OUP user menu

Genotype and mating type analysis of Cryptococcus neoformans and Cryptococcus gattii isolates from China that mainly originated from non-HIV-infected patients

Xiaobo Feng, Zhirong Yao, Daming Ren, Wanqing Liao, Jingsong Wu
DOI: http://dx.doi.org/10.1111/j.1567-1364.2008.00422.x 930-938 First published online: 1 September 2008


Cryptococcosis has been reported to be mostly associated with non-HIV-related patients in China. However, little is known about the molecular characteristics of clinical isolates from the Cryptococcus neoformans species complex in this country. In this study, 115 clinical isolates were included. Molecular type VNI was the most representative (n=103), followed by VGI (n=8), VNIII (n=2), VNIV (n=1), and VGII (n=1). With the exception of a serotype D mating type a isolate, all possessed the MATα locus. Multilocus sequence typing (MLST) revealed that most Cryptococcus gattii isolates from China shared identical MLST profiles with the most common MLST genotype reported in the VGI group, and the only one VGII isolate resembled the Vancouver Island outbreak minor genotype. The C. gattii strains involved in this study were successfully grouped according to their molecular type and mating types by PCR-restriction fragment length polymorphism (RFLP) analysis of the GEF1 gene. Our results suggest that (1) in China, cryptococcosis is mostly caused by C. neoformans var. grubii (molecular type VNI), and mating type α; (2) The most common causative agents of C. gattii infection in China are closely related to a widely distributed MLST genotype; and (3) The PCR-RFLP analysis of the GEF1 gene has the potential to identify the molecular and mating types of C. gattii simultaneously.

  • Cryptococcus neoformans
  • cryptococcosis
  • molecular epidemiology
  • mating type
  • genotype
  • restriction fragment length polymorphism


The Cryptococcus neoformans species complex is the causative agent of cryptococcosis and has been classified into two species recently, namely C. neoformans and Cryptococcus gattii, based on phenetic, biological and molecular differences. Within C. neoformans, C. neoformans var. grubii (serotype A), C. neoformans var. neoformans (serotype D) and AD hybrids are included, whereas within C. gattii (serotypes B and C), no varieties have been designated (Kwon-Chung et al., 2002; Kwon-Chung & Varma et al., 2006). Cryptococcus neoformans is distributed globally, mainly responsible for infections in immuno-compromised patients. Cryptococcus gattii, however, mainly infects immuno-competent individuals and has been considered to be restricted to tropical and subtropical areas (Casadevall & Perfect et al., 1998), but its enlarged distribution has been reported in recent studies (Fraser et al., 2003, 2005; Kidd et al., 2005, 2007). In molecular typing studies, eight major molecular types have been established, namely VNI to VNIV and VGI to VGIV. VNI to VNIV belong to C. neoformans and VGI to VGIV to C. gattii (Boekhout et al., 2001; Meyer et al., 2003; Bovers, 2008a, b). In addition, hybrids between C. neoformans (VNIV) and C. gattii (VGI) have been recovered from clinical sources recently, indicating the existence of this hybrid in nature (Bovers et al., 2006, 2008a, b). As differences have been reported in the biology, virulence and epidemiology of the molecular types, more studies on the population structure and molecular epidemiology of this pathogen have been performed globally (Boekhout et al., 2001; Casali et al., 2003; Fraser et al., 2003, 2005; Meyer et al., 2003; Campbell et al., 2005; Jain et al., 2005; Litvintseva et al., 2005; Tay et al., 2006).

A bipolar mating system, consisting of mating types α and a, has been found in both C. neoformans and C. gattii. Compared with strains of mating type a, strains of mating type α occur more predominantly in environmental and clinical sources (Kwon-Chung & Bennett et al., 1978; Casali et al., 2003; Fraser et al., 2003; Kidd et al., 2004; Campbell et al., 2005; Chaturvedi et al., 2005; Litvintseva et al., 2005). Strains of mating types α have been found to be more virulent than congenic a strains in serotype D (Kwon-Chung et al., 1992). Conversely, the serotype A congenic strains of mating type α and a exhibited equivalent virulence in animal models. However, it was demonstrated that MATα strains preferentially disseminated to the central nervous system during coinfection with MATa strains (Nielsen et al., 2003, 2005).

