. Holloway, M. J. Gutie´rrez, F. J. Recio, and A. Espinel-Ingroff. crob. Agents Chemother. 50:2009–2015. 2000. Correlation of fluconazole MICs with clinical outcome in cryptococcal 19. Salam, A. H. A. 2005. In vitro susceptibility of Cryptococcus neoformans infection. Antimicrob. Agents Chemother. 44:1544–1548. clinical isolates from Egypt to seven antifungal drugs. Mycoses 48:327–332. 2. Barchiesi, F., E. Spreghini, A. M. Schimizzi, M. Maracci, D. Giannini, F. 20. Schmitt-Hoffmann, A., B. Roos, M. Heep, M. Schleimer, E. Weidekamm, T. Carle, and G. Scalise. 2004. Posaconazole and amphotericin B combination Brown, M. Roehrle, and C. Beglinger. 2006. Single-ascending-dose pharma- therapy against Cryptococcus neoformans infection. Antimicrob. Agents Che- cokinetics and safety of the novel broad-spectrum antifungal triazole mother. 48:3312–3316. BAL4815 after intravenous infusions (50, 100, and 200 milligrams) and oral 3. Brandt, M. E., M. A. Pfaller, R. A. Hajjeh, R. J. Hamill, P. G. Pappas, A. L. administrations (100, 200, and 400 milligrams) of its prodrug, BAL8557, in Reingold, D. Rimland, and D. W. Warnock for The Cryptococcal Disease healthy volunteers. Antimicrob. Agents Chemother. 50:279–285. Active Surveillance Group. 2001. Trends in antifungal drug susceptibility of 21. Schmitt-Hoffmann, A., B. Roos, J. Maares, M. Heep, J. Spickerman, E. Cryptococcus neoformans isolates in the United States: 1992 to 1994 and 1996 Weidekamm, T. Brown, and M. Roehrle. 2006. Triazole BAL4815 after to 1998. Antimicrob. Agents Chemother. 45:3065–3069. intravenous infusion and oral administration of its prodrug, BAL8557, in 4. Brouwer, A. E., A. Rjanuwong, W. Chierakul, G. E. Griffin, R. A. Larsen, healthy volunteers. Antimicrob. Agents Chemother. 50:286–293. N. J. White, and T. S. Harrison. 2004. Combination antifungal therapies for 22. Schwarz, P., G. Janbon, F. Dromer, O. Lortholary, and E. Dannaoui. 2007. HIV-associated cryptococcal meningitis: a randomised trial. Lancet 363: Combination of amphotericin B with flucytosine is active in vitro against 1764–1767. flucytosine-resistant isolates of Cryptococcus neoformans. Antimicrob. 5. Chen, S. C., and T. C. Sorrell. 2007. Antifungal agents: new drugs, old drugs. Agents Chemother. 51:383–385. Med. J. Aust. 187:404–409. 23. Seifert, H., U. Aurbach, D. Stefanik, and O. Cornely. 2007. In vitro activities 6. Courtney, R., S. Pai, M. Laughlin, J. Lim, and V. Batra. 2003. Pharmaco- of isavuconazole and other antifungal agents against Candida bloodstream kinetics, safety, and tolerability of oral posaconazole administered in single isolates. Antimicrob. Agents Chemother. 51:1818–1821. and multiple doses in healthy adults. Antimicrob. Agents Chemother. 47: 24. Souza, L. K., O. F. L. Fernandes, C. C. Kobayashi, X. S. Passos, C. R. Costa, 2788–2795. J. A. Lemos, A. H. Souza-Juńior, and M. R. Silva. 2005. Antifungal suscep- 7. Dannaoui, E., M. Abdul, M. Arpin, A. Michel-Nguyen, M. A. Piens, A. Favel, tibilities of clinical and environmental isolates of Cryptococcus neoformans in O. Lortholary, and F. Dromer for the French Cryptococcosis Study Group. Goiaˆnia City, Goia´s, Brazil. Rev. Inst. Med. Trop. Sao Paulo 47:253–256. 2006. Results obtained with various antifungal susceptibility testing methods 25. te Dorsthorst, D. T. A., P. E. Verweij, J. Meletiadis, M. Bergervoet, N. C. do not predict early clinical outcome in patients with cryptococcosis. Anti- Punt, J. F. Meis, and J. W. Mouton. 2002. In vitro interaction of flucytosine microb. Agents Chemother. 50:2464–2470. combined with amphotericin B or fluconazole against thirty-five yeast iso- 8. Fernańdez, C. M., M. Gonza´lez, M. T. Illnait, and G. F. Martıńez. 1998. lates determined by both the fractional inhibitory concentration index and Determinacioń de la concentracioń mıńima inhibitoria de anfotericina B en the response surface approach. Antimicrob. Agents Chemother. 46:2982– levaduras de intere´s me´dico. Rev. Cuba. Med. Trop. 50:48–53. 2989. 9. Hazen, K. C., and S. A. Howell. 2003. Candida, Cryptococcus, and other 26. Torres, H. A., R. Y. Hachen, R. F. Chemaly, D. P. Kontoyiannis, and I. Raad. yeasts of medical importance, p. 1693–1711. In P. R. Murray, E. J. Baron, 2005. Posaconazole: a broad-spectrum triazole antifungal. Lancet Infect. J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of clinical Dis. 5:775–785. microbiology, 8th ed., vol. 2. ASM Press, Washington, DC. 27. Ullmann, A. J., O. A. Cornely, A. Burchardt, R. Hachem, D. P. Kontoyiannis, 10. Herbrecht, R., Y. Nivoix, C. Fohrer, S. Natarajan-Ame, and V. Letscher-Bru. K. Topelt, R. Courtney, D. Wexler, G. Krishna, M. Martinho, G. Corcoran, 2005. Management of systemic fungal infections: alternatives to itraconazole. and I. Raad. 2006. Pharmacokinetics, safety, and efficacy of posaconazole in J. Antimicrob. Chemother. 56:39–48. patients with persistent febrile neutropenia or refractory invasive fungal 10a.Illnait-Zaragozi, M., I. Curfs-Breuker, G. Martinez, C. Fernandez, T. Boek- infection. Antimicrob. Agents Chemother. 50:658–666.

64 Chapter 6 In vitro antifungal susceptibilities and Amplified Fragment Length Polymorphism genotyping of a worldwide collection of 350 clinical, veterinary and environmental Cryptococcus gattii isolates

Published in Antimicrobial Agents and Chemotherapy 2010;54:5139-45. Chapter 6

In Vitro Antifungal Susceptibilities and Ampliﬁed Fragment Length Polymorphism Genotyping of a Worldwide Collection of 350 Clinical, Veterinary, and Environmental Cryptococcus gattii Isolates� Ferry Hagen,1,2 Maria-Teresa Illnait-Zaragozi,3 Karen H. Bartlett,4 Danie¨lle Swinne,5 Erik Geertsen,6 Corne´ H. W. Klaassen,6 Teun Boekhout,1,2 and Jacques F. Meis6* CBS-KNAW Fungal Biodiversity Centre, Department of Yeast and Basidiomycete Research, Utrecht, Netherlands1; University Medical Center Utrecht, Department of Internal Medicine and Infectious Diseases, Eijkman Winkler Institute, Utrecht, Netherlands2; Tropical Medicine Institute Pedro Kouri, Department of Mycology and Bacteriology, Havana, Cuba3; University of British Columbia, School of Occupational and Environmental Hygiene, Vancouver, British Columbia, Canada4; Scientiﬁc Institute of Public Health, Mycology Section, Brussels, Belgium5; and Canisius Wilhelmina Hospital, Department of Medical Microbiology and Infectious Diseases, Nijmegen, Netherlands6

Received 2 June 2010/Returned for modiﬁcation 15 August 2010/Accepted 10 September 2010

The in vitro susceptibilities of a worldwide collection of 350 Cryptococcus gattii isolates to seven antifungal drugs, including the new triazole isavuconazole, were tested. With amplified fragment length polymorphism (AFLP) fingerprinting, human, veterinary, and environmental C. gattii isolates were subdivided into seven AFLP genotypes, including the interspecies hybrids AFLP8 and AFLP9. The majority of clinical isolates (n � 215) comprised genotypes AFLP4 (n � 76) and AFLP6 (n � 103). The clinical AFLP6 isolates had significantly higher geometric mean MICs for flucytosine and fluconazole than the clinical AFLP4 isolates. Of the seven

antifungal compounds examined in this study, isavuconazole had the lowest MIC90 (0.125 �g/ml) for all C. gattii isolates, followed by a 1 log2 dilution step increase (MIC90, 0.25 �g/ml) for itraconazole, voriconazole, and posaconazole. Amphotericin B had an acceptable MIC90 of 0.5 �g/ml, but ﬂuconazole and ﬂucytosine had relatively high MIC90sof8�g/ml.

The basidiomycetous yeast Cryptococcus gattii is responsible With the increasing use of molecular techniques, such as PCR for life-threatening invasive disease in apparently healthy hu- fingerprinting, restriction fragment length polymorphism mans and animals (7, 19). A typical C. gattii infection is ac- (RFLP) analysis of the PLB1 and URA5 loci, and amplified quired through the respiratory tract, from which it can further fragment length polymorphism (AFLP) fingerprint analysis, as disseminate to the central nervous system, resulting in fatal well as several multilocus sequence typing (MLST) ap- meningitis (7, 19, 32). Cryptococcosis caused by the primary proaches, it became clear that C. gattii could be divided into pathogenic yeast C. gattii was, until a decade ago, a rarely five distinct genotypes, named AFLP4/VGI, AFLP5/VGIII, encountered infection outside tropical and subtropical regions AFLP6/VGII, AFLP7/VGIV, and AFLP10 (the last one of (17, 26, 27). However, this changed due to an unprecedented which is a recently observed novel genotype) (2, 6, 7, 13, 16, 20, outbreak that emerged in the temperate climate of Vancouver 21, 23). Until recently, a serotype agglutination assay was Island (British Columbia, Canada) that subsequently expanded widely used to distinguish C. neoformans (serotypes A and D) farther into the Pacific Northwest (1, 8, 10, 16). Its sibling from C. gattii (serotypes B and C) (7, 27). In general, serotype species, Cryptococcus neoformans, differs ecologically and epi- B strains are found in each of the five C. gattii AFLP genotypes, demiologically from C. gattii since it occurs on a global scale but it seems that C. gattii serotype C strains are restricted to and is linked with disease occurring in immunocompromised genotypes AFLP5/VGIII and AFLP7/VGIV (2, 6, 16, 21, 27). individuals, such as HIV-positive patients and transplant pa- In addition, it was found that C. gattii and C. neoformans can tients who receive immune-suppressive medicines (7, 10, 18, form interspecies hybrids, named genotype AFLP8 (C. neofor- 19, 31). mans var. neoformans AFLP2/VNIII serotype D � C. gattii Cryptococcus gattii can be discerned from C. neoformans AFLP4/VGI serotype B) and AFLP9 (C. neoformans var. grubii using a wide range of microbiological and molecular tech- AFLP1/VNI serotype A � C. gattii AFLP4/VGI serotype B). niques (7, 20). A convenient method is the use of canavanine- These interspecies hybrids have, until now, been isolated only glycine-bromothymol blue (CGB) medium, which allows C. from clinical samples, and they might have a higher virulence gattii but not C. neoformans to grow and which changes the pH potential than regular C. gattii or C. neoformans isolates (4, 5; indicator in the medium from green-yellowish to blue (18). F. Hagen, K. Tintelnot, and T. Boekhout, unpublished data). Treatment of cryptococcosis depends on, besides the immune status of the patient, the severity and localization of the * Corresponding author. Mailing address: Canisius Wilhelmina infection (11). Severe cases of cryptococcosis in immunocom- Hospital, Department of Medical Microbiology and Infectious Dis- eases, Weg door Jonkerbos 100, Nijmegen, 6532 SZ Netherlands. petent and -compromised patients are treated according to the Phone: 31243657514. Fax: 31243657516. E-mail: [email protected]. guidelines of the Infectious Diseases Society of America, ac- � Published ahead of print on 20 September 2010. cording to which treatment consists of an induction therapy for

66 Antifungal susceptibilities and genotyping of a global collection of Cryptococcus gattii isolates

TABLE 1. Distribution of Cryptococcus gattii strains on the basis of AFLP genotype proﬁles, geographical origin, and clinical, veterinary, or environmental source

No. (%) of strainsa AFLP Geographical origin Source of isolation genotype proﬁle North South Total Africa Asia Australia Europe Unknown Clinical Veterinary Environmental Unknown America America

AFLP4 16 (11.7) 14 (10.2) 8 (5.8) 74 (54.0) 9 (6.6) 15 (10.9) 1 (0.7) 76 (55.5) 29 (21.2) 30 (21.9) 2 (1.5) 137 (39.1) AFLP5 1 (4.5) 2 (9.1) 10 (45.5) 7 (31.8) 2 (9.1) 15 (68.2) 5 (22.7) 2 (9.1) 22 (6.3) AFLP6 3 (1.8) 6 (3.6) 11 (6.5) 19 (11.2) 75 (44.4) 55 (32.5) 103 (60.9) 27 (16.0) 38 (22.5) 1 (0.6) 169 (48.3) AFLP6A 4 (5.5)* 1 (1.4)* 5 (6.8)* 56 (76.7)* 7 (9.6)* 36 (49.3)* 18 (24.7)* 18 (24.7)* 1 (1.4)* 73 (43.2)* AFLP6B 3 (3.4)* 2 (2.3)* 10 (11.4)* 14 (15.9)* 11 (12.5)* 48 (54.5)* 61 (69.3)* 7 (8.0)* 20 (22.7)* 88 (52.1)* AFLP6C 8 (100)* 6 (75.0)* 2 (25.0)* 8 (4.7)* AFLP7 7 (63.6) 2 (18.2) 2 (18.2) 10 (90.9) 1 (9.1) 11 (3.1) AFLP8 2 (33.3) 2 (33.3) 2 (33.3) 6 (100) 6 (1.7) AFLP9 2 (66.7) 1 (33.3) 3 (100) 3 (0.9) AFLP10 2 (100) 2 (100) 2 (0.6)

Total 28 (8.0) 22 (6.3) 20 (5.7) 101 (28.9) 95 (27.1) 81 (23.1) 3 (0.9) 215 (61.4) 73 (20.9) 57 (16.3) 5 (1.4) 350 (100)

a An asterisk indicates strains that belong to one of the subgenotypes within AFLP6, namely, AFLP6A, AFLP6B, or AFLP6C; these strains are the cumulative number of strains in the main genotype AFLP6 group.

2 weeks with a combination of amphotericin B and flucytosine, mation of a polysaccharide capsule. Approximately 150 �l of C. gattii cells was followed by a 10-week consolidation therapy with fluconazole incubated for 3 h in 1.6 ml urea solution (10.7 M) at room temperature. After (11, 24). centrifugation at 4,000 � g for 2 min, the pellet was resuspended in 750 �l of lysis buffer (10% [wt/vol] sodium dodecyl sulfate, 10% [wt/vol] sodium Sarkosyl in TE Cryptococcus neoformans has been extensively studied for its [Tris-EDTA] buffer, pH 8.0) and 750 �l phenol-chloroform-isoamyl alcohol 6 in vitro susceptibility to a wide variety of antifungal com- (25:24:1, pH 8.0). The cells were disrupted by bead beating at full speed pounds, including the new triazoles posaconazole, voricon- (TissueLyser; Qiagen, Venlo, Netherlands) for 5 min. After centrifugation at azole, ravuconazole, and isavuconazole (12, 14, 28, 29, 33). 17,000 � g for 15 min at 4°C, approximately 750 �l supernatant was added to an Despite the ongoing C. gattii outbreak, only a few studies using equal volume of ice-cold 96.2% ethanol and 100 �l ice-cold 3.0 M sodium acetate. Precipitation of genomic DNA was accelerated by incubating the solu- relatively small sets of C. gattii isolates have been performed to tion at �20°C for 1 h. Genomic DNA was precipitated by centrifugation at investigate their in vitro susceptibilities to amphotericin B, 17,000 � g for 15 min at 4°C, the pellet obtained was dissolved in 100 �l flucytosine, fluconazole, and the new triazole antifungals (12, Tris-EDTA buffer (pH 8.0), and 1 �l RNase (Invitrogen, Breda, Netherlands) 15, 28–30). A few studies divided the C. gattii isolates into was added, followed by an incubation step at 37°C to degrade any remaining groups according to their serotype or genotype (15, 29). RNA. The quality and quantity of genomic DNA were measured using an ND-1000 spectrophotometer (Nanodrop, Wilmington, DE), and final working Therefore, we studied the in vitro susceptibilities of each of solutions, each with a concentration of 100 ng/�l, were prepared. the C. gattii genotypes from a large worldwide collection, sub- Mating typing, serotyping, and genotyping. The mating types of all isolates divided by AFLP genotyping, to amphotericin B, flucytosine, were determined by partial amplification of the STE12 locus and either the fluconazole, itraconazole, voriconazole, posaconazole, and the MATa or the MAT� allele using two different primer sets, as described by Bovers new experimental broad-spectrum antifungal triazole isavu- et al. (5). Briefly, 1 �l genomic DNA (100 ng/�l) was added to 24 �l PCR mixture conazole. consisting of 17.8 �l double-distilled H2O, 0.75 �l MgCl2 (50 mM; Bioline, London, United Kingdom), 2.5 �l 10� PCR buffer (Bioline), 1.9 �l deoxynucleoside triphosphates (1 mM; Bioline), 0.1 �l Taq DNA polymerase (5 U/�l; Bio- MATERIALS AND METHODS line), and 0.5 �l of the forward and reverse primers (Biolegio, Nijmegen, Neth- erlands) for either the STE12� allele (forward primer STE12�-F809 [5�-TTGA Strains and media. A worldwide collection of 350 C. gattii strains, represented CCTTTTRTTCCGCAATG-3�] and reverse primer STE12�-R1607 [5�-TTTCT by 215 clinical isolates (61.4%), 57 veterinary isolates (16.3%), 73 environmental TCTCCCCTGTTTATAGGC-3�]) or the STE12a allele (forward primer isolates (20.9%), and 5 isolates from an unknown source (1.4%), were included STE12a-F537 [5�-GTTCTTTGGAATGGCTTATTTCATAT-3�] and reverse in this study. Strains were obtained from the public culture collections of the primer STE12a-R1299 [5�-GMCTTGCGTGGATCATATCTA-3�]). PCRs were CBS-KNAW Fungal Biodiversity Centre (Utrecht, Netherlands), BCCM-IHEM carried out with an initial denaturation step at 94°C for 5 min, followed by 35 Scientific Institute of Public Health (Brussels, Belgium), as well as the working cycles of denaturation at 94°C for 30 s, annealing at 58°C for 30 s, and extension group collections of the authors. According to their geographical origins, the C. at 72°C for 1 min and then 72°C for 5 min. The reaction mixture was finally gattii strains were grouped as being of African (n � 28; 8.0%), Asian (n � 22; 6.3%), Australian (n � 20; 5.7%), European (n � 101; 28.9%), North American cooled to 21°C. As a control, C. gattii mating type � strains CBS6956 and WM276 (n � 95; 27.1%), South American (n � 81; 23.1%), and unknown (n � 3; 0.9%) (CBS10510) and mating-type a strains CBS1930 and E566 (CBS11233) were origin. The group of North American strains contained a set of 75 (24.1%) C. included. gattii strains related to the ongoing Vancouver Island outbreak. The distributions A Cryptocheck agglutination test kit (Iatron Laboratories, Tokyo, Japan) was of the C. gattii strains, according to their geographical origins and sources of used to determine the serotypes of those C. gattii isolates with contradictory isolation, are listed in Table 1. The strains were checked for purity and were serotype data. cultivated on malt extract agar medium (MEA; Oxoid, Basingstoke, United The genotype of each C. gattii isolate was determined using AFLP fingerprint Kingdom). After inoculation, the cultures were incubated for 2 days at 30°C. A analysis, as described previously by Boekhout et al. (2). C. gattii isolates WM178 working collection was made by growing the C. gattii strains on MEA slants for (CBS10082; AFLP6/VGII), A1M-R265 (CBS10514; AFLP6A/VGIIa), A1M- 2 days at 30°C, after which the strains were stored at 4°C. R272 (CBS10865; AFLP6B/VGIIb), A6M-R38 (CBS11545; AFLP6C/VGIIc), Genomic DNA extraction. An extraction procedure previously described by WM161 (CBS10081; AFLP5/VGIII), WM179 (CBS10078; AFLP4/VGI), Bolano et al. (3) was further optimized to obtain high-quality genomic DNA. C. WM779 (CBS10101; AFLP7/VGIV), CBS10488 (AFLP8), CBS10496 (AFLP9), gattii strains were subcultured for 2 days at 30°C on yeast extract (1%)–peptone and IHEM14941S (CBS11687; AFLP10) were used as controls to assign the (1%)–D-glucose (2%) agar supplemented with 0.5 M NaCl to prevent the for- AFLP genotypes to all C. gattii isolates (2, 8, 13, 20, 21).

67 Chapter 6

In vitro susceptibility testing. The MICs of amphotericin B (Bristol Myers for itraconazole and voriconazole, and � 0.016 to 0.5 �g/ml Squibb, Woerden, Netherlands), flucytosine (Valeant Pharmaceuticals, Zoeter- for posaconazole and isavuconazole. The geometric mean meer, Netherlands), fluconazole and voriconazole (Pfizer Central Research, Sandwich, Kent, United Kingdom), itraconazole (Janssen Cilag, Tilburg, Neth- MIC values varied widely when they were calculated for all erlands), posaconazole (Schering-Plough Corp., Kenilworth, NJ), and isavucon- 350 C. gattii strains (Table 2). The highest geometric mean azole (Basilea Pharmaceutica, Basel, Switzerland) were determined using the MIC was 3.573 �g/ml for fluconazole, followed by geometric broth microdilution method, in accordance with the guidelines in CLSI docu- mean MICs of 2.896 �g/ml for flucytosine and 0.272 �g/ml ment M27-A3 (9). Drug-free and sterile controls were included. Paecilomyces for amphotericin B and much lower geometric mean MICs variotii (ATCC 22319) and Candida krusei (ATCC 6258, CBS573) were used for quality control. The microtiter plates were incubated for 72 h at 35°C under for itraconazole and the new triazoles voriconazole, aerobic conditions. posaconazole, and isavuconazole (0.124, 0.108, 0.113, and

Data analysis. Values for the MIC50 and MIC90 were obtained by ordering the 0.051 �g/ml, respectively). MIC data for each antifungal in ascending arrays and selecting the median and Comparison of geometric mean MIC values of the C. gattii 90th quantile, respectively, of the MIC distribution. Geometric mean MICs were strains obtained from clinical, environmental, and veteri- computed using Microsoft Office Excel 2007 software, for which purpose values of �x were set equal to 0.5x and values more than �y were set equal to 2y (x and nary sources revealed that the last category of strains had y are the lowest and highest MICs, respectively, in the range). A two-tailed decreased susceptibility to all seven antifungal drugs investigated Mann-Whitney-Wilcoxon test was applied to compare the MIC values between except flucytosine in comparison to the susceptibilities of the different C. gattii genotypes; and a multivariate analysis of variance (MANOVA) clinical and environmental strains. When the susceptibilities of test was applied to compare geographical origins, AFLP genotype profiles, and MIC values (StatistiXL software, version 1.8, StatistiXL, Nedland, WA, Austra- the clinical versus veterinary C. gattii strains and the environmen- lia). A P value of �0.05 was considered statistically significant. tal versus veterinary C. gattii strains were compared, these differences were significant for itraconazole (P � 0.026 and P � 0.02, respectively) and the novel triazoles posaconazole (P � 0.008 and RESULTS P � 0.02, respectively), voriconazole (P � 0.013 and P � 0.008, Mating typing, serotyping, and genotyping. On the basis of respectively), and isavuconazole (P � 0.004 and P � 0.002, re- the AFLP genotypes, the 350 C. gattii isolates were divided spectively). into five haploid AFLP genotypes (n � 341; 97.4%) and two A MANOVA showed that there was a significant difference hybrid AFLP genotypes (n � 9; 2.6%), namely, AFLP4 (n � (P � 0.001) between the susceptibility values obtained for the 137; 39.1%), AFLP5 (n � 22; 6.3%), AFLP6 (n � 169; seven antifungal compounds and the geographical origins of 48.3%), AFLP7 (n � 11; 3.1%), AFLP8 (n � 6; 1.7%), the C. gattii strains (Table 3). It was observed that the MIC AFLP9 (n � 3; 0.9%), and AFLP10 (n � 2; 0.6%). The values of amphotericin B and flucytosine contributed to this majority of the 215 clinical C. gattii isolates belonged to significant difference (P � 0.003 and P � 0.02, respectively). either genotype AFLP4 (n � 76; 35.4%) or genotype AFLP6 This difference was caused by C. gattii strains from Europe, (n � 103; 47.9%). Strains that clustered within genotype which differed significantly in their susceptibility to amphoter- AFLP6 could be further divided into the three subgenotypes icin B compared to the susceptibilities of strains from Africa AFLP6A (n � 73; 43.2%), AFLP6B (n � 88; 52.1%), (P � 0.002), Asia (P � 0.014), Australia (P � 0.041), North and AFLP6C (n � 8; 4.7%) (Table 1). America (P � 0.001), and South America (P � 0.019), which The serotypes of the genotype AFLP7 strains were investi- were found to be less susceptible. For flucytosine, North Amer- gated in more detail, since several of these isolates were, ac- ican C. gattii strains had a significantly higher geometric mean cording to the literature, known to be serotypes B and C (6, 21, MIC than African (P � 0.001), Asian, (P � 0.001), European 23). However, we determined that all C. gattii AFLP7 isolates (P � 0.001), and South American (P � 0.001) strains (Table 3). were serotype C. The two genotype AFLP10 strains were de- When all C. gattii genotype AFLP4 and AFLP6 strains (Table termined to be serotype B. Within the other AFLP genotypes, 2) were analyzed by MANOVA, it was observed that it was not no contradictions from the literature were found, and serotyp- the geographical origin but the genotype that contributed to ing was omitted for these groups. the significant differences in antifungal susceptibility patterns The mating types for all C. gattii isolates were determined by (P � 0.001). The antifungal drugs amphotericin B (P � 0.001), amplification of either the STE12a or the STE12� allele. This fluconazole (P � 0.001), itraconazole (P � 0.038), voricon- revealed that within the group of 341 haploid C. gattii strains azole (P � 0.001), and isavuconazole (P � 0.001) were found (genotypes AFLP4 to AFLP7 and AFLP10), the mating type � to be highly significant contributors to this phenotypic differ- strains (n � 296; 86.8%) were dominant over the mating type ence between AFLP4 and AFLP6 C. gattii strains. a strains (n � 45; 13.2%). Clinical strains that belonged to the genotype AFLP6 cluster

In vitro susceptibility testing. MIC ranges, MIC50s, geomet- had a signiﬁcantly (P � 0.05) higher geometric mean MIC than ric mean MICs, and MIC90s related to AFLP genotype are strains of genotype AFLP4 for six of the seven antifungal presented in Table 2. For each antifungal for each AFLP compounds tested. However, both genotypes had almost iden-

genotype examined, the MIC50 and geometric mean MIC val- tical geometric mean MICs for posaconazole (0.112 and 0.124 ues differed by �1 log2 dilution step, indicating that in all cases �g/ml, respectively; P � 0.391). Strikingly, clinical AFLP6 iso-

the MIC50s obtained by inspection reasonably reflect the cen- lates had high geometric mean MICs of 4.961 and 5.638 �g/ml tral tendency of antifungal susceptibility of the population. for flucytosine and fluconazole, respectively, compared to

Isavuconazole had the lowest MIC90s for all C. gattii iso- those for clinical isolates within genotype AFLP4 (1.401 and

lates surveyed by at least 1 log2 dilution step. The overall 2.467 �g/ml, respectively). ranges of MICs for each of the seven drugs were �0.016 to C. gattii strains within the Vancouver outbreak-related ge- 1 �g/ml for amphotericin B, 0.125 to �64 �g/ml for flucy- notype, genotype AFLP6, were divided into the three defined tosine, 0.25 to �64 �g/ml for fluconazole, �0.016 to 1 �g/ml subgenotypes AFLP6A (n � 73; 43.2%), AFLP6B (n � 88;

68 Antifungal susceptibilities and genotyping of a global collection of Cryptococcus gattii isolates

TABLE 2. MIC ranges, MIC50s, MIC90s, and geometric mean MICs for all 350 C. gattii strains and AFLP genotypes MIC (�g/ml) Strain Drug Range 50% Geometric mean 90% All C. gattii strains (n � 350) Amphotericin B �0.016–1 0.25 0.272 0.5 Flucytosine 0.125–�64 2 2.896 8 Fluconazole 0.25–64 4 3.573 8 Itraconazole �0.016–1 0.125 0.124 0.25 Voriconazole �0.016–1 0.125 0.108 0.25 Posaconazole �0.016–0.5 0.125 0.113 0.25 Isavuconazole �0.016–0.5 0.063 0.051 0.125

C. gattii AFLP4 (n � 137) Amphotericin B �0.016–0.5 0.25 0.238 0.5 Flucytosine 0.125–16 1 1.667 4 Fluconazole 0.25–16 2 2.550 8 Itraconazole �0.016–0.5 0.125 0.119 0.25 Voriconazole �0.016–0.25 0.063 0.090 0.25 Posaconazole �0.016–0.5 0.125 0.123 0.25 Isavuconazole �0.016–0.5 0.031 0.044 0.125

C. gattii AFLP5 (n � 22) Amphotericin B 0.125–0.5 0.25 0.214 0.5 Flucytosine 0.5–4 2 1.938 4 Fluconazole 0.25–8 2 2.064 4 Itraconazole �0.016–0.25 0.063 0.078 0.125 Voriconazole �0.016–0.125 0.063 0.059 0.125 Posaconazole 0.031–0.25 0.063 0.076 0.125 Isavuconazole �0.016–0.063 0.031 0.023 0.063

C. gattii AFLP6 (n � 169) Amphotericin B 0.125–1 0.25 0.312 0.5 Flucytosine 0.5–�64 4 5.033 8 6 Fluconazole 0.5–64 4 5.137 16 Itraconazole �0.016–1 0.125 0.148 0.5 Voriconazole �0.016–1 0.125 0.142 0.25 Posaconazole �0.016–0.5 0.125 0.119 0.25 Isavuconazole �0.016–0.5 0.063 0.067 0.25

C. gattii AFLP6A (n � 73) Amphotericin B 0.125–0.5 0.25 0.299 0.5 Flucytosine 1–16 4 5.268 8 Fluconazole 0.5–32 4 4.235 8 Itraconazole �0.016–1 0.125 0.123 0.5 Voriconazole �0.016–1 0.125 0.117 0.25 Posaconazole �0.016–0.5 0.125 0.105 0.25 Isavuconazole �0.016–0.5 0.063 0.050 0.25

C. gattii AFLP6B (n � 88) Amphotericin B 0.125–1 0.25 0.319 0.5 Flucytosine 0.5–�64 4 4.609 8 Fluconazole 1–16 8 5.438 16 Itraconazole �0.016–0.5 0.25 0.162 0.25 Voriconazole �0.016–0.5 0.125 0.152 0.25 Posaconazole �0.016–0.25 0.125 0.124 0.25 Isavuconazole �0.016–0.25 0.063 0.078 0.25

C. gattii AFLP6C (n � 8) Amphotericin B 0.25–0.5 0.25 0.354 0.5 Flucytosine 4–16 8 8.724 16 Fluconazole 8–64 16 16.000 64 Itraconazole 0.25–0.5 0.25 0.297 0.5 Voriconazole 0.25–1 0.25 0.354 1 Posaconazole 0.125–0.5 0.25 0.229 0.5 Isavuconazole 0.063–0.5 0.25 0.193 0.5

C. gattii AFLP7 (n � 11) Amphotericin B 0.125–0.5 0.5 0.343 0.5 Flucytosine 1–8 2 2.269 8 Fluconazole 0.5–32 4 4.537 16 Itraconazole �0.016–0.25 0.125 0.104 0.25 Voriconazole 0.031–0.25 0.125 0.117 0.25 Posaconazole 0.031–0.25 0.125 0.110 0.25 Isavuconazole �0.016–0.25 0.063 0.052 0.125

C. gattii AFLP8–10 (n � 11) Amphotericin B 0.125–0.5 0.25 0.220 0.25 Flucytosine 1–8 1 1.656 8 Fluconazole 0.5–8 2 2.130 8 Itraconazole �0.016–0.125 0.063 0.043 0.125 Voriconazole �0.016–0.25 0.031 0.059 0.125 Posaconazole �0.016–0.125 0.063 0.043 0.125 Isavuconazole �0.016–0.063 0.016 0.020 0.063

69 Chapter 6

TABLE 3. In vitro susceptibilities of the C. gattii strains according to geographical origin

MIC (�g/ml) C. gattii strain origin Drug Range 50% Geometric mean 90% Africa (n � 28) Amphotericin B 0.125–0.5 0.25 0.320 0.5 Flucytosine 1–16 2 2.378 8 Fluconazole 0.25–32 4 4.308 8 Itraconazole �0.016–0.25 0.125 0.111 0.25 Voriconazole �0.016–0.25 0.125 0.114 0.25 Posaconazole �0.016–0.25 0.125 0.122 0.25 Isavuconazole �0.016–0.5 0.063 0.058 0.125

Asia (n � 22) Amphotericin B 0.125–0.5 0.25 0.312 0.5 Flucytosine 0.5–16 2 2 8 Fluconazole 0.5–16 2 2.741 8 Itraconazole 0.031–0.5 0.125 0.1 0.250 Voriconazole 0.031–0.25 0.063 0.08 0.250 Posaconazole 0.063–0.5 0.125 0.107 0.250 Isavuconazole 0.016–0.25 0.031 0.044 0.125

Australia (n � 20) Amphotericin B 0.125–0.5 0.25 0.297 0.5 Flucytosine 1–32 2 3.138 16 Fluconazole 0.5–16 2 2.639 8 Itraconazole 0.063–0.5 0.125 0.134 0.25 Voriconazole 0.031–0.25 0.125 0.091 0.25 Posaconazole 0.031–0.5 0.125 0.121 0.25 Isavuconazole 0.016–0.125 0.063 0.049 0.125

Europe (n � 101) Amphotericin B �0.016–1 0.25 0.223 0.5 Flucytosine 0.125–�64 2 2.263 8 Fluconazole 0.5–32 4 3.301 8 Itraconazole �0.016–0.5 0.125 0.137 0.25 Voriconazole �0.016–1 0.125 0.112 0.25 Posaconazole �0.016–0.5 0.125 0.131 0.25 Isavuconazole �0.016–0.5 0.063 0.055 0.125

North America (n � 95) Amphotericin B 0.125–0.5 0.25 0.300 0.5 Flucytosine 1–16 4 4.334 8 Fluconazole 0.5–64 4 4.240 8 Itraconazole �0.016–1 0.125 0.121 0.25 Voriconazole �0.016–1 0.125 0.114 0.25 Posaconazole �0.016–0.5 0.125 0.103 0.25 Isavuconazole �0.016–0.5 0.063 0.047 0.25

South America (n � 81) Amphotericin B 0.125–1 0.25 0.277 0.5 Flucytosine 0.5–16 4 2.865 8 Fluconazole 0.25–16 4 3.548 16 Itraconazole �0.016–1 0.125 0.126 0.25 Voriconazole �0.016–0.5 0.125 0.111 0.25 Posaconazole �0.016–0.5 0.125 0.104 0.25 Isavuconazole �0.016–0.5 0.063 0.051 0.25

52.1%), and AFLP6C (n � 8; 4.7%). Significant differences in DISCUSSION the geometric mean MICs between C. gattii AFLP6A and AFLP6B strains were found for all antifungal drugs except Despite the fact that C. gattii has become an important amphotericin B (P � 0.398) and flucytosine (P � 0.091); how- primary pathogen due to the expanding outbreak in British Columbia (Canada) and the Pacific Northwest (United States), ever, the geometric mean MICs were within 1 log2 dilution step. There was also no significant difference in amphotericin B there is a lack of in-depth studies on the susceptibilities of large MICs when AFLP6A and AFLP6C strains (P � 0.353) and sets of C. gattii strains to conventional and new antifungal AFLP6B and AFLP6C strains (P � 0.607) were compared, compounds (10). In particular, the relation between suscepti- while all other antifungal compounds tested showed significant bility profiles and the different genotypes within the C. gattii differences between these subgenotypes (Table 2). However, complex has rarely been studied and has been studied with only the number of strains within subgenotype AFLP6C is too small low numbers of isolates (15). In this study, we have compared to compare its antifungal susceptibility characteristics with the susceptibilities of 350 C. gattii isolates collected from dif- those of the larger group of strains within genotypes AFLP6A ferent geographical regions and sources (e.g., clinical, veteri- and AFLP6B. nary, and environmental) to four clinically used antifungal

70 Antifungal susceptibilities and genotyping of a global collection of Cryptococcus gattii isolates

compounds, i.e., amphotericin B, flucytosine, fluconazole, and significantly (P � 0.05) higher geometric mean MICs for six of itraconazole, and the three novel triazoles voriconazole, the antifungal compounds tested than strains belonging to ge- posaconazole, and isavuconazole. The older antifungal drugs, notype AFLP4 (n � 76). However, both genotypes had almost amphotericin B, flucytosine, and fluconazole, had lower levels identical geometric mean MICs for posaconazole (0.112 and of activity against all C. gattii isolates than itraconazole and the 0.124 �g/ml, respectively). Strikingly, clinical AFLP6 isolates novel triazoles, as visualized by their geometric mean MIC had higher geometric mean MICs of 4.961 and 5.638 �g/ml for values (Table 2). The observation that flucytosine and flucon- flucytosine and fluconazole, respectively, than clinical isolates azole had poorer antifungal activity against C. gattii, as well as within genotype AFLP4 (1.401 and 2.467 �g/ml, respectively). against its sibling, C. neoformans, was a consistent finding in Similar results were recently reported for a collection of 43 previous studies (12, 14, 15, 25, 29, 30). For testing of C. North American C. gattii isolates (15). In the current study, neoformans susceptibility to amphotericin B, flucytosine, and fluconazole and flucytosine had the lowest levels of activity fluconazole, so-called surrogate breakpoints were introduced against the C. gattii strains. Itraconazole, which has been used by Pfaller et al. (25). C. neoformans was regarded to be highly clinically to treat cryptococcosis, had a MIC90 of only 0.25 susceptible to amphotericin B when the MIC in vitro was �1 �g/ml, which is 1 or 2 log2 dilutions lower than the MIC90s �g/ml. In this study, none of the 350 C. gattii strains showed previously found by others (15, 30). Isavuconazole, posacon- higher in vitro MICs, and thus, all isolates remained highly azole, and voriconazole demonstrated excellent potency susceptible. Similar findings were reported in three other stud- against each isolate and all AFLP genotypes, including isolates ies, which tested lower numbers of isolates (15, 29, 30). On the with reduced fluconazole susceptibilities. contrary, decreased in vitro susceptibility (�16 �g/ml) to flu- The results for voriconazole (MIC90, 0.25 �g/ml) were com- conazole was observed for 26 strains (7.4%), and only one parable to those of other studies (15, 29, 30) and 1 log2 dilution clinical isolate within the AFLP6C cluster appeared to have higher than that found in one study (22). Posaconazole MICs high-level resistance (MICs, �64 �g/ml) to fluconazole in vitro. differed by a maximum of 2 log2 dilution steps between this and

With respect to ﬂucytosine, the number of strains within the previous studies. Isavuconazole had the lowest MIC90 (0.125 group of isolates with decreased susceptibility (MICs, �8 �g/ �g/ml) in this large study, and the MIC90 was within 1 log2 ml) was 84 (24.3%), of which 9 (2.57%) belong to the category dilution of that recently reported (MIC , 0.063 �g/ml) by 6 90 of high-level-resistant isolates (MICs, �32 �g/ml). Thompson et al. (29). When isolates were tested by Etest,

The majority of C. gattii isolates in the present study belong these authors also found a MIC90 of 0.125 �g/ml (28). Isavu- to either genotype AFLP4 (n � 137; 39.1%) or AFLP6 (n � conazole has even better in vitro activity against C. neoformans, 169; 48.3%), while the remaining 44 (12.6%) C. gattii isolates a finding which also holds true for the other antifungal com- belong to one of the minor genotypes (AFLP5 or AFLP7 to pounds (14). AFLP10) that are more restricted to certain geographical re- All antifungal compounds tested were active against all C. gions and which less frequently cause infections (7, 27). The gattii genotypes; however, some strains showed high MICs, as preponderance toward C. gattii strains with an AFLP4 or was the case for amphotericin B (up to 1 �g/ml), flucytosine AFLP6 genotype can be interpreted as being the result of the (�64 �g/ml), fluconazole (64 �g/ml), and itraconazole and analysis of a nonrepresentative selection of strains. This ap- voriconazole (1 �g/ml). The two novel triazoles isavuconazole parent overrepresentation might be due to the inclusion of and posaconazole had relatively low MIC values compared to preserved culture collection strains from well-studied geo- those of the other antifungal compounds tested. We conclude graphical localities, such as Australia and the C. gattii outbreak that amphotericin B, itraconazole, voriconazole, posaconazole, region in North America (27). These localities were repeatedly and isavuconazole were the drugs that were the most active found to harbor more genotype AFLP4 or AFLP6 strains against C. gattii isolates in vitro. Isavuconazole, with its attrac- rather than strains of genotype AFLP5 or AFLP7 (15, 16, 27). tive pharmacological and toxicological properties and avail- The South American continent has been shown to be a melting ability in oral and parenteral formulations, may be a welcome pot of known haploid AFLP genotypes compared to the diver- addition to the clinician’s antifungal armamentarium to fight sity of genotypes in other localities (21, 23, 27). However, this pathogen. preserved C. gattii strains from the less well studied regions, Africa, Asia, and Europe, were found to follow the trend that ACKNOWLEDGMENTS strains with an AFLP4 genotype profile were more frequently isolated than others (13, 27). Hence, it appears that C. gattii We thank Kathrin Tintelnot (Robert Koch Institute, Berlin, Ger- many) for serotyping some isolates; Arjan Burggraaf and William AFLP4 and AFLP6 strains are collected and preserved more Chew for help with AFLP genotyping; Ilse Breuker-Curfs and Sanne frequently and that the availability and exchangeability of Eijkemans for assistance with susceptibility testing; and the members strains can influence studies toward a specific genotype (13, of the European C. gattii Epidemiology Study Group (Francoise Dro- 27). It might influence the outcome of susceptibility values mer, Institut Pasteur, Paris, France; Kathrin Tintelnot, Robert Koch when only the geographical origin is taken into consideration, Institute, Berlin, Germany; Roberta Iatta and Maria Teresa Mon- tagna, Universita` degli Studi di Bari, Bari, Italy; Josep M. Torres- while a combined analysis of the geographical origin with a Rodriguez, IMIM Autonomous University of Barcelona, Barcelona, genotyping method (e.g., RFLP, AFLP, or MLST) will reveal Spain; Maria Anna Viviani, Universita` degli Studi di Milano, Milan, more subtle differences, as has been shown in the current Italy; Aristea Velegraki, University of Athens, Athens, Greece; Maria study. Francisca Colom Valiente, Universidad Miguel Herna`ndez, Alicante, C. gattii Spain; Martine Gari-Toussaint, Centre Hospitalier Universitaire, Nice, When only the clinical strains from the two major France; Manuel Cuenca-Estrella, Instituto de Salud Carlos III, Ma- genotypes, genotypes AFLP4 and AFLP6, were compared, it drid, Spain; and Jesu´s Guinea, Hospital General Universitario Grego- appeared that strains belonging to AFLP6 (n � 103) had rio Maranõń, Madrid, Spain); Wieland Meyer (University of Sydney,

71 Chapter 6

Sydney, Australia) and Eddie Byrnes and Joe Heitman (Duke Univer- susceptibilities of Cryptococcus gattii from the Pacific Northwest of the sity, Durham, NC) for providing C. gattii isolates. United States. J. Clin. Microbiol. 48:539–544. F.H. was funded by the Odo van Vloten Foundation, Netherlands, 16. Kidd, S. E., F. Hagen, R. L. Tscharke, M. Huynh, K. H. Bartlett, M. Fyfe, L. and M.-T.I-Z. was sponsored by the International Society for Human Macdougall, T. Boekhout, K. J. Kwon-Chung, and W. Meyer. 2004. A rare and Animal Mycology. This study has been partly sponsored by a grant genotype of Cryptococcus gattii caused the cryptococcosis outbreak on Van- couver Island (British Columbia, Canada). Proc. Natl. Acad. Sci. U. S. A. from Basilea Pharmaceutica, Basel, Switzerland. 101:17258–17263. J.F.M. has been a consultant to Astellas, Basilea, Merck, and Scher- 17. Kwon-Chung, K. J., and J. E. Bennett. 1984. Epidemiologic differences ing-Plough and received speaker’s fees from Gilead, Janssen Pharma- between the two varieties of Cryptococcus neoformans. Am. J. Epidemiol. ceutica, Merck, Pfizer, and Schering-Plough. C.H.W.K. received a 120:123–130. grant from Pfizer. None of the other authors has a potential conflict of 18. Kwon-Chung, K. J., I. Polacheck, and J. E. Bennett. 1982. Improved diag- interest. The sponsors of the research played no decision-making role nostic medium for separation of Cryptococcus neoformans var. neoformans in the design, execution, analysis, or reporting of the research. (serotypes A and D) and Cryptococcus neoformans var. gattii (serotypes B and C). J. Clin. Microbiol. 15:535–537. REFERENCES 19. Ma, H., and R. C. May. 2009. Virulence in Cryptococcus species. Adv. Appl. Microbiol. 67:131–190. 1. Bartlett, K. H., S. E. Kidd, and J. W. Kronstad. 2008. The emergence of 20. Meyer, W., D. M. Aanensen, T. Boekhout, M. Cogliati, M. R. Diaz, M. C. Cryptococcus gattii in British Columbia and the Pacific Northwest. Curr. Esposto, M. Fisher, F. Gilgado, F. Hagen, S. Kaocharoen, A. P. Litvintseva, Infect. Dis. Rep. 10:58–65. T. G. Mitchell, S. P. Simwami, L. Trilles, M. A. Viviani, and K. J. Kwon- 2. Boekhout, T., B. Theelen, M. Diaz, J. W. Fell, W. C. Hop, E. C. Abeln, F. Chung. 2009. Consensus multi-locus sequence typing scheme for Cryptococ- Dromer, and W. Meyer. 2001. Hybrid genotypes in the pathogenic yeast cus neoformans and Cryptococcus gattii. Med. Mycol. 47:561–570. Cryptococcus neoformans. Microbiology 147:891–907. 3. Bolano, A., S. Stinchi, R. Preziosi, F. Bistoni, M. Allegrucci, F. Baldelli, A. 21. Meyer, W., A. Castan˜eda, S. Jackson, M. Huynh, E. Castan˜eda, and the Martini, and G. Cardinali. 2001. Rapid methods to extract DNA and RNA IberoAmerican Cryptococcal Study Group. 2003. Molecular typing of from Cryptococcus neoformans. FEMS Yeast Res. 1:221–224. IberoAmerican Cryptococcus neoformans isolates. Emerg. Infect. Dis. 9:189– 4. Bovers, M., F. Hagen, E. E. Kuramae, H. L. Hoogveld, F. Dromer, G. 195. St.-Germain, and T. Boekhout. 2008. AIDS patient death caused by novel 22. Morera-Lo´pez, Y., J. M. Torres-Rodríguez, T. Jimeńez-Cabello, and T. Baro´- Cryptococcus neoformans � C. gattii hybrid. Emerg. Infect. Dis. 14:1105– Toma´s. 2005. Cryptococcus gattii: in vitro susceptibility to the new antifungal 1108. albaconazole versus fluconazole and voriconazole. Med. Mycol. 43:505–510. 5. Bovers, M., F. Hagen, E. E. Kuramae, M. R. Diaz, L. Spanjaard, F. Dromer, 23. Ngamskulrungroj, P., F. Gilgado, J. Faganello, A. P. Litvintseva, A. L. Leal, H. L. Hoogveld, and T. Boekhout. 2006. Unique hybrids between the fungal K. M. Tsui, T. G. Mitchell, M. H. Vainstein, and W. Meyer. 2009. Genetic pathogens Cryptococcus neoformans and Cryptococcus gattii. FEMS Yeast diversity of the Cryptococcus species complex suggests that Cryptococcus Res. 6:599–607. gattii deserves to have varieties. PLoS One 4:e5862. 6. Bovers, M., F. Hagen, E. E. Kuramae, and T. Boekhout. 2008. Six mono- 24. Perfect, J. R., W. E. Dismukes, F. Dromer, D. L. Goldman, J. R. Graybill, phyletic lineages identified within Cryptococcus neoformans and Cryptococcus R. J. Hamill, T. S. Harrison, R. A. Larsen, O. Lortholary, M. H. Nguyen, gattii by multi-locus sequence typing. Fungal Genet. Biol. 45:400–421. P. G. Pappas, W. G. Powderly, N. Singh, J. D. Sobel, and T. C. Sorrell. 2010. 7. Bovers, M., F. Hagen, and T. Boekhout. 2008. Diversity of the Cryptococcus Clinical practice guidelines for the management of cryptococcal disease: neoformans-Cryptococcus gattii species complex. Rev. Iberoam. Micol. 25:S4– 2010 update by the Infectious Diseases Society of America. Clin. Infect. Dis. S12. 50:291–322. 8. Byrnes, E. J., R. J. Bildfell, S. A. Frank, T. G. Mitchell, K. A. Marr, and 25. Pfaller, M. A., S. A. Messer, L. Boyken, C. Rice, S. Tendolkar, R. J. Hollis, J. Heitman. 2009. Molecular evidence that the range of the Vancouver G. V. Doern, and D. J. Diekema. 2005. Global trends in the antifungal Island outbreak of Cryptococcus gattii infection has expanded into the Pacific susceptibility of Cryptococcus neoformans (1990 to 2004). J. Clin. Microbiol. Northwest in the United States. J. Infect. Dis. 199:1081–1086. 43:2163–2167. 9. Clinical and Laboratory Standards Institute. 2008. Reference method for 26. Sorrell, T. C. 2001. Cryptococcus neoformans variety gattii. Med. Mycol. broth dilution antifungal susceptibility testing of yeasts. Approved standard 39:155–168. M27-A3. Clinical and Laboratory Standards Institute, Wayne, PA. 27. Springer, D. J., and V. Chaturvedi. 2010. Projecting global occurrence of 10. Datta, K., K. H. Bartlett, R. Baer, E. J. Byrnes, E. Galanis, J. Heitman, L. Cryptococcus gattii. Emerg. Infect. Dis. 16:14–20. Hoang, M. J. Leslie, L. MacDougall, S. S. Magill, M. G. Morshed, K. A. 28. Thompson, G. R., III, A. W. Fothergill, N. P. Wiederhold, A. C. Vallor, B. L. Marr, and the Cryptococcus gattii Working Group of the Pacific Northwest. Wickes, and T. F. Patterson. 2008. Evaluation of Etest method for deter- 2009. Spread of Cryptococcus gattii into Pacific Northwest region of the mining isavuconazole MICs against Cryptococcus gattii and Cryptococcus United States. Emerg. Infect. Dis. 15:1185–1191. neoformans. Antimicrob. Agents Chemother. 52:2959–2961. 11. Dromer, F., C. Bernede-Bauduin, D. Guillemot, O. Lortholary, and the 29. Thompson, G. R., III, N. P. Wiederhold, A. W. Fothergill, A. C. Vallor, B. L. French Cryptococcosis Study Group. 2008. Major role for amphotericin Wickes, and T. F. Patterson. 2009. Antifungal susceptibilities among differ- B-flucytosine combination in severe cryptococcosis. PLoS One 3:e2870. ent serotypes of Cryptococcus gattii and Cryptococcus neoformans. Antimi- 12. Gomez-Lopez, A., O. Zaragoza, M. Dos Anjos Martins, M. C. Melhem, J. L. crob. Agents Chemother. 53:309–311. Rodriguez-Tudela, and M. Cuenca-Estrella. 2008. In vitro susceptibility of 30. Trilles, L., B. Fernandez-Torres, M. dos Santos Lazera, B. Wanke, and J. Cryptococcus gattii clinical isolates. Clin. Microbiol. Infect. 14:727–730. Guarro. 2004. In vitro antifungal susceptibility of Cryptococcus gattii. J. Clin. 13. Hagen, F., A. G. Burggraaf, A. Kamermans, J. Sweere, K. Tintelnot, D. Microbiol. 42:4815–4817. Swinne, F. Dromer, R. Iatta, M. T. Montagna, J. M. Torres-Rodriguez, M. A. 31. Van Elden, L. J., A. M. Walenkamp, M. M. Lipovsky, P. Reiss, J. F. Meis, S. Viviani, A. Velegraki, M. F. Colom Valiente, M. Gari-Toussaint, and T. de Marie, J. Dankert, and A. I. Hoepelman. 2000. Declining number of Boekhout. 2009. The occurrence of the primary pathogenic yeast Cryptococ- patients with cryptococcosis in the Netherlands in the era of highly active cus gattii in Europe. Mycoses 52(Suppl. 1):57. antiretroviral therapy. AIDS 14:2787–2788. 14. Illnait-Zaragozi, M. T., G. F. Martínez, I. Curfs-Breuker, C. M. Fernańdez, 32. Velagapudi, R., Y. P. Hsueh, S. Geunes-Boyer, J. R. Wright, and J. Heitman. T. Boekhout, and J. F. Meis. 2008. In vitro activity of the new azole isavu- 2009. Spores as infectious propagules of Cryptococcus neoformans. Infect. conazole (BAL4815) compared with six other antifungal agents against 162 Immun. 77:4345–4355. Cryptococcus neoformans isolates from Cuba. Antimicrob. Agents Che- 33. Yamazumi, T., M. A. Pfaller, S. A. Messer, A. Houston, R. J. Hollis, and mother. 52:1580–1582. R. N. Jones. 2000. In vitro activities of ravuconazole (BMS-207147) against 15. Iqbal, N., E. E. DeBess, R. Wohrle, B. Sun, R. J. Net, A. M. Ahlquist, T. 541 clinical isolates of Cryptococcus neoformans. Antimicrob. Agents Che- Chiller, and S. R. Lockhart. 2010. Correlation of genotype and in vitro mother. 44:2883–2886.

72 Chapter 7 Cryptococcus neoformans- Cryptococcus gattii species complex: an international study of wild-type susceptibility endpoint distributions and epidemiological cutoff values for fluconazole, itraconazole, posaconazole, and voriconazole

Published in Antimicrobial Agents and Chemotherapy 2012;56:5898-906. Chapter 7

Cryptococcus neoformans-Cryptococcus gattii Species Complex: an International Study of Wild-Type Susceptibility Endpoint Distributions and Epidemiological Cutoff Values for Fluconazole, Itraconazole, Posaconazole, and Voriconazole

A. Espinel-Ingroff,a A. I. Aller,b E. Canton,c L. R. Castañón-Olivares,d A. Chowdhary,e S. Cordoba,f M. Cuenca-Estrella,g A. Fothergill,h J. Fuller,i N. Govender,j F. Hagen,k M. T. Illnait-Zaragozi,l E. Johnson,m S. Kidd,n C. Lass-Flörl,o S. R. Lockhart,p M. A. Martins,q J. F. Meis,r M. S. C. Melhem,q L. Ostrosky-Zeichner,s T. Pelaez,t M. A. Pfaller,u W. A. Schell,v G. St-Germain,w L. Trilles,x and J. Turnidgey VCU Medical Center, Richmond, Virginia, USAa; Hospital Universitario de Valme, Sevilla, Spainb; Unidad de Microbiologia Experimental, Hospital Universitario La Fe, Valencia, Spainc; Universidad Nacional Autónoma de Mexico, Distrito Federal, Méxicod; Vallabhbhai Patel Chest Institute, University of Delhi, Delhi, Indiae; Departamento de Micología, Instituto Nacional de Enfermedades Infecciosas, ANLIS “Dr. Carlos G. Malbrán,” Buenos Aires, Argentinaf; Servicio de Micología, Centro Nacional de Microbiología, Instituto de Salud Carlos III, Madrid, Spaing; University of Texas Health Science Center, San Antonio, Texas, USAh; The University of Alberta, Edmonton, Alberta, Canadai; National Institute for Communicable Diseases, National Health Laboratory Service, Johannesburg, South Africaj; Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, Nijmegen, Netherlandsk; Institute of Tropical Medicine Pedro Kouri, Havana, Cubal; The HPA Mycology Reference Laboratory, Kingsdown, Bristol, United Kingdomm; Women’s and Children’s Hospital, Adelaide, South Australia, Australian; The Innsbruck Medical University, Innsbruck, Austriao; Mycotic Diseases Branch, Centers for Disease Control and Prevention, Atlanta, Georgia, USAp; The Adolfo Lutz Institute Public Health Reference Center, São Paulo and Rio Claro, Brazilq; Department of Medical Microbiology, Radboud University Medical Centre, Canisius Wilhelmina Hospital, Nijmegen, the Netherlandsr; University of Texas Health Science Center, Houston, Texas, USAs; Hospital General Universitario Gregorio Marañón, Faculty of Medicine-Universidad Complutense, Madrid, Spaint; University of Iowa, Iowa City, Iowa, USAu; Duke University Medical Center, Durham, North Carolina, USAv; Laboratoire de Santé Publique du Québec, Canadaw; Instituto de Pesquisa Clinica Evandro Chagas-FIOCRUZ, Rio de Janeiro, Brazilx; and University of Adelaide, Adelaide, South Australia, Australiay

Epidemiological cutoff values (ECVs) for the Cryptococcus neoformans-Cryptococcus gattii species complex versus ﬂuconazole, itraconazole, posaconazole, and voriconazole are not available. We established ECVs for these species and agents based on wild- type (WT) MIC distributions. A total of 2,985 to 5,733 CLSI MICs for C. neoformans (including isolates of molecular type VNI [MICs for 759 to 1,137 isolates] and VNII, VNIII, and VNIV [MICs for 24 to 57 isolates]) and 705 to 975 MICs for C. gattii (including 42 to 260 for VGI, VGII, VGIII, and VGIV isolates) were gathered in 15 to 24 laboratories (Europe, United States, Argen- tina, Australia, Brazil, Canada, Cuba, India, Mexico, and South Africa) and were aggregated for analysis. Additionally, 220 to 359 MICs measured using CLSI yeast nitrogen base (YNB) medium instead of CLSI RPMI medium for C. neoformans were evaluated. CLSI RPMI medium ECVs for distributions originating from at least three laboratories, which included >95% of the modeled WT population, were as follows: ﬂuconazole, 8 �g/ml (VNI, C. gattii nontyped, VGI, VGIIa, and VGIII), 16 �g/ml (C. neoformans nontyped, VNIII, and VGIV), and 32 �g/ml (VGII); itraconazole, 0.25 �g/ml (VNI), 0.5 �g/ml (C. neoformans and C. gattii nontyped and VGI to VGIII), and 1 �g/ml (VGIV); posaconazole, 0.25 �g/ml (C. neoformans nontyped and VNI) and 0.5 �g/ml (C. gattii nontyped and VGI); and voriconazole, 0.12 �g/ml (VNIV), 0.25 �g/ml (C. neoformans and C. gattii nontyped, VNI, VNIII, VGII, and VGIIa,), and 0.5 �g/ml (VGI). The number of laboratories contributing data for other molecular types was too low to ascertain that the differences were due to factors other than assay variation. In the absence of clinical breakpoints, our ECVs may aid in the detection of isolates with acquired resistance mechanisms and should be listed in the revised CLSI M27-A3 and CLSI M27-S3 documents.

he Cryptococcus neoformans-Cryptococcus gattii complex is the 31, 51). C. neoformans comprises molecular types VNI and VNII Tmost common species of non-Candida yeasts recovered from (both serotype A), VNIII (serotype A/D hybrid), and VNIV (se- clinical specimens (32.9% of 8,717 isolates) (41). In addition, rotype D), while C. gattii comprises VGI, VGII, and VGIV (all cryptococcal disease has been reported as the second most com- serotype B), VGII, and VGIV (both serotype C) (25). Molecular mon severe fungal infection in certain regions (41). Infections type VGII has been of particular interest in recent years due to the caused by C. neoformans var. grubii (serotype A) and, to a lesser emergence of the novel subtypes VGIIa, VGIIb, and VGIIc; mo- degree, by C. neoformans var. neoformans (serotype D) are seen worldwide among immunocompromised hosts (4). The more geographically restricted C. gattii (serotypes B and C) causes in- Received 29 May 2012 Returned for modification 18 August 2012 fections among immunocompromised as well as nonimmuno- Accepted 27 August 2012 compromised patients, and the infections are more difficult to Published ahead of print 4 September 2012 treat (27, 38). Irrespective of the species, cryptococcal disease is Address correspondence to A. Espinel-Ingroff, [email protected]. associated with high mortality rates (�12.7%) (15, 17, 38). Using Copyright © 2012, American Society for Microbiology. All Rights Reserved. molecular methodologies, eight major molecular types have been doi:10.1128/AAC.01115-12 identified among the four serotypes and their hybrids (4, 6, 8, 25,

74 Wild-type susceptibility endpoint distributions and epidemiological cutoff values

lecular types VNI to VNIV and VGI to VGIV also have been des- MATERIALS AND METHODS ignated AFLP1 to AFLP3 and AFLP4 to AFLP7, respectively (4, 6, Isolates. Each isolate originated from a unique clinical specimen. The 7). Identification of these different molecular types has been asso- MICs of the four triazoles used in the present study for ECV definition ciated with differences in antifungal susceptibility and virulence were obtained at one of the following medical centers: VCU Medical Cen- (4, 6–11, 22, 25, 33, 51). ter, Richmond, VA; Unidad de Microbiologia Experimental, Hospital In addition to the different amphotericin B formulations, flu- Universitario La Fe, Valencia, Spain; Universidad Nacional Autónoma de conazole and itraconazole are recommended as primary alterna- México, Mexico; Vallabhbhai Patel Chest Institute, University of Delhi, tive induction treatments for infections caused by C. neoformans Delhi, India; Departamento de Micología, INEI, ANLIS “Dr. Carlos G. and C. gattii and voriconazole and posaconazole as salvage con- Malbrán,” Buenos Aires, Argentina; Servicio de Micología, Centro Nacio- nal de Microbiología, Instituto de Salud Carlos III, Madrid, Spain; solidation therapies (35, 38); fluconazole is also the drug of choice University of Texas Health Science Center, San Antonio, TX; The Univer- for lifelong suppressive (maintenance) therapy or primary ther- sity of Alberta, Edmonton, Alberta, Canada; National Institute for Com- apy in some areas (24). The azoles block the pathway of ergosterol municable Diseases, National Health Laboratory Service, Johannesburg, biosynthesis by inhibiting the 14-�-lanosterol demethylase en- South Africa; Department of Medical Microbiology and Infectious zyme, which is coded by the CYP51 gene in C. neoformans Diseases, Canisius-Wilhelmina Hospital, Nijmegen, the Netherlands; In- (CnCYP51; also called ERG11)(46). The wide use of fluconazole stitute of Tropical Medicine Pedro Kouri, Havana, Cuba; The HPA My- and other triazoles has led to in vitro resistance among Candida cology Reference Laboratory, Kingsdown, Bristol, United Kingdom; and Cryptococcus isolates, especially to fluconazole more than to Women’s and Children’s Hospital, Adelaide, South Australia, Australia; the newer triazoles, voriconazole, and posaconazole (38). Two The Innsbruck Medical University, Innsbruck, Austria; Centers for Dis- azole resistance mechanisms have been identified in C. neofor- ease Control and Prevention, Atlanta, GA; The Adolfo Lutz Institute Pub- mans (18, 32, 46, 54). lic Health Reference Center, São Paulo and Rio Claro, Brazil; Hospital Universitario de Valme, Sevilla, Spain; University of Texas Health Science The use of standard testing methods has allowed the recog- Center, Houston, TX; Hospital General Universitario Gregorio Marañón, nition of antifungal resistance, as well as the proposal of clinical Faculty of Medicine-Universidad Complutense, Madrid, Spain; Univer- breakpoints (CBPs) for Candida spp. and epidemiological cut- sity of Iowa, Iowa City, IA; Duke University Medical Center, Durham, off values (ECVs) for Aspergillus spp. by the Clinical and Lab- NC; Institut national de santé publique du Québec, Laboratoire de santé oratory Standards Institute (CLSI) and the European Commit- publique du Québec, Canada; and Instituto de Pesquisa Clinica Evandro tee for Antimicrobial Susceptibility Testing (AFST-EUCAST) Chagas-FIOCRUZ, Rio de Janeiro, Brazil. Species and molecular type (19–21, 40, 42, 44, 47). More recently, CLSI amphotericin B identification were performed at each medical center using standard and flucytosine ECVs were defined for the Cryptococcus neofor- methodologies (8, 11, 25, 28, 33, 36, 51). We have aggregated the maxi- mans-Cryptococcus gattii species complex (22). CBPs are based mum available CLSI data from each laboratory and agent as follows: 2,985 7 on MIC distributions, pharmacokinetic and pharmacody- posaconazole, 4,019 itraconazole, 4,693 voriconazole, and 5,733 flucona- namic (PK/PD) parameters, animal studies, and clinical out- zole MICs for C. neoformans (including MICs for the VNI to VNIV isolates) and 705 posaconazole, 828 itraconazole, 923 voriconazole, and 975 comes to therapy, while the ECV is based mostly on MIC dis- fluconazole MICs for C. gattii (including those for the VGI to VGIV iso- tributions. Numerous surveys indicate that CLSI MICs are �16 lates) (8, 11, 24, 25, 26, 33, 36, 51). Sets of 220 (itraconazole) to 359 �g/ml (fluconazole) and �0.5 �g/ml (the other three triazoles) (fluconazole) MICs obtained using CLSI YNB broth instead of the CLSI for most C. neoformans and C. gattii isolates (5, 9, 10, 16, 24, 43, RPMI medium (12) for C. neoformans were also available and analyzed 50). In the last few years, azole (mostly fluconazole) antifungal separately. One or both quality control (QC) isolates (Candida parapsilo- susceptibility differences have been reported for these two spe- sis ATCC 22019 and Candida krusei ATCC 6258) were used by the partic- cies and for their molecular types and serotypes (10, 11, 25, 33, ipant laboratories (Table 1)(12, 13). 50, 51). However, CBPs or ECVs based on data from multiple Antifungal susceptibility testing. In order to include MIC results in laboratories are not available for either C. neoformans or C. the total set of available aggregated CLSI data from the participant labo- gattii versus the triazoles. ECVs defined in the present study ratories (Tables 2 to 5), triazole MICs were obtained at each center by could help to characterize the susceptibility of these species and following the CLSI M27-A3 broth microdilution method (standard RPMI 1640 broth [0.2% dextrose], final inoculum concentrations that ranged to monitor the emergence of strains with mutations that could from 0.4 � 103 to 5 � 103 CFU/ml, and 72 h of incubation); MICs were lead to reduced antifungal susceptibility to fluconazole, itra- the lowest drug concentrations that produced �50% growth inhibition conazole, posaconazole, and voriconazole. compared to the growth control (12). Testing conditions and interpre- The purpose of the study was dual: (i) to define wild-type (WT; tation of MICs were otherwise identical for those isolates grown in population of isolates in a species-drug combination with no de- CLSI YNB broth. MIC data for the two QC reference strains, utilized tectable acquired resistance mechanisms) (14, 52) susceptibility during the years of testing in each center, were obtained each time that endpoint distributions of each species/molecular type and triazole a set of isolates was tested following the CLSI M27-A3 broth microdi- combination originating from at least 3 laboratories and (ii) to lution method (12, 13). propose ECVs (highest WT susceptibility endpoint) of four tria- Definitions. The ECV, a term previously used in similar fungal reports zoles. We aggregated the CLSI RPMI broth microdilution MICs of and also known as the COWT, is the highest wild-type (WT) cutoff sus- fluconazole, itraconazole, posaconazole, and voriconazole ob- ceptible value (19–22, 44). ECVs are based on MIC distributions where tained in 15 to 24 laboratories (2,985 to 5,733 MICs for C. neofor- there are two distinct populations: (i) the WT population of isolates/MICs with no detectable acquired or mutational resistance to the drug being mans and 705 to 975 MICs for C. gattii [species/molecular type evaluated and (ii) the non-WT population or isolates that harbor one or and agent/combination dependent]) in Europe, the United States, more resistance markers (14, 52). A non-WT organism shows reduced Argentina, Australia, Canada, Cuba, Brazil, India, Mexico, and susceptibility to the agent being evaluated compared to the WT popula- South Africa. The 220 to 359 MICs that were obtained using the tion, but it may or may not respond to treatment to the drug being eval- alternative CLSI yeast nitrogen base (YNB) broth (12, 13) for C. uated. When data pertaining to the establishment of clinical breakpoints neoformans were analyzed separately. (CBPs) are not available, ECVs serve as an early indication of emerging

75 Chapter 7

TABLE 1 MICs for QC strains used in 11 to 15 laboratories according to the CLSI broth microdilution method at 48 ha QC MIC range (mode) % MICs within MIC range in Antifungal in �g/ml (CLSI range (CLSI �g/ml QC isolate agent document) document) (present study) Modes in �g/ml (present study)b Candida parapsilosis Fluconazole 1–4 (2) 98.1 1–4 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 4, 4, ND ATCC 22019 Itraconazole 0.06–0.5 (0.25)b 97.5 0.03–0.5c 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.25, 0.25, ND, ND, ND, ND Posaconazole 0.03–0.25 (0.12) 98.8 0.03–0.25 0.06, 0.06, 0.06, 0.06, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.25, ND, ND, ND, ND, ND, ND, ND Voriconazole 0.03–0.25 (0.06) 100 0.01–0.25c 0.03, 0.03, 0.03, 0.03, 0.03, 0.03, 0.06, 0.06, 0.06, 0.06, 0.06, 0.06, 0.12, 0.12, 0.12, ND, ND, ND, ND Candida krusei Fluconazole 16–128 (32) 100 2–64d 16, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 64, ND, ND, ND ATCC 6258 Itraconazole 0.25–1 (0.5) 100 0.03–1d 0.12, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.5, 0.5, 0.5, 0.5, ND, ND, ND, ND, ND, ND Posaconazole 0.12–1 (0.5) 99.6 0.12–2d 0.12, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, ND, ND, ND, ND, ND, ND, ND Voriconazole 0.12–1 (0.5) 100 0.12–0.5 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.5, 0.5, 0.5, 0.5, ND, ND, ND, ND a MICs were determined as described by the CLSI M27-A3 document using RPMI 1640 broth (12). b Modes were not listed for 9 of the 25 laboratories because only 24-h (from 4 to 5 laboratories) or a single (from two to four laboratories) MIC(s) for QC isolates (C. parapsilosis and C. krusei) were reported for small subsets of isolates (8 to 84 isolates). However, MICs were in range (95 to 100%) and available 24-h modes were within 1 to 2 dilutions. ND, not determined or obtained using the CLSI YNB broth. c MICs were outside the range in one (itraconazole [4.5%]) and two (voriconazole [2 to 3%]) laboratories for QC isolate C. parapsilosis ATCC 22019. d MICs were outside the range in two (ﬂuconazole [0.3 to 5%] and posaconazole [1.5 to 2%]) to four (itraconazole [0.8 to 5%]) laboratories for QC isolate C. krusei ATCC 6258.

changes in the patterns of susceptibility of organisms to the agent being lates (here referred to as nontyped isolates), (ii) CLSI RPMI medium evaluated. aggregated data from 2 to 6 laboratories for VNI to VNIV molecular types Data analysis. MIC distributions from each laboratory (laboratories (here referred to as typed isolates), and (iii) single-laboratory CLSI YNB were coded) for each available combination of triazole and species or data. The ﬁnal WT distributions for C. gattii (Tables 2 to 5) were as fol- molecular type were listed in Excel spreadsheets before they were aggre- lows: (i) CLSI RPMI medium aggregated data of each of the four triazoles gated for the analysis. The MIC distributions of each of the four triazoles from 4 to 6 laboratories for nontyped isolates and (ii) CLSI RPMI medium for C. neoformans (Tables 2 to 5) were as follows: (i) CLSI RPMI medium aggregated data from 1 to 7 laboratories for VGI to VGIV (also referred to aggregated data from 11 to 18 laboratories for nonmolecular typed iso- as typed isolates). The aggregated MIC distributions obtained in at least

TABLE 2 WT ﬂuconazole MIC distributions for the Cryptococcus neoformans-Cryptococcus gattii species complexa No. of isolates for which the MIC (�g/ml) was:e Species or No. of Total no. molecular type Mediumb Genotypedc labsd of isolates �0.12 0.25 0.5 1 2 4 8 16 32 �64 C. neoformans RPMI N 18 4,446 92 133 258 543 1,225 1,372 569 180 52 22 VNI Y 6 1,137 4 12 40 127 376 456 89 20 10 3 VNII Y 4 42 1 1 1 7 13 12 5 2 VNIII Y 4 54 1 3 9 9 22 811 VNIV Y 4 54 2 17 19 6622 All isolates Both 24 5,733 97 149 319 705 1,629 1,868 668 206 67 25

C. neoformans YNB N 1 359 1 5 13 23 73 84 126 26 8

C. gattii RPMI N 6 137 1 8 19 71 25 6 4 3 VGI Y 7 260 1 18 39 69 101 29 3 VGIIf Y 4 101 3 3 26 32 17 13 7 VGIIa Y 3 200 2 1 16 93 80 6 2 VGIIb Y 2 106 3 9 31 54 9 VGIIc Y 2 42 1 2 21 17 1 All VGII isolates Y 6 449 2 7 28 151 168 53 32 8 VGIII Y 3 43 1 1 6 12 18 41 VGIV Y 3 86 2 3 21 19 32 54 All isolates Both 10 975 2 24 63 149 360 258 68 40 11 a Fluconazole MICs were determined by the CLSI broth microdilution method (12). b RPMI and YNB, RPMI 1640 and yeast nitrogen base, respectively, as described by the CLSI M27-A3 document (12). c Y, yes; N, no. d Number of laboratories contributing data to each MIC distribution. e The modal MIC for each distribution is underlined. f Isolates identiﬁed as belonging to the VGII molecular type and not being one of the VGIIa, VGIIb, or VGIIc subtypes examined as a separate group.

76 Wild-type susceptibility endpoint distributions and epidemiological cutoff values

Table 3 WT itraconazole MIC distributions for Cryptococcus neoformans-Cryptococcus gattii species complexa No. of isolates for which the MIC (�g/ml) was:e Species or No. of Total no. molecular type Mediumb Genotypedc labsd of isolates �0.008 0.016 0.03 0.06 0.12 0.25 0.5 1 2 �4 C. neoformans RPMI N 11 2,731 106 174 377 949 865 222 32 4 2 VNI Y 6 1,145 25 54 147 315 447 144 13 VNII Y 4 32 1 3 3 15 721 VNIII Y 4 54 1 8 20 14 4 7 VNIV Y 4 57 17 15 9 6 6 4 All isolates Both 20 4,019 26 181 347 736 1,423 1,021 247 32 4 2

C. neoformans YNB N 1 220 1 5 22 38 91 61 1 1

C. gattii RPMI N 5 70 4 9 16 25 12 4 VGI Y 6 257 3 11 42 77 103 20 1 VGIIf Y 3 47 2 13 15 10 6 1 VGIIa Y 2 176 7 15 44 63 41 6 VGIIb Y 2 106 4 5 34 44 14 4 1 VGIIc Y 2 42 1 9 24 71 All VGII isolates Y 4 371 4 9 33 94 126 85 18 2 VGIII Y 3 44 1 1 3 14 17 62 VGIV Y 3 86 1 1 4 22 24 29 32 All isolates Both 9 828 8 5 31 98 232 282 144 24 4 a Itraconazole MICs were determined by the CLSI broth microdilution method (12). b RPMI and YNB, RPMI 1640 and yeast nitrogen base, respectively, as described by the CLSI M27-A3 document (12). c Y, yes; N, no. d Number of laboratories contributing data to each MIC distribution. e The modal MIC for each distribution is underlined. f Isolates identiﬁed as belonging to the VGII molecular type and not being one of the VGIIa, VGIIb, or VGIIc subtypes examined as a separate group. three laboratories were used to calculate ECVs by the statistical method calculate the MIC that captures at least 95%, 97.5%, and 99% of the 7 (52), where the modeled population is based on ﬁtting a normal distribu- modeled WT population (19–22). A search for outlier laboratories in each tion at the lower end of the MIC range, working out the mean and stan- distribution was also performed, and only one outlier set of voriconazole dard deviation of that normal distribution, and using those parameters to MICs was excluded from the analysis. In addition, ECVs were not esti-

Table 4 WT posaconazole MIC distributions for Cryptococcus neoformans-Cryptococcus gattii species complexa No. of isolates for which the MIC (�g/ml) was:e Species or Total no. No. of molecular type Mediumb Genotypedc of labsd isolates �0.008 0.016 0.03 0.06 0.12 0.25 0.5 1 2 �4 C. neoformans RPMI N 13 2,120 52 152 418 938 440 92 23 5 VNI Y 3 759 21 49 108 334 203 42 2 VNII Y 2 24 1 15 521 VNIII Y 2 35 2 19 12 2 VNIV Y 2 47 18 11 11 4 3 All isolates Both 15 2,985 21 122 305 780 1,149 486 94 23 5

C. neoformans YNB ND ND

C. gattii RPMI N 4 68 1 8 8 17 20 12 1 1 VGI Y 3 182 3 6 37 64 58 13 1 VGIIf Y 1 5 11111 VGIIa Y 2 176 7 18 51 62 35 3 VGIIb Y 2 106 5 6 60 31 4 VGIIc Y 2 42 1 3 6 21 11 All VGII isolates Y 2 329 5 1 8 25 115 100 61 14 VGIII Y 2 42 2 14 7 8 10 1 VGIV Y 2 84 2 17 25 28 12 All isolates Both 6 705 9 9 26 93 223 203 113 29 a Posaconazole MICs were determined by the CLSI broth microdilution method (12). b RPMI and YNB, RPMI 1640 and yeast nitrogen base, respectively, as described by the CLSI M27-A3 document (12). c Y, yes; N, no. d Number of laboratories contributing data to each MIC distribution. e The modal MIC for each distribution is underlined. f Isolates identiﬁed as belonging to the VGII molecular type and not being one of the VGIIa, VGIIb, or VGIIc subtypes examined as a separate group.

77 Chapter 7

Table 5 WT voriconazole MIC distributions for Cryptococcus neoformans-Cryptococcus gattii species complexa No. of isolates for which the MIC (�g/ml) was:e Species or Total no. No. of molecular type Mediumb Genotypedc of labsd isolates �0.008 0.016 0.03 0.06 0.12 0.25 0.5 1 2 �4 C. neoformans RPMI N 12 3,473 34 412 814 1,090 748 252 92 20 6 5 VNI Y 5 1,089 35 48 100 385 376 119 19 5 1 1 VNII Y 3 27 5 6 9 34 VNIII Y 3 51 9 7 19 12 3 1 VNIV Y 3 53 13 19 12 6 2 1 All isolates Both 22 4,693 69 487 946 1,515 1,145 380 113 25 7 6

C. neoformans YNB N 1 281 35 78 105 53 10

C. gattii RPMI N 5 98 2 7 13 37 23 11 5 VGI Y 6 258 3 13 38 69 90 45 VGIIf Y 3 96 1 1 29 38 23 4 VGIIa Y 3 197 6 3 47 113 23 3 2 VGIIb Y 2 106 3 2 6 46 47 2 VGIIc Y 2 42 1 2 7 19 11 2 All VGII isolates Y 5 441 3 7 7 84 204 112 20 4 VGIII Y 2 42 1 3 12 21 5 VGIV Y 2 84 8 28 29 16 3 All isolates Both 9 923 8 28 69 230 367 189 28 4 a Voriconazole MICs were determined by the CLSI broth microdilution method (12). b RPMI and YNB, RPMI 1640 and yeast nitrogen base, respectively, as described by the CLSI M27-A3 (12). c Y, yes; N, no. d Number of laboratories contributing data to each MIC distribution. e The modal MIC for each distribution is underlined. f Isolates identiﬁed as belonging to the VGII molecular type and not being one of the VGIIa, VGIIb, or VGIIc subtypes examined as a separate group.

mated when (i) the distribution was grossly skewed, which precluded especially fluconazole, are primary or salvage therapeutic agents statistical fitting, or (ii) there were data from less than 3 laboratories or the for cryptococcal infections (2, 38). total number of isolates in a molecular type was less than 50. Despite standardization efforts, variability is common during interlaboratory comparisons of MIC results (12, 13). Table 1 de- RESULTS AND DISCUSSION picts the MIC data obtained in participant laboratories for one or The ultimate goal of susceptibility testing is to predict with some both QC isolates (overall range and individual modes), C. parap- reliability the clinical outcome when an infected patient is treated silosis ATCC 22019 and C. krusei ATCC 6258, when clinical iso- with the specific agent evaluated and is referred to as the CBP (14, lates were tested. As in previous CLSI ECV studies (19–22), the 30). CBPs are based on the correlation of in vitro data, or the MIC majority of MIC ranges were within the CLSI established limits for for the infecting pathogen, with clinical and microbiological out- the two QC strains (�98%), with a certain degree of interlabora- comes in order to differentiate an organism as treatable or non- tory modal variability (mostly �2-fold dilution); inconsistencies treatable (14, 30). On the other hand, and as previously defined, were more frequent when testing itraconazole, where MICs were ECVs are based on MIC distributions that comprise the WT and outside the range of 0.8 to 5% for C. parapsilosis ATCC 22019 (1 non-WT populations; the ECV is the highest MIC that belongs to laboratory) and C. krusei ATCC 6258 (4 laboratories). Intralabo- the WT population. While CBPs predict clinical outcome of ther- ratory reproducibility was excellent in the laboratory using the apy, the role of the ECV is to detect emerging resistance or those CLSI YNB medium (data not shown). Similar modal variability non-WT strains with reduced susceptibility (due to mutations) to was also observed among the laboratories for the clinical Crypto- the agent being evaluated. Attempts to correlate in vitro flucona- coccus spp. isolates in this and previous multicenter studies estab- zole results with clinical outcome have not been successful (15, 55) lishing ECVs of several antifungal agents for Aspergillus spp. and and, to our knowledge, have not been reported for the other three Cryptococcus spp. (19–22). The most probable reason is the indi- triazoles. Therefore, CBPs for either C. neoformans or C. gatti in- vidual interpretations of MIC endpoints, and/or it may be due to fections versus any antifungal agent are not available due to the the use of different lots of antifungal powders. paucity of data on PK/PD and clinical outcomes compared to The aggregated MIC distributions of the four triazoles for C. MICs. However, the ECVs of the four triazoles established in the neoformans and C. gattii (for nontyped and typed isolates) are present study for C. neoformans and C. gattii may identify the depicted in Tables 2 to 5, which also list single-laboratory distri- non-WT clinical isolates and serve as an early indication of emerg- butions for C. neoformans (CLSI YNB data). Fluconazole modal ing changes in the susceptibility patterns of these organisms. Even MICs measured in RPMI medium were 1 to 4 �g/ml for C. neo- though cryptococcal meningitis and other infections decreased formans typed and nontyped isolates, with the lowest mode for with the use of antiretroviral therapies in developed countries, VNIV. Modes were consistently higher among the C. gattii groups these infections are still a major problem among immunosup- (4 to 16 �g/ml), with the highest for VGIIc (Table 2). Fluconazole pressed patients and in certain geographical areas. The triazoles, modes from individual contributing laboratories were either 2 or

78 Wild-type susceptibility endpoint distributions and epidemiological cutoff values

TABLE 6 ECVs and percentages of isolates of the Cryptococcus neoformans-Cryptococcus gattii species complex above each triazole WT distribution obtained in 3 to 18 laboratories by the CLSI M27-A3 broth microdilution method ECV (�g/ml) (% of observations above each statistical ECV or non-WT)c Antifungal Mode Statistical ECV Statistical ECV Statistical ECV Species Genotypea Agent (�g/ml)b �95% �97.5% �99% C. neoformans Nontyped isolates Fluconazole 4 16 (1.7) 16 (1.7) 32 (0.5) Itraconazole 0.12 0.5 (1.1) 1 (0.2) 1 (0.2) Posaconazole 0.12 0.25 (5.7) 0.5 (1.3) 0.5 (1.3) Voriconazole 0.06 0.25 (3.5) 0.25 (3.5) 0.5 (0.9) VNI Fluconazole 4 8 (2.9) 8 (2.9) 16 (1.1) Itraconazole 0.12 0.25 (1.1) 0.5 (0) 0.5 (0) Posaconazole 0.06 0.25 (0.3) 0.25 (0.3) 0.25 (0.3) Voriconazole 0.06 0.25 (2.4) 0.25 (2.4) 0.5 (0.6) VNIII Fluconazole 4 16 (1.9) 16 (1.9) 32 (0) Voriconazole 0.06 0.25 (2) 0.25 (2) 0.5 (0) VNIV Voriconazole 0.03 0.12 (5.7) 0.12 (5.7) 0.12 (5.7)

C. gattii Nontyped isolates Fluconazole 4 8 (9.5) 8 (9.5) 16 (5.1) Itraconazole 0.12 0.5 (0) 0.5 (0) 1 (0) Posaconazole 0.12 0.5 (1.5) 0.5 (1.5) 1 (0) Voriconazole 0.06 0.25 (5.1) 0.5 (0) 0.5 (0) VGI Fluconazole 4 8 (1.2) 16 (0) 16 (0) Itraconazole 0.25 0.5 (0.4) 0.5 (0.4) 1 (0) Posaconazole 0.12 0.5 (0.5) 0.5 (0.5) 1 (0) Voriconazole 0.12 0.5 (0) 0.5 (0) 1 (0) VGII Fluconazole 8 32 (6.9) 32 (6.9) 64 (1) Itraconazole 0.12 0.5 (2.1) 0.5 (2.1) 1 (0) Voriconazole 0.12 0.25 (4.1) 0.5 (0) 0.5 (0) VGIIa Fluconazole 4 8 (4) 16 (1) 16 (1) Voriconazole 0.12 0.25 (2.5) 0.25 (2.5) 0.25 (2.5) 7 VGIII Fluconazole 4 8 (2.3) 8 (2.3) 16 (0) Itraconazole 0.25 0.5 (4.5) 1 (0) 1 (0) VGIV Fluconazole 8 16 (4.7) 32 (0) 32 (0) Itraconazole 0.5 1 (2.3) 1 (2.3) 2 (0) a Data from laboratories using the CLSI RPMI broth (12). b Mode, MIC most frequently obtained for each distribution. c Calculated ECVs comprising �95%, �97.5%, or �99% of the statistically modeled population for which MIC distributions originated in at least three laboratories.

4 �g/ml for C. neoformans nontyped and most typed isolates; the (voriconazole mode, 0.25 �g/ml) and VGIIc and VGIV (itracona- exception was the mode (1 �g/ml) from two of the four laborato- zole and posaconazole mode, 0.5 �g/ml) (Tables 3 to 5). Reflect- ries that contributed data for VNIV, but the number of isolates ing fluconazole distributions, modes from individual participant was small. Overall, our values reflect previous fluconazole MIC90s laboratories were within one dilution of each other (modes, 0.12 and modes for similar sets of typed or nontyped C. neoformans and to 0.25 �g/ml and 0.06 to 0.12 �g/ml for itraconazole and both

C. gattii isolates and the fact that MICs are usually higher (MIC90s posaconazole and voriconazole, respectively). Based on these data and modes, 2 to 8 �g/ml versus 8 to 32 �g/ml) for C. gattii inde- and the wide geographical range over which the MICs have been pendent of testing conditions (10, 16, 23, 50). However, the MIC90 collected in the present study, we surmise that we are presenting has also been higher (16 �g/ml) for C. neoformans nontyped isolates valid data. (5, 39), but some of those results were obtained by the EUCAST Table 6 depicts the proposed fluconazole, itraconazole, po- method (39). The modal MIC for C. neoformans measured in CLSI saconazole, and voriconazole ECVs for the aggregated distribu- YNB medium was higher (mode, 8 �g/ml) than those measured in tions of C. neoformans and C. gattii (typed or nontyped isolates). CLSI RPMI medium; this discrepancy has been reported previously ECVs were proposed when the data originated from at least three (10, 55). These results further underline the variability of susceptibil- laboratories using the statistical methodology that comprised ity test results using different methodologies. �95% of the modeled population instead of the observed popu- Modes for the other three triazoles were more species-, molec- lation; the values that comprised �97.5 and �99% of the popu- ular type- and medium-dependent (Tables 3 to 5). In agreement lation were also obtained. Some of the proposed ECVs were based with previous data (10, 16, 23, 50), the MICs of itraconazole, po- on small numbers of isolates (Tables 2 to 5, n �100), and those saconazole, and voriconazole for C. gattii also tended to be higher could be considered tentative values. Although three other distri- (modes, 0.06 to 0.5 �g/ml, typed and nontyped isolates) than butions (fluconazole versus VNIV and itraconazole versus VNIII those for C. neoformans (modes, 0.016 to 0.12 �g/ml, typed and and VNIV) originated in at least three laboratories, statistical fit- nontyped isolates), with the highest modes for VGIIb and VGIIc ting was not possible and ECVs were not proposed. The CLSI

79 Chapter 7

fluconazole ECVs for the different subsets of C. neoformans iso- be species-specific and, for this fungal group, molecular type-spe- lates were either 8 �g/ml (VNI) or 16 �g/ml (nontyped and VNIII cific. This information is important; it has been reported that 20% isolates). Fluconazole ECVs for C. gattii ranged from 8 �g/ml of cryptococcal infections are caused by multiple strains or mo- (nontyped, VGI, VGIIa, and VGIII isolates) to 32 �g/ml (VGII). It lecular types, but generally only one isolate per infection is tested is interesting that a fluconazole MIC of �16 �g/ml is anecdotally (17). believed to be the resistance cutoff, but this value can be perceived During the 1990s, the early years of fluconazole therapy, re- here as a non-WT value for some of the subsets of the two species. lapses were not associated with in vitro resistance due to the higher So far mutations have not been observed when the fluconazole initial fluconazole doses (38). Since then, relapses have been at- MIC is �16 �g/ml (49). ECVs of the three other triazoles were also tributed to the excessive use of this agent as primary and prophy- species-, agent-, and molecular type-dependent as follows: (i) itra- lactic therapies, as well as the use of suboptimal doses for long conazole ECVs of 0.25 �g/ml (VNI), 0.5 �g/ml (nontyped isolates periods of time, especially among AIDS patients. As relapses began of both species and VGI to VGIII), and 1 �g/ml (VGIV); (ii) to be associated with high fluconazole MICs and/or a 4-fold MIC posaconazole ECVs of 0.25 �g/ml (C. neoformans nontyped and increase (1, 24, 37, 38), several aspects of sterol metabolism in VNI) and 0.5 �g/ml (C. gattii nontyped and VGI); and (iii) vori- these isolates were investigated to determine azole resistance conazole ECVs of 0.12 �g/ml (VNIV), 0.25 �g/ml (nontyped iso- mechanisms in C. neoformans. A point mutation in the ERG11 lates of both species, VNI, VNIII, VGII, and VGIIa), and 0.5 �g/ml (CnCYP51) gene resulting in a G484S substitution in the target (VGI). The overall trend was that ECVs for C. neoformans molec- enzyme has been observed in C. neoformans strains with high flu-

ular types were lower than those for C. gattii. Previous MIC90s and conazole MICs (�32 �g/ml); this mutation alters the affinity of modes of these three triazoles have been similar (9, 10, 23, 44, 50). the drug for the target enzyme (32, 48, 54). The overexpression of Because the distributions in CLSI YNB originated from a single the gene CnAFR1 that encodes the membrane efflux pumps has laboratory, ECVs were not proposed for these distributions (Ta- been found to reduce cell drug (fluconazole) content in C. neofor- bles 2, 3, and 5). Tentative values of 16 �g/ml for fluconazole mans (29, 45). Azole cross-resistance among cryptococcal strains (encompassing 97.8% of the isolates), 2 �g/ml for itraconazole is rare; the lack of cross-resistance between itraconazole and flu- (encompassing 99.5% of the isolates), and 0.5 �g/ml for voricona- conazole has been attributed to the dual itraconazole target in C. zole (encompassing 100% of the isolates) can be suggested for neoformans, which appears to block both the lanosterol 14-�-de- CLSI YNB MICs (data not shown). Additional data from other methylase and the NADH-dependent 3-ketosteroid reductase laboratories using either CLSI YNB or CLSI RPMI medium (at (37). This was recently confirmed in a C. neoformans isolate, where least three laboratory subsets), or for subsets comprising at least cross-resistance was evident between fluconazole (MICs � 64 �g/ 100 isolates, should corroborate/extend our tentative values. ml) and voriconazole (MICs � 2 �g/ml) but not with itracona- ECVs encompassing both �97.5% and �99% of the modeled zole. In contrast, increased itraconazole susceptibility (lower population were mostly the same or one dilution higher (Table 6); MIC) and no change in the posaconazole MIC were observed (49). the same applied when all isolates of each species were pooled A missense mutation (Y145F substitution) in the ERG11 gene together, e.g., all VGIIs (data not shown). Although the proposed (P450Dm, ERG11) led to fluconazole-voriconazole cross-resistance. ECVs in the present study are not predictors of clinical outcome, In addition, a high level of heteroresistance to fluconazole, more these critical drug concentrations may aid in identifying those commonly observed in C. gattii than in C. neoformans isolates, was cryptococcal strains with decreased susceptibility to the triazole observed in the same isolate (49, 53). Some of these findings ex- being evaluated. plain the different rates of non-WT isolates found in the present The frequency of MICs above the ECV (non-WT) varied ac- study. cording to the distribution analyzed (Table 6). The rate of flu- Based on small numbers of patients/isolates and the use of conazole non-WT MICs was higher (1.7 to 9.5%) than those of the different methodologies, correlations between a variety of flu- other three triazoles (0 to 5.7%). Between the two species, the conazole MICs and clinical outcomes (1, 3, 34) have been re- fluconazole and itraconazole rates of non-WT MICs were usually ported: more rapid CSF sterilization and infection eradication has lower for all C. neoformans isolates (1.1 to 2.9%) than for all C. been associated with MICs of 4 to 8 �g/ml and clinical failure or gatti isolates (0 to 9.5%) and almost the same for voriconazole. In recurrence with MICs of �16 �g/ml. In addition, either a lack of contrast, the rate was lower for posaconazole (0.3 to 5.7% versus correlation (15) or a relationship that was dependent on the com- 0.5 to 1.5%). However, C. gattii MIC distributions were small bination of high MICs with other factors have been documented (Tables 2 to 5). Although cryptococcal infections caused by both (55). A more useful MIC cutoff that would predict azole failure in species are clinically similar, there is more of a delayed treatment treatment of cryptococcal infections has also been hampered by response and other complications in infections caused by C. gattii other factors that influence outcome to therapy in addition to than those due to C. neoformans (38). In vitro resistance to flu- microbiological clearance (17, 37). Conversely, our ECVs could conazole (3.1% to 46%) and itraconazole (0.5% to 17.4%) has aid in the separation of WT strains from those with reduced azole been reported for C. neoformans isolates; although the itracona- susceptibility or non-WT isolates. However, more information is zole cutoff was �1 �g/ml, it was variable for fluconazole (�16 needed regarding azole resistance in C. neoformans and C. gattii, �g/ml and �64 �g/ml) (5, 16, 39). Based on arbitrary voricona- especially for posaconazole and voriconazole, as well as the rela- zole cutoffs of �2 �g/ml for both Cryptococcus spp., in vitro resis- tionship between non-WT strains and resistance mechanisms. tance to this agent has been low or absent (0% to 0.9%) (23, 24, 43, In conclusion, the ECVs of fluconazole (8 to 32 �g/ml), itra- 50); the same applies to posaconazole (23, 24, 43). Species identi- conazole (0.25 to 1 �g/ml), posaconazole (0.25 to 0.5 �g/ml), and fication and antifungal susceptibility testing emphasize the utility voriconazole (0.12 to 0.25 �g/ml) proposed in this study for the C. of WT cutoffs as a practical tool to detect triazole resistance among neoformans-C. gattii species complex are species- and molecular Cryptococcus isolates. Our results also point out that ECVs should type-specific, but these differences were mostly within one dilu-

80 Wild-type susceptibility endpoint distributions and epidemiological cutoff values

tion. Further investigation should determine the relationship be- tional supplement. CLSI document M27-S3. Clinical and Laboratory tween molecular mechanisms of triazole resistance and our pro- Standards Institute, Wayne, PA. posed non-WT values. Some of the distributions were small 14. Dalhoff A, Ambrose PG, Mouton JW. 2009. A long journey from minimum inhibitory concentration testing to clinically predictive break- (especially for various C. neoformans molecular types), and con- points: deterministic and probabilistic approaches in deriving break- tinuing surveillance should either corroborate or expand the in- points. Infection 37:296–305. formation provided in the present study. In the absence of CBPs, 15. Dannaoui E, et al. 2006. Results obtained with various antifungal suscep- these ECVs may be clinically useful in detecting non-WT isolates tibility testing methods do not predict early outcome in patients with cryptococcosis. Antimicrob. Agents Chemother. 50:2264–2470. that have reduced susceptibility to itraconazole, posaconazole, 16. DeBedout C, et al. 1999. In vitro antifungal susceptibility of clinical and voriconazole or that harbor fluconazole resistance mecha- isolates of Cryptococcus neoformans var. neoformans and C. neoformans nisms. ECVs should be included in the revised version of the CLSI var. gattii. Rev. Iberoam. Micol. 16:36–39. M27-A3 document. 17. Desnos-Ollivier M, et al. 2010. Mixed infections and in vivo evolution in the human fungal pathogen Cryptococcus neoformans. mBio 1:e00091–10. doi:10.1128/mBio.00091–10. ACKNOWLEDGMENTS 18. Espinel-Ingroff A. 2008. Mechanisms of resistance to antifungal agents: We acknowledge all members of the Pacific Northwest Cryptococcus gattii yeasts and filamentous fungi. Rev. Iberoam. Micol. 25:101–106. Public Health Working Group and the members of the Group for 19. Espinel-Ingroff A, et al. 2010. Wild-Type MIC distributions and epide- Enteric, Respiratory and Meningeal Disease Surveillance in South Af- miological cutoff values for the triazoles and six Aspergillus spp. for the rica (GERMS-SA) for their contribution, i.e., submission of isolates. We CLSI broth microdilution method (M38–A2). J. Clin. Microbiol. 48: 3251–3257. also acknowledge the efforts of the clinicians and laboratory technicians 20. Espinel-Ingroff A, et al. 2011. Wild-Type MIC distributions and epide- who provided isolates for this study. We thank all technical personnel in miological cutoff values for caspofungin and Aspergillus spp. for the CLSI the participating laboratories for generating the data used in this study. broth microdilution method (M38-A2 document). Antimicrob. Agents We also thank Naureen Iqbal and Carol Bolden at the CDC and María de Chemother. 55:2855–2859. los Angeles Martínez at Escuela Nacional de Ciencias Biológicas, IPN, 21. Espinel-Ingroff A, et al. 2011. Wild-Type MIC distributions and epide- México, for their technical assistance. miological cutoff values for amphotericin B and Aspergillus spp. for the The findings and conclusions of this article are those of the authors CLSI broth microdilution method (M38-A2 document). Antimicrob. and do not necessarily represent the views of the Centers for Disease Agents Chemother. 55:5150–5154. Control and Prevention (CDC). 22. Espinel-Ingroff A, et al. 2012. Cryptococcus neoformans-Cryptococcus gattii complex: An international study of wild-type susceptibility endpoint distributions and epidemiological cutoff values for amphotericin B and REFERENCES flucytosine. Antimicrob. Agents Chemother. 56:3107–3113. 1. Aller AI, et al. 2000. Correlation of fluconazole MICs with clinical out- 23. Gomez-Lopez A, et al. 2008. In vitro susceptibility of Cryptococcus gattii come in cryptococcal infection. Antimicrob. Agents Chemother. 44: clinical isolates. Clin. Microbiol. Infect. 14:727–730. 7 1544–1548. 24. Govender NP, Patel J, van Wyk M, Chiller TM, Lockhart SR. 2011. 2. Berg J, Clancy CJ, Nguyen MH. 1998. The hidden danger of fluconazole Trends in antifungal drug susceptibility of Cryptococcus neoformans iso- primary prophylaxis for patients with AIDS. Clin. Infect. Dis. 26:186–187. lates obtained through population-based surveillance in South Africa in 3. Bicanic T, Harrison A, Niepieklo A, Dyakopu N, Meintjes G. 2006. 2002–2003 and 2007–2008. Antimicrob. Agents Chemother. 55:2606– Symptomatic relapse of HIV-associated cryptococcal meningitis after ini- 2611. tial fluconazole monotherapy: the role of fluconazole resistance and im- 25. Hagen F, et al. 2010. In vitro antifungal susceptibilities and amplified mune reconstitution. Clin. Infect. Dis. 43:1069–1073. fragment length polymorphism genotyping of a worldwide collection of 4. Bovers M, Hagen F, Boekhout T. 2008. Diversity of the Cryptococcus 350 clinical, veterinary, and environmental Cryptococcus gattii isolates. neoformans-Cryptococcus gattii species-complex. Rev. Iberoam. Micol. 25: Antimicrob. Agents Chemother. 54:5139–5149. S4–S12. 26. Hagen F, et al. 2012. Extensive genetic diversity within the Dutch clinical 5. Brandt ME, et al. 2001. Trends in antifungal drug susceptibility of Cryp- Cryptococcus neoformans population. J. Clin. Microbiol. 50:1918–1926. tococcus neoformans isolates in the United States: 1992 to 1994 and 1996 to 27. Harris JR, et al. 2011. Cryptococcus gattii in the United States: clinical 1998. Antimicrob. Agents Chemother. 45:3065–3069. aspects of infection with an emerging pathogen. Clin. Infect. Dis. 53: 6. Byrnes EJ, et al. 2010. Emergence and pathogenicity of highly virulent 1188–1195. Cryptococcus gattii genotypes in the northwest United States. PLoS Pathog. 28. Illnait-Zaragozi MT, et al. 2008. In vitro activity of the new azole isavu- 6:e1000850. doi:10.1371/journal.ppat.1000850. conazole (BAL4815) compared with six other antifungal agents against 7. Byrnes EJ, et al. 2011. A diverse population of Cryptococcus gattii molec- 162 Cryptococcus neoformans isolates from Cuba. Antimicrob. Agents ular type VGIII in Southern Californian HIV/AIDS patients. PLoS Pathog. Chemother. 52:1580–1582. 7:e1002205. doi:10.1371/journal.ppat.1002205. 29. Joseph-Horne T, Halloman DW, Loeffin R, Kelly SL. 1995. Cross- 8. Castañón-Olivares LR, et al. 2009. Genotyping of Mexican Cryptococcus resistance to polyene and azole drugs in Cryptococcus neoformans. Antimi- neoformans and C. gattii isolates by PCR-fingerprinting. Med. Mycol. 47: crob. Agents Chemother. 39:1526–1529. 713–721. 30. Kahlmeter G, et al. 2003. European harmonization of MIC breakpoints 9. Chandenier JK, et al. 2004. In vitro activity of amphotericin B, flucona- for antimicrobial susceptibility testing of bacteria. J. Antimicrob. Che- zole and voriconazole against 162 Cryptococcus neoformans isolates from mother. 52:143–148. Africa and Cambodia. Eur. J. Clin. Microbiol. Infect. Dis. 23:505–508. 31. Kidd SE, et al. 2004. A rare genotype of Cryptococcus gattii caused the 10. Chong HS, Dagg R, Malik R, Chen S, Carter D. 2010. In vitro suscep- cryptococcosis outbreak on Vancouver Island (British Columbia, Can- tibility of the yeast pathogen Cryptococcus and other azoles varies with ada). Proc. Natl. Acad. Sci. U. S. A. 101:1758–1763. molecular genotype. J. Clin. Microbiol. 48:4115–4120. 32. Lamb DC, Corran A, Baldwin BC, Kwon-Chung J, Kelly SL. 1995. 11. Chowdhary A, et al. 2011. In vitro antifungal susceptibilities and geno- Resistant P45051 AI activity in azole antifungal tolerant Cryptococcus neo- types of 308 clinical and environmental isolates of Cryptococcus neofor- formans from AIDS patients. FEBS Lett. 368:326–330. mans var. grubii and C. gattii serotype B from North West India. J. Med. 33. Lockhart SR, et al. 2012. Epidemiological cutoff values for triazole drugs Microbiol. 60:961–967. in Cryptococcus gattii: correlation of molecular type and in vitro suscepti- 12. Clinical and Laboratory Standards Institute. 2008. Reference method for bility. Diagn. Microbiol. Infect. Dis. 73:144–148. broth dilution antifungal susceptibility testing of yeasts, third edition. 34. Menichetti F, et al. 1996. High-dose fluconazole therapy for cryptococcal CLSI document M27-A3. Clinical and Laboratory Standards Institute, meningitis in patients with AIDS. Clin. Infect. Dis. 22:838–840. Wayne, PA. 35. Ostrosky-Zeichner L, Marr KA, Rex JH, Cohen SH. 2003. Amphotericin 13. Clinical and Laboratory Standards Institute. 2008. Reference method for B: time for a new “gold standard.” Clin. Infect. Dis. 37:415–425. broth dilution antifungal susceptibility testing of yeasts, third informa- 36. Pan W, et al. 2012. Resistance of Asian Cryptococcus neoformans serotype

81 Chapter 7

A is confined to few microsatellite genotypes. PLoS One 7:e32868. doi: coccus neoformans ATP binding cassette (ABC) transporter-encoding 10.1371/journal.pone.0032868. gene, CnAFR1, involved in the resistance to fluconazole. Mol. Microbiol. 37. Perfect JR, Cox JM. 1999. Drug resistance in Cryptococcus neoformans. 47:357–371. Drug Resist. Updat. 2:259–269. 46. Revankar SG, et al. 2004. Cloning and characterization of the lanosterol 38. Perfect JR, et al. 2010. Clinical practice guidelines for the management of 14-�-demethylase (ERG11) gene in Cryptococcus neoformans. Biochem. cryptococcal disease: 2010 update by the Infectious Diseases Society of Biophys. Res. Commun. 324:719–728. America. Clin. Infect. Dis. 50:291–322. 47. Rodriguez-Tudela JL, et al. 2008. Epidemiological cutoffs and cross- 39. Perkins A, Gomez-Lopez A, Mellado E, Rodriguez-Tudela JL, Cuenca- resistance to azole drugs in Aspergillus fumigatus. Antimicrob. Agents Estrella M. 2005. Rates of antifungal resistance among Spanish clinical Chemother. 52:2468–2472. isolates of Cryptococcus neoformans var. neoformans. J. Antimicrob. Che- 48. Sheng C, et al. 2009. Three-dimensional model of lanosterol-�- mother. 56:1144–1147. demethylase from Cryptococcus neoformans: active-site characterization 40. Pfaller MA, et al. 2009. Wild type MIC distribution and epidemiological and insights into azole binding. Antimicrob. Agents Chemother. 53:3487– cutoff values for Aspergillus fumigatus and three triazoles as determined by 3495. the Clinical and Laboratory Standards Institute broth microdilution 49. Sionov E, et al. 2012. Identification of a Cryptococcus neoformans cyto- methods. J. Clin. Microbiol. 47:3142–3146. chrome P450 lanosterol 14-�-demethylase (Erg11) residue critical for dif- 41. Pfaller MA, et al. 2009. Results from the ARTEMIS DISK Global Anti- ferential susceptibility between fluconazole/voriconazole and itracona- fungal Surveillance Study, 1997 to 2007: 10.5-year analysis of susceptibil- zole/posaconazole. Antimicrob. Agents Chemother. 56:1162–1169. ities of noncandidal yeast species to fluconazole and voriconazole deter- 50. Thompson GR, et al. 2009. Antifungal susceptibilities among different mined by the CLSI standardized disk diffusion testing. J. Clin. Microbiol. serotypes of Cryptococcus gattii and Cryptococcus neoformans. Antimicrob. 47:117–123. Agents Chemother. 53:309–311. 42. Pfaller MA, Andes D, Diekema DJ, Espinel-Ingroff A, Sheehan D. 2010. 51. Trilles L, Meyer W, Wanke B, Guarro J, Lazera M. 2012. Correlation of Wild-type MIC distributions, epidemiological cutoff values and species- antifungal susceptibility and molecular type within Cryptococcus neofor- specific clinical breakpoints for fluconazole and Candida: time to harmo- mans/C. gattii species complex. Med. Mycol. 50:328–332. nization of CLSI and EUCAST broth microdilution methods. Drug Resist. 52. Turnidge J, Kahmeter G, Kronvall G. 2006. Statistical characterization of Updat. 14:180–195. bacterial wild-type MIC value distributions and the determination of ep- 43. Pfaller MA, Castanheira M, Diekema DJ, Messer SA, Jones RN. 2011. idemiological cut-off values. Clin. Microbiol. Infect. 12:418–425. Wild-type MIC distributions and epidemiological cutoff values for flu- 53. Varma A, Kwon-Chung KJ. 2010. Heteroresistance of Cryptococcus gattii conazole, posaconazole and voriconazole when testing Cryptococcus neo- to fluconazole. Antimicrob. Agents Chemother. 54:2303–2311. formans by CLSI broth microdilution methods. Diagn. Microbiol. Infect. 54. Venkateswarlu K, Taylor M, Manning NJ, Rinaldi MG, Kelly SL. 1997. Dis. 71:252–259. Fluconazole tolerance in clinical isolates of Cryptococcus neoformans. An- 44. Pfaller MA, et al. 2012. Wild-type MIC distributions and epidemiological timicrob. Agents Chemother. 41:748–751. cutoff values for amphotericin B, flucytosine, and itraconazole and Can- 55. Witt MD, et al. 1996. Identification of patients with acute AIDS- dida spp. as determined by CLSI broth microdilution. J. Clin. Microbiol. associated cryptococcal meningitis who can be effectively treated with 50:2040–2046. fluconazole: the role of antifungal susceptibility testing. Clin. Infect. Dis. 45. Posteraro B, et al. 2003. Identification and characterization of a Crypto- 22:322–328.

82 Chapter 8 Environmental isolation and characterisation of Cryptococcus species from living trees in Havana city, Cuba

Published in Mycoses 2012;55:e138-44. Chapter 8

Environmental isolation and characterisation of Cryptococcus species from living trees in Havana city, Cuba

M. T. Illnait-Zaragozı´,1 G. F. Martı´nez-Machı´n,1 C. M. Ferna´ ndez-Andreu,1 M. R. Perurena-Lancha,1 B. Theelen,2 T. Boekhout,2 J. F. Meis3,4 and C. H. Klaassen4 1Instituto Pedro Kourı´, Havana, Cuba, 2CBS-KNAW Fungal Biodiversity Centre, Utrecht, The Netherlands, 3Department of Medical Microbiology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands and 4Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, Nijmegen, The Netherlands

Summary Cryptococcus isolates from Cuban patients were identified as C. neoformans var. grubii. Although this species has since long been associated with bird droppings, a recent genotyping study provided strong evidence for additional origins of exposure. We sampled different species of trees in Havana, Cuba to identify other potential sources of exposure to this fungus. A total of 662 samples were collected from 331 trees and cacti from Havana, Cuba. Initial selection of the isolates was carried out by conventional techniques. Isolates were further characterised using a combination of AFLP analysis and DNA sequence analysis. Identification by conventional methods yielded 121 C. neoformans and 61 C. gattii isolates. Molecular analyses showed that none of these isolates was C. gattii and only one isolate proved to be C. neoformans var. grubii. A total of 27 different other species were identified. The most prevalent species was C. heveanensis (33%). Sixty-five unidentifiable isolates segregated into ten potentially novel species. Conventional cultivation methods have a low specificity for C. neoformans complex and molecular analyses need to be applied to confirm identification of isolates from environmental sources. Environmental niches responsible for most of human cryptococcal infections in Cuba remain to be identified.

Key words: Cryptococcus, identiﬁcation, environment, AFLP.

The genus Cryptococcus encompasses over hundred Introduction species,4 but the C. neoformans ⁄ C. gattii species complex Since the mid 1980s, cryptococcosis emerged as a life- is the most common associated with human and animal threatening infection in patients with AIDS, where diseases.4,5 At present, mycosis due to this fungus ranks dissemination is commonly fatal even when treated as one of the three most common life-threatening with the most effective antifungal drugs.1,2 The limited opportunistic infections in persons with impaired T-cell empirical evidence for direct person to person transmis- function, particularly in patients with AIDS.5 sion of this infection have led to the belief that infections This saprophytic yeast was ﬁrst cultivated by Sanfelice in humans are predominantly caused by inhalation of from fermenting peach juice in 1894 and during the basidiospores from environmental sources.3 following years, the fungus was isolated from lesions or secretions of humans and animals. More than 50 years later, Emmons reported its isolation from the environ- Correspondence: C. H. Klaassen, Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, Nijmegen, The Nether- ment and until now it has been recognised in many 1,5–7 lands. areas of the world. On the basis of differences in Tel.: +31243657514. Fax: +31243657516. ecology, physiology, capsular polysaccharide, clinical E-mail: [email protected] manifestation and genetics, two species with two varieties and ﬁve serotypes have been recognised. C. neofor- Submitted for publication 28 November 2011 Revised 27 December 2011 mans var. grubii (serotype A), that are responsible of Accepted for publication 31 December 2011 more than 90% of all human cryptococcal infections;

84 Environmental isolation and characterisation of Cryptococcus

C. neoformans var. neoformans (serotype D) causes fewer agar (SDA) and yeast growth was confirmed by micro- cases of disease and is being considered less pathogenic; scopic examination. Initial identification was based on hybrid serotype AD strains have been isolated from the colony- and cellular morphology, phenoloxidase activity environment and from patients in North America and (characteristic brown appearance on CAA), urease Europe and C. gattii, (serotypes B and C) which tend to production, growth at 37 �C, and growth on canava- infect immunocompetent individuals.1,5–7 nine-glycine-bromothymol blue medium (CGB).14 Previous epidemiological studies on clinical and environmental isolates from pigeon droppings, carried DNA isolation out in different regions along the Cuban island, showed that all studied isolates represented C. neoformans var. A full loop of freshly grown cells was resuspended in a grubii and that only few microsatellite genotypes from vial containing lysis buffer (350 ll) and ceramic beads environmental isolates could be linked to genotypes (Roche Diagnostics, Almere, The Netherlands) and from human isolates, suggesting that certain clinical subjected to mechanical lysis in a Magnalyzer instru- isolates may originate from additional ecological ment (Roche Diagnostics) for 30 s at 6500 rpm. niche(s).8 At present, it has become clear that Crypto- Following centrifugation for 2 min at 12 000 · g, coccus isolates, in addition to bird excreta,9 are present 200 ll of supernatant was combined with proteinase on trees and cacti.1,7,10,11 Here, we report the results of K (30 ll) (Roche Diagnostics) and incubated 10 min at a study on the environmental occurrence of this yeast in 65 �C and for 10 min at 95 �C. Finally, the DNA was relation to different species of trees and cacti. purified using a MagNA Pure LC instrument (Roche Diagnostics) in combination with MagNA Pure DNA Isolation Kit III, according to the manufacturer�s Materials and methods instructions. DNA yield and purity were estimated by UV absorbance measurements using standard proce- Tree sampling dures. Plants were sampled along the northern coast of Havana, Cuba between April and June, which corresponds to the AFLP analysis flowering season. Urban areas with large numbers and variety of plants species were chosen. Three hundred and Approximately 50 ng of genomic DNA was subjected to twenty-nine adult plants were sampled: 65 Indian a combined restriction–ligation procedure containing 8 almond trees (Terminalia cattapa), 65 flamboyant trees 5 pmol of EcoRI adapter, 50 pmol MseI adapter, 2 U of (Delonix regia), 65 Ficus spp. (mainly Indian laurel trees), EcoRI (New England Biolabs, Beverly, MA, USA), 2 U of 60 cacti of different species, 60 casuarins (Casuarina MseI (New England Biolabs) and 1 U of T4 DNA ligase equisetifolia), five blue mahoe trees (Talipariti elatum), (Promega, Leiden, The Netherlands) in a total volume of three ceiba (Ceiba pentandra), two seagrape (Coccoloba 20 llof1· reaction buffer for 1 h at 20 �C. Next, the uvifera), one copey (Clusia resea), one yellow Oleander mixture was diluted five times with 10 mmol l)1 (Thevetia peruviana), one yellow flamboyant (Peltophorum Tris ⁄ HCl pH 8.3 buffer. Adapters were made by mixing adratum), one ballnut (Calophyllum inophyllum), one equimolar amounts of complementary oligonucleotides caribbean pine (Pinus caribaea) and one roble (Tabebuia (5¢-CTCGTAGACTGCGTACC-3¢ and 5¢-AATTGGTACG- pentaphylle). From each tree and cactus, two samples CAGTC-3¢ for EcoRI; 5¢-GACGATGAGTCCTGAC-3¢ and were obtained. Sampling was done by rubbing the inner 5¢-TAGTCAGGACTCAT-3¢ for MseI) and heating to surface of hollows or fissures with a sterile cotton-tipped 95 �C, subsequently cooling slowly to ambient temper- swab wetted in sterile saline.12 After sampling, each ature. One microlitre of the diluted restriction–ligation swab was placed in a sterile glass tube with 2 ml of saline mixture was used for amplification in a volume of 25 ll and transported to the mycology laboratory where they under the following conditions: 1 lmol l)1 EcoRI primer were processed within 4 h. Swabs were directly inocu- with two selective residues (5¢-Flu-GACTGCGTACCA- lated onto caffeic acid agar plates (CAA),13 incubated at ATTCAC-3¢), 1 lmol l)1 MseI primer with one selective 28 �C and inspected daily for 7–10 days. residues (5¢-GATGAGTCCTGACTAAG-3¢), 0.2 mmol l)1 each dNTP and 1 U of Taq DNA polymerase (Roche Diagnostics) in 1· reaction buffer containing Conventional identification )1 1.5 mmol l MgCl2. After an initial denaturation step All pigmented colonies suspicious of representing Cryp- for 4 min at 94 �C in the first 20 cycles, a touch-down tococcus were isolated separately on Sabouraud dextrose procedure was applied: 15 s denaturation at 94 �C; 15 s

85 Chapter 8

annealing at 66 �C with the temperature for each Data analysis successive cycle lowered by 0.5 �C and 1 min of extension at 72 �C. Cycling was then continued for 30 AFLP data were analysed using BioNumerics version cycles with an annealing temperature of 56 �C. After 5.1 (Applied Maths, Sint-Martens-Latem, Belgium) completion of the cycles, incubation at 72 �C for 10 min using the Pearson correlation coefficient and UPGMA was performed before the reactions were cooled to room clustering for DNA fragments in the range of 60 bp and temperature. The amplicons were then combined with to 500 bp. the ET550-R size standard (GE Healthcare, Diegem, Belgium) and analysed on a MegaBACE 500 automated Results DNA analysis platform equipped with a 48 capillary array (GE Healthcare), according to the manufacturer�s According to the conventional methods, 34% of the instructions. sampled trees (111 ⁄ 331) yielded one or more smooth and variably mucoid yeast-like colonies exhibiting some chocolate brown shades after 2–4 days of incubation on DNA sequence analysis CAA medium. Multiple colonies on one plate were Part of the ribosomal gene cassette containing the analysed separately. A total of 198 isolates suspected to ITS1 ⁄ 5.8S ⁄ ITS2 and D1–D2 regions was amplified by represent Cryptococcus spp. were analysed. Initial iden- PCR. The 50 ll amplification reaction contained 1 ll of tification results of all samples are shown in Table 1. DNA (approx. 20–100 ng), 0.5 lmol l)1 of amplifica- More than one-third of all CAA positive isolates (73 out tion primers V9 (5¢-TGCGTTGATTACGTCCCTGC-3¢ and of 198) were also CGB positive. All Cryptococcus-like RLR3R 5¢-GGTCCGTGTTTCAAGAC-3¢), 0.2 mmol l)1 isolates from trees were subjected to AFLP analysis. The )1 dNTP�s, 2 mmol l MgCl2 and 1 U of FastStart Taq resulting dendrogram showed 28 clusters of AFLP DNA polymerase (Roche Diagnostics) in 1· reaction fingerprints (Fig. 1). From each ten isolates in a cluster, buffer. After an initial denaturation for 10 min at 94 �C, one random isolate was identified by sequencing the 35 cycles of 30 s 94 �C, 30 s 52 �C and 2 min 72 �C ITS1 ⁄ 5.8S ⁄ ITS2 and D1–D2 regions. Molecular identi- were applied. After an additional incubation for 10 min fication showed that the most prevalent species was at 72 �C, the reactions were cooled to room tempera- C. heveanensis (36%). Eighteen isolates had sequences ture. Amplified products were purified by SPRI chemistry (AMPure; Beckman Coulter, Mijdrecht, The Table 1 Number of isolates obtained from tree samples inoculated Netherlands) and subjected to bidirectional sequence on CAA and distribution according to their behaviour on CGB analysis using primers 5¢-ITS1 (TCCGTAGGTGAACCT- medium. GCGG-3¢) and ITS4 (5¢-TCCTCCGCTTATTGATATGC-3¢) Plant species (nr. positive for the ITS1 ⁄ 5.8S ⁄ ITS2 region and primers NL1 (5¢- samples ⁄ total nr. of samples) CGB (+) CGB ()) Total GCATATCAATAAGCGGAGGAAAAG-3¢) and RLR3R for the D1–D2 regions respectively. Sequence reactions Almond (41 ⁄ 65) 31 48 79 were performed using the DYEnamic ET Dye Terminator Casuarins (35 ⁄ 60) 42 11 53 Flamboyants (25 ⁄ 65) 36 9 45 Kit (GE Healthcare) according to the recommendations Cacti (10 ⁄ 60) 14 5 19 of the manufacturer. Sequence reaction products were Others* (0 ⁄ 81) 0 0 0 purified using SPRI chemistry (CleanSeq; Beckman Total 123 73 196 Coulter) and analysed using a MegaBACE 500 platform as recommended. Obtained DNA sequences were com- CAA, caffeic acid agar plates; CGB, canavanine-glycine- pared with the public DNA databases using BLAST bromothymol blue medium. analysis15 and with the in-house yeast database of the *Ficus spp. (n = 65), blue mahoe trees (n = 5), ceiba (n = 3), sea CBS-KNAW Fungal Biodiversity Centre. grape (n = 2), copey, yellow oleander, yellow flamboyant, ballnut, caribbean pine and roble (one each).

Figure 1 Dendrogram of AFLP fingerprints restricted to DNA fragments in the range of 60–500 bp based on UPGMA cluster analysis and the Pearson correlation coefficient. The scale bar represents the percentage of similarity. Coloured squares under A indicate the species of tree from which the isolates were obtained (yellow: casuarina; red: almond; blue: flamboyant). Coloured squares under B indicate the growth on CGB medium (cyan: negative; purple: positive).

86 Environmental isolation and characterisation of Cryptococcus

87 Chapter 8

Table 2 Identiﬁcation results of environmental Cryptococcus iso- predominantly associated with one species of tree. lates (n = 198) collected from Cuban trees and cacti based on Within multiple species, both CGB positive and negative AFLP analysis and ITS - D1 D2 sequences. Cluster IDs correspond ⁄ isolates were found, indicating CGB positivity to be a to those in Fig. 1. poor parameter for fungal identiﬁcation of environmen- # Isolates # AFLP clusters (ID) tal Cryptococcus isolates.

Cryptococcus heveanensis 71 1 (22) 1 Unknown sequences 65 10 (1,2,3,4,10, Discussion 14,15,17,24,28) Cryptococcus flavescens 18 4 (7,8,12,13) A recent study has shown that all clinical cryptococcal 2 Unidentified sequence 17 3 (5,6,9) isolates from Cuban patients represented C. neoformans Cryptococcus 10 2 (26,27) var. grubii. Molecular characterisation by genotyping albidosimilis ⁄ liquefaciens Kwoniella mangroviensis 6 1 (16) suggests that isolates obtained from bird droppings play Bullera dendrophila 2 1 (18) only a minor role in the exposure to this pathogen and Bullera pseudoalba 2 1 (23) that many clinical isolates may originate from addi- Cryptococcus flavus 2 1 (11) tional ecological niche(s).8 In our search for cryptococ- Sterigmatomyces elviae 2 1 (20) Candida sp. 1 1 (21) cal isolates from the environment, we used two Cryptococcus albidus 1 1 (25) screening media (CAA and CGB) to facilitate presump- Cryptococcus neoformans 1 1 (19) tive recognition of either C. neoformans or C. gattii. var. grubii Pigmentation on media with phenolic substrates is 1Sequence not matching to any Genbank entry. widely used for isolation and identification of these 2 yeasts due to its high sensitivity when compared with Sequence matching to a Genbank entry, but with no organism 16–20 name provided. other mycological media. Our primary isolations were initially selected by the presence of pigmented matching C. flavescens, but these segregated into four colonies on CAA. This medium generally performs well different AFLP clusters, suggesting that these isolates with C. neoformans ⁄ C. gattii species complex isolated represent four different closely related yeast species. Ten from clinical samples, but we19 and others21–24 expe- isolates were identified as C. liquefaciens or C. albidosim- rienced that isolates from the environment and some ilis. The distinction between these two species is difficult clinical isolates may display variability in pigmentation to make based on the ITS and D1 ⁄ D2 species, but these intensity. In addition, it has been shown that UV isolates clearly segregated into two AFLP clusters with radiation may induce mutations affecting melanogene- four and six isolates respectively. Seventeen isolates sis.25 This might also occur in the environment due to (three AFLP clusters) had sequences matching Genbank sunlight exposure. Therefore, we selected colonies that entries, but corresponded to yet unnamed species. exhibited any degree of pigmentation (even beige or tiny Additional identified species represented Kwoniella mang- brown) for further identification by conventional and roviensis, Bullera dendrophila, B. pseudoalba, C. flavus, molecular methods. Sterigmatomyces elviae, C. albidus and one Candida sp. The CGB medium was proposed in 1982 by Kwon- Approximately one-third of all isolates (n = 65) distrib- Chung et al. [14] to distinguish C. neoformans from uted over ten AFLP clusters could not be identified based C. gattii. In the present study, although 33% of primar- on ITS and D1–D2 sequences and may represent novel ily selected colonies on CAA yielded a positive reaction fungal taxa. Only one strain isolated from an almond on CGB medium, molecular analyses showed that none tree was identified as C. neoformans var. grubii. The of these isolates was actually C. gattii. Due to the genotype of this isolate was determined using microsat- growing number of environmental studies and the ellite analysis and corresponded to those of isolates development of more reliable identification methods, previously recovered from pigeon droppings (results not many new Cryptococcus species have been described in shown). An overview of molecular identification results recent years.4 However, C. neoformans ⁄ C. gattii species is shown in Table 2. In Cuba, Cryptococcus species seem complex remains the major responsible agent causing to be more associated with almond trees and this is the human and animal cryptococcosis. Typical colour first time that they are found in relation with casuarina change on CGB medium is widely used in mycology and flamboyant trees. Although many Indian laurel laboratories for the differentiation of both species of the trees were sampled, C. neoformans species complex-like complex,26,27 suggesting that it can be used as a organisms were not found on this tree. Most non- presumptive identification test. However, there are C. neoformans species identified in this study were several other microorganisms beside C. gattii, which

88 Environmental isolation and characterisation of Cryptococcus

show a positive reaction on CGB medium24,28–31 and source for melanin substrates from the environment.37 other species different from the C. neoformans ⁄ C. gattii We propose the hypothesis that other non-neoformans species complex are rarely involved in human infection. species also could use similar ways as survival mech- This is in marked contrast with the high diversity of anism in the environment, which must be verified. species found in the environment, which are more likely Due to the increase in yeast infections in different to increase the chance of misidentification when using human diseases, renewed attention is focused on the conventional techniques such as pigmentation on CAA question of the epidemiology and origin of these fungi. and growth on CGB media. These results highlight the In this context, the question of the importance of the need for molecular confirmation for reliable identifica- different environmental sources as reservoir of Crypto- tion of cryptococcal isolates from the environment. coccus species less virulent than C. neoformans complex AFLP has been reported before as a powerful method has been neglected so far.4 Analysing the structure of for inter- and ⁄ or intraspecific discrimination between the environmental population of C. neoformans and closely related fungal species from various genera of C. gattii can significantly increase our knowledge of its filamentous fungi such as Aspergillus,32 Scedospori- ecology, evolution and epidemiology. However, the low um ⁄ Pseudallescheria33 and Apophysomyces34 but also number of reports of the prevalence of this yeast in the from yeasts such as Candida.35 Our results show that environment does not allow a generally valid epidemi- AFLP is a highly efficient and powerful method as a first ological evaluation, and more epidemiological studies line screening tool to distinguish between different should be performed to determine the exact distribution fungal species in large collections of isolates from of C. neoformans and C. gattii in Cuba. environmental sources such as Cryptococcus spp. It has been suggested that the primary habitat of References C. neoformans could be vegetations and decaying wood, 1 Kidd SE, Chow Y, Mak S et al. Characterization of environ- not being associated with a particular tree, but rather mental sources of the human and animal pathogen Crypto- with a specialised niche resulting from the natural coccus gattii in British Columbia, Canada, and the Pacific 11,27,28 biodegradation of wood. Because of previous Northwest of the United States. Appl Environ Microbiol 2007; reports of isolation of Cryptococcus from almond, laurel 73: 1433–43. trees and cactus,7,10,36 these species were selected while 2 Olszewski MA, Hang Y, Huffnagle GB. Mechanisms of cryptococcal virulence and persistence. Future Microbiol 2010; 5: the criterion for inclusion of the other tree species was 1269–88. their narrow geographical relation with the first ones. 3 Hiremath SS, Chowdhary A, Kowshik T, Randhawa HS, Sun 8 Our results are in agreement with those reported by Tay S, Xu J. Long-distance dispersal and recombination in envi- et al. [29] who found non-neoformans ⁄ gattii cryptococ- ronmental populations of Cryptococcus neoformans var. grubii cal isolates during an environmental sampling study, from India. Microbiology 2008; 154: 1513–24. 4 Fonseca A, Boekhout T, Fell JW. Cryptococcus Vuillemin while looking for C. gattii in Klang Valley, Malaysia. (1901). In: Kurtzman CP, Fell JW, Boekhout T (eds), The Interestingly, a total of 29 (76.3%) C. laurentii isolates Yeasts: A Taxonomic Study. Amsterdam: Elsevier Science & from their collection produced blue colours on CGB Technology, 2011: 1665–741. agar. 5 Byrnes EJ 3rd, Marr KA. The outbreak of Cryptococcus gattii in Although many authors prefer gathering debris from Western North America: epidemiology and clinical issues. Curr Infect Dis Rep 2011; 13: 256–61. decayed wood in tree trunk hollows, we selected 6 Litvintseva AP, Kestenbaum L, Vilgalys R, Mitchell TG. swabbing as sampling method, which lately has been Comparative analysis of environmental and clinical popula- regarded as more useful compared with conventional tions of Cryptococcus neoformans. J Clin Microbiol 2005; 43: approaches. This method not only consistently yielded 556–64. more frequent isolations of fungi, less false-negative 7 Chowdhary A, Randhawa HS, Prakash A, Meis JF. Environ- mental prevalence of Cryptococcus neoformans and C. gattii in results and much higher mean colony counts, but it was India: an update. Crit Rev Microbiol 2012; 38: 1–16. also a more simple, less time-consuming and less 8 Illnait-Zaragozı´ MT, Martıńez-Machıń GF, Fernańdez-Andreu 7,12 hazardous technique. For the C. neoformans ⁄ C. gattii CM, Boekhout T, Meis JF, Klaassen CH. Microsatellite typing of species complex, the scavenging capacity of chemical clinical and environmental Cryptococcus neoformans var grubii compounds made by other microbes suggests that this isolates from Cuba shows multiple genetic lineages. PLoS ONE 2010; 5: e9124. doi:10.1371. organism gains the benefit of melanisation by having 9 Kielstein P, Hotzel H, Schmalreck A, Khaschabi D, Glawisch- the machinery for melanin synthesis and remodelling nig W. Occurrence of Cryptococcus spp. in excreta of pigeons without having to invest in biosynthetic pathways. This and pet birds. Mycoses 2000; 43: 7–15. fact establishes a precedence for a new substrate for 10 Randhawa HS, Kowshik T, Chowdhary A et al. The expanding melanisation in this yeast that could provide a new host tree species spectrum of Cryptococcus gattii and Crypto-

89 Chapter 8

coccus neoformans and their isolations from surrounding soil in 25 Kwon-Chung KJ, Polacheck I, Popkin TJ. Melanin-lacking India. Med Mycol 2008; 46: 823–33. mutants of Cryptococcus neoformans and their virulence for 11 Hagen F, Boekhout T. The search for the natural habitat of mice. J Bacteriol 1982; 150: 1414–21. Cryptococcus gattii. Mycopathologia 2010; 70: 209–11. 26 Pedroso RS, Ferreira JC, Candido RC. The isolation and 12 Randhawa HS, Kowshik T, Khan ZU. Efficacy of swabbing characterization of virulence factors of Cryptococcus spp. from versus a conventional technique for isolation of Cryptococcus saprophytic sources in the city of Ribeirao Preto, Sao Paulo, neoformans from decayed wood in tree trunk hollows. Med Brazil. Microbiol Res 2009; 164: 221–7. Mycol 2005; 43: 67–71. 27 Lazera MS, Salmito MA, Londero AT, Trilles L, Nishikawa 13 Vidotto V, Aoki S, Ponton J, Quindos G, Koga-Ito CY, Pugliese MM, Wanke B. Possible primary ecological niche of Crypto- A. A new caffeic acid minimal synthetic medium for the rapid coccus neoformans. Med Mycol 2000; 38: 379–83. identification of Cryptococcus neoformans isolates. Rev Iberoam 28 Lazera MS, Pires FD, Camillo-Coura L et al. Natural habitat of Micol 2004; 21: 87–89. Cryptococcus neoformans var. neoformans in decaying wood 14 Kwon-Chung KJ, Polacheck I, Bennett JE. Improved diagnostic forming hollows in living trees. J Med Vet Mycol 1996; 34: medium for separation of Cryptococcus neoformans var. neo- 127–31. formans (serotypes A and D) and Cryptococcus neoformans var. 29 Tay ST, Na SL, Tajuddin TH. Natural occurrence and growth gattii (serotypes B and C). J Clin Microbiol 1982; 15: 535–7. reaction on canavanine-glycine-bromothymol blue agar of 15 Altschul SF, Madden TL, Scha¨ffer AA et al. Gapped BLAST and non-neoformans Cryptococcus spp. in Malaysia. Mycoses 2008; PSI-BLAST: a new generation of protein database search 51: 515–9. programs. Nucleic Acids Res 1997; 25: 3389–402. 30 Polacheck I, Kwon-Chung KJ. Canavanine resistance in 16 Nandhakumar B, Kumar CP, Prabu D, Menon T. Mustard seed Cryptococcus neoformans. Antimicrob Agents Chemother 1986; agar, a new medium for differentiation of Cryptococcus neo- 29: 468–73. formans. J Clin Microbiol 2006; 44: 674. 31 Klein KR, Hall L, Deml SM, Rysavy JM, Wohlfiel SL, Wenge- 17 Tendolkar U, Tainwala S, Jog S, Mathur M. Use of a new nack NL. Identification of Cryptococcus gattii by use of L-ca- medium – tobacco agar, for pigment production of Crypto- navanine glycine bromothymol blue medium and DNA coccus neoformans. Indian J Med Microbiol 2003; 21: 277–9. sequencing. J Clin Microbiol 2009; 47: 3669–72. 18 Stepanovic´ S, Vukovic´ D, Radonjic´ I, Dimitrijevic´ V, Sˇvabiæ- 32 Klaassen CH, Osherov N. Aspergillus strain typing in the ge- Vlahovic´ M. Ground red hot pepper agar in the isolation and nomics era. Stud Mycol 2007; 59: 47–51. presumptive identification of Cryptococcus neoformans. Mycoses 33 Lackner M, Rezusta A, Villuendas MC, Palacian MP, Meis JF, 2002; 45: 384–8. Klaassen CH. Infection and colonization due to Scedosporium 19 Hernańdez IC, Machıń GM, Andreu CM, Illnait MT. Pigmen- in Northern Spain. An in vitro antifungal susceptibility and tacioń de cepas de Cryptococcus neoformans sobre agar semilla molecular epidemiology study of 60 isolates. Mycoses 2011; de girasol. Rev Cubana Med Trop 2003; 55: 119–20. 54: 12–21. 20 Pedroso RS, da Costa KR, Ferreira JC, Candido RC. Evaluation 34 Chakrabarti A, Shivaprakash MR, Curfs-Breuker I, Baghela A, of melanin production by Cryptococcus species in four different Klaassen CH, Meis JF. Apophysomyces elegans: epidemiology, culture media. Rev Soc Bras Med Trop 2007; 40: 566–8. amplified fragment length polymorphism typing and in vitro 21 Golubev WI, Staib F. Green and brown color effects in trem- antifungal susceptibility pattern. J Clin Microbiol 2010; 48: ellaceous yeast fungi on Staib agar. Mycoses 2000; 43: 1–5. 4580–5. 22 Wang HS, Zeimis RT, Roberts GD. Evaluation of a caffeic acid- 35 Borst A, Theelen B, Reinders E, Boekhout T, Fluit AC, Save- ferric citrate test for rapid identification of Cryptococcus neo- lkoul PH. Use of amplified fragment length polymorphism formans. J Clin Microbiol 1977; 6: 445–9. analysis to identify medically important Candida spp., includ- 23 Ikeda R, Sugita T, Jacobson ES, Shinoda T. Laccase and mel- ing C. dubliniensis. J Clin Microbiol 2003; 41: 1357–62. anization in clinically important Cryptococcus species other 36 Loperana Y, Ren P, Li X et al. Genotipic characterization of than Cryptococcus neoformans. J Clin Microbiol 2002; 40: environmental isolates of Cryptococcus gattii from Puerto Rico. 1214–8. Mycopathologia 2010; 170: 279–85. 24 McTaggart L, Richardson SE, Seah C, Hoang L, Fothergill A, 37 Frases S, Salazar A, Dadachova E, Casadevall A. Cryptococcus Zhang SX. Rapid identification of Cryptococcus neoformans var. neoformans can utilize the bacterial melanin precursor ho- grubii, C. neoformans var. neoformans, and C. gattii using rapid mogentisic acid for fungal melanogenesis. Appl Environ biochemical tests, differential media, and DNA sequencing. Microbiol 2007; 73: 615–21. J Clin Microbiol 2011; 49: 2522–7.

90 Chapter 9 Reactivation of a Cryptococcus gattii infection in a cheetah (Acinonyx jubatus) held in the National Zoo, Havana, Cuba

Published in Mycoses 2011;54:e889-92. Chapter 9

Reactivation of a Cryptococcus gattii infection in a cheetah (Acinonyx jubatus) held in the National Zoo, Havana, Cuba

M. T. Illnait-Zaragozı´,1 F. Hagen,2 C. M. Ferna´ ndez-Andreu,1 G. F. Martı´nez-Machı´n,1 J. L. Polo-Leal,3 T. Boekhout,2 C. H. Klaassen4 and J. F. Meis4 1Bacteriology and Mycology Department, Tropical Medicine Institute �Pedro Kourı´�, Havana, Cuba, 2Yeast and Basidiomycete Research, Fungal Biodiversity Centre (CBS-KNAW), Utrecht, The Netherlands, 3Department of Microbiology, National Zoo, Havana, Cuba and 4Department of Medical Microbiology and Infectious Diseases, Canisius-Wilhelmina Hospital, Nijmegen, The Netherlands

clear that C. neoformans and C. gattii could be divided Introduction into distinct genotypes named AFLP1 ⁄ VN1, AFLP2 ⁄ Cryptococcosis is a major cause of mycosis with a VNIV and AFLP3 ⁄ VNIII for C. neoformans and genotype worldwide distribution found in the human and warm- AFLP4 ⁄ VGI, AFLP5 ⁄ VGIII,AFLP 6 ⁄ VGII, AFLP7 ⁄ VGIV blooded animal population. Its causative agents are and AFLP10 ⁄ VGIV for C. gattii (Franzot SP et al., J Clin Cryptococcus neoformans and C. gattii, which are the Microbiol 1999; 37: 838–40; Boekhout T et al., Micro- most important pathogens among the basidiomycetous biology 2001; 147: 891–907; Hagen F et al., Antimicrob fungi (Caban˜es FJ, Rev Iberoam Micol 2008; 25: S1–S3). Agents Chemother 2010; 54: 5139–45). The genus includes over 80 species; however, only Cryptococcus neoformans has a global distribution and C. neoformans and C. gattii are considered pathogenic infects predominantly immunocompromised animal and [Fonseca A et al., Cryptococcus Vuilemin. In: The Yeasts: human hosts, whereas C. gattii was thought to be A Taxonomic Study, 5th edn. Kurtzman CP et al. (eds), restricted to tropical and subtropical regions and has 1661–737. Elsevier, Amsterdam, 2011]. The C. neofor- not been associated with an impaired immune system mans ⁄ C. gattii species complex formerly included the (Bovers M et al., FEMS Yeast Res 2006; 6: 599–607; two varieties C. neoformans var. neoformans (serotypes A Bovers M et al., Rev Iberoam Micol 2008; 25: S4–S12.). and D) and C. neoformans var. gattii (serotypes B and C). Up to now, only C. neoformans var. grubii has been Currently, C. neoformans var. gattii has been raised to routinely isolated from environmental and clinical the species level as C. gattii whereas C. neoformans was samples in Cuba (Illnait-Zaragozi MT et al., PloS One separated into the two different varieties C. neoformans 2010; 5: e9124). Herein, we characterise a yeast strain var. grubii (serotype A) and C. neoformans var. neofor- isolated from a captivated cheetah (Acinonyx jubatus) mans (serotype D) (Kwon-Chung KJ et al., Taxon 2002; that was imported from South Africa and kept at the 51: 804–6). National Zoo in Havana, Cuba. Until recently an agglutination assay was widely used to determine the serotypes of C. neoformans ⁄ C. gattii Methods species complex (Bovers M et al., FEMS Yeast Res 2006; 6: 599–607). With the emergence of molecular tech- A yeast-like fungus was isolated from nasal discharge of niques, such as PCR fingerprinting, restriction fragment a cheetah and was subsequently identified using mac- length polymorphism (RFLP) analysis of the PLB1 and romorphological and micromorphological characteris- URA5 locus, amplified fragment length polymorphism tics including growth at 37 �C, absence of germ tube (AFLP) fingerprint analysis as well as several multi- formation, morphology on cornmeal agar with Tween locus sequence typing (MLST) approaches, it became 80, pigmentation on caffeic acid agar, urea hydrolysis, myo-inositol assimilation, growth on canavanine- Correspondence: J. F. Meis MD, PhD, Department of Medical Microbiology glycine-bromothymol blue medium and serological and Infectious Diseases, Canisius-Wilhelmina Hospital, PO Box 9015, 6500 agglutination using the Crypto Check system (Iatron GS Nijmegen, The Netherlands. Laboratories, Tokyo, Japan). The strain was deposited in Tel.: +31 24 365 7514. Fax: +31 24 365 7516. the public culture collection of the CBS-KNAW Fungal E-mail: [email protected] Biodiversity Centre (Utrecht, The Netherlands) under Accepted for publication 11 April 2011 accession number CBS11421.

92 Reactivation of a Cryptococcus gattii infection in a cheetah

Genomic DNA was extracted as described previously at 72 h were read optically at 420 nm after agitation. (Hagen F et al., Antimicrob Agents Chemother 2010; 54: Candida parapsilosis ATCC 22019 and Candida krusei 5139–45). The mating-type was determined using two ATCC 6258 were used as quality control. PCRs that specifically amplify the mating-type a and a allele of the STE12 locus (Bovers M et al., FEMS Yeast Results and discussion Res 2006; 6: 599–607). Reference strains WM779 (mating-type a, genotype AFLP7 ⁄ VGIV) and CBS1930 Cryptococcus has the potential to infect a wide range of (mating-type a, genotype AFLP6 ⁄ VGII) were included animals throughout the world, including wild animals, in the analysis. AFLP analysis was carried out as farm animals, domestic animals, aquatic animals and previously described (Boekhout T et al., Microbiology several varieties of birds (Duncan C et al., Can Vet J 2001; 147: 891–907). Strains WM179 (AFLP4 ⁄ VGI), 2006; 47: 993–8). Even though this is the most WM161 (AFLP5 ⁄ VGIII), WM178 (AFLP 6 ⁄ VGII) and frequent systemic fungal disease described in felines, WM779 (AFLP7 ⁄ VGIV) served as reference isolates and subclinical infection with spontaneous elimination of a phenetic tree was calculated using the single linkage cryptococci is much more common. Factors mediating clustering algorithm in combination with Pearson progression to infection have been linked to the pres- correlation using the software package Bionumerics ence of concurrent disease or other immunosuppressive version 4.6.1 algorithm (Applied Maths, Sint-Martens- conditions. (Duncan C et al., Med Mycol 2005; 43: 663– Latem, Belgium). 6; Duncan C et al., Med Mycol 2005; 43; 511–6; Multi-locus sequence typing was performed using the Duncan C et al., Can Vet J 2006; 47: 993–8; MacDou- seven standard loci: CAP59, GPD1, IGS1, LAC1, PLB1, gall L et al., Emerg Infect Dis 2007; 13: 42–50). SOD1 and URA5 according to Meyer W et al.(Med The cheetah has been considered a paradigm of Mycol 2009; 47: 561–70). In addition, the CAP10, vulnerability to infectious diseases because the species MPD1 and TEF1a loci were sequenced to compare the lacks heterogeneity at MHC loci that encodes peptides strain with a recently published large C. gattii MLST mediating immune responsiveness to pathogens (Mill- dataset (Fraser JA et al., Nature 2005; 437: 1360–4). ward IR and Williams MC, J S Afr Vet Assoc 1999; 70: Sequences have been deposited in GenBank under 35–9; Miller-Edge MA and Worley MB, Vet Immunol accession numbers HQ733955–HQ734184. lmmunopathol 1992; 30: 261–74). Cryptococcosis in Susceptibility testing using a broth microdilution these felines has been described in captive animals from method was performed in accordance with Clinical South Africa, especially due to C. gattii (Beehler WA, J and Laboratory Standards Institute document M27-A3 Am Vet Med Assoc 1982; 181: 1400–1; Berry WL et al., guidelines (Reference Method for Broth Dilution Antifungal J Zoo Wildl Med 1997; 28: 485–90; Bolton LA et al., JS Susceptibility Testing of Yeasts. Approved Standard, 3rd Afr Vet Assoc 1999; 70: 35–9; Millward IR and Williams edn, M27-A3. Wayne, PA, 2008). Standard antifungal MC, J S Afr Vet Assoc 1999; 70: 35–9). 9 powders of amphotericin B (Sigma, Zwijndrecht, The In the present case, the cheetah was imported from Netherlands), flucytosine (Valeant Pharmaceuticals, South Africa to Cuba at the age of approximately Zoetermeer, The Netherlands), itraconazole (Janssen ⁄ 1 year, where it was subsequently displayed at the Cilag, Tilburg, The Netherlands), fluconazole (Pfizer carnivores� area of the National Zoo. After 16 months in Central Research, Sandwich, Kent, UK), voriconazole captivity, the cheetah showed signs of sudden loss of (Pfizer Central Research), posaconazole (Schering weight, fatigue, anorexia, shortness of breath and nasal Plough, Kenilworth, NJ, USA) and isavuconazole discharge. The animal was taken to the veterinary clinic (Basilea Pharmaceuticals, Basel, Switzerland) were for sampling and died before any diagnosis could be used. The stock solutions of the drugs were prepared made (Polo-Leal JL et al., Rev Cub Med Trop 2010; 62: in the appropriate solvent. The final concentrations of 257–60). the antifungal agents were 0.016–8 lg ml)1 for Using AFLP and MLST analysis the strain was amphotericin B, itraconazole, voriconazole, and posa- classified as genotype AFLP4 ⁄ VGI. Additional mating- conazole; 0.063–32 lg ml)1 for flucytosine and fluco- type analysis revealed that the STE12a allele was nazole; and 0.004–4.00 lg ml)1 for isavuconazole. present and previous identification of the yeast as a C. After 72 h incubation at 35 �C the MIC was defined gattii strain serotype B based on conventional tech- as the lowest concentration of drug showing absence of niques and serology was confirmed. growth for amphotericin B and a prominent reduction There are several arguments that support the of growth (‡50%) for the other antifungal compounds hypothesis that the animal had a dormant infection at compared with the drug-free growth control. The MICs the time of entry at the zoo. First, so far all Cryptococcus

93 Chapter 9

isolates obtained from clinical and environmental sam- lineage that can be found in the tropical and subtropical ples in Cuba have been identified as C. neoformans var. regions of Africa, Asia, Australia, Central America and grubii and despite an extensive search, no C. gattii has South America (Hagen F et al., Antimicrob Agents been found (Martıńez-Machıń G et al., Rev Cub Med Trop Chemother 2010; 54: 5139–45; Springer DJ and Cha- 2004; 56: 77–9; Illnait-Zaragozi MT et al., PloS One turvedi V, Emerg Infect Dis 2010; 16: 14–20). The usage 2010; 5: e9124). Secondly, it has been shown before of large MLST datasets has recently been proven to be that the interval between exposure and development of an excellent approach to trace a clinical C. gattii strain clinical disease is highly variable. At one extreme, acute to its geographic origin (Lindberg J et al., Emerg Infect Dis infections have been reported while reactivations of long 2007; 13: 178–9; Hagen F et al., Med Mycol 2010; 48: silent subclinical infections are evident (Lindberg J et al., 528–31). MLST data of the cheetah C. gattii strain was Emerg Infect Dis 2007; 13: 178–9; Hagen F et al., Med compared with a large C. gattii MLST dataset and a subset Mycol 2010; 48: 528–31). We hypothesise that cap- of these strains was finally used to show the relationship tivity could be a very stressful factor for these animals between the cheetah C. gattii strain and environmental, that have normally a solitary life style and are adapted veterinary and clinical C. gattii AFLP4 ⁄ VGI strains that to running. This might have contributed to the poor are most closely related (Fig. 1). This revealed that the C. health of the cheetah and reactivation of the dormant gattii strain from the cheetah was nearly identical to a infection after being held in captivity during 1 year at strain isolated from an unknown source in Kenya, the zoo. whereas the cheetah strain was far more distantly related And finally, other studies have shown that there is a to strains from Central and South America (Fig. 1). This regional correlation among environmental and clinical strongly suggests that the cheetah had a dormant or isolates from Africa, France, Japan and the USA where subclinical C. gattii infection, acquired in Africa, which this aspect has been studied (Sorrell T et al., J Clin was activated during its captivity in the National Zoo in Microbiol 1996; 34: 1253–60). In the current case, Havana, Cuba. molecular biological characterisation showed that the Although some veterinarians prefer surgery when strain belonged to C. gattii genotype AFLP4 ⁄ VGI, a possible or euthanasia, a review of the literature showed

Figure 1 Cryptococcus gattii isolated from the cheetah is presented in bold; underlined strain numbers representing the four strains WM179 (AFLP4 ⁄ VGI), WM161 (AFLP5 ⁄ VGIII), WM178 (AFLP6 ⁄ VGII) and WM779 (AFLP7 ⁄ VGIV) were used as references. A set of 18 C. gattii AFLP4 ⁄ VGI strains is included to represent genetic lineages from other continents. The evolutionary history was inferred using the Maximum Parsimony method in MEGA version 4.1 (http://www.megasoftware. net). Values shown next to the branches are percentages (>75%) of replicated trees in which the associated taxa clustered together in the bootstrap test (1000 repli- cates).

94 Reactivation of a Cryptococcus gattii infection in a cheetah

that antifungal compounds in the group of azoles, 1544–8). In agreement with previously published especially ketoconazole and itraconazole, have been studies, which have shown high in vitro activity of the widely used in the treatment of animals affected with new azole isavuconazole for Cryptococcus species (Illnait- cryptococcosis, although the optimum treatment strat- Zaragozı´ MT et al., Antimicrob Agents Chemother 2008; egy is still not clear (Beehler BA, J Am Vet Med Assoc 52: 1580–2; Hagen F et al., Antimicrob Agents Chemo- 1982; 181: 1400–1; Medleau L et al., Am J Vet Res ther 2010; 54: 5139–45), we found that isavuconazole 1990; 51: 1454–8). was more potent (0.016 lg ml)1) than voriconazole The CLSI M27-A3 document recommends that yeast and itraconazole (0.031 lg ml)1 each one) for this C. strains with flucytosine MICs ‡32 lg ml)1 can be gattii strain. Unfortunately, the cheetah died before the regarded as resistant to this drug (CLSI, Reference diagnosis was made and no treatment was attempted. Method for Broth Dilution Antifungal Susceptibility Testing This is the first documented case of C. gattii in Cuba, of Yeasts. Approved Standard, 3rd edn, M27-A3. Wayne, although genotypic analysis suggests that the animal PA, 2008). At present there has been no amphotericin B had a dormant or subclinical infection at the moment of breakpoint established for the C. neoformans ⁄ C.gattii entrance into the zoo. Investigation of cryptococcal complex; therefore, we defined a strain being resistant cases using molecular tools may improve our under- with a MIC ‡2 lg ml)1, which was found to be standing of the risk factors for this disease, and we associated with therapeutic failure (Lozano-Chiu M et suggest performing active search for cryptococcal col- al., J Clin Microbiol 1998; 36; 2817–22). By these onisation ⁄ infection in these captive animals which definitions the studied isolate was susceptible to flucy- could lead to the formulation of preventive strategies. tosine and amphotericin B (8 and 0.125 lg ml)1, respectively). The MIC of 2 g ml)1 for fluconazole l Acknowledgment was the highest in vitro activity among the azoles. Previous reports have also demonstrated a low activity We thank Dr Alf Botha (University of Stellenbosch, of fluconazole against C. neoformans strains and suggest Stellenbosch, South Africa) and Dr Kathrin Tintelnot that good therapeutic results obtained with this agent (Charite University, Berlin, Germany) for providing are due to its high concentrations in cerebrospinal fluid different sets of C. gattii DNA to complete the MLST (Aller AI et al., Antimicrob Agents Chemother 2000; 44: phylogenetic analysis.

Chapter Fatal Cryptococcus gattii genotype10 AFLP5 infection in an immunocompetent Cuban patient

Published in Medical Mycology Case Reports 2013;2:48-51. Chapter 10

Fatal Cryptococcus gattii genotype AFLP5 infection in an immunocompetent Cuban patient

Marı´a T. Illnait-Zaragozı´ a, Lilia M. Ortega-Gonzalez b, Ferry Hagen c, Gerardo F. Martı´nez-Machin a, Jacques F. Meis c,d,n

a Department of Microbiology, Tropical Medicine Institute ‘‘Pedro Kouri’’, P.O. Box 601, Havana, Cuba b Intensive Care Unit, Tropical Medicine Institute ‘‘Pedro Kouri’’, P.O. Box 601, Havana, Cuba c Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, PO Box 9015, Nijmegen 6500 GS, The Netherlands d Department of Medical Microbiology, Radboud University Nijmegen Medical Centre, PO Box 9101, Nijmegen 6500 HB, The Netherlands

articleinfo abstract

Article history: We describe the ﬁrst clinical case of cryptococcosis due C. gattii in a Cuban immunocompetent patient Received 31 December 2012 who had a traveling history two years before to Central America. Molecular characterization of the Received in revised form isolate showed it to be genotype AFLP5 of which MLST sequences clustered with clinical and 1 February 2013 environmental strains from Colombia. The patient died one year after the diagnosis despite a prolonged Accepted 1 February 2013 treatment with (liposomal) amphotericin B, ﬂuconazole, voriconazole and gamma interferon. & 2013 International Society for Human and Animal Mycology. Published by Elsevier B.V All rights Keywords: reserved. Cryptococcal meningitis C. gattii AFLP5 Imported mycoses Treatment

1. Introduction human clinical samples in Cuba [5,6]. Here we report a fatal case of C. gattii meningitis in Cuba in an immunocompetent patient. Cryptococcosis is a life-threatening mycosis that develops following the inhalation and dissemination of fungal conidia and/or desiccated yeast forms. The Cryptococcus neoformans/ 2. Case Cryptococcus gattii species complex are free-living saprophytic yeasts which can survive in diverse niches [1]. The complex A 56-year-old white Cuban male from Villa Clara, a city in the includes two species with their varieties and serotypes: C. neofor- central part of the island, who had no significant medical history, mans var. grubii (serotype A; genotype AFLP1), C. neoformans var. was admitted on July 5th 2011 because of mild fever, myalgia, neoformans (serotype D; genotype AFLP2) and the intervarietal malaise and diffuse headache. He was a moderate smoker and state C. neoformans serotype AD (genotype AFLP3), which mostly used only sporadically alcohol. His most recent travel history was affect immunocompromised patients worldwide; the other to Honduras (2003–2005) and Guatemala (2007–2009) where he species, C. gattii (serotypes B and C; genotypes AFLP4-7 and worked as a general practitioner. Physical examination revealed a AFLP10) mainly affects individuals without apparent alterations mild fever of 38 1C, signs of lung consolidation on the lower right of the immune system and used to have a more restricted side, and fine crackles with chest auscultation. A chest radiograph distribution to tropical and subtropical climates, although recent showed an inflammatory lesion on the lower right side consistent findings indicates a drastic change in the adaptation of this with pneumonia. No cultures or clinical chemical investigations organism to other environments [2–4]. Up to now, only C. neoformans were performed. The patient was treated empirically with peni- var. grubii has been routinely isolated from environmental and cillin but after seven days (July 12th, 2011) he was admitted to the local hospital because of worsening of his general condition and development of a severe headache. Hematologic and serologic tests, including HIV antibody screening, were negative. A

lumbar puncture showed a high opening pressure (40 cm H2O) n Corresponding author at: Department of Medical Microbiology and Infectious and cerebrospinal ﬂuid (CSF) analysis revealed pleocytosis (215 Diseases, Canisius Wilhelmina Hospital, PO Box 9015, Nijmegen 6500 GS, 3 The Netherlands. Tel.: 31 243657514. cells/mm predominantly lymphocytes), reduced glucose level þ E-mail address: [email protected] (J.F. Meis). (2.6 mg/dL) and elevated protein level (480 mg/dL) (Table 1).

98 Fatal Cryptococcus gattii genotype AFLP5 infection in an immunocompetent Cuban patient

Table 1 1st, 2012, due to progressive deterioration, clinical signs of intracra- CSF findings during one year of C. gattii meningoencephalitis nial hypertension, seizures and papilledema. This improved slightly after CSF puncture but after seven days receiving amphotericin B Date Results deoxycholate (1 mg/kg/day) and fluconazole 800 mg/d he died on OP (cm Clinical chemistry Microbiology July 8th, 2012, one year after the start of the clinical presentation of H2O) the disease. An autopsy was not performed. Protein Glucose Cells India Culture Identification of the isolated yeasts as C. gattii was performed (mg/dL) (mg/dL) (mm3) ink by urease, inositol fermentation test and growth on canavanine- 12/7/2011 40 480 2.6 70 ( ) C. gattii glycine bromothymol blue (CGB) agar to differentiate C. neofor- þ 20/7/2011 35 171 2.4 215 ( ) C. gattii mans from C. gattii. Molecular identification as C. gattii genotype þ 25/8/2011 32 40.2 3.4 17.5 ( )() þ � AFLP5 was done by using amplified fragment length polymorphism 5/9/2011 28 20 3.8 12 ( )() þ � (AFLP) fingerprinting, as described previously [7,8]. The mating- 19/9/2011 25 30 3.8 10 ( )() þ � 3/10/2011 26 35 4 11 ( )() type was determined using a conventional PCR using STE12a- and þ � 13/10/2011 38 28 4.18 13 ( )() STE12a-specific primers [8] that revealed the presence of the þ � 23/10/2011 25 17.8 3.49 10 ( )() þ � mating-type a allele for both isolates. The two Cuban C. gattii 9/11/2011 35 17.8 3.49 10 ( )() þ � AFLP5 isolates were sequenced for the seven nuclear loci CAP10, 5/12/2011 22 38 4.1 13.5 ( )() þ � GPD1 LAC1, MPD1, PLB1 TEF1 1/7/2012 50 45 2.5 15 ( )() , IGS1, and to compare them to the þ � data of other isolates with the genotype AFLP5, as published by Hagen et al. (2012) and Byrnes et al. (2011) [3,9]. Sequences were Sputum and CSF was cultured and treatment was started with deposited in Genbank and can be accessed via the accession ceftriaxone (4 g/d). After isolation of Klebsiella pneumoniae from numbers KC424622–KC424635. Isolates were deposited in the sputum cultures, amikacin (15 mg/kg/d during seven days) was CBS Fungal Biodiversity collection in Utrecht, the Netherlands as added to the treatment. The patient did not improve satisfactorily CBS 12755 (CUBA 250) and CBS 12754 (CUBA 350). A bootstrap and therefore a computerized tomography (CT) scan and mag- Maximum Likelihood phylogenetic analysis was performed using netic resonance imaging (MRI) were performed which were the settings as being used for a previously published C. gattii MLST unremarkable. To reduce suspected intracranial hypertension, study [3] (Fig. 1). the lumbar puncture was repeated a week later (July 20th, 2011). Opening pressure was still increased with signs of inflammation (Table 1). Direct microscopic examination of CSF with 3. Discussion India ink showed the presence of a large number of rounded and encapsulated yeasts compatible with Cryptococcus. The treatment C. gattii differs from the more commonly isolated C. neofor- was changed to amphotericin B deoxycholate (0.7 mg/kg) intrave- mans in clinical aspects, ecological niche and genetic makeup. nously and 400 mg (b.i.d.) of fluconazole orally. Flucytosine was Unlike C. neoformans, it most often affects immunocompetent not available in Cuba. After an initial improvement with antifungal patients exposed to an environmental source causing pulmonary treatment the patient experienced worsening pulmonary functions, and cerebral infections. Furthermore C. gattii is known to be more intense headache, neurological impairment, decreased visual acuity resistant to antifungal agents. In the Caribbean, cryptococcosis and hearing loss on the right side. Therefore on August 23th, 2011, appears to be a relatively rare disease with an estimated 7800 the patient was transferred to an infectious diseases reference center new patients each year [10]. In Cuba, the disease was first in Havana where the isolate was identified as Cryptococcus gattii. In reported in the early 1950s [11]. Since then sporadic cases were vitro susceptibility testing showed low MICs of amphotericin B associated with alcoholism, organ transplantations and immuno- (0.125 mg/L), fluconazole (2 mg/L), flucytosine (0.5 mg/L), itracona- logical disorders. In 1986, with the beginning of the AIDS zole (0.125mg/L), posaconazole (0.125mg/L), voriconazole epidemic in Cuba, the amount of individuals suffering from (0.063 mg/L) and isavuconazole (0.016 mg/L). cryptococcosis had increased over the years with an average of So far the patient had received a cumulative dose of 1.500 mg eight new cases each year [12]. This might still be an under- amphotericin B deoxycholate and his kidney functions had representation since a retrospective autopsy study of HIV/AIDS become slightly disturbed (creatinine 130 mmol/L, ureum patients performed during a 10-year period showed that disse- 10 10.6 mmol/L). Mayor findings at physical examination revealed minated or central nervous system cryptococcosis was a serious a right VI cranial nerve palsy, blurred vision and diplopia in and common disorder in 29% of the cases [13]. The clinical addition to a non-resolving pneumonia. Repeat CT scan with and presentation of cryptococcal infection varies from asymptomatic without contrast did not show signs of cerebral edema or other pulmonary colonization to severe pneumonia with respiratory abnormalities. Because of the poor clinical response and worsen- failure and life-threatening meningitis [14,15]. Mostly infections ing kidney function amphotericin B deoxycholate was changed to due to C. gattii genotypes AFLP4 and AFLP6 usually occurs in liposomal amphotericin B (3 mg/kg) and fluconazole 800 mg/d patients without detectable predisposing factors while genotypes was continued from August 29th to October 6th, 2011, but again AFLP5, AFLP7 and AFLP10 appear to be associated with an without clinical improvement. Several lumbar punctures were impaired immune system (mainly AIDS, but also any other necessary to control high intracranial pressure. All samples immunosuppressive condition including solid organ transplanta- remained microscopically positive for cryptococci (Table 1). In tion) which is similar to C. neoformans [3,16]. Interestingly none this period the patient became blind and lost his hearing. After a of these underlying diseases, previously associated with C. gattii total dose of 8 g liposomal amphotericin B, fluconazole was AFLP5 infection, could be demonstrated in our patient. replaced by voriconazole (800 mg/d orally) and the patient was About 65%–77% of cases with C. gattii infections presents discharged on December 5th, 2011. During the last 2-month more often with lung involvement [1]. Initial symptoms of our period the patient also received recombinant IFNg (200 mg) patient fitted with a community-acquired pneumonia and sputum subcutaneously thrice weekly. Because voriconazole was no cultures only demonstrated Klebsiella. It might be hypothesized longer available he was treated with fluconazole 400 mg from that the development of a community-acquired pneumonia could March 21st until July 1st 2012. The patient’s blindness and loss of generate appropriated conditions favoring rapid multiplication and hearing remained permanent. He was re-admitted to hospital on July dissemination of C. gattii from its dormant state. Cryptococcal

99 Chapter 10

infection involves any part of the central nervous system and leads to a wide spectrum of neurological symptoms such as headache, altered consciousness, seizures, and cranial nerve palsies. Seaton et al. [17] reviewed the ophthalmic findings in 82 immunocompetent patients in whom C. gattii was involved. A high rate of visual loss was observed in 52.6% of survivors similar what had occurred in our patient. Environmental exposure is known to be the dominant risk factor for C. gattii infection, but so far all pathogenic clinical and environmental Cryptococcus isolates from Cuba have been identified as C. neoformans var. grubii [5]. However a single case of a fatal veterinary C. gattii genotype AFLP4 infection was described recently in a cheetah imported from South Africa. Molecular epidemiologic studies strongly suggested that the animal was not infected in Cuba, but most likely in Africa [18]. It has been demonstrated before that C. gattii infections can be imported subclinically and reactivate after being dormant for many years [3]. There are three main arguments that support the hypothesis that our patient had a dormant infection when he returned to Cuba from Central America. First, there is no evidence suggesting environmental presence of this species despite multiple attempts to recover C. gattii from plants, trees and cacti in Cuba [6], which is in strong contrast to for example India where C. gattii can be isolated from up to 50% of investigated trees [16]. Secondly, it has been ample demonstrated that the time interval between exposure to C. gattii and the development of the disease is highly variable; at one extreme, acute infection have been reported while reactivation of long silent subclinical infections is also possible depending of the immunological status of the patient [3,19]. We have no indication that a hereditary immune disorder might have been involved but there is in vitro evidence that clinical C. gattii isolates induced a more pronounced inflammatory response compared to other Cryptococcus species and environmental C. gattii [20]. This might explain the severe immunopathology seen in human (and animal) C. gattii infections. And finally, molecular biological characterization showed that the strain belonged to C. gattii genotype AFLP5. Although this genotype was originally reported from India and the USA, current literature show that it has spread throughout Latin-American countries (Argentina, Brazil, Colombia, Guatemala Mexico, and Venezuela), the west coast of the USA., as well as over Asia and Australia [3,16]. Comparison with a large C. gattii genotype AFLP5 MLST dataset [3,9] yielded a high similarity degree to strains from clinical and environmental sources in Colombia. The patient’s travel history to Central America suggests acquisition during this time period. The neurological picture in conjunction with the presence of yeasts in the CSF let to suspect cryptococcal infection in our patient initiating treatment according to management guidelines [21] and availability of drugs. For meningoencephalitis and disseminated disease due to C. gattii, recommended induction, consolidation, and suppressive treatment are basically the same as for C. neoformans. Because C. gattii infection is associated with more neurological complications and delayed response to therapy, as demonstrated in this case, a longer course of therapy, increasing dose and/or recombinant IFNg as salvage therapy for patients unresponsive to multiple antifungal drugs has been suggested [21]. The latter salvage therapy was successful in patients with C. neoformans infections [22]. For persistent and relapse isolates it is highly recommended to look for changes in Fig. 1. A 1000 bootstrap Maximum Likelihood phylogenetic analysis of the two the MIC from the original isolate (Z3-dilution difference suggests � Cuban strains (CBS 12755 and CBS 12754), compared to C. gattii AFLP5 strains development of direct drug resistance) [13,21]. Unfortunately, from previously published MLST studies [3,9]. Both Cuban strains (marked in red) although the direct examination of CSF stayed positive during the clustered together with clinical and environmental strains from Colombia. whole disease course, cultures remained negative. Our case is a Numbers next to branches represent bootstrap values 75. (For interpretation Z quick alert to suspect C. gattii AFLP5 infection also in HIV negative of the references to color in this figure legend, the reader is referred to the web version of this article.) patients, to consider the growing possibility of imported

100 Fatal Cryptococcus gattii genotype AFLP5 infection in an immunocompetent Cuban patient

infections from endemic regions and the need for more effective [9] Byrnes EJ, Li W, Ren P, Lewit Y, Voelz K, Fraser JA, et al. A diverse population alternatives for treating this mycosis. of Cryptococcus gattii molecular type VGIII in southern Californian HIV/AIDS patients. PLoS Pathogens 2011;7:e1002205. [10] Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, Chiller TM. Estimation of the current global burden of cryptococcal meningitis among Conflict of interest persons living with HIV/AIDS. AIDS 2009;23:525–530. [11] Curbelo A. Septicemia y meningitis a Cryptococcus neoformans. Archivos del JFM received grants from Astellas, Merck, Pfizer, and Schering- Hospital Universitario (Cuba) 1951;9:324–330. [12] Illnait MT, Valde´s IC, Martıńez GF, Fernańdez CM. Criptococcosis. Una alerta Plough. He has been a consultant to Basilea and Merck and necesaria. Revista Cubana de Medicina Tropical 2001;53:224–225. received speaker fees from Merck and Gilead. All other authors: [13] Illnait-Zaragozı´ MT, Martıńez GF, Fernańdez CM, Hagen F, Boekhout T, no potential conflicts of interest. Klaassen CH, et al. Microsatellite typing and susceptibilities of serial Cryptococcus neoformans isolates from Cuban patients with recurrent cryptococcal meningitis. BMC Infectious Diseases 2010;10:289–296. References [14] Goldman JD, Vollmer ME, Luks AM. Cryptococcosis in the immunocompetent patient. Respiratory Care 2010;55:1499–1503. [15] Kim YS, Lee IH, Kim HS, Jin SS, Lee JH, Sung-Kyoung K, et al. Pulmonary [1] Kwon-Chung KJ, Boekhout T, Wickes BL, Fell JW. Systematics of the genus cryptococcosis mimicking primary lung cancer with multiple lung metas- Cryptococcus and its type species C. neoformans. In: Heitman J, editor. tases. Tuberculosis and Respiratory Diseases 2012;73:182–186. Cryptococcus: from human pathogen to model yeast. Washington, DC: [16] Chowdhary A, Prakash A, Randhawa1 HS, Hagen F, Klaassen CH, Meis JF. First ASM Press; 2011. p. 3–15. environmental isolation of Cryptococcus gattii, genotype AFLP5, from India [2] Chowdhary A, Randhawa HS, Boekhout T, Hagen F, Klaassen CH, Meis JF. and a global review. Mycoses 2013:56, http://dx.doi.org/10.1111/myc.12039. Temperate climate niche for Cryptococcus gattii in Northern Europe. Emerging [17] Seaton RA, Verma N, Naraqi S, Wembri JP, Warrell DA. Visual loss in Infectious Diseases 2012;8:172–174. immunocompetent patients with Cryptococcus neoformans var. gattii menin- [3] Hagen F, Colom MF, Swinne D, Tintelnot K, Iatta R, Montagna MT, et al. gitis. Transactions of the Royal Society of Tropical Medicine and Hygiene Autochthonous and dormant Cryptococcus gattii infections in Europe. Emer- ging Infectious Diseases 2012;18:1618–1624. 1997;91:44–49. [4] Kidd SE, Hagen F, Tscharke RL, Huynh M, Bartlett KH, Fyfe M, et al. A rare [18] Illnait-Zaragozı´ MT, Hagen F, Fernańdez CM, Martıńez GF, Polo JL, Boekhout T, genotype of Cryptococcus gattii caused the cryptococcosis outbreak on et al. Reactivation of a Cryptococcus gattii infection in a cheetah (Acinonyx Vancouver Island (British Columbia, Canada). Proceedings of the National jubatus) held in the National Zoo, Havana, Cuba. Mycoses 2011;54: Academy of Sciences USA 2004;101:17258–17263. e889–e892. [5] Illnait-Zaragozi MT, Martıńez-Machıń GF, Fernańdez-Andreu CM, Boekhout T, [19] Hagen F, van Assen S, Luijckx GJ, Boekhout T, Kampinga GA. Activated Meis JF, Klaassen CH. Microsatellite typing of clinical and environmental dormant Cryptococcus gattii infection in a Dutch tourist who visited Vancou- Cryptococcus neoformans var. grubii isolates from Cuba shows multiple ver Island (Canada): a molecular epidemiological approach. Medical Mycol- genetic lineages. PLoS ONE 2010;5:e9124. ogy: 2010;48:528–531. [6] Illnait-Zaragozı´ MT, Martıńez-Machıń GF, Fernańdez-Andreu CM, Perurena- [20] Schoffelen T, Illnait-Zaragozı´ MT, Joosten LAB, Netea MG, Boekhout T, Meis JF. Lancha MR, Theelen B, Boekhout T, et al. Environmental isolation and Cryptococcus gattii induces a cytokine pattern that is distinct from other characterization of Cryptococcus species from living trees in Havana City, cryptococcal species. PLoS ONE 2013;8:e55579. Cuba. Mycoses 2012;55:138–144. [21] Perfect JR, Dismukes WE, Dromer F, Goldman DL, Graybill JR, Hamill RJ, et al. [7] Boekhout T, Theelen B, Diaz M, Fell JW, Hop WC, Dromer F, et al. Hybrid Clinical practice guidelines for the management of cryptococcal disease: genotypes in the pathogenic yeast Cryptococcus neoformans. Microbiology 2010 update by the Infectious Diseases Society of America. Clinical Infectious 2001;147:891–907. Diseases 2010;50:291–322. [8] Hagen F, Illnait-Zaragozi MT, Bartlett KH, Swinne D, Geertsen E, Klaassen CH, [22] Netea MG, Brouwer AE, Hoogendoorn EH, Van der Meer JW, Koolen M, et al. In vitro antifungal susceptibilities and AFLP genotyping of a worldwide Verweij PE, et al. Two patients with cryptococcal meningitis and idiopathic collection of 350 clinical, veterinary and environmental Cryptococcus gattii CD4 lymphopenia: defective cytokine production and reversal by recombinant isolates. Antimicrobial Agents and Chemotherapy 2010;54:5139–5145. interferon- gamma therapy. Clinical Infectious Diseases 2004;39:e83–e87. 10

101

Chapter Cryptococcus gattii induces a11 cytokine pattern that is distinct from other cryptococcal species

Published in PLoS ONE 2013;8:e55579. Chapter 11

Cryptococcus gattii Induces a Cytokine Pattern That Is Distinct from Other Cryptococcal Species

Teske Schoffelen1, Maria-Teresa Illnait-Zaragozi2, Leo A. B. Joosten1, Mihai G. Netea1, Teun Boekhout3,4, Jacques F. Meis5, Tom Sprong1,5,6* 1 Department of Medicine and Nijmegen Institute for Infection, Inflammation & Immunity (N4i), Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, 2 Department of Bacteriology and Mycology, Instituto Pedro Kourı´, Havana, Cuba, 3 CBS-KNAW Fungal Biodiversity Centre, Utrecht, The Netherlands, 4 Shanghai Key Laboratory for Molecular Medical Mycology, Changzheng Hospital, Second Military Medical University, Shanghai, China, 5 Department of Medical Microbiology and Infectious Diseases, Canisius-Wilhelmina Hospital, Nijmegen, The Netherlands, 6 Department of Medicine, Canisius-Wilhelmina Hospital, Nijmegen, The Netherlands

Abstract Understanding more about the host’s immune response to different Cryptococcus spp. will provide additional insight into the pathogenesis of cryptocococcis. We hypothesized that the ability of C. gattii to cause disease in immunocompetent humans depends on a distinct innate cytokine response of the host to this emerging pathogen. In the current study we assessed the cytokine profile of human peripheral blood mononuclear cells (PBMCs) of healthy individuals, after in vitro stimulation with 40 different well-defined heat-killed isolates of C. gattii, C. neoformans and several hybrid strains. In addition, we investigated the involvement of TLR2, TLR4 and TLR9 in the pro-inflammatory cytokine response to C. gattii. Isolates of C. gattii induced higher concentrations of the pro-inflammatory cytokines IL-1b, TNF-a and IL-6 and the Th17/22 cytokine IL-17 and IL-22 compared to C. neoformans var neoformans and C. neoformans var grubii. In addition, clinical C. gattii isolates induced higher amounts of cytokines than environmental isolates. This difference was not observed in C. neoformans var. grubii isolates. Furthermore, we demonstrated a likely contribution of TLR4 and TLR9, but no role for TLR2, in the host’s cytokine response to C. gattii. In conclusion, clinical heat-killed C. gattii isolates induced a more pronounced inflammatory response compared to other Cryptococcus species and non-clinical C. gattii. This is dependent on TLR4 and TLR9 as cellular receptors.

Citation: Schoffelen T, Illnait-Zaragozi M-T, Joosten LAB, Netea MG, Boekhout T, et al. (2013) Cryptococcus gattii Induces a Cytokine Pattern That Is Distinct from Other Cryptococcal Species. PLoS ONE 8(1): e55579. doi:10.1371/journal.pone.0055579 Editor: Maria Leite de Moraes, CNRS, France Received October 10, 2012; Accepted January 3, 2013; Published January 31, 2013 Copyright: � 2013 Schoffelen et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: M.G.N. was supported by a Vici Grant of the Netherlands Organization for Scientific Research. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction cytokines excreted by Th17 cells, have several pro-inflammatory functions, one of which is eliciting defensin production by The incidence of cryptococcosis has increased dramatically over epithelial cells [7]. Previous studies have shown a crucial role of the past decades, due in a large part to the global HIV pandemic. Th17 cells in human antifungal defense against mucosal Candida More than 600,000 deaths are estimated to occur each year as a albicans infections [8–10]; but the role of this particular Th- result of cryptococcal meningoencephalitis [1]. The species C. lymphocyte subset in anti-cryptococcal defense is not clear. neoformans is an opportunistic pathogen mainly affecting immuno- Which cytokines are released depends on recognition of compromised hosts. In contrast, C. gattii mainly causes disease in microbial components by pattern recognition receptors (PRRs) apparently immunocompetent hosts at lower incidence [2,3]. C. on the cells of the innate immune system. Toll-like receptors gattii is emerging over the past decade as a pathogen in the Pacific (TLRs), a well-defined set of PRRs, are expressed on a variety of North-West of North America and has caused a large outbreak on cells and are important mediators of pro-inflammatory cytokine Vancouver Island [4,5]. This outbreak was mainly caused by a release. However, their role in mediating cytokine response to single, hypervirulent genotype of C. gattii, namely AFLP6A/VGIIa Cryptococcus spp. is being debated [11–15]. [6]. Understanding more about the host’s immune response to Cells of the innate immune system are important for initial different Cryptococcus spp, will provide additional insight into the defense against pathogens. Upon contact with pathogens, they pathogenesis of cryptocococcis. We hypothesized that the ability of produce pro-inflammatory cytokines such as tumor necrosis factor C. gattii to cause disease in immunocompetent humans depends on (TNF)-a, Interleukin (IL)-1b and IL-6, thereby initiating a specific a distinct innate cytokine response of the host to this emerging adaptive cellular immune response. Anti-inflammatory cytokines pathogen. Therefore, in the current study we assessed the cytokine such as IL-1RA are also produced and act as downregulators of profile of human peripheral blood mononuclear cells (PBMCs) of this immune response. Of particular interest for fungal infections, healthy individuals, after in vitro stimulation with well-defined heat- the cytokines IL-1b and IL-6 in the presence of IL-23 induce the killed isolates of C. gattii, C. neoformans and several hybrids. In development of T-helper (Th)17 cells. IL-17 and IL-22, the major

104 Cryptococcus gattii induces a cytokine pattern that is distinct from other cryptococcal species

addition, we investigated the involvement of TLR2, TLR4 and cytokines using a 100-fold lower concentration of Candida albicans TLR9 in the pro-inflammatory cytokine response to C. gattii. was much higher (data not shown). Results for the induction of T- cell derived cytokines IL-17 and IL-22 after 7 days of incubation Results are shown in Figure 1. It appeared that the studied Cryptococcus strains induce low amounts of IL-17 but substantial quantities of Quantitative comparison of cytokine induction between IL-22, again with significant inter-strain variation in the produc- different Cryptococcus spp. tion of these cytokines. We determined the concentration of several cytokines produced Figure 2 shows a quantitative comparison of cytokine induction by PBMCs upon stimulation with 40 different heat-killed between two varieties of C. neoformans, C. gattii and various hybrid Cryptococcus species complex isolates in order to elucidate the isolates. C. gattii was a more potent inducer of the pro- cytokine milieu in cryptococcal infection and to explore differences inflammatory cytokines TNF-a, IL-1b, IL-6 and the T-cell between the species. In preliminary experiments, we determined cytokines IL-17 and IL-22, compared to both C. neoformans that the minimal concentration of yeasts necessary to induce varieties. The different species did not differ with regard to IL- cytokine production is 107 microorganisms/mL (data not shown). 1Ra induction. Interestingly, the interspecies hybrids containing C. There was substantial inter-strain variation in the production of gattii as a partner of the mating pair induced significantly higher the pro-inflammatory cytokines IL-1b, TNF-a, IL-6 and the anti- cytokine production than hybrids which were the result of mating inflammatory cytokine IL-1Ra. TNF-a and IL-1b were induced in between the two varieties of C. neoformans. This suggests that an low amounts (up to 300 pg/mL). Interestingly, production of these

Figure 1. All forty Cryptococcus strains induce low amounts of IL-17, but high amounts of IL-22. IL-17 and IL-22 production after 7 d by PBMCs stimulated with RPMI+, either one of 40 different heat-killed Cryptococcus strains [107 microorganisms/mL] or heat-killed Candida albicans [105 microorganisms/mL] is shown respectively. Mean values 6 SE (n = 5) of three independent experiments are presented. doi:10.1371/journal.pone.0055579.g001

105 Chapter 11

Figure 2. Comparison of C. gattii isolates and interspecies hybrids with C. neoformans isolates and hybrids between both C. neoformans varieties. The forty heat-killed Cryptococcus isolates are grouped according to (sub)species. Cytokine production by human PBMCs after 24 h (IL-1b, TNF-a, IL-6 and IL-1Ra) and 7 d (IL-17 and IL-22) incubation is shown. Mean values (n = 5 to 7) 6 SE of three independent experiments are presented. *, p 0.01 to 0.05; **, p 0.001 to 0.01; ***, p,0.001. The horizontal line represents the lower detection limit. doi:10.1371/journal.pone.0055579.g002

inheritable factor is responsible for the difference in cytokine VGII which was involved in the Vancouver Island outbreak, were production. compared to four environmental C. gattii isolates (isolates 13,15,16,17), as well as to four clinical C. neoformans isolates Quantitative comparison of cytokine induction between (isolates 1,4,5,9), with regard to the cytokine induction (Figure 3). environmental and clinical strains within the Clinical C. gattii isolates induced significantly higher IL-1b and IL- Cryptococcus species complex 6 amounts compared to environmental isolates. Moreover, clinical b a Sixteen clinical C. gattii isolates (isolates 10,12,14,18,19–21,23– C. gattii isolates also induced higher IL-1 , IL-6, TNF- , IL-1Ra 29,39,40), of which six isolates belonging to the genotype AFLP6/ and IL-17 than clinical C. neoformans isolates. The C. gattii genotype

106 Cryptococcus gattii induces a cytokine pattern that is distinct from other cryptococcal species

Figure 3. Comparison of cytokine production by PBMCs induced by clinical or environmental cryptococcal isolates. Heat killed clinical isolates of C. gattii are compared to environmental C. gattii isolates and to clinical isolates of C. neoformans. The clinical isolates of C. gattii genotype AFLP6/VGII are depicted separately. Mean values (n = 5 to 7) 6 SE values of three independent experiments are presented. *, p 0.01 to 0.05; **, p 0.001 to 0.01; ***, p,0.001. The horizontal line represents the lower detection limit. doi:10.1371/journal.pone.0055579.g003

AFLP6/VGII, however, induced no higher amounts of other low levels of cytokines by C. neoformans var grubii isolates, as seen in cytokines compared to the other clinical C. gattii isolates. the panel of 40 isolates, was confirmed. In a different panel of Cuban C. neoformans var grubii isolates, comparison of clinical with environmental isolates showed no Involvement of different Pattern Recognition Receptors significant difference (P value for IL-6 and IL-22: 0.19 and 0.07 (PRRs) in cytokine production induced by C. gattii respectively) in cytokine production (Figure 4). The induction of To assess which PRRs are involved in recognizing C. gattii, we performed experiments in which PBMCs were preincubated for

107 Chapter 11

Figure 4. Comparison of cytokine production by PBMCs induced by clinical or environmental C. neoformans var grubii isolates. Cytokine production by human PBMCs after 24 h (IL-6) and 7 d (IL-22) incubation with heat-killed isolates is shown. Mean values (n = 7) 6 SE of three independent experiments are presented. ns, not significant. doi:10.1371/journal.pone.0055579.g004

one hour with specific PRR blocking reagents prior to stimulation without mediating IL-17 and seems to be critical in differentiation with heat-killed C. gattii or, as a control, culture medium. of naı¨ve T cells to Th22 cells [16]. IL-22 is a unique cytokine in Stimulation with culture medium showed undetectable levels for that it acts only on non-immune cells including keratinocytes, all cytokines (not shown). Blocking TLR2 had no effect on myofibroblasts and epithelial cells in tissues of the respiratory cytokine production by C. gattii, whereas this antibody significantly inhibited IL-1beta production after stimulation with Pam3cys (a known TLR-2 ligand) (Figure S1). Blocking TLR4 significantly diminished IL-1b induction by C. gattii, with a trend towards significance for TNF-a (P = 0.06). Interestingly, blocking TLR9 led to significantly higher concentrations of IL-1b induced by C. gattii compared to its control, and a trend towards significance (P = 0.06) was found for TNF-a. Blocking TLR9 had a negative effect (P = 0.03) on IL-17 production induced by C. gattii (Figure 5 for the effect on IL-1b and IL-17). We performed these experiments also with C. neoformans var grubii (H99). The latter isolate did not elicit a substantial pro- inflammatory cytokine response in PBMCs, as shown in previous experiments with other strains. Moreover, we did not observe an increase in IL-1b and TNF-a production induced by C. neoformans var grubii when blocking TLR9 (results not shown).

Discussion In the present study we investigated the in-vitro cytokine production of human PBMCs incubated with 40 different heat- killed isolates of the Cryptococcus neoformans species complex. We demonstrate that C. gattii isolates induces higher concentrations of pro-inflammatory and Th17/22 cytokines compared to C. neoformans var. neoformans and C. neoformans var. grubii. In addition, we found that clinical C. gattii isolates were able to induce higher amounts of cytokines than environmental isolates or clinical C. neoformans isolates. Furthermore, we demonstrated a contribution of TLR4 and TLR9, but no role for TLR2, in the host’s cytokine response to C. gattii. Figure 5. The role of TLR2, TLR4 and TLR9 in IL-1b and IL-17 induction by C. gattii. Cytokine production by human PBMCs Our results indicate that Cryptococcus neoformans species complex preincubated for one hour with culture medium (white bar) or PRR seems to induce mainly a IL-22 response, with surprisingly low IL- blocking reagents (dark gray bars) or their control (light gray bar) prior 17 production. This argues against a Th17 response to crypto- to stimulation with heat-killed C. gattii (strain B5742) [107/ml]. IL-1b is coccal infection as we hypothesized, but rather to an exclusively determined after 24 h incubation, IL-17 is determined after 7 d IL-22 producing subset of Th-cells. A candidate for this areTh22 incubation. Mean values 6 SE of eight individuals in 4 independent experiments (IL-17) or six individuals in 5 independent experiments (IL- cells. These cells have not been found in mice so far; therefore only 1b) (with exclusion of additional four individuals with undetectable studies with human cells can be used to determine the role of this cytokine induction by C. gattii) are presented. *, p 0.01 to 0.05. The Th subset in host defense against cryptococcal infections. The aryl horizontal line represents the lower detection limit. hydrocarbon receptor is identified to mediate IL-22 production doi:10.1371/journal.pone.0055579.g005

108 Cryptococcus gattii induces a cytokine pattern that is distinct from other cryptococcal species

s- , 1992a , 1992a , 2001 , 1997 , 1997 , 1997 , 2001 , 1997 , 1997 , 1997 , 1997 , 2001 , 2001 , 2001 , 2001 , 2002 et al. et al. , 2003 , 2003 , 2003 , 1999 , 2004 et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. , 2005 , 2004 ., 2004 et al. et al. et al. et al. et al. et al. et al et al. et al. Reference/Source Boekhout Katsu Boekhout Boekhout Boekhout Meyer Boekhout Kidd Kidd Meyer Boekhout Meyer Boekhout Meyer Boekhout Boekhout Boekhout Boekhout Boekhout Kwon-Chung Kwon-Chung , CBS7000 6 Cryptococcus Cryptococcus tree, Brazil wood from hollow, (H99) Cryptococcus bacillisporus , Mt. Annan, New South Cassia grubii var. sp. debris from car park of zoo, San NIH433) 6 (RV 20186) Eucalyptus Diego, USA Eucalyptus camaldulensis reference strain of molecularDiego, type USA VGIII, San Human, California, USA Eucalyptus tereticornis Wales, Australia Human, Auckland, New Zealand Detritus of almond tree, Colombia Immunocompetent male, Duncan, Vancouver Island, Canada Immunocompetent male, Victoria, CanadaSick goat, Aruba Immunocompetent human, Seattle, USA, Kidd Boekhout Immunocompetent human, reference strain of molecular type VGI, Sydney, Australia Human, type strain of California, USA Immunocompetent human, lung, reference strain of molecular type VGII, Sydney, Australia Human, Thailand Subculture of type straingattii of HIV positive human, referencemolecular strain type of VNIV, Melbourne, Australia. Clinical isolate, Argentina Cryptococcal meningitis patient, Tanzania Lengeler AIDS patient, Zimbabwe Decaying wood of Genetic offspring of CBS6885 ( = NIH12 Congenic pair with JEC21only that in differs mating type Subculture of type strainneoformans of Congenic pair with JEC20in that mating differs type only 5B 5C 4 5 5A 6 6 6 6 4 5B 5C 6 4 4 2 1 1 1 1 2 2 1 2 B B B C B B B B B B C B B B D A A A A D D A D neoformans grubii grubii grubii grubii neoformans neoformans grubii neoformans var. var. var. var. var. var. var. var. var. VGIII VGI VGIII VGIII VGII VGII VGII VGII VGI VGIII VGIII AFLP5 = VGIII C VGII VGI VGI C. gattii Species and varieties Sero-type AFLP-genotype Origin C. gattii C. gattii C. gattii C. gattii C. gattii C. gattii C. gattii C. gattii C. neoformans C. gattii C. gattii C. gattii C. gattii C. gattii C. gattii C. neoformans C. neoformans C. neoformans C. neoformans C. neoformans C. neoformans C. neoformans C. neoformans 11 CBS10079 CBS10078 CBS10081 NIH191, ATCC32608 IFM50894, CBS10082 NIH-B-3939 CBS10511, NIH-B4476 CBS10513, NIH-B4500 (T) (R) (R) (R) (R) (T) CBS6955 WM728 CBS8755 HOO58-I-682 WM276CN043 CBS10510 A1M R265 CBS10514 A1M R368 A1M-R376 CBS1930 CBS6956WM178 NIH444, ATCC32609 WM161 CBS6993 NIH18 CBS6998CBS8273 NIH365 WM179 CBS6289, RV20186, 125.91 CBS10512 P152 CBS8336 B-3501 CBS6900 JEC20 CBS8710CBS996 CBS10515, H99 JEC21 WM629 Details of the 40 cryptococcal isolates. 18 17 15 13 14 20 21 22 23 24 16 19 10 11 12 Table 1. No. in experiment Isolate Other specification 1 5 2 6 7 3 4 8 9

109 Chapter 11 , 2001 , 1997 , 2002 ., 2001 ., 2001 ., 2001 2012 2012 2012 2012 , 2003 ., 2006 ., 2008 ., 2006 , 2004 et al. et al et al et al et al. et al. et al et al et al et al. et al. et al. et al. et al. et al. Boekhout Reference/Source Bovers Boekhout Boekhout Bovers Boekhout Meyer Hagen Hagen Katsu Hagen Hagen Boekhout Diaz and Fell, 2005 Latouche , peach, Italy C. neoformans HIV-positive human, The Netherlands HIV-positive human, Canada, visited Mexico Bovers Type strain - - HIV-negative human, The Netherlands - Cheetah, reference strain oftype molecular VGIV, Johannesburg, South Africa Milk from mastitic cow, Switzerland HIV- patient from Mexico, Spain HIV- patient from Mexico, Spain HIV positive patient, India Clinical, Johannesburg, South Africa Human, Punjab, India HIV-negative human, Greece HIV-negative human, Greece 8 9 3 3 3 8 3 7 2 10 10 7 7 7 6B 6A BD BD AD AD AD BD AD C D B B B C C B B neoformans C. neoformans C. neoformans C. neoformans var. 6 6 6 AFLP4 AFLP4 VGIV AFLP4 VGIV VGIV VGIV VGII VGII AFLP2 C. gattii AFLP1 - - - AFLP2 C. gattii - Species and varieties Sero-type AFLP-genotype Origin C. gattii C. gattii C. neoformans C. gattii C. gattii C. gattii C. gattii C. gattii C. gattii C. gattii 308 # IFM50896, CBS10101 CBS10089 CBS10090 (R) CBS10496 LSPQ CBS132 RV52733 NYJ40 CBS10489 AMC2010404 RV52755 CBS10488 AMC770616 CBS5467 WM779 IHEM14941 Slimy RV 63979, IHEM14941, CBS11687 IHEM14941 WhiteRV 63979, IHEM14941 M27055 B5748 B5742 AV54 AV55 Cont. 33 34 36 35 32 37 31 38 30 Table 1. No. in experiment Isolate Other specification 39 40 doi:10.1371/journal.pone.0055579.t001 29 28 27 26 25

110 Cryptococcus gattii induces a cytokine pattern that is distinct from other cryptococcal species

Table 2. Details of 11 additional C. neoformans var grubii isolates, arranged by Microsatellite Complex (MC) [29].

Number in experiment Isolate Other specification Species Serotype MC Origin Reference/Source

I 37-07-17 Cuba 617-05 C. neoformans var grubii A MC1 Clinical Illnait Zaragozi et al., 2010 II 44-08-52 Cuba CA 1-5 C. neoformans var grubii A MC1 Environmental Illnait Zaragozi et al., 2010 III 37-07-03 Cuba 24-2b C. neoformans var grubii A MC1 Environmental Illnait Zaragozi et al., 2010 IV 36-10-01 Cuba CH-2 C. neoformans var grubii A MC2 Environmental Illnait Zaragozi et al., 2010 V 44-08-16 Cuba 569-06 C. neoformans var grubii A MC2 Clinical Illnait Zaragozi et al., 2010 VI 36-09-16 Cuba 225-99 C. neoformans var grubii A MC2 Clinical Illnait Zaragozi et al., 2010 VII 36-09-32 Cuba 227-01 C. neoformans var grubii A MC3 Clinical Illnait Zaragozi et al., 2010 VIII 36-09-57 Cuba 0119 C. neoformans var grubii A MC3 Clinical Illnait Zaragozi et al., 2010 IX 36-10-46 Cuba 30-2D C. neoformans var grubii A MC3 Environmental Illnait Zaragozi et al., 2010 X 36-10-56 Cuba 315-01 C. neoformans var grubii A MC4 Clinical Illnait Zaragozi et al., 2010 XI 36-09-53 Cuba 098 C. neoformans var grubii A MC6 Clinical Illnait Zaragozi et al., 2010

doi:10.1371/journal.pone.0055579.t002 ystem, skin and digestive tract, which express receptors for this adequate immune response to C. gattii. These cells reflect the cytokine [17]. IL-22 promotes the production of antimicrobial second line of defense when the yeast enters the host after agents called b-defensins by epithelial cells and serves in mucosal inhalation. Our results showed a less optimal recognition and defenses against pathogens. It is tempting to speculate that IL-22 is initial cytokine induction of C. neoformans var. grubii and var. important for the initial anti-cryptococcal defense because it has a neoformans, which suggests that in a host with inadequate cellular function at the place of entrance of this yeast, namely the epithelial immunity this less optimal innate cytokine response leads more surface of the respiratory system. However, to confirm these easily to infection with C. neoformans var. grubii and var. neoformans speculations, further research should be attempted to identify the compared to C. gattii. Clinical data support this, since infections of role of IL-22 and Th22 cells in clinical patients with cryptococ- immunocompromised hosts with C. neoformans var grubii is far more cosis. prevalent than infection with C. gattii [20]. Higher amounts of the pro-inflammatory cytokines IL-1b, A potential limitation of our study is that heat-killed instead of TNFa, IL-6, IL-17 and IL-22 by human PBMCs were induced by live cryptococci were used. However, at the temperatures used for C. gattii compared to both varieties of C. neoformans, indicating that heat-killing, most virulence factors (capsular polysaccharide, certain (virulence) factors of C. gattii are responsible for a more lipoproteins) are retained. Moreover, in number of previous pronounced inflammatory reaction. This finding is strengthened as studies, heat-killed cryptococci were used and significant inflam- the same trend was seen in the hybrids containing C. gattii as a matory responses specific for capsulated and unencapsulated partner of the mating pair. Therefore we suggest that an cryptococci were found [21,22]. One study investigated lympho- inheritable factor is responsible for the difference in cytokine cyte proliferation after stimulation with live and heat-killed production. cryptococci and found no difference [23]. Thus, we feel that in Our finding that C. gattii induces a more powerful pro- this study, the use of heat-killed crytococci is justified. inflammatory response aimed at more efficient defense against Our experiments using a virulent C. gattii strain in stimulating the pathogen is supported by the work of Ngamskulrungroj et al PBMCs that were pre-incubated with specific PRR-blocking [18]. The authors compared the pathogenesis of the two reagents indicate a role for TLR4 and TLR9 in recognizing Cryptococcus species in mice using an inhalation model and they Cryptococcus and subsequently modulation of the pro-inflammatory found that in naive mice, C. gattii grew significantly slower in blood cytokine response. TLR4 seemed to be involved in mounting a than C. neoformans. Infection with C. gattii was restricted to the pro-inflammatory cytokine response. Previous studies suggest that lungs, while C. neoformans dissimenated to the brain causing glucuronoxylomannan, the major capsular component [15] or meningoencephalitis. When mice were infected intravenously with other cryptococcal cell wall elements [24] are involved in binding 11 low inoculums of yeast, C. neoformans was more virulent than C. to TLR4. In this study we did not design experiments in order to gattii. Apparently, in this murine model the host’s peripheral identify which cell wall components are involved in the initial immune cells are able to clear C. gattii infection more efficiently, cytokine response. Cytokine responses appeared to be independent probably by a more adequate cytokine response. However, in of TLR2 recognition, since blocking of this receptor had no effect humans, C. gattii species seem to be more virulent, as they are able on cytokine concentrations. This contrasted with what is found in to cause disease in apparently immunocompetent hosts. Large- mice by Biondo et al. who demonstrated a key role of TLR2, but scale environmental colonization for C. gattii was found during the not of TLR4 [12]. Other studies, however, found no major role for Vancouver Island outbreak, whereas only relatively few people TLR2 in survival of cryptococcal infections in a murine model developed overt disease [6]. It can be hypothesized that a specific [11,13]. defect in the innate immune system of affected hosts predisposes Based on our results, a special role in Cryptococcus recognition them to infection with C. gattii. Furthermore, other factors such as can be ascribed to TLR9. Unmethylated CpG-rich DNA is the intracellular survival, outgrowth or dissemination may also be best-known ligand for this receptor. Nakamura et al. have shown important for virulence of C. gattii, independent of the initial pro- that TLR9 recognizes cryptococcal DNA [14]. We found that this inflammatory cytokine response [19]. In our experiments we used receptor mediates IL-17 production, without any effect on IL-22. PBMCs of healthy individuals who are expected to have an Conversely, blockade of TLR9 resulted in increased IL-1b

111 Chapter 11

production in response to C. gattii. The latter effect opposes the murine IgG1 k isotype (Biolegend, San Diego, CA, USA) as possible effect of TLR4. However, a specific combination of PRRs control. TLR9 inhibitory oligonucleotides ODN TTAGGG (anti that bind available fungal PAMPs lead to pathways that interact TLR9) [28] and its negative control were obtained from with each other because of a limited set of shared adaptor molecules InvivoGen (San Diego, CA, USA). and transcription factors, and converge to a tailored response [25]. Likely, TLR9 and TLR4 work together in recognizing Cryptococcus Isolation and stimulation of PBMCs and their signaling pathways interact downstream. Interestingly, we Human peripheral blood mononuclear cells (PBMCs) were did not see TLR9 dependent negative modulation of C. neoformans collected from buffy coats of healthy donors after written informed var. grubii, indicating that the TLR9 dependent recognition of consent had been obtained. PBMCs were isolated using density Cryptococcus is species-dependent. Negative modulation of immune gradient centrifugation on Ficoll-Hypaque (GE Healthcare, responses to fungal pathogens mediated by TLR9 have been Uppsala, Sweden). The cells from the interphase were aspirated observed in other studies [26]. As the host’s response to C. gattii relies and washed three times in sterile PBS and resuspended in culture on an initial pro-inflammatory cytokine response more than in C. medium RPMI 1640 Dutch modification (Sigma-Alderich, St neoformans infections, it can be speculated that susceptibility to C. Louis, MO, USA) supplemented with 1% L-glutamine, 1% gattii is influenced by subtle TLR polymorphisms and not necessarily pyruvate and 1% gentamicin. Cells were counted in a Coulter by a defective adaptive immune response. Counter ZH (Beckman Coulter, Fullerton, CA, USA), and adjusted In the present study we investigated the in-vitro cytokine to 56106 cells/ml. Thereafter, they were incubated in a round- production of human PBMCs incubated with 40 different heat- bottom 96-wells plate (volume 200 ml/well) at 37uC and 5% CO2 killed isolates of Cryptococcus neoformans species complex. We with either one of the heat-killed cryptococcal strains (final demonstrated that isolates of C. gattii induce higher concentrations concentration of 107/ml), or heat-killed C. albicans (final concen- of the pro-inflammatory cytokines IL-1b, TNF-a and IL-6 and the tration of 105/mL, which is known to induce substantial amounts Th17/22 cytokines IL-17 and IL-22 compared to C. neoformans var of cytokines) or culture medium alone. After 24 hours or 7 days (in neoformans and C. neoformans var grubii. In addition, we found that the presence of 10% human pool serum) supernatants were clinical C. gattii isolates induced higher amounts of IL-1beta and collected and stored at 220uC until being assayed. IL-6 than environmental isolates. Furthermore, we demonstrated In a subsequent experiment, PBMCs were preincubated for one a likely contribution of TLR4 and TLR9, but no role for TLR2, in hour with inhibitory ligand for TLR4 (Bartonella quintana LPS the host’s cytokine response to C. gattii. In conclusion, clinical C. (200 ng/ml) or culture medium as control, anti-TLR2 or control gattii isolates induced a more pronounced inflammatory cytokine antibody (10 mg/ml), TLR9 inhibitory oligonucleotides and its response compared to other Cryptococcus species and non-clinical C. negative control (25 mg/ml). After preincubation, C. gattii B5742, gattii that is dependent on TLR4 and TLR9 as cellular receptors. isolate 27 in the previous experiment, or specific TLR ligands were added, such as Pam3cys or E.coli LPS (10 mg/ml and 10 ng/ml Materials and Methods respectively] and PBMCs were incubated as described. Cryptococcal strains Forty cryptococcal isolates from the CBS Fungal Biodiversity Cytokine assays Centre (Utrecht, the Netherlands) were used in this study. These Tumor necrosis factor-a (TNF-a), Interleukin-1b (IL-1b), IL-6 isolates were obtained from laboratory, clinical, environmental and IL-1 receptor antagonist (IL-1Ra) concentrations were deter- and veterinary sources. A detailed overview of the origin, sero- and mined from the culture supernatant after 24 hours of incubation AFLP genotype of these isolates is provided in Table 1. Twenty- using commercially available ELISA kits (TNF-a, IL-1b and IL- three isolates were identified as C. gattii, 5 C. neoformans var. 1Ra: R&D systems, Minneapolis, MN, USA. IL-6: Sanquin,Am- neoformans,5C. neoformans var. grubii and 7 hybrids, 3 of which were sterdam, the Netherlands) according to the manufacturer’s instruc- interspecies hybrids between C. gattii and C. neoformans var. tions. T-cell derived cytokines IL-17 and IL-22 concentrations were neoformans and 4 hybrids between both C. neoformans varieties. In determined in the supernatant after 7 days of incubation using addition, 11 Cuban isolates were used in separate experiments, all ELISA kits (R&D systems). Lower detection limits were 78 pg/ml, identified as C. neoformans var grubii (Table 2). 39 pg/ml, 15 pg/ml, 200 pg/ml, 40 pg/ml and 78 pg/ml for Prior to the experiments, the strains were freshly grown on TNF-a, IL-1b, IL-6, IL-1Ra, IL-17 and IL-22 respectively. Sabouraud dextrose agar plates. A suspension of each strain was prepared in sterile phosphate buffered saline (PBS), heat-killed Ethics statement overnight at 56uC and quantified by spectrophotometry at a Written informed consent of healthy donors was provided. The wavelength of 530 nm. The suspensions were checked for fungal study was approved by the Medical Ethical Committee Arnhem- and bacterial growth on a Sabouraud dextrose agar plate and a Nijmegen in the Netherlands. blood agar plate respectively. No growth was observed after 5 days. All strains were stored at 4uC until used. Statistical analysis Results from at least three different experiments with a range of Candida strain 5–7 donors were pooled and analyzed using GraphPad Prism 5 Heat-killed Candida albicans ATCC MYA-3573 (UC 820), a well software (GraphPad, San Diego, CA). Data are given as mean 6 described clinical isolate, suspended in sterile PBS, was used as a SE. The Mann-Whitney U-test for unpaired, nonparametrical positive control. data was used to compare differences in cytokine production between two groups. The Kruskal-Wallis test with Dunn’s multiple Reagents and antibodies comparison test was used when more than two groups were Bartonella LPS, a penta-acylated LPS which is an antagonist of compared. The Wilcoxon matched-pairs signed rank test was used TLR4-dependent signaling, was obtained as previously described to analyze differences in cytokine production between inhibitors [27]. An anti-TLR2 monoclonal antibody from eBioscience (San and their controls in the inhibition experiments. The level of Diego, CA, USA) was used, and an irrelevant isotype-matched significance was set at p,0.05.

112 Cryptococcus gattii induces a cytokine pattern that is distinct from other cryptococcal species

Figure S1

Supporting Information Acknowledgments Figure S1 IL-1b induction by Pam3cys and E. coli LPS The authors thank Ferry Hagen for providing cryptococcal strains. after blocking of TLR2 and TLR4 respectively. IL-1b production by human PBMCs is shown (A) induced by pam3cys Author Contributions [10 mg/ml] after preincubated for one hour with anti-TLR2 or Conceived and designed the experiments: T. Schoffelen LABJ MGN JFM control antibody [10 mg/ml] and (B) by E. coli LPS [10 ng/ml] T. Sprong. Performed the experiments: T. Schoffelen LABJ MGN T. after preincubation for one hour with TLR4 antagonist Bartonella Sprong. Analyzed the data: T. Schoffelen LABJ MGN JFM T. Sprong. quintana LPS [200 ng/ml] or culture medium. Mean values Contributed reagents/materials/analysis tools: MTIZ TB JFM. Wrote the (n = 10) 6 SE of five independent experiments are presented. paper: T. Schoffelen MTIZ LABJ MGN TB JFM T. Sprong. (TIF)

References 1. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, et al. (2009) 16. Trifari S, Kaplan CD, Tran EH, Crellin NK, Spits H (2009) Identification of a Estimation of the current global burden of cryptococcal meningitis among human helper T cell population that has abundant production of interleukin 22 persons living with HIV/AIDS. AIDS 23: 525–530. and is distinct from T(H)-17, T(H)1 and T(H)2 cells. Nat Immunol 10: 864–871. 2. Speed B, Dunt D (1995) Clinical and host differences between infections with the 17. Eyerich S, Eyerich K, Pennino D, Carbone T, Nasorri F, et al. (2009) Th22 cells two varieties of Cryptococcus neoformans. Clin Infect Dis 21: 28–34; discussion 35- represent a distinct human T cell subset involved in epidermal immunity and 26. remodeling. J Clin Invest 119: 3573–3585. 3. Chayakulkeeree M, Perfect JR (2006) Cryptococcosis. Infect Dis Clin North Am 18. Ngamskulrungroj P, Chang Y, Sionov E, Kwon-Chung KJ (2012) The primary 20: 507–544, v–vi. target organ of Cryptococcus gattii is different from that of Cryptococcus neoformans in a 4. Byrnes EJ, Li W, Lewit Y, Ma H, Voelz K, et al. (2010) Emergence and murine model. MBio 3. pathogenicity of highly virulent Cryptococcus gattii genotypes in the northwest 19. Ma H, Hagen F, Stekel DJ, Johnston SA, Sionov E, et al. (2009) The fatal fungal United States. PLoS Pathog 6: e1000850. outbreak on Vancouver Island is characterized by enhanced intracellular 5. Chaturvedi V, Chaturvedi S (2011) Cryptococcus gattii: a resurgent fungal parasitism driven by mitochondrial regulation. Proc Natl Acad Sci U S A 106: pathogen. Trends Microbiol 19: 564–571. 12980–12985. 6. Kidd SE, Hagen F, Tscharke RL, Huynh M, Bartlett KH, et al. (2004) A rare 20. Brown GD, Netea MG, SpringerLink (Online service) (2007) Immunology of genotype of Cryptococcus gattii caused the cryptococcosis outbreak on Vancouver Fungal Infections. Dordrecht: Springer. Island (British Columbia, Canada). Proc Natl Acad Sci U S A 101: 17258– 21. Levitz SM, Farrell TP, Maziarz RT (1991) Killing of Cryptococcus neoformans by 17263. human peripheral blood mononuclear cells stimulated in culture. J Infect Dis 7. Eyerich S, Eyerich K, Cavani A, Schmidt-Weber C (2010) IL-17 and IL-22: 163: 1108–1113. siblings, not twins. Trends Immunol 31: 354–361. 22. Siddiqui AA, Shattock RJ, Harrison TS (2006) Role of capsule and interleukin-6 8. van de Veerdonk F, Gresnigt M, Kullberg B, van der Meer J, Joosten L, et al. in long-term immune control of Cryptococcus neoformans infection by specifically (2009) Th17 responses and host defense against microorganisms: an overview. activated human peripheral blood mononuclear cells. Infect Immun 74: 5302– BMB Rep 42: 776–787. 5310. 9. van de Veerdonk FL, Kullberg BJ, Netea MG (2010) Pathogenesis of invasive 23. Mody CH, Syme RM (1993) Effect of polysaccharide capsule and methods of candidiasis. Curr Opin Crit Care 16: 453–459. preparation on human lymphocyte proliferation in response to Cryptococcus 10. Eyerich K, Foerster S, Rombold S, Seidl HP, Behrendt H, et al. (2008) Patients neoformans. Infect Immun 61: 464–469. with chronic mucocutaneous candidiasis exhibit reduced production of Th17- associated cytokines IL-17 and IL-22. J Invest Dermatol 128: 2640–2645. 24. Doering TL (2009) How sweet it is! Cell wall biogenesis and polysaccharide 11 11. Yauch LE, Mansour MK, Shoham S, Rottman JB, Levitz SM (2004) capsule formation in Cryptococcus neoformans. Annu Rev Microbiol 63: 223–247. Involvement of CD14, toll-like receptors 2 and 4, and MyD88 in the host 25. van de Veerdonk F, Kullberg B, van der Meer J, Gow N, Netea M (2008) Host- response to the fungal pathogen Cryptococcus neoformans in vivo. Infect Immun 72: microbe interactions: innate pattern recognition of fungal pathogens. Curr Opin 5373–5382. Microbiol 11: 305–312. 12. Biondo C, Midiri A, Messina L, Tomasello F, Garufi G, et al. (2005) MyD88 26. Kasperkovitz PV, Khan NS, Tam JM, Mansour MK, Davids PJ, et al. (2011) and TLR2, but not TLR4, are required for host defense against Cryptococcus Toll-like receptor 9 modulates macrophage antifungal effector function during neoformans. Eur J Immunol 35: 870–878. innate recognition of Candida albicans and Saccharomyces cerevisiae. Infect Immun 13. Nakamura K, Miyagi K, Koguchi Y, Kinjo Y, Uezu K, et al. (2006) Limited 79: 4858–4867. contribution of Toll-like receptor 2 and 4 to the host response to a fungal 27. Popa C, Abdollahi-Roodsaz S, Joosten LA, Takahashi N, Sprong T, et al. (2007) infectious pathogen, Cryptococcus neoformans. FEMS Immunol Med Microbiol 47: Bartonella quintana lipopolysaccharide is a natural antagonist of Toll-like receptor 148–154. 4. Infect Immun 75: 4831–4837. 14. Nakamura K, Miyazato A, Xiao G, Hatta M, Inden K, et al. (2008) 28. Stunz LL, Lenert P, Peckham D, Yi AK, Haxhinasto S, et al. (2002) Inhibitory Deoxynucleic acids from Cryptococcus neoformans activate myeloid dendritic cells oligonucleotides specifically block effects of stimulatory CpG oligonucleotides in via a TLR9-dependent pathway. J Immunol 180: 4067–4074. B cells. Eur J Immunol 32: 1212–1222. 15. Shoham S, Huang C, Chen JM, Golenbock DT, Levitz SM (2001) Toll-like 29. Illnait-Zaragozi MT, Martıńez-Machıń GF, Fernańdez-Andreu CM, Boekhout receptor 4 mediates intracellular signaling without TNF-alpha release in T, Meis JF, et al. (2010) Microsatellite typing of clinical and environmental response to Cryptococcus neoformans polysaccharide capsule. J Immunol 166: Cryptococcus neoformans var. grubii isolates from Cuba shows multiple genetic 4620–4626. lineages. PLoS One 5: e9124.

113

Chapter General discussion and 12 future perspectives Chapter 12

Currently the genus Cryptococcus includes about 100 species but the C. neoformans/C. gattii species complex is most prominently associated with human disease (Fonseca et al., 2011). The expanding knowledge of these yeasts has allowed to identify the existence of two species comprising seven serotypes and different genetic groups:C. neoformans has been segregated into C. neoformans var. grubii (serotype A; AFLP1/VNI, AFLP1A/ VNB/VNII, and AFLP1B/VNII), C. neoformans var. neoformans (serotype D; AFLP2/ VNIV) and the intervarietal serotype C. neoformans AD (AFLP3/VNIII); the other species, C. gattii, has been divided into five distinct genotypes (AFLP4/VGI, AFLP5/ VGIII, AFLP6/VGII, AFLP7/VGIV, and AFLP10/VGIV). While serotype B strains are found in each of the five C. gattii AFLP genotypes, serotype C strains are restricted to genotypes AFLP5/VGIII and AFLP7/VGIV. It was found that C. gattii and C. neoformans can also form interspecies hybrids: genotype AFLP8 (C. neoformans var. neoformans AFLP2/VNIII serotype D × C. gattii AFLP4/VGI serotype B) and AFLP9 (C. neoformans var. grubii AFLP1/VNI serotype A × C. gattii AFLP4/VGI serotype B) (Bovers et al., 2006; Bovers et al., 2008a; Bovers et al., 2008b; Meyer et al., 2009; Ngamskulrungroj et al., 2009; Springer and Chaturvedi, 2010). These aspects, although basic, are vital for understanding the disease.

The natural history of infection due to the C. neoformans/C. gattii species complex has been well documented. Although there are occasional reports of cases that took up to 22 years of evolution, it leads inexorably to death if appropriate and timely treatment is not administrated. Once symptoms emerge, 86% of the patients will die during the first year of which 70% during the first three months even when properly treated (Casadevall and Perfect, 1998, Friedman et al., 2005). These alarming statistics about the poor prognosis of the disease become even more worrisome with the knowledge that the estimated annual incidence exceeds one million cases (Armstrong et al., 2014; Park et al., 2009).

This background was enough motivation to write this thesis, compilation of eleven publications grouped in three main parts. This last chapter contains a comprehensive view of the work included in the thesis: 1. molecular characterization of clinical and environmental C. neoformans species complex isolates from Cuba and The Netherlands, 2. the antifungal susceptibility of Cuban and worldwide cryptococcal strains, 3. the analysis of C. neoformans/C. gattii species complex isolated from humans, animal and the environment and 4. attempts to elucidate some pathogenic differences between C. neoformans and C. gattii.

Molecular characterization of clinical and environmental C. neoformans species complex isolated from Cuba and the Netherlands So far, attempts to characterize Cuban Cryptococcus isolates were restricted to conventional and serological typing studies by means of the commercial kit Crypto Check (Iatron Laboratories, Tokyo, Japan) in a limited number of specimens. The collaborative work between Instituto Pedro Kouri (IPK), Canisius Wilhelmina Ziekenhuis (CWZ) and Centraalbureau voor Schimmelcultures (CBS) allowed a step

116 General discussion and future directions forward since all specimens of this genus in the collection at the IPK in Havana were re- identified and their genotypic patterns were defined using high resolution techniques. These criteria were used for comparative analysis of isolates according to their origin (clinical versus environmental) and to assess the isolates of patients with recurrent neurocryptococcosis. The re-identification by AFLP and sequencing D1/D2 and ITS1/5.8S/ITS2 fragments allowed confirmation of the dominance of C. neoformans var. grubii in Cuba. The mating-type of Cuban C. neoformans var. grubii isolates was determined for the first time and results showed that all isolates were MATα (Illnait- Zaragozí et al., 2010). This result coincides with the current international literature.

It is generally recognized that the C. neoformans/C. gattii species complex can exist as both the yeast and filamentous phase without being regarded as a dimorphic pathogen. Hyphae are not considered essential for virulence. The role of these infectious particles remains unclear; they are either the dried yeasts or the sexual basidiospores. However, the mating-type analysis in these organisms is important since it has been shown that MATα cells, present in both clinical and environmental samples in a greater proportion, are associated with increased virulence and are prone to spread to the CNS. Molecular analyzes indicate that genes involved in sexual reproduction of C. neoformans/C. gattii species complex are closely related to those encoding for virulence markers (capsular polysaccharide, melanin synthesis, thermotolerance, etc) (Litvintseva and Mitchell, 2009; Lin, 2009; Metin et al., 2010; del Poeta and Casadevall, 2011).

AFLP has been well standardized (Boekhout et al., 2001; Meyer et al., 2009). It was used for the initial characterization of all isolates (190 from Cuba and 300 from the Netherlands) and as reference in the evaluation of a set of nine microsatellite markers selected from the nucleotide sequence of the type strain H99. The analysis of Cuban clinical and environmental C. neoformans var. grubii suggests the existence of alternative sources of infection other than pigeon droppings. Clinical Cuban C. neoformans complex isolates showed a high genetic diversity in 19 of the 20 microsatellite clusters. In contrast, specimens obtained from pigeons droppings showed a low genetic variability (80% of them gathered in two of the 20 clusters) (Illnait-Zaragozí et al., 2010). There are few publications related to this topic. They generally show a close relationship between clinical and environmental genotypes (Boekhout et al., 2001; Trilles et al., 2003). Findings of other investigators are in line with our conclusions (Zhu et al., 2010).

Although studies suggests that serotype D is the most prevalent in Europe (Viviani et al., 2006), the majority of the Dutch studied isolates were serotype A and MATα, just 12 as in Cuba, and microsatellite typing revealed a high genetic diversity among isolates belonging to this serotype. The genetic diversity of the Dutch clinical C. neoformans var. grubii population was even slightly higher than that observed in Cuba with 196 genotypes among 259 studied isolates. This result could be due the time span of the study was 30 years (which is longer than the 20-year time frame of the Cuban study). The composition of the Dutch population, characterized by extensive immigration, which includes people from the former overseas colonies of Surinam, Netherlands Antilles,

117 Chapter 12 and Indonesia, as well as immigrants from northern Africa and eastern Europe, may be another explanation for these surprising results (Hagen et al., 2012 (a)).

When multiple isolates (sampled on the same day but from different clinical specimens or at different time points) from 38 Dutch and 7 Cuban patients were analyzed, microsatellite profiles of most of them were found to be identical. However, some Dutch patients and 70% of the Cuban patients were found to be infected by multiple C. neoformans isolates including interspecies hybrid isolates of C. neoformans var. neoformans and C. gattii (genotype AFLP8) (Illnait-Zaragozí et al., 2010; Hagen et al., 2012 (a)). These findings confirm the broad genetic variability among clinical isolates from both countries and urges to hypothesize about the possible mechanism of recurrence of this fungal pathogen due to multiple genotypes. Acquisition of several genotypes at the initial infection or during the course of the disease. Previous studies have shown the ubiquity of the C. neoformans/C. gattii species complex that may coexist in the same ecological niche and, in turn, may result in simultaneous infection by different isolates (Lazera et al., 2000; Mandal et al., 2005; Kidd et al., 2007). Another mechanism that can explain this observation is a possible microevolution process since it is known that microorganisms constantly adjust and modulate their behavior in order to survive in different environments, thereby generating diversity in clonal populations. All these aspects together may stimulate the selection of certain variants which favors chronic or recurrent infection (Jain and Fries, 2008).

The use of selected markers for microsatellite typing has a high discriminatory power and serves as a technique of genetic characterization (de Valk et al., 2009; Illnait-Zaragozí et al., 2010; Hagen et al., 2012 (a)). High resolution technologies such as MLST, are currently available and have been proposed as reference method in genotypic characterization of the C. neoformans/C. gattii complex. However, its complex implementation that requires the sequencing of at least seven loci, makes it impractical for most laboratories (Meyer et al., 2009).

In conclusion we demonstrated for the first time that the prevalence of C. neoformans var. grubii MATα isolates in Cuba and its wide distribution in both clinical and environmental samples. The present studies provide new insights into the pathogenesis and epidemiology of cryptococcal disease and its causative agent. The genetic segregation between clinical and environmental isolates suggests the presence of additional hitherto unstudied niches of C. neoformans var. grubii. It can also be used to identify the recurrence or persistence of a cryptococcal infection caused by co- infection with different genotypes. In this sense, the panel of selected markers showed a high discriminatory power and the microsatellite typing revealed its superiority as genetic characterization technique with respect to AFLP for molecular epidemiology studies for C. neoformans var. grubii. To expand these studies the development of in vitro and in vivo pathogenicity experiments will contribute significantly to corroborate the results.

118 General discussion and future directions

Antifungal susceptibility of Cuban and worldwide cryptococcal strains During the last decades the incidence of fungal infections has raised, which has favored the frequent use of systemic antifungals and the occurrence of isolates with decreased susceptibility or even resistance to antifungal drugs (Vandenbossche et al., 2002; Cuenca-Estrella, 2010). Most of the studied C. neoformans species complex strains from Cuba and the Nederlands showed quite uniform patterns of susceptibility to the tested antifungal agents. Both species, regardless to their origin, were susceptible to amphotericin B, itraconazole, voriconazole, posaconazole, and isavuconazole. When all of the isolates were considered overall, the lowest MICs ranges were found for isavuconazole, voriconazole, and posaconazole while the widest ranges and highest MICs were seen for flucytosine and fluconazole (Illnait-Zaragozí et al., 2008; Illnait- Zaragozí et al., 2010; Hagen et al., 2012 (a)).

It is becoming evident that differences exist between genetic groups among C. neoformans and C. gattii species. The current study shows that there is a trend toward lower MIC values of amphotericin B for C. neoformans genotype AFLP2 isolates, similarly to that observed in Croatian and Spanish studies (Guinea et al., 2010; Mlinaric-Missoni et al., 2011). On the other hand, C. gattii strains belonging to AFLP6 had significantly (P ≤ 0.05) higher geometric mean MICs than strains belonging to genotype AFLP4 for six of the antifungal compounds tested. However, these genotypes had almost identical geometric mean MICs of posaconazole. These results are comparable with those for a recently reported collection of 43 North American C. gattii isolates (Iqbal et al., 2010).

The development of new antifungal drugs, such as the novel triazoles and the reported increase in antifungal resistance, highlight the importance of monitoring the susceptibility profiles of clinical isolates. Most of the studied C. neoformans and C. gattii isolates, cultured either at the same time from different body sites or at different time points generally showed the same pattern of MIC values and were not resistant for the seven antifungal compounds tested. However, in serial isolates from two Cuban and two Dutch patients, MICs values of triazoles increased 4-5 log2 dilutions over time. Failures could be attributed to the excessive use of this agent, as well as to the use of suboptimal doses for long periods of time, especially among AIDS patients (Govender et al., 2011).

Its excellent pharmacokinetic and toxicity profile were the bases for the extensive use of fluconazole in some groups of patients. This has undoubtedly resulted in an increase of fluconazole-resistant strains of C. neoformans/C. gattii species complex as 12 illustrated by case reports on therapeutic failures in AIDS patients who undergo long- term maintenance therapy (Pfaller et al., 2004; Cuenca-Estrella et al., 2010). In line with previous studies, the current work corroborated that fluconazole and flucytosine have the lowest levels of activity against C. neoformans and C. gattii isolates. Among the studied isolates, environmental isolates seem to be less susceptible to fluconazole and flucytosine than the clinical ones, in harmony with previous reports (Brandt et al., 2001; Pfaller et al., 2004; Souza et al., 2005). Isolates obtained from HIV-negative patients

119 Chapter 12

showed lower MIC90s compared to those from AIDS patients (Illnait-Zaragozí et al., 2008; Illnait-Zaragozí et al., 2010; Hagen et al., 2012 (a)).

Azole resistance has been primarily reported in patients with systemic mycosis who received long-term azole therapy. Microorganisms may develop resistance through three major pathways: (i) reduction of the accumulation of the drug within the fungal cell, (ii) decrease of the affinity of the drug to its target, and (iii) change in the metabolism to counterbalance the drug effect (Cuenca-Estrella, 2010; Vandeputte et al., 2012). In some highly resistant isolates, several mechanisms appear to act simultaneously. The increasing resistance during antifungal treatment has been attributed to the sequential acquisition of different mechanisms (Cannon et al., 2009). It is also known that some isolates may carry heteroresistance, especially against fluconazole; particular subpopulations within each colony display a greater tolerance to high concentrations of antifungal. This event is independent of prior exposure to antifungals and has been described for both species of the C. neoformans/C. gattii complex (Sionov et al., 2009; Varma and Kwon-Chung, 2010).

Additionally, for bacteria it is known that significant risks are involved in the environmental route of development of resistance. In fungi, the relation between the use of antimicrobial agents outside human medicine and the development of resistance to clinically used compounds has been shown. Similarly, the widespread use of azole- based fungicides, herbicides, and plant growth regulators is a potential risk factor for the development of resistance in human medical practice (Seifert et al., 2007). Although the hypothesis of a fungicide-driven route of azole resistance development in fungi is controversial, there is some evidence that such a route may exist. Therefore, regular surveillance of cultures as well as in vitro susceptibility testing of human isolates is strongly recommended (Snelders et al., 2012).

Each of the different classes of antifungals has limitations in comparison with an ideal agent. The polyenes are limited by their toxicity; echinocandins are only available for intravenous therapy and lack activity against Cryptococcus as well as endemic mycoses, and azoles are limited by their drug-drug interactions and development of resistance. These limitations promote searching to identify new targets of antifungal drugs (Cuenca-Estrella et al., 2010). Several published antifungal susceptibility studies, including this thesis, indicate that the newer azoles have excellent potency against each isolate and all AFLP genotypes, inclusive of those with reduced fluconazole susceptibilities. Both voriconazole and posaconazole show strong antifungal activity against all Cryptococcus strains, while isavuconazole displays even lower MIC values (Illnait-Zaragozí et al., 2008; Illnait-Zaragozí et al., 2010; Hagen et al., 2012). The latter, a water-soluble triazole suitable for oral and intravenous administration, has shown in vitro a broad spectrum of activity against all major opportunistic and true pathogenic fungi. Its pharmacokinetic and pharmacodynamic (PK/PD) characteristics, are slow elimination, low plasma clearance, and extensive tissue distribution (Odds et al., 2006; Seifert et al., 2007; Schmitt-Hoffmann et al., 2006(a); Schmitt-Hoffmann et al., 2006(b)),

120 General discussion and future directions which makes isavuconazole a promising candidate to replace the current drugs, given their potential side effects.

The main goal of susceptibility testing is to predict with some reliability the clinical outcome when an infected patient is treated with the specific agent evaluated. This in vitro susceptibility is expressed as the clinical breakpoints (CBPs). Such values are based on the correlation of in vitro data with clinical and microbiological outcomes in order to differentiate between a treatable or non-treatable organism (Dalhoff et al., 2009). So far, a major concern has been the lack of CBPs of any antifungal agent for both C. neoformans and C. gattii infections due to the paucity of data on PK/PD and clinical outcomes in comparison with measured MICs. Species identification and antifungal susceptibility testing emphasize the utility of wild type (WT) cutoffs as a practical tool to detect triazole resistance among Cryptococcus isolates. Recently the epidemiologic cutoff values (ECVs) of amphotericin B (0.5 to 1 µg/ml) and flucytosine (4 to 16 µg/ml) versus both species were established (Espinel-Ingroff, et al., 2012 (a)). With these data it has become possible to relate the ECVs of these agents to those of fluconazole (8 to 32 µg/ml), itraconazole (0.25 to 1 µg/ml), posaconazole (0.25 to 0.5 µg/ml) and voriconazole (0.12 to 0.25 µg/ml). Proposed ECVs for the C. neoformans/C. gattii species complex appear to be species- and molecular type-specific, but these differences are mostly within one dilution. Further investigations should determine the relationship between molecular mechanisms of triazole resistance and the proposed non-WT values (Espinel-Ingroff, et al., 2012 (b)).

In conclusion, this thesis offers the largest study of antifungal susceptibility on Cuban isolates where the MIC values of seven antifungal agents used in first-line treatment were compared with a substantial representation of isolates collected over 25 years. Data obtained from the Cuban strains were included in the study of a large collection of C. gattii which helped to establish ECVs for this species complex against the major triazoles. These promising results still need to be correlated with clinical outcome. Continuous surveillance should corroborate and expand the information provided in the present study.

Analysis of C. neoformans/C. gattii species complex isolated from humans, animal and the environment in Cuba and pathogenic differences between C. neoformans and C. gattii isolates In Cuba, cryptococcosis was first reported in the early 1950s (Curbelo, 1951). Since then sporadic cases were associated with alcoholism, organ transplants and immunological 12 disorders. In 1986 with the beginning of the AIDS epidemic in Cuba, the amount of individuals suffering from this mycosis had increased with an average of eight new cases each year (Illnait-Zaragozi et al., 2010). This still might be an underrepresentation since a retrospective autopsy study of HIV/AIDS patients performed during a 10-year period showed that disseminated or central nervous system cryptococcosis was a serious disorder in 29% of the cases (Arteaga et al., 1998). These results concur with those reported by Cox et al. who conducted a six month prospective study to describe

121 Chapter 12 and compare the clinical and autopsy-verified causes of death in hospitalized HIV- infected and HIV-uninfected patients in Uganda. The study showed that cryptococcal infection was after tuberculosis the most frequent cause of death reaching more than 20% among the 53 autopsies performed (Cox et al., 2012).

As shown before, epidemiological studies on clinical- and pigeon droppings isolates, carried out in different regions in Cuba, revealed that most of were C. neoformans var. grubii. It was also established that only few microsatellite genotypes from environmental isolates were linked to genotypes from human isolates, suggesting that some clinical isolates may originate from additional ecological niche(s) (Illnait-Zaragozi et al., 2010). At present, it has become clear that Cryptococcus isolates, in addition to bird excreta (Kielstein et al., 2000) are present on trees and cacti (Chowdhary et al., 2012; Hagen et al., 2010 (b)). Hence, a search for the occurrence of this yeast in relation to different species of trees and cacti in Havana, Cuba was carried out.

A total of 662 samples were collected from 331 trees and cacti. Initial selection of the isolates was done by conventional techniques. They were further characterized using a combination of AFLP and DNA sequence analysis. Identification by conventional methods yielded 121 C. neoformans and 61 C. gattii isolates while molecular analyses showed that none of these isolates was C. gattii and only one proved to be C. neoformans var. grubii MATα. A total of 27 different other species were identified, and the most prevalent being C. heveanensis (33%). Moreover, 65 unidentifiable isolates were segregated into ten potentially novel species (Illnait-Zaragozí et al., 2012). This finding is in strong contrast with India, for example, where C. gattii can be isolated from up to 50% of investigated trees (Chowdhary et al., 2013).

Analyzing the structure of the environmental population of C. neoformans and C. gattii can significantly increase our knowledge of their ecology, evolution and epidemiology (Hagen et al., 2010 (b)). However, the low prevalence of this yeast found on the studied plants does not allow a valid epidemiological evaluation and environmental niches responsible for human cryptococcal infections in Cuba remain to be identified. Additionally, these results suggest that conventional methods have a low specificity for the C. neoformans/C. gattii species complex and molecular analyses need to be applied to confirm identification of isolates from environmental sources.

Although there was no evidence for the environmental presence of C. gattii in Cuba, we included two reports of fatal infections due to this species, one from a cheetah at the National Zoo in Havana and one from a HIV-negative patient. The animal was imported from South Africa and the patient had a history of travel to Central America. Both died due to cryptococcal infection after 16 months and 2 years, respectively (Polo et al., 2010; Illnait-Zaragozi et al., 2011; Illnait-Zaragozi et al., 2013).

There are three main arguments to support the hypothesis that both, animal and human patient, had a dormant infection: i) there was no evidence suggesting environmental

122 General discussion and future directions presence of this species in Cuba despite multiple attempts to recover C. gattii from trees and cacti (Illnait-Zaragozí et al., 2012); ii) it has been demonstrated that the time interval between exposure to C. gattii and the development of the disease is highly variable being either an acute infection or reactivation of a longstanding subclinical infection depending on the immunological status of the patient (Hagen et al., 2010 (a); Hagen et al., 2012 (b)); iii) both had a history of travel to C. gattii endemic regions and AFLP and the bootstrap Maximum Likelihood phylogenetic analysis after sequencing of the molecular analysis (Hagen et al., 2012 (b)) showed that the C. gattii strain from the cheetah was genotype AFLP4, nearly identical to a strain isolated from an unknown source in Kenya, and the C. gattii strain from the patient was genotype AFLP5 also closely related with two clinical and two environmental strains from Colombia (Illnait- Zaragozi et al., 2011; Illnait-Zaragozi et al., 2013). Although genotype AFLP5 was originally reported from India, the current literature suggests that it has spread throughout Latin- American countries (Argentina, Brazil, Colombia, Guatemala Mexico, and Venezuela), the west coast of the U.S.A., as well as over Asia and Australia (Chowdhary et al., 2013).

Infections due to C. gattii genotypes AFLP4 and AFLP6 usually occur in patients without detectable predisposing factors while genotypes AFLP5, AFLP7 and AFLP10 appear to be more associated with an impaired immune system, similar to C. neoformans (Hagen et al., 2010 (a); Hagen et al., 2012 (b)). Interestingly no underlying diseases, previously associated with C. gattii genotype AFLP5 infections, could be demonstrated in our patient.

Cryptococcus gattii infection usually requires longer treatment, is associated with a poor therapeutic response, development of neurological complications and high mortality (Kozubowski and Heitman, 2011; Kwon-Chung et al ., 2011, Marr et al., 2011). The human patient passed away one year after the diagnosis was made despite a prolonged treatment with amphotericin B, fluconazole, voriconazole and gamma interferon even when the isolate did not show in vitro resistance to these antifungal agents (Illnait-Zaragozí et al., 2013). The animal died before any diagnosis could be made (Polo et al., 2010).

Until a decade ago, cryptococcosis caused by the primary pathogenic yeast C. gattii was a rarely encountered infection outside tropical and subtropical regions (Springer and Chaturvedi, 2010). However, this has changed due to an unprecedented outbreak that emerged in the temperate climate of Vancouver Island (British Columbia, Canada) and subsequently expanded further into the Pacific Northwest (Byrnes et al., 2009). 12 Even though C. neoformans and C. gattii share most of the properties that confers pathogenicity, the lather species seems to be more virulent, as it has been held responsible for disease in apparently immunocompetent hosts (Illnait-Zaragozí et al., 2013; Hagen et al., 2010 (a); Hagen et al., 2012 (b)). Although significant environmental colonization by C. gattii was found during the Vancouver Island outbreak, relatively few people developed overt disease (Kidd et al., 2007). This led to the hypothesis that a specific defect in the innate immune system predisposed the patients to infection

123 Chapter 12 with this agent. In addition, other factors such as intracellular survival, outgrowth or dissemination may play an important role in their virulence (Ma et al., 2009).

Our study on the in vitro cytokine production of human PBMCs demonstrated that i) Cuban clinical and environmental C. neoformans var. grubii isolates did not show significant difference in cytokine induction; ii) C. gattii induce higher concentrations of the pro-inflammatory cytokines (IL-1β, TNF-α and IL-6 and the Th17/22 cytokines IL-17 and IL-22) compared to C. neoformans var. neoformans and C. neoformans var. grubii; iii) clinical C. gattii isolates induced higher amounts of IL-1β and IL-6 than environmental isolates; iiii) a contribution of TLR4 and TLR9, but not of TLR2 was observed in the ex-vivo cytokine response to C. gattii (Stoffelen et al., 2013).

The current study was not designed to identify which components of the cell wall are involved in the initial cytokine response. This reaction appeared to be independent of TLR2 recognition, since blocking of this receptor had no effect on cytokine concentrations (Stoffelen et al., 2013). This contrasted with what Biondo et al. found in mice; they demonstrated a key role of TLR2, but not of TLR4 (Biondo et al., 2005). Conversely, other studies found no major role of TLR2 in survival of cryptococcal infections in a murine model (Yauch et al., 2004; Nakamura et al., 2006).

To summarize it was not possible to demonstrate the presence of local isolates of C. gattii in Cuba. Attempts to recover this species from environmental samples have been fruitless and isolates from clinical cases (human and animal) seem to originate from dormant infections imported from endemic regions. These findings agree with evidence points at the relationship between development of infection and a visit to areas with higher C. gattii prevalence. A comprehensive prospective study in Cuba is necessary to get an estimation of the environmental prevalence of C. neoformans and C. gattii, not only in decayed wood but also in other natural substrates. Finally, knowledge of the host’s immune response to different Cryptococcus spp. can provide further insight that will lead to a better understanding of the pathogenesis of cryptococcosis and its management.

124 General discussion and future directions

Final remarks It is widely accepted that fungal pathogens have an enormous influence on plant and animal life. Recent reports described the extraordinary and frightening impact of these agents on species extinction, food security, and ecosystem disturbances. In contrast, the effect that fungal infections have on human health is not widely recognized and deaths resulting from these infections are often overlooked. During their life most people will suffer from a superficial fungal infection. Such infections are generally easy to cure, but millions of individuals worldwide will contract life-threatening invasive infections that are much harder to diagnose and to treat. Of particular concern is the high mortality rate associated with such infections. It often exceeds 50% despite the availability of several antifungal drugs (Brandt and Park, 2013; Brown et al., 2012; Frey- Klett et al., 2011; Wang et al., 2011). Cryptococcal infections are not an exception in this context. Our studies, on molecular, clinical-epidemiological and in vitro susceptibility aspects in relation with the C. neoformans/C. gattii species, have been conducted to improve our understanding of the disease and its causative agent.

125 Chapter 12

References

Armstrong-JD, Meintjes G, Brown GD. A neglected epidemic: fungal infections in HIV/AIDS. Trends Microbiol 2014; 22: 120-7.

Arteaga E, Capó De Paz V, Pérez-Fernández-Terán ML. Micosis oportunistas invasivas en el sida. Un estudio de 211 autopsias. Rev Iberoam Micol 1998; 15: 33-5.

Biondo C, Midiri A, Messina L, Tomasello F, Garufi G, Catania MR, Bombaci M, Beninati C, Teti G, Mancuso G. MyD88 and TLR2, but not TLR4, are required for host defense against Cryptococcus neoformans. Eur J Immunol 2005; 35: 870-8.

Boekhout T, Theelen B, Diaz M, Fell JW, Hop WC, Dromer F, Meyer W. Hybrid genotypes in the pathogenic yeast Cryptococcus neoformans. Microbiology 2001; 147: 891-907.

Bovers M, Hagen F, Boekhout T. Diversity of the Cryptococcus neoformans-Cryptococcus gattii species complex. Rev Iberoam Micol 2008; 25: S4-12. (a)

Bovers M, Hagen F, Kuramae EE, Diaz MR, Spanjaard L, Dromer F, Hoogveld HL, Boekhout T. Unique hybrids between the fungal pathogens Cryptococcus neoformans and Cryptococcus gattii. FEMS Yeast Res 2006; 6: 599-607.

Brandt ME, Park BJ. Think fungus–Prevention and control of fungal Infections. Emerg Infect Dis 2014; 19: 1688-9.

Brandt ME, Pfaller MA, Hajjeh RA, Hamill RJ, Pappas PG, Reingold AL, Rimland D, Warnock DW, Cryptococcal disease active surveillance group. Trends in antifungal drug susceptibility of Cryptococcus neoformans isolates in the United States: 1992 to 1994 and 1996 to 1998. Antimicrob Agents Chemother 2001; 45: 3065-9.

Brown GD, Denning DW, Gow NAR, Levitz SM, Netea MG, White TC. Hidden killers: human fungal infections. Sci Transl Med 2012; 4. doi: 10.1126/scitranslmed.3004404.

Byrnes EJ, Bildfell RJ, Frank SA, Mitchell TG, Marr KA, Heitman J. Molecular evidence that the range of the Vancouver Island outbreak of Cryptococcus gattii infection has expanded into the Pacific Northwest in the United States. J Infect Dis 2009; 199: 1081-6.

Cannon RD, Lamping E, Holmes R, Niimi K, Baret PV, Keniya MV, Tanabe K, Niimi M, Goffeau A, Monk BC. Efflux-mediated antifungal drug resistance. Clinical Microbiol Rev 2009; 22: 291-321.

Casadevall A, Perfect JR. Cryptococcus neoformans. ASM Press, Washington DC 1998.

Chowdhary A, Prakash A, Randhawa HS, Kathuria S, Hagen F, Klaassen CH, Meis JF. First

126 General discussion and future directions environmental isolation of Cryptococcus gattii, genotype AFLP5, from India and a global review. Mycoses 2013; 56: 222-8.

Chowdhary A, Randhawa HS, Prakash A, Meis JF. Environmental prevalence of Cryptococcus neoformans and C. gattii in India: an update. Crit Rev Microbiol 2012; 38: 1-16.

Cox JA, Lukande RL, Nelson AM, Mayanja-Kizza H, Colebunders R, van Marck E, Manabe YC. An autopsy study describing causes of death and comparing clinico-pathological findings among hospitalized patients in Kampala, Uganda. PLoS One 2012; 7: e33685.

Cuenca-Estrella M. Antifúngicos en el tratamiento de las infecciones sistémicas: importancia del mecanismo de acción, espectro de actividad y resistencias. Rev Esp Quimioter 2010; 23:169-76. Curbelo A. Septicemia y meningitis a Cryptococcus neoformans. Arch Hosp Univ (Cuba) 1951; 9:324-30.

Dalhoff A, Ambrose PG, Mouton JW. A long journey from minimum inhibitory concentration testing to clinically predictive breakpoints: deterministic and probabilistic approaches in deriving breakpoints. Infection 2009; 37: 296-305. de Valk HA, Meis JF, Bretagne S, Costa JM, Lasker BA, Balajee SA, Pasqualotto AC, Anderson MJ, Alcázar-Fuoli L, Mellado E, Klaassen CH. Interlaboratory reproducibility of a microsatellite- based typing assay for Aspergillus fumigatus through the use of allelic ladders: Proof of concept. Clin Microbiol Infect 2009; 15: 180-7. del Poeta M, Casadevall A. Ten challenges on Cryptococcus and Cryptococcosis. Mycopathologia 2012;173:303-10.

Doering TL. How sweet it is! Cell wall biogenesis and polysaccharide capsule formation in Cryptococcus neoformans. Annu Rev Microbiol 2009; 63: 223-47.

Espinel-Ingroff A, Aller AI, Canton E, Castañón-Olivares LR, Chowdhary A, Cordoba S, Cuenca- Estrella M, Fothergill A, Fuller J, Govender N, Hagen F, Illnait-Zaragozi MT, Johnson E, Kidd S, Lass-Flörl C, Lockhart SR, Martins MA, Meis JF, Melhem MSC, Ostrosky-Zeichner L, Pelaez T, Pfaller MA, Schell WA, St-Germain G, Trilles L, Turnidge J. Cryptococcus neoformans- Cryptococcus gattii species complex: an international study of wild-type susceptibility endpoint distributions and epidemiological cutoff values for fluconazole, itraconazole, posaconazole and voriconazole. Antimicrob Agents Chemother 2012; 56: 5898-906. (b)

Espinel-Ingroff A, Chowdhary A, Cuenca-Estrella M, Fothergill A, Fuller J, Hagen F, Govender N, Guarro J, Johnson E, Lass-Flörl C, Lockhart SR, Martins MA, Meis JF, Melhem MSC, Ostrosky- Zeichner L, Pelaez T, Pfaller MA, Schell WA, Trilles L, Kidd S, Turnidge J. Cryptococcus neoformans- 12 Cryptococcus gattii species complex: an international study of wild-type susceptibility endpoint distributions and epidemiological cutoff values for amphotericin B and flucytosine. Antimicrob Agents Chemother 2012; 56: 3107-13. (a)

Fonseca A, Boekhout T, Fell JW. Cryptococcus Vuillemin (1901). In: Kurtzman CP, Fell JW, Boekhout T. (Eds.) The yeasts: a taxonomic study. Elsevier Science & Technology, Amsterdam, 2011; p: 1665-741.

127 Chapter 12

Frey-Klett P, Burlinson P, Deveau A, Barret M, Tarkka M, Sarniguet A. Bacterial-fungal interactions: hyphens between agricultural, clinical, environmental, and food microbiologists. Microbiol Mol Biol Rev 2011; 75: 583-609.

Friedman GD, Fessel WJ, Udaltsova NV, Hurley LB. Cryptococcosis: the 1981–2000 epidemic. Mycoses 2005; 48: 122-5.

Govender NP, Patel J, van Wyk M, Chiller TM, Lockhart SR. Trends in antifungal drug susceptibility of Cryptococcus neoformans isolates obtained through population-based surveillance in South Africa in 2002–2003 and 2007–2008. Antimicrob Agents Chemother 2011; 55: 2606-11.

Guinea J, Hagen F, Peláez T, Boekhout T, Tahoune H, Torres-Narbona M, Bouza E. Antifungal susceptibility, serotyping, and genotyping of clinical Cryptococcus neoformans isolates collected during 18 years in a single institution in Madrid, Spain. Med Mycol 2010; 48: 942-8. Hagen F, Boekhout T. The search for the natural habitat of Cryptococcus gattii. Mycopathologia 2010; 70: 209–11 (b).

Hagen F, Illnait-Zaragozí MT, Meis JF, Chew WH, Curfs-Breuker I, Mouton JW, Hoepelman AI, Spanjaard L, Verweij PE, Kampinga GA, Kuijper EJ, Boekhout T, Klaassen CH. Extensive genetic diversity within the Dutch clinical Cryptococcus neoformans population.J Clin Microbiol 2012; 50: 1918-26 (a).

Hagen F, van Assen S, Luijckx GJ, Boekhout T, Kampinga GA. Activated dormant Cryptococcus gattii infection in a Dutch tourist who visited Vancouver Island (Canada): a molecular epidemiological approach. Med Mycol 2010; 48: 528-31 (a).

Illnait-Zaragozí MT, Hagen F, Fernández CM, Martínez GF, Polo JL, Boekhout T, Klaassen CH, Meis JF. Reactivation of a Cryptococcus gattii infection in a cheetah (Acinonyx jubatus) held in the National Zoo, Havana, Cuba. Mycoses 2011; 54: e889-92 (b).

Illnait-Zaragozi MT, Martínez-Machín GF, Curfs-Breuker I, Fernández-Andreu CM, Boekhout T, Meis JF. In vitro activity of the new azole isavuconazole (BAL4815) compared with six other antifungal agents against 162 Cryptococcus neoformans isolates from Cuba. Antimicrob Agents Chemother 2008; 52: 1580-2.

Illnait-Zaragozi MT, Martínez-Machín GF, Fernández-Andreu CM, Boekhout T, Meis JF, Klaassen CH. Microsatellite typing of clinical and environmental Cryptococcus neoformans var. grubii isolates from Cuba shows multiple genetic lineages. Plos One 2010; 5: e9124.

Illnait-Zaragozí MT, Martínez-Machín GF, Fernández-Andreu CM, Perurena-Lancha MR, Theelen B, Boekhout T, et al. Environmental isolation and characterization of Cryptococcus species from living trees in Havana City, Cuba. Mycoses 2012; 55: 138-44.

Illnait-Zaragozí MT, Ortega-Gonzalez LM, Hagen F, Martínez-Machin GF, Meis JF. Fatal

128 General discussion and future directions

Cryptococcus gattii genotype AFLP5 infection in an immunocompetent Cuban patient. Med Mycol Case Report 2013;2:48-51.

Iqbal N, DeBess EE, Wohrle E, Sun B, Net RJ, Ahlquist AM, Chiller T, Lockhart SR. Correlation of genotype and in vitro susceptibilities of Cryptococcus gattii from the Pacific Northwest of the United States. J Clin Microbiol 2010; 48:539-44.

Jain N, Fries BC. Phenotypic switching of Cryptococcus neoformans and Cryptococcus gattii. Mycopathologia 2008; 166: 181-8.

Kidd SE, Chow Y, Mak S, Bach PJ, Chen H, Hingston AO, Kronstad JW, Bartlett KH. Characterization of environmental sources of the human and animal pathogen Cryptococcus gattii in British Columbia, Canada, and the pacific northwest of the United States. Appl Environ Microbiol 2007; 73: 1433-43.

Kielstein P, Hotzel H, Schmalreck A, Khaschabi D, Glawischnig W. Occurrence of Cryptococcus spp. in excreta of pigeons and pet birds. Mycoses 2000; 43: 7-15.

Kozubowski L, Heitman J. Profiling a killer, the development ofCryptococcus neoformans. FEMS Microbiol Rev 2011; doi:10.1111/j.1574-6976.2011.

Lazera MS, Salmito MA, Londero AT, Trilles L, Nishikawa MM, Wanke B. Possible primary ecological niche of Cryptococcus neoformans. Med Mycol 2000; 38: 379-83.

Lewis RE, Pharm D. Current concepts in antifungal pharmacology. Mayo Clin Proc 2011; 86: 805- 817.

Lin X. Cryptococcus neoformans: Morphogenesis, infection, and evolution. Infect Genet Evol 2009; 9: 401-16.

Litvintseva AP, Mitchell TG. Most environmental isolates of Cryptococcus neoformans var. grubii (serotype A) are not lethal for mice. Infect Immun 2009; 77: 3188-95.

Ma H, Hagen F, Stekel DJ, Johnston SA, Sionov E, Falk R, Polacheck I, Boekhout T, May RC. The fatal fungal outbreak on Vancouver Island is characterized by enhanced intracellular parasitism driven by mitochondrial regulation. Proc Natl Acad Sci U S A 2009; 106: 12980-5. 12 Mandal P, Banerjee U, Casadevall A, Nosanchuk JD. Dual infections with pigmented and albino strains of Cryptococcus neoformans in patients with or without human immunodeficiency virus infection in India. J Clin Microbiol 2005; 43: 4766-72.

Marr KA, Datta K, Pirofski LA, Barnes R. Cryptococcus gattii infection in healthy hosts: a sentinel for subclinical immunodeficiency? Clin Infect Dis 2012 54: 153-4.

Metin B, Findley K, Heitman J. The mating type locus (MAT) and sexual reproduction of

129 Chapter 12

Cryptococcus heveanensis: Insights into the evolution of sex and sex-determining chromosomal regions in fungi. PLos Genet 2010; 6: e1000961.

Mlinaric-Missoni E, Hagen F, Chew WH, Vazic-Babic V, Boekhout T, Begovac J. In vitro antifungal susceptibilities and molecular typing of sequentially isolated clinical Cryptococcus neoformans strains from Croatia. J Med Microbiol 2011; 60:1487-95.

Nakamura K, Miyagi K, Koguchi Y, Kinjo Y, Uezu K, Kinjo T, Akamine M, Fujita J, Kawamura I, M, Adachi Y, Ohno N, Takeda K, Akira S, Miyazato A, Kaku M, Kawakami K. Limited contribution of Toll-like receptor 2 and 4 to the host response to a fungal infectious pathogen, Cryptococcus neoformans. FEMS Immunol Med Microbiol 2006; 47: 148-54.

Ngamskulrungroj P, Gilgado F, Faganello J, Litvintseva AP, Leal AL, Tsui KM, Mitchell TG, Vainstein MH, Meyer W. Genetic diversity of the Cryptococcus species complex suggests that Cryptococcus gattii deserves to have varieties. Plos One 2009; 4: e5862. Odds, F. C. Drug evaluation: BAL-8557, a novel broad-spectrum triazole antifungal. Curr. Opin. Investig. Drugs 2006; 7: 766-72.

Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, Chiller TM. Estimation of the current global burden of cryptococcal meningitis among persons living with HIV/AIDS. AIDS 2009; 23: 525-30.

Polo JL, Fernández CM, Martínez GF, Illnait MT, Perurena MR. Cryptococcus gattii aislado de un guepardo (Acinonyx jubatus) del Parque Zoológico Nacional de Cuba. Rev Cubana Med Trop 2010; 62: 257-60.

Schmitt-Hoffmann A, Roos B, Heep M, Schleimer M, Weidekamm E, Brown T, Roehrle M, Beglinger C. Single-ascending-dose pharmacokinetics and safety of the novel broad-spectrum antifungal triazole BAL4815 after intravenous infusions (50, 100, and 200 milligrams) and oral administrations (100, 200, and 400 milligrams) of its prodrug, BAL8557, in healthy volunteers. Antimicrob Agents Chemother 2006; 50: 279-85. (a)

Schmitt-Hoffmann A, Roos B, Maares J, Heep M, Spickerman J, Weidekamm E, Brown T, Roehrle M. Triazole BAL4815 after intravenous infusion and oral administration of its prodrug, BAL8557, in healthy volunteers. Antimicrob Agents Chemother 2006; 50:286-93. (b)

Schoffelen T, Illnait-Zaragozi MT, Joosten LAB, Netea MG, Boekhout T, Meis JF, Sprong T. Cryptococcus gattii induces a cytokine pattern that is distinct from other cryptococcal species. Plos ONE 2013; 8(1):e55579.

130 General discussion and future directions

Seifert H, Aurbach U, Stefanik D, Cornely O. In vitro activities of isavuconazole and other antifungal agents against Candida bloodstream isolates. Antimicrob Agents Chemother 2007; 51: 1818-21.

Shoham S, Huang C, Chen JM, Golenbock DT, Levitz SM. Toll-like receptor 4 mediates intracellular signaling without TNF-alpha release in response to Cryptococcus neoformans polysaccharide capsule. J Immunol 2001; 166: 4620-6.

Sionov E, Chang YC, Garraffo HM, Kwon-Chung KJ. Heteroresistance to fluconazole in Cryptococcus neoformans is intrinsic and associated with virulence. Antimicrob Agents Chemother 2009; 53: 2804-15.

Snelders E, Camps SMT, Karawajczyk A, Schaftenaar G, Kema GHJ, van der Lee HA, Klaassen CH, Melchers WJ, Verweij PE. Triazole fungicides can induce cross-resistance to medical triazoles in Aspergillus fumigatus. Plos One 2012; 7: e31801.

Souza LK, Fernándes OFL, Kobayashi CC, Passos XS, Costa CR, Lemos JA, Souza-Júnir AH, Silva MR. Antifungal susceptibilities of clinical and environmental isolates of Cryptococcus neoformans in Goiânia City, Goiás, Brazil. Rev Inst Med Trop Sao Paulo 2005; 47: 253-6.

Springer DJ, Chaturvedi V. Projecting global occurrence of Cryptococcus gattii. Emerg Infect Dis 2010; 16: 14-20.

Trilles L, Lazera M, Wanke B, Theelen B, Boekhout T. Genetic characterization of environmental isolates of the Cryptococcus neoformans species complex from Brazil. Med Mycol 2003; 41: 383- 90.

Vandenbossche I, Vaneechoutte M, Vandevenne M, De Baere T, Verschraegen G. Susceptibility testing of fluconazole by the NCCLS broth macrodilution method, E-test, and disk diffusion for application in the routine laboratory. J Clin Microbiol 2002; 40: 918-21.

Vandeputte P, Ferrari S, Coste AT. Antifungal resistance and new strategies to control fungal infections. Int J Microbiol 2012; Article ID 713687, doi:10.1155/2012/713687.

Varma A, Kwon-Chung KJ. Heteroresistance of Cryptococcus gattii to fluconazole. Antimicrob Agents Chemother 2010; 54: 2303-11.

Viviani MA, et al. Molecular analysis of 311 Cryptococcus neoformans isolates from a 30-month ECMM survey of cryptococcosis in Europe. FEMS Yeast Res 2006; 6: 614-19.

Wang X, Li W, Sun S, Kozubowski L, Lee SC, Feretzaki M, Heitman J. Know your enemy: how to build and vanquish a global fungal scourge. Mycopathologia 2012;173: 295-301. 12 Yauch LE, Mansour MK, Shoham S, Rottman JB, Levitz SM. Involvement of CD14, toll-like receptors 2 and 4, and MyD88 in the host response to the fungal pathogen Cryptococcus neoformans in vivo. Infect Immun 2004; 72: 5373-82.

Zhu J, Kang Y, Uno J, Taguchi H, Liu Y, Ohata M, Tanaka R, Moretti MK, Mikami Y. Comparison of genotypes between environmental and clinical isolates of Cryptococcus neoformans var. grubii based on microsatellite patterns. Mycopathologia 2010; 169: 47-55.

131

Chapter 3 Summary 1

Samenvatting

Resumen Chapter 13

Summary

Cryptococcosis ranks among the three most common life-threatening opportunistic infections in persons with AIDS. Other patient groups with impaired T-cell function have an up to 6% lifetime risk to develop clinical cryptococcosis. Although about 100 species belonging to the genus Cryptococcus have been described, members of the C. neoformans/C. gattii species complex are associated with the majority of human infections. This complex contains two pathogenic species: C. neoformans (serotypes A, D and AD), and C. gattii (serotypes B and C) and at least 10 monophyletic lineages representing these species and their serotypes as well as some hybrids. This thesis includes three main parts that compile eleven publications aimed to contribute to a better understanding of cryptococcosis and it causative agent.

Studies of local and global epidemiology of Cryptococcus spp. have been hampered by the lack of rapid, discriminatory, and exchangeable typing methods. Part I of the thesis describes the molecular characterization of clinical and environmental C. neoformans species complex isolated from Cuba and the Netherlands. For this purpose a set of 190 Cuban strains (122 clinical and 68 from pigeon guano) as well as 300 Dutch clinical strains were investigated by determining the mating type, serotype, AFLP and microsatellite genotype. For the last method nine microsatellite markers from the genome of C. neoformans var. grubii H99 were selected using the Tandem Repeats Finder software. All the Cuban and most of the Dutch isolates were serotype αA (demonstrated for the first time in Cuba). In general the most frequent genotype was AFLP1, distantly followed by AFLP2 and AFLP3. Microsatellite typing revealed multiple clusters of related genotypes which were assigned as microsatellite complexes. Dutch strains showed a high genetic diversity among serotype A isolates but a lower diversity within serotype D; most of the Cuban environmental isolates (70%) fell into one microsatellite complex containing only few clinical isolates (49 environmental versus 2 clinical) while clinical isolates were segregated over multiple microsatellite complexes. These findings led us to conclude that the selected panel of microsatellite markers was an excellent tool to study the epidemiology of C. neoformans var. grubii; that a large genotypic variation among these clinical isolates exist; and that the occurrence of additional source(s) of human cryptococcal infections beside pigeon guano in Cuba is possible. AFLP genotyping revealed that one of the Dutch patients was infected by multiple genotypes, while genotyping with microsatellites of 19 isolates from seven Cuban patients with recurrent cryptococcal meningitis displayed 14 distinctive profiles: in three patients the infection was associated with identical isolates and the remaining patients had relapse isolates which were genotypically different. Antifungal susceptibility for amphotericin B, fluconazole, flucytosine, itraconazole, voriconazole, posaconazole and isavuconazole was tested by CLSI M27-A3 broth microdilution. Among the Dutch studied strains, one had high flucytosine MIC (>64 µg/mL), while decreased susceptibility (>16 µg/mL) of flucytosine and fluconazole was found in 9 and 10 C. neoformans isolates, respectively. None of the Cuban strains were resistant to the studied drugs. These findings suggest that recurrence of cryptococcal meningitis in these patients was not associated with drug resistance of the original strain but with an initial infection with different strains or a re-infection with a new strain.

134 Summary - Resumen

The management of cryptococcosis has been one of the most studied mycotic diseases but there are still many unresolved questions. Despite medical advances, the mortality rate during the first three months of treatment exceeds 20% and approaches 100% in patients who do not receive proper therapy. Part II of the thesis was devoted to the study of antifungal activity of isavuconazole (a new triazole) and six other antifungal drugs on a wide collection of cryptococcal strains isolated from human, veterinary, and environmental sources. One hundred sixty-five Cuban C. neoformans var. grubii and a worldwide collection of 350 C. gattii isolates were tested against amphotericin B, flucytosine, fluconazole, itraconazole, voriconazole, posaconazole and isavuconazole by CLSI M27-A3 broth microdilution. Of the seven antifungal compounds, voriconazole and posaconazole showed strong in vitro antifungal activity (up to 0.25 µg/mL) while isavuconazole displayed even lower MIC90 (up to 0.125 µg/mL). Amphotericin B had an acceptable MIC90 (up to 0.5 µg/mL), but flucytosine and fluconazole had higher MIC90s (up to 8 µg/ mL). The results indicate that, given the oral bioavailability and the well-tolerated nature of the new azoles, they might become an important addition to the armamentarium of antifungal agents against Cryptococcus strains. To the best of our knowledge this was the first report evaluating isavuconazole against Cryptococcus spp. which was the most effective antifungal drug in terms of in vitro activity. In addition C. gattii isolates were divided into seven AFLP genotypes, including the interspecies hybrids AFLP8 and AFLP9. Most of clinical isolates (n = 215) comprised genotypes AFLP4 (n = 76) and AFLP6 (n = 103). The latter had significantly higher geometric mean MICs of flucytosine and fluconazole than the first ones which suggested that susceptibility to antifungal drugs is species- and genotype-specific. This was corroborated by the international study of wild type (WT) susceptibility endpoint distributions and epidemiological cutoff values (ECV). Such parameters were recently established by Espinel-Ingroff and co-workers for amphotericin B andfl ucytosine versus C. neoformans/C. gattii species complex and here it was established for fluconazole, itraconazole, posaconazole, and voriconazole. The results highlight the utility of cutoffs as a practical tool to detect resistance among Cryptococcus isolates and that such values are species- and molecular type-specific: fluconazole (8 to 32 µg/mL), itraconazole (0.25 to 1 µg/mL), posaconazole (0.25 to 0.5 µg/mL), and voriconazole (0.12 to 0.25 µg/ mL). It was suggested to include these ECVs in the revised version of the CLSI M27-A3 document and to continue susceptibility surveillance in order to corroborate or expand the information provided in this study as well as to start investigations to determine the relationship between molecular mechanisms of triazole resistance and the proposed ECV. C. neoformans and C. gattii have been worldwide isolated from environmental sources. Although the latter one used to have a more restricted distribution to tropical and subtropical climates the outbreaks in North America indicates the adaptation of this organism to new environments. This, together with the highly variable interval between exposure and development of clinical disease, makes it more difficult to arrive 13 to conclusions of its epidemiology. Part III of the thesis involves the analysis of C. neoformans/C. gattii species complex isolated from humans, animal and plants in Cuba as well as studies about pathogenic differences between both species.

135 Chapter 13

Routinely only C. neoformans var. grubii used to be isolated from clinical samples and pigeon droppings and previous work provided strong evidence of an additional origin of exposure to this yeast in Cuba. This led to conduct an environmental study of 662 samples collected from 331 trees and cacti in Havana, Cuba. Initial selection of Cryptococcus spp. isolates by conventional techniques yielded 121 C. neoformans and 61 C. gattii. Further characterization by AFLP and DNA sequence analysis showed that none of them was C. gattii; only one proved to be C. neoformans var. grubii; 27 were identified as other different species; and 65 unidentifiable isolates segregated into ten potentially novel species. Up to now attempts to recover C. gattii from environmental samples in Cuba have been fruitless, so these results highlight the need of more epidemiological studies to determine the exact distribution of C. neoformans and C. gattii in Cuba as well as the importance to apply molecular methods to confirm identification of isolates from environmental sources. Part III also includes the description of the first two documented cases of C. gattii in Cuba: from a cheetah (Acinonyx jubatus) held at the National Zoo, in Havana and from a Cuban immunocompetent patient. Both cases had a travel history (South Africa and Central America respectively) and molecular analysis showed that the strains from the animal belong to genotype AFLP4 and those of the human to AFLP5. MLST sequences clustered with strains from Kenya and Colombia respectively which strongly suggests that both cases had a dormant infection when arrived to Cuba and that this method is an excellent approach to trace a clinical strain to its geographic origin. Although C. neoformans and C. gattii share most of the properties that confer pathogenicity, the latter species seems to be more virulent, as it has been the main causative agent of disease in apparently immunocompetent hosts. At the end of Part III a series of experiments is described aimed to demonstrate that the ability of C. gattii to cause disease in humans depends on a distinct innate cytokine response of the host. For this purpose an ex-vivo cytokine profile of human peripheral blood mononuclear cells of healthy individuals was determined, after stimulation with 40 different well- defined heat-killed C. gattii, C. neoformans and several hybrid isolates. Furthermore, the involvement of TLR2, TLR4 and TLR9 in the pro-inflammatory cytokine response to C. gattii was investigated. The results suggested that clinical C. gattii isolates induced a more pronounced inflammatory response (IL-1b, TNFα and IL-6) and Th17/22 cytokine response compared to non-clinical C. gattii and other Cryptococcus species which is dependent on TLR4 and TLR9 as cellular receptors. In this study it was not identified which cell wall components were involved in the initial cytokine response. Since about a million persons worldwide contract cryptococcosis and its mortality exceeds 50% despite the availability of several antifungal drugs, we hope that these studies of the C. neoformans/C. gattii species complex have contributed to more understanding of these pathogenic yeasts and the infection they cause.

136 Summary - Resumen

Samenvatting

Cryptococcose behoort tot de drie meest voorkomende levensbedreigende opportunistische infecties bij mensen met HIV/AIDS. Andere patiëntengroepen met een verminderde T-celfunctie hebben een tot 6% hoger risico om gedurende hun leven cryptococcose te ontwikkelen. Hoewel er ongeveer 100 soorten gisten zijn beschreven die tot het genus Cryptococcus behoren, zijn soorten uit het C. neoformans/C. gattii species complex grotendeels verantwoordelijk voor infecties bij mensen en dieren. Het complex bevat twee pathogene soorten: C. neoformans (serotypen A, D en AD ) en C. gattii (serotypen B en C ) en valt uiteen in tenminste 10 monofyletisch clusters, daarnaast zijn er enkele hybriden beschreven. Dit proefschrift bestaat uit drie onderdelen die elf publicaties omvatten en die mogelijk bijdragen aan een beter begrip van de cryptococcose en de verwekker.

Onderzoek naar de lokale en globale epidemiologie van Cryptococcus soorten werd in het verleden gehinderd door een gebrek aan snelle, onderscheidende en uitwisselbare typeringstechnieken. Deel I van dit proefschrift beschrijft de moleculaire karakterisatie van klinische en omgevingsisolaten uit het C. neoformans species complex geïsoleerd in Cuba en in Nederland. Hiervoor is een set van 190 Cubaanse stammen (122 klinische en 68 uit duivenontlasting) en 300 Nederlandse klinische stammen onderzocht voor wat betreft het mating-type, serotype, AFLP en microsatelliet genotype. Voor de laatst genoemde methode werden negen microsatellietmarkers uit het genoom van C. neoformans var. grubii stam H99 geselecteerd met behulp van het softwarepakket Tandem Repeats Finder. Alle Cubaanse en de meeste Nederlandse isolaten waren mating- en serotype αA (dit is voor het eerst dat dit werd beschreven voor Cuba). Het meest voorkomende genotype bleek AFLP1 te zijn, gevolgd door AFLP2 en AFLP3. Microsatelliet typering liet meerdere clusters van verwante genotypes zien die microsatelliet-complexen worden genoemd. Nederlandse stammen hadden een hoge genetische diversiteit onder serotype A isolaten maar een lagere diversiteit binnen serotype D; de meeste Cubaanse omgevingsisolaten (70%) vielen in een microsatelliet- complex met slechts weinig klinische isolaten (49 omgeving versus 2 klinische) terwijl klinische isolaten verspreid bleken te zijn over meerdere microsatelliet-complexen. Deze bevindingen leidde tot de conclusie dat het gekozen panel van microsatellietmarkers een uitstekend hulpmiddel is om de epidemiologie van C. neoformans var. grubii te onderzoeken; dat er grote genotypische verschillen tussen klinische isolaten bestaan; en dat het mogelijk is dat er extra bron(nen) van menselijke Cryptococcus-infecties zijn in Cuba, buiten de al bekende duivenontlasting.

Genotypering middels AFLP liet zien dat één van de Nederlandse patiënten geïnfecteerd was met verschillende genotypen, terwijl microsatelliet genotypering van 19 isolaten van zeven Cubaanse patiënten met terugkerende cryptokokkenmeningitis 14 moleculaire types liet zien: bij drie patiënten was de infectie geassocieerd met identieke 13 isolaten terwijl de andere patiënten geinfecteerd waren met isolaten die genotypisch verschillend waren. Antifungale gevoeligheid voor amfotericine B, fluconazol, flucytosine, itraconazol, voriconazol, posaconazol en isavuconazole werd getest met de CLSI M27-A3 bouillon microdilutie-methode. Onder de Nederlandse stammen

137 Chapter 13 was er slechts één met een hoge flucytosine MIC (> 64 µg/ml), terwijl verminderde gevoeligheid (> 16 µg/ml) van flucytosine en fluconazol werd gezien in respectievelijk 9 en 10 C. neoformans isolaten. Geen van de Cubaanse stammen waren resistent tegen antifungale geneesmiddelen. Deze bevindingen suggereren dat een hernieuwde infectie van cryptokokkenmeningitis bij deze patiënten niet geassocieerd is met resistentie van de oorspronkelijke stam, maar met een initiële infectie met verschillende stammen of herinfectie met een nieuwe stam.

Cryptococcose is een van de meest bestudeerde schimmelziekten, maar er zijn nog veel onopgeloste vragen. Ondanks de medische vooruitgang is de sterfte in de eerste drie maanden van de behandeling nog steeeds meer dan 20% en loopt op tot bijna 100% bij patiënten die niet de juiste therapie krijgen. Deel II van dit proefschrift is gewijd aan de studie van de antifungale activiteit van isavuconazole (een nieuw triazool) en zes andere antimycotica op een brede verzameling van Cryptococcus stammen geïsoleerd van humane, veterinaire en ecologische bronnen .

Honderdvijfenzestig Cubaanse C. neoformans var. grubii isolaten en een wereldwijde verzameling van 350 C. gattii isolaten werden getest tegen amfotericine B, flucytosine, fluconazol, itraconazol, voriconazol, posaconazol en isavuconazole met de CLSI M27- A3 bouillon microdilutie-methode. Van de zeven antischimmelmiddelen vertoonden voriconazol en posaconazol een sterke in vitro antifungale activiteit (tot 0,25 µg/ml) terwijl isavuconazole zelfs nog lagere MIC90 (tot 0,125 µg/ml) had. Amfotericine B had een aanvaardbare MIC90 (tot 0,5 µg/ml), maar flucytosine en fluconazol hadden hogere MIC90s (tot 8 µg/ml). De resultaten laten zien dat de nieuwe azolen, mede door de goede orale biobeschikbaarheid en goede verdraagzaamheid, een belangrijke toevoeging zijn aan het arsenaal van antischimmelmiddelen tegen Cryptococcus- infecties. Dit onderzoek is de eerste evaluatie van isavuconazole tegen Cryptococcus spp. waarbij bleek dat het hoogste in vitro activiteit van alle antimycotica had. Tevens werden C. gattii isolaten verdeeld in zeven AFLP-genotypen, waaronder de interspecies hybriden met het genotype AFLP8 en AFLP9. De meeste klinische isolaten (n = 215) omvatten de genotypen AFLP4 (n = 76) en AFLP6 (n = 103). Deze laatste hadden significant hogere geometrisch gemiddelde MICs van flucytosine en fluconazol dan de eersten, wat suggereerde dat de gevoeligheid voor antimycotica soorten genotype specifiek is. Dit werd bevestigd door een internationaal onderzoek van wild-type (WT) gevoeligheidsverdelingen en epidemiologische cut-off waarden (ECV). Dergelijke parameters zijn onlangs vastgesteld voor amfotericine B en flucytosine versus het C. neoformans/C. gattii species complex en hier beschreven is voor fluconazol, itraconazol, posaconazol en voriconazol. De resultaten benadrukken het nut van ECVs als een praktisch hulpmiddel om resistentie te detecteren voor Cryptococcus isolaten en dat deze waarden species- en genotype specifiek zijn: fluconazol (8 tot 32 µg/ml), itraconazol (0,25-1 µg/ml), posaconazol (0,25-0,5 µg/ml) en voriconazol (0,12-0,25 µg/ ml). Er werd voorgesteld om deze ECVs op te nemen in de herziene versie van CLSI M27-A3 document en door te gaan met gevoeligheidssurveillance om deze studie resultaten te bevestigen of uit te breiden en onderzoek te gaan doen naar de relatie tussen moleculaire mechanismen van triazool resistentie en de voorgestelde ECV .

C. neoformans en C. gattii zijn wereldwijd geïsoleerd uit het milieu. Hoewel C. gattii een meer beperkte distributie in tropische en subtropische klimaten had, laten de uitbraken

138 Summary - Resumen in Noord-Amerika zien dat dit micro-organisme zich goed aan heeft weten te passen aan nieuwe omgevingen. Samen met het zeer variabele interval tussen blootstelling en ontwikkeling van klinische ziekte, bemoeilijkt dit om harde conclusies te trekken over de epidemiologie. Deel III van dit proefschrift omvat een analyse van het C. neoformans/C. gattii species complex geïsoleerd van patienten, dieren en planten in Cuba alsmede studies over verschillen in pathogenese tussen beide soorten.

In het verleden werd alleen C. neoformans var. grubii routinematig geïsoleerd uit klinisch materiaal en uitwerpselen van duiven, eerder uitgevoerd onderzoek levert sterke aanwijzingen van een extra bron van blootstelling aan deze gist in Cuba. Dit gaf aanleiding tot een groot onderzoek van 662 monsters genomen van 331 bomen en cactussen in Havana, Cuba. Een eerste selectie van Cryptococcus spp. isolaten gedaan met conventionele technieken leverde 121 C. neoformans en 61 C. gattii stammen op. Verdere analyse van AFLP en DNA-sequentie data toonden aan dat er geen C. gattii aanwezig was en slechts één isolaat bleek C. neoformans var. grubii te zijn: 27 isolaten werden geïdentificeerd als andere soorten en en 65 niet-geïdentificeerde isolaten bestonden uit tien potentieel nieuwe soorten. Tot nu toe zijn pogingen om C. gattii te isoleren uit de omgeving in Cuba vruchteloos gebleken. Dit benadrukt de noodzaak van meer epidemiologisch onderzoek naar de exacte verspreiding van C. neoformans en C. gattii in Cuba en het belang van moleculaire methoden voor identificatie van omgevingsisolaten.

Deel III beschrijft ook de eerste twee gedocumenteerde gevallen van C. gattii in Cuba: een cheetah (Acinonyx jubatus) gehouden in de dierentuin in Havana en een Cubaanse immunocompetente patiënt. Beide gevallen hadden een relevante reisgeschiedenis (Zuid-Afrika en Midden-Amerika) en moleculaire analyse toonde aan dat de stam van het dier behoorde tot genotype AFLP4 en die van de patient tot AFLP5. MLST analyse toonde aan dat deze Cubaanse C. gattii-stammen clusterden met stammen uit respectievelijk Kenia en Colombia, wat suggereert dat beide gevallen mogelijk met een subklinische infectie naar Cuba zijn gereisd. MLST is een uitstekende benadering om van een klinische stam de geografische oorsprong op te sporen.

Hoewel C. neoformans en C. gattii beiden in het bezit zijn van virulentiefactoren die pathogeniciteit veroorzaken, lijkt de laatste soort meer virulent te zijn, omdat het de belangrijkste verwekker van de ziekte is in de immunocompetente gastheer. Aan het slot van deel III is een reeks experimenten beschreven met het doel om aan te tonen dat het vermogen van C. gattii om ziekte bij de mens te veroorzaken afhangt van een speciale aangeboren cytokine respons bij de gastheer. Daarom is een ex-vivo cytokine profiel bepaald van humane perifere bloed mononucleaire cellen van gezonde individuen na stimulatie met 40 verschillende goed gedefinieerde, door verhitting gedode, C. gattii, C. neoformans en verschillende hybride isolaten. Bovendien is de betrokkenheid van TLR2, TLR4 en TLR9 in de pro-inflammatoire cytokine respons tegen C. gattii onderzocht. De resultaten suggereerden dat klinische C. gattii isolaten een meer uitgesproken ontstekingsreactie (IL-1b, αTNF en IL-6) en Th17/22 cytokine 13 respons induceerden dan niet-klinische C. gattii en andere Cryptococcus soorten die afhankelijk zijn van TLR4 en TLR9 als cellulaire receptoren. In dit onderzoek werden niet de celwand componenten geanalyseerd die betrokken zijn bij de oorspronkelijke cytokine respons.

139 Chapter 13

Ongeveer een miljoen mensen wereldwijd lopen cryptococcose op en de sterfte is hoger dan 50%, ondanks de beschikbaarheid van verschillende antimycotica. Ik hoop dat dit onderzoek naar het C. neoformans/C. gattii species complex zal bijdragen tot meer kennis van deze pathogene gisten en de infectie die ze veroorzaken.

140 Summary - Resumen

Resumen

La criptococosis constituye una de las tres infecciones oportunistas potencialmente mortales más frecuentes en las personas con sida. Otros grupos de pacientes con la función de las células T alterada tienen un riesgo de hasta el 6 % de desarrollar la forma clínica de esta infección. Aunque se han descrito alrededor de 100 especies pertenecientes al género Cryptococcus, los miembros del complejo de especies C. neoformans/C. gattii son los más frecuentemente asociados a infecciones humanas. Este complejo contiene dos especies patógenas: C. neoformans (serotipos A, D y AD) y C. gattii (serotipos B y C) y al menos 10 linajes filogenéticos que representan estas especies y sus serotipos, así como algunos híbridos. Esta tesis incluye tres partes principales que recopilan once publicaciones destinadas a contribuir a una mejor comprensión de la criptococosis y su agente causal.

Los análisis de epidemiología molecular local y global de Cryptococcus spp., se han visto obstaculizados por la falta de métodos de caracterización rápidos, discriminatorios e intercambiables. La Parte I de la tesis describe la caracterización molecular de aislamientos clínicos y ambientales de C. neoformans aislados en Cuba y los Países Bajos. Con este propósito fue investigada una colección de 190 aislamientos cubanos (122 clínicos y 68 recuperados de excretas de palomas), así como 300 aislamientos clínicos holandeses. A los mismos se les determinó el serotipo y el tipo de apareamiento mediante reacción en cadena de la polimerasa (PCR, de sus siglas en inglés), así como genotipificación mediante AFLP y fragmentos cortos repetitivos. Para este último se seleccionaron nueve marcadores microsatélites del genoma de C. neoformans var. grubii H99 utilizando el programa Tandem Repite Finder. Todos los aislamientos cubanos y la mayoría de los holandeses correspondieron con el serotipo Aα (demostrado por primera vez en Cuba). En general, el genotipo más frecuente resultó el AFLP1, seguido a gran distancia por AFLP2 y AFLP3. La tipificación empleando los microsatélites reveló múltiples grupos de genotipos relacionados que fueron denominados complejos de microsatélites. Los aislamientos holandeses mostraron una alta diversidad genética entre los especímenes del serotipo A, la cual fue menor para el serotipo D; la mayoría de aislamientos ambientales cubanos (70 %) quedaron agrupados en un complejo de microsatélites en el que sólo resultaron representados unos pocos aislamientos clínicos (49 ambientales vs 2 clínicos), mientras que los aislamientos clínicos se encontraban segregados en múltiples complejos de microsatélites. Estos hechos llevaron a la conclusión de que el panel seleccionado de marcadores de microsatélites resulta una excelente herramienta para el estudio de la epidemiología molecular de C. neoformans var. grubii; que existe una gran variación genotípica entre estos aislamientos clínicos; y a enunciar la hipótesis sobre la existencia de una fuente de criptococosis humanas 13 adicional al guano de paloma en Cuba.

La caracterización mediante AFLP reveló que uno de los pacientes holandeses estaba infectado por múltiples genotipos, mientras que el empleo de microsatélites

141 Chapter 13 demostró en una serie de 19 aislamientos de siete pacientes cubanos con meningitis criptocócica recurrente, 14 perfiles distintivos: en tres pacientes la infección se asoció con aislamientos idénticos y los restantes pacientes tuvieron recaídas por aislamientos que eran genotípicamente diferente. Se determinó la sensibilidad antifúngica frente a anfotericina B, fluconazol, 5-fluorocitosina, itraconazol, voriconazol, posaconazol e isavuconazol por el método de microdilución en caldo según el documento M27-A3 del Instituto de Estándares para Laboratorios Clínicos (CLSI, de sus siglas en inglés). Entre todos los aislamientos holandeses estudiados, uno mostró alta concentración inhibitoria mínima (CIM) para 5-fluorocitosina (> 64 g/mL) y disminución dela sensibilidad (> 16 g/mL) para 5-fluorocitosina y fluconazol en 9 y 10 aislamientos de C. neoformans, respectivamente. Ninguno de los aislamientos cubanos resultó resistente a los fármacos estudiados. Estos hallazgos sugieren que la recurrencia de la meningitis criptocócica en estos pacientes no se asoció con resistencia del aislamiento original a los medicamentos, sino a una infección inicial con diferentes aislamientos o una reinfección con un aislamiento genotípicamente dierente.

El manejo de la criptococosis ha sido uno de los más estudiados entre las enfermedades fúngicas; no obstante, todavía hay muchos aspectos sin resolver. A pesar de los avances médicos, la tasa de mortalidad durante los primeros tres meses de tratamiento es superior al 20 % y se acerca al 100 % en los pacientes que no reciben la terapia adecuada. La Parte II de la tesis se dedica al estudio de la susceptibilidad de una colección amplia de aislamientos de Cryptococcus spp. de origen humano, veterinario y ambiental frente al isavuconazol (triazol de generación reciente) y otros seis fármacos antifúngicos.

Ciento sesenta y cinco C. neoformans var. grubii cubanos y una colección mundial de 350 aislamientos de C. gattii fueron enfrentados a anfotericina B, 5-fluorocitosina, fluconazol, itraconazol, voriconazol, posaconazol e isavuconazol por el método de microdilución en caldo según el documento M27-A3 del CLSI. De los siete compuestos estudiados, voriconazol y posaconazol mostraron fuerte actividad antifúngica in vitro

(hasta 0,25 g/mL) mientras que isavuconazol exhibió CIM90 aún más baja (hasta 0,125 g/mL). La anfotericina B presentó un valor de CIM90 aceptable (hasta 0,5 g/mL), pero

5-fluorocitosina y fluconazol mostraron MIC90s de hasta 8 g/mL. Los resultados indican que, dada la naturaleza de los nuevos azoles (voriconazol, posaconazol e isavuconazol), su biodisponibilidad oral y buena tolerancia, estos podrían llegar a ser una adición importante al arsenal de agentes antifúngicos para el tratamiento de la criptococosis. Hasta donde sabemos, estos constituyen los primeros informes de evaluación de isavuconazol frente a Cryptococcus spp., fármaco antifúngico que resultó más eficaz en términos de actividad in vitro.

Adicionalmente los aislamientos de C. gattii fueron divididos en siete genotipos mediante AFLP, incluyendo los híbridos interespecíficos AFLP8 y AFLP9. La mayoría de los aislamientos clínicos (n = 215) correspondieron a genotipos AFLP4 (n = 76) y AFLP6 (n = 103). Los correspondientes a AFLP6 mostraron una media geométrica de los valores de CIM significativamente superiores para 5-fluorocitosina y fluconazol en comparación

142 Summary - Resumen con los del genotipo AFLP4. Esto sugiere que la susceptibilidad a los antifúngicos debe ser de tipo especie- y genotipo- específica. Esto fue corroborado por el estudio internacional de las distribuciones de punto final de la susceptibilidad antifúngica y los valores de corte epidemiológicos (ECV, de sus siglas en inglés) de aislamientos salvajes (WT, de sus siglas en inglés). Tales parámetros se establecieron recientemente por Espinel - Ingroff y colaboradores para la anfotericina B y 5-fluorocitosina frente al complejo de especies C. neoformans/C. gattii y en el presente trabajo se establecen para fluconazol, itraconazol, posaconazol y voriconazol. Los resultados ponen de manifiesto la utilidad de los puntos de corte como una herramienta práctica para detectar la resistencia entre los aislamientos de Cryptococcus, y se corrobora que dichos valores son específicos para cada especie y tipo molecular: fluconazol (8 a 32 µg/mL), itraconazol (0,25 a 1 µg/mL), posaconazol (0,25 a 0,5 µg/mL) y voriconazol (0,12 a 0,25 µg/mL). Se sugirió incluir estos ECV en la versión revisada del documento M27-A3 del CLSI y mantener la vigilancia de la susceptibilidad a fin de corroborar o ampliar los datos ofrecidos en el estudio, así como llevar a cabo una investigación para determinar la relación entre los mecanismos moleculares de resistencia a los triazoles y el ECV propuesto.

C. neoformans y C. gattii han sido ampliamente aislados a partir de fuentes naturales. Aunque el último se caracteriza por tener una distribución más restringida a los climas tropicales y subtropicales, el brote ocurrido en América del Norte indica la adaptación de este organismo a nuevos ambientes. Esto, junto con el intervalo altamente variable entre la exposición al microorganismo y el desarrollo de la enfermedad clínica, hace más difícil llegar a conclusiones sobre la epidemiología de la criptococosis. La Parte III de la tesis consiste en el análisis de aislamientos del complejo de especies C. neoformans/C. gattii recuperados de humanos, animales y plantas en Cuba, así como estudios de patogenicidad ex-vivo de ambas especies.

Rutinariamente sólo C. neoformans var. grubii suele ser aislado a partir de muestras clínicas y de las excretas de palomas en Cuba. Adicionalmente, evidencias moleculares recientes sugirieron la presencia de otras fuentes naturales de infección por esta levadura además de las investigadas. Estos hechos motivaron el estudio de 662 muestras obtenidas de 331 árboles y cactus de La Habana, Cuba. La selección inicial de los aislamientos de Cryptococcus spp. se realizó mediante técnicas convencionales y arrojó 121 C. neoformans y 61 C. gattii. La caracterización posterior mediante AFLP y secuenciación de ADN, demostró que ninguno de los aislamientos correspondía a C. gattii; sólo uno resultó ser C. neoformans var. grubii; 27 fueron identificados como otras especies diferentes; y 65 aislamientos no identificables resultaron segregados en diez grupos genéticos no descritos anteriormente. Hasta ahora los intentos de recuperación de C. gattii en muestras ambientales en Cuba han sido infructuosos, por lo que estos resultados destacan la necesidad de continuar los estudios epidemiológicos para 13 determinar la distribución exacta de C. neoformans y C. gattii en la isla, así como la importancia de aplicar los métodos moleculares para confirmar la identificación de los aislamientos recuperados de fuentes ambientales.

143 Chapter 13

La Parte III también incluye la descripción de los dos primeros casos de C. gattii documentados en Cuba: a partir de un guepardo (Acinonyx jubatus) del Zoológico Nacional, en La Habana y de un paciente inmunocompetente cubano. Ambos casos tenían antecedentes de viaje (África del Sur y América Central, respectivamente). El análisis molecular mostró que el aislamiento del animal pertenece al genotipo AFLP4 y el del humano a AFLP5. Las secuencias genómicas obtenidas por secuenciación multilocus (MLST, de sus siglas en inglés) mostraron la semejanza con cepas de Kenia y Colombia, respectivamente, lo que sugiere que ambos casos tenían una infección latente en el momento de su arribo a Cuba y que este método es excelente para trazar el origen geográfico de los aislamientos.

Aunque C. neoformans y C. gattii comparten la mayoría de las propiedades que les confieren patogenicidad, esta última especie parece ser más virulentas, ya que es la principal responsable de la enfermedad en huéspedes aparentemente inmunocompetentes. Al final de la Parte III se describe una serie de ensayos con el objetivo de demostrar que la capacidad de C. gattii para causar enfermedad en los seres humanos depende de una respuesta innata distintiva de citoquinas. Con este fin, se evaluó el perfil de citoquinas producidas por células mononucleares de humanos obtenidas de sangre periférica de individuos sanos las cuales fueron estimuladas con 40 aislamientos de C. gattii, C. neoformans y varios híbridos inactivados por calor. Además, se investigó la participación de TLR2, TLR4 y TLR9 en la respuesta de citoquinas pro- inflamatorias durante la estimulación con C. gattii. Los resultados sugirieron que los aislamientos clínicos de C. gattii son capaces de inducir una respuesta de citoquinas inflamatorias (IL-1b, TNF α e IL-6) y Th17/22 más pronunciada en comparación con aislamientos de C. gattii no clínicos y otras especies de Cryptococcus, la cual parece ser dependiente de los receptores celulares TLR4 y TLR9. En este estudio no fue objetivo identificar qué componentes de la pared celular están implicados en la respuesta de citoquinas inicial lo que debe ser investigado más adelante.

A pesar de los avances en el manejo de los pacientes con sida y el desarrollo de potentes antifúngicos, la tasa de morbimortalidad de la criptococosis a nivel mundial se mantiene por encima de límites aceptables. Esperamos que los estudios en relación con el complejo de especies C. neoformans/C. gattii comprendidos en esta tesis, contribuyan a mejorar la comprensión de estas levaduras patógenas y la infección que causan.

144 Chapter 4 Acknowledgments, Curriculum1 vitae and List of publications Chapter 14

Acknowledgments

The acknowledgment pages would exceed the thesis itself if I would mention all the names of those persons who somehow have contributed to the completion of this work. However, there are some names which I cannot fail to mention.

To the wonderful researchers from across the Atlantic, that even without knowing me have opened their doors and gave me this wonderful opportunity: To my mentor Dr. Jacques F. Meis, for his wisdom, friendship, carefulness, hospitality without limits, and confidence. To Andreas and Teun, thanks a lot for such a wonderful and productive collaboration. I sincerely hope that the conclusion of this chapter does not mean the end of the relationship and collaboration.

To my teachers, colleagues and friends Gerardo and Carlos, from whom I learned not only mycology. To you, my eternal gratitude, admiration and affection. To the rest of my team: Mayda, Iraida and Ernesto, to all of you that have been always present and who have offered me experience and youths.

Ferry and Ilse, two indispensable persons along this race. Thank you for your guidance, invaluable help, warm friendship and my first Dutch words (goedemorgen, dank u wel, alsjeblieft, doei…). You will be forever in my thoughts and my heart.

To Iliana, Dihadenys and Gilda because without your unconditional support and solidarity, not only during the course of this work, would have been much more difficult to exist. Thanks for knowing me so well and make a blind eye to my forgetfulness, for your words of encouragement in each of my lost battles.

And what I would have been without Gato? Thanks for being, for sharing and for getting me out of whatever trouble there were.

To persons who join me right now or not, but whose timely and inestimable friendship have enriched my spiritual life and already are part of who I am: Kenneth & Xenia, Corina, Debby, Adriana, Hanny, Jannie, Brenda, Tom, Teske, Tina and personnel of the clinical microbiology department at CWZ. Like stars on the firmament all of you have your own brightness on my personal sky.

Can’t find precise words to define what Dagmar, Sander, Rachid & family, Philip and Hans & Esther means to me; thank you for being, for the privilege to let me meet you and for making me feel safe and warm at home while I was away from mine.

I am particularly indebted to the Cuban Health Ministry, to the General Directorate and Academic Department of the Tropical Medicine Institute “Pedro Kourí” headed by Dr. Jorge Pérez and Dra. Nereyda Cantelar respectively, with members of the Bacteriology- Mycology Department and my IPK-family, for their support, continued interest and contributions to make this dream come true; thanks a lot!!

146 Acknowledgments, Curriculum vitae and List of publications

To a teacher’s teacher, Dra. Alina Llop and our unforgettable Professor Dr. Gustavo Kouri, from whom I have learned to love not only science, but also humanity. At this point I pause to confess, I feel lucky to have had the opportunity to somehow coincide in time and space with you. Not only my family and friends have made me grow as a person, your presence during this short time in my life, dear Professors, will mark the rest of it.

To my parents and children, responsible of my existence and the desire to be every day better; thank you for your limitless love without which would not have been possible to achieve this worthy goal. I feel immensely fortunate to have the most wonderful possession: My family.

To everyone who have encouraged me, have been complicit and have told me ... yes you can!, I don’t know if a simple “thanks” will be enough to express my gratitude to everything you have done for me.

147 Chapter 14

Curriculum vitae

María Teresa Illnait Zaragozí was born on March 10th, 1967 in Havana, Cuba. In this city she followed all her studies including the University at the High Institute of Medical Sciences. In 1991 she became a Medical Doctor and started specialization as Clinical Microbiologist (1991-1995) at the Tropical Medicine Institute “Pedro Kouri” (IPK) Havana, Cuba. In 2005 she came in contact with researchers from the Canisius Wilhelmina Hospital in Nijmegen to continue her ongoing experiments on Cryptococcus and cryptococcosis. The results obtained during her visits to the Canisius-Wilhelmina Hospital and Radboudumc, Nijmegen are part of this thesis. During her career as Mycologist she received national awards from the Cuban Science Academy (in 2008, 2011, 2013) and the Annual Health Ministry Prize (in 2007, 2011) as well as the Medals of the 50th, 60th and 70th Anniversary of IPK among others. She has obtained different scientific degrees and categories: Master of Bacteriology- Mycology (1996), Auxiliary professor (2006) and Titular Researcher (2012). Since April 2012, she is acting head of the Bacteriology-Mycology Department at IPK. She is the proud parent of two sons, Jose Carlos and Carlitos.

148 Acknowledgments, Curriculum vitae and List of publications

List of publications

1. Badali H, Fernández-González M, Mousavi B, Illnait Zaragozí MT, González- Rodríguez JC, de Hoog GS, Meis JF. Chromoblastomycosis due to Fonsecaea pedrosoi and F. monophora in Cuba. Mycopathologia 2013; 175(5-6):439-44.

2. Schoffelen T, Illnait Zaragozi MT, Joosten LAB, Netea MG, Boekhout T, Meis JF, Sprong T. Cryptococcus gattii induces a cytokine pattern that is distinct from other cryptococcal species. Plos One 2013; 8(1):e55579.

3. Illnait Zaragozí MT, Ortega-Gonzalez LM, Hagen F, Martínez-Machin GF, Meis JF. Fatal Cryptococcus gattii genotype AFLP5 infection in an immunocompetent Cuban patient. Med Mycol Case Rep. 2013; 2:48-51

4. Illnait Zaragozí MT, Martínez GF, Fernández CM, Marchena JJ, Perurena MR. Contribution to the study of cryptococcosis in children in Cuba. Rev Cub Med Trop 2013; 65(1): 6-63. Spanish

5. Espinel-Ingroff A, Aller AI, Canton E, Castañón-Olivares LR, Chowdhary A, Cordoba S, Cuenca-Estrella M, Fothergill A, Fuller J, Govender N, Hagen F, Illnait Zaragozí MT, Johnson E, Kidd S, Lass-Flörl C, Lockhart SR, Martins MA, Meis JF, Melhem MS, Ostrosky-Zeichner L, Pelaez T, Pfaller MA, Schell WA, St-Germain G, Trilles L, Turnidge J. Cryptococcus neoformans-Cryptococcus gattii species complex: an international study of wild-type susceptibility endpoint distributions and epidemiological cutoff values for fluconazole, itraconazole, posaconazole, and voriconazole. Antimicrob Agents Chemother 2012;56(11):5898-906.

6. Hagen F, Illnait Zaragozí MT, Meis JF, Chew WH, Curfs-Breuker I, Mouton JW, Hoepelman AI, Spanjaard L, Verweij PE, Kampinga GA, Kuijper EJ, Boekhout T, Klaassen CH. Extensive genetic diversity within the Dutch clinical Cryptococcus neoformans population. J Clin Microbiol 2012;50(6):1918-26.

7. Illnait Zaragozí MT, Martínez GF, Fernández CM, Perurena MR, Theelen B, Boekhout T, Meis JF, Klaassen CH. Environmental isolation and characterisation of Cryptococcus species from living trees in Havana city, Cuba. Mycoses 2012;55(3):e138-44.

8. Illnait Zaragozí MT, Martínez GF, Fernández CM, Perurena MR, Rodríguez I, Jerez LE, Monroy EX, Valdés EA. Cryptococcosis and its causal agent. Contributions to the epidemiology and pathogenesis. Boletín CNSCS 2011; No. 2. ISSN 2073-9281. Spanish 14

149 Chapter 14

9. Illnait Zaragozí MT, Hagen F, Fernández CM, Martínez GF, Polo JL, Boekhout T, Klaassen CH, Meis JF. Reactivation of a Cryptococcus gattii infection in a cheetah (Acinonyx jubatus) held in the National Zoo, Havana, Cuba. Mycoses 2011;54(6):e889-92.

10. Illnait Zaragozí MT, Gato R, Martínez GF, Otero A, Sarracent J, Rodríguez H, Fernández CM, Valdés IC. Effect of 4B3 monoclonal antibody on the experimental Cryptococcus neoformans infection. Rev Cub Med Trop 2011;63(2):123-9. Spanish

11. Illnait Zaragozí MT, Martínez GF, Fernández CM, Hagen F, Boekhout T, Klaassen CH, Meis JF. Microsatellite typing and susceptibilities of serial Cryptococcus neoformans isolates from Cuban patients with recurrent cryptococcal meningitis. BMC Infect Dis 2010;10:289-96.

12. Hagen F, Illnait Zaragozí MT, Bartlett KH, Swinne D, Geertsen E, Klaassen CH, Boekhout T, Meis JF. In vitro antifungal susceptibilities and amplified fragment length polymorphism genotyping of a worldwide collection of 350 clinical, veterinary, and environmental Cryptococcus gattii isolates. Antimicrob Agents Chemother 2010;54(12):5139-45.

13. Illnait Zaragozí MT, Martínez GF, Fernández CM, Boekhout T, Meis JF, Klaassen CH. Microsatellite typing of clinical and environmental Cryptococcus neoformans var. grubii isolates from Cuba shows multiple genetic lineages. PLoS One 2010;5(2):e9124.

14. Najafzadeh MJ, Badali H, Illnait Zaragozí MT, De Hoog GS, Meis JF. In vitro activities of eight antifungal drugs against 55 clinical isolates of Fonsecaea spp. Antimicrob Agents Chemother 2010;54(4):1636-8.

15. Illnait Zaragozí MT, Illnait-Ferrer J, Blanco A. Antifungal effect of Petiveria alliacea L extract. Rev CENIC de Ciencias Biológicas 2010;41(1):79-82. Spanish

16. Fernández CM, Martínez GF, Illnait Zaragozí MT, Perurena MR, González L. Occupational histoplasmosis outbreaks in Havana province. Rev Cub Med Trop 2010;62(1):68-72. Spanish

17. Polo JL, Fernández CM, Martínez GF, Illnait Zaragozí MT, Perurena MR. Cryptococcus gattii isolated from a cheetah (Acinonyx jubatus) of Cuba’s National Zoo. Rev Cub Med Trop 2010;62(3):257-60. Spanish

18. Prieto LM, Díaz LA, Illnait Zaragozí MT, Perurena MR, Cantelar N, Fernández CM, Martínez GF. Susceptibility to nystatin of oral Candida isolates and their correlation with treatment response. Rev Cub Med Trop 2010;62(3):237-44. Spanish

150 Acknowledgments, Curriculum vitae and List of publications

19. Fernández CM, Martínez GF, Illnait Zaragozí MT, Perurena MR. Animal histoplasmosis. Actualización. Cuba Zoo 2009;19(2):45-54. Spanish

20. De Barros JD, Do Nascimento SM, De Araújo FJ, Braz Rde F, Andrade VS, Theelen B, Boekhout T, Illnait Zaragozí MT, Gouveia MN, Fernandes MC, Monteiro MG, De Oliveira MT. Kodamaea (Pichia) ohmeri fungemia in a pediatric patient admitted in a public hospital. Med Mycol 2009;47(7):775-9.

21. Illnait Zaragozí MT, Martínez GF, Curfs-Breuker I, Fernández CM, Boekhout T, Meis JF. In Vitro activity of the new azole isavuconazole (BAL4815) compared with six other antifungal agents against 162 Cryptococcus neoformans isolates from Cuba. Antimicrob Agents Chemother 2008;52(4):1580-2.

22. Verduyn Lunel FM, Curfs-Breuker I, Illnait Zaragozí MT, Meis JF. Fluconazol- en voriconazol gevoegheid van klinische gist isolaten. Resulten van negen jaar Nijmeegs onderzoek. Ned Tijdschr Med Microbiol 2008;16(1):12-9. Dutch

23. Fernández CM, Martínez GF, Illnait Zaragozí MT, Perurena MR, Águila A, Brito M. In vitro susceptibility of Candida strains against fluconazole and amphotericin B. Rev Cub Med Trop 2007;59(2):113-8. Spanish

24. Gato R, Martínez GF, Rodríguez H, Otero A, Sarracent J, Illnait Zaragozí MT. Recognition of intact cells from Cryptococcus neoformans by the anti-glucuronoxylomannan 4B3 monoclonal antibody. Rev Cub Med Trop 2006;58(2):162-4. Spanish

25. Fernández CM, Martínez GF, Perurena MR, Illnait Zaragozí MT, Valdés IC. The collection of fungi cultures of “Pedro Kourí” Institute of Tropical Medicine. Rev Cub Med Trop 2005;57(3):214-8. Spanish

26. Rodríguez H, Pupo M, Illnait Zaragozí MT, Otero A, Martínez GF. Monoclonal antibodies that recognize Cryptococcus neoformans capsular polysaccharide. Rev Cub Med Trop 2005;57(2):162-4. Spanish

27. Martínez GF, Barrial L, Illnait Zaragozí MT, Valdés IC, Fernández CM, Perurena MR, Polo JL, Mendoza D. Usefulness of D-proline in the differentiation of of Cryptococcus neoformans varieties. Rev Cub Med Trop 2004;56(1):77-9. Spanish.

28. Illnait Zaragozí MT, Toirac YM, Valdés Hernández IC, Machín GM, Fernández CM, Perurena Lancha MR, Llanes DM. Obtention of specific antisera for the serotyping of Cryptococcus neoformans var. neoformans. Rev Cub Med Trop 2004;56(2):142-4. Spanish 14

151 Chapter 14

29. Illnait Zaragozí MT, Valdés IC, Fernández CM, Perurena MR, Martínez GF. LD50 determination of Cryptococcus neoformans strain using different inoculation rate. Rev Pan Infectol 2004;6(1):8-11. Spanish

30. Illnait Zaragozí MT, Tamargo I, Martínez GF, Fuentes K, Valdés IC, Rodriguez F. Central nervous system infection by Cryptococcus neoformans and Streptococcus pneumoniae in a seronegative negative HIV patient. Rev Pan Infectol 2004;6(2):54- 56. Spanish

31. Fernández CM, Martínez A, Echemendia Y, Martínez GF, Perurena MR, Illnait Zaragozí MT. Histoplasma capsulatum var. capsulatum in vitro susceptibility to amphotericin B, ketoconazole, itraconazole and fluconazole. Rev Cub Med Trop 2003;55(2):76-82. Spanish

32. Valdés IC, Martínez GF, Fernández CM, Illnait Zaragozí MT. Pigmentation study of Cryptococcus neoformans strains on sunflower seed agar. Rev Cub Med Trop 2003;55(2):119-20. Spanish

33. Illnait Zaragozí MT, Vilaseca JC, Fernández CM, Martínez GF. Enzyme- linked immunosorbent assay for detection and quantification of Cryptococcus neoformans antigen. Mem Inst Oswaldo Cruz 2001;96:241-245.

34. Illnait Zaragozí MT, Valdés IC, Martínez GF, Fernández CM. Cryptococcosis. A necessary alert. Rev Cub Med Trop 2001;53(3):224-5. Spanish

35. Martínez GF, Illnait Zaragozí MT. Opportunistic mycoses in AIDS patients. In Microbiología y Parasitología Medicas Section IV, Tomo I, Editorial Ciencias Médicas 2001 pp. 541-50. Spanish

36. Fernández CM, Martínez GF, Illnait Zaragozí MT, Perurena MR, González M. The identification of Cryptococcus neoformans var. neoformans in Cuban clinical isolates. Rev Cubana Med Trop 1998;50(2):167-9. Spanish.

37. Fernández CM, González M, Illnait Zaragozí MT, Martínez GF. Minimum inhibitory concentration of amphotericin B in yeasts of medical interest. Rev Cubana Med Trop. 1998;50(1):48-53. Spanish.

38. Illnait Zaragozí MT, Martínez GF, Fernández CM. Isolation of anti-Cryptococcus neoformans immunoglobulin G. Rev Cubana Med Trop 1998;50(1):27-30. Spanish

152