Cryptococcus neoformans is associated with avian feces and bird migration, but the sibling species C. gattii has been considered to inhabit Eucalyptus and other tree species and causes endemic infection in tropical and subtropical areas (Callejas et al., 1998; Casadevall & Perfect et al., 1998; Trilles et al., 2003; Kidd et al., 2004; Litvintseva et al., 2007; Nielsen et al., 2007). The recent outbreak of C. gattii in Vancouver Island (BC, Canada), which is located in a temperate climatic zone (Hoang et al., 2004), has attracted greater attention to this primary pathogen. This outbreak was found to be mainly caused by strains of genotype VGIIa (=AFLP genotype 6A) and mating type α (Kidd et al., 2004). To further understand the origin and extent of this outbreak, epidemiologic studies have been carried out to elucidate the route of transmission and survey the spread of this pathogen (Kidd et al., 2005; Lindberg et al., 2007; MacDougall et al., 2007). In those studies, identical genotypes were determined from different regions by multilocus sequence typing (MLST), suggesting extensive strain dispersal (Fraser et al., 2005; Kidd et al., 2005; MacDougall et al., 2007).

Because of the ongoing HIV pandemic, the morbidity and mortality of cryptococcosis are increasing, especially in South Asia and Africa (Bogaerts et al., 1999; Banerjee et al., 2001). As a result, several studies focused on this pathogen among HIV-infected individuals. Cryptococcus neoformans var. grubii was found to be the major (more than 99%) causative agent (Mitchell & Perfect et al., 1995; Casadevall & Perfect et al., 1998), whereas C. gattii was recovered recently as well from these patients (Chaturvedi et al., 2005; Litvintseva et al., 2005). In China, unlike other parts of the world, cases of cryptococcal infection have been reported mostly from non-HIV-infected patients (Lu et al., 2005; Yao et al., 2005; Chen et al., 2008). Until recently, epidemiologic surveys were restricted mainly to serotype analysis by homemade antisera (Li et al., 1987). In another study, Chen (2008) studied 129 clinical Cryptococcus isolates from China, of which C. neoformans var. grubii, molecular type VNI, possessed the majority of the isolates studied (93.0%), while the remaining strains were C. gattii (VGI, 7.0%).

In order to further analyse the population structure and the molecular epidemiology of the C. neoformans species complex in China, a large-scale collection of clinical isolates from six regions was undertaken. Our objectives were to demonstrate the population structure of clinical representatives of the C. neoformans species complex, and to reveal epidemiological links among C. gattii isolates from China and other geographic areas.

Materials and methods

Cryptococcal isolates from China

One hundred and fifteen clinical isolates were collected from 14 hospitals from six regions in China. The regional distribution of the isolates was as follows: East China (n=77), South China (n=21), Central China (n=7), North China (n=6), South West China (n=3), and North East China (n=1) (Fig. 1). Among these isolates, 105 were from subtropical and 10 from temperate regions. The isolates were cultured from 109 patients (Note: six were sequential isolates), and 75 out of 78 (96.2%) of the isolates were from HIV-negative patients. Among the 115 isolates, 94 (81.7%) were sampled from CSF, nine (7.8%) from blood, four (3.5%) from sputum, and eight (7.0%) from other tissues (see Supporting Information, Table S1). All isolates were identified as C. neoformans or C. gattii by morphology, biochemical and physiological tests, such as microscopic observation of India ink preparations, colony appearance on l-3,4-dihydroxyphenylalanine (l-DOPA) agar, positive urease reaction, and the ability to grow at 37 °C. Canavanine–glycine–bromothymol blue agar was used to differentiate C. gattii from C. neoformans as described previously (Kwon-Chung et al., 1982).


The distribution of cryptococcal isolates from China involved in this study. The constitution of molecular types from six regions of China is shown in brackets. Geographic distribution of subtropical, tropical and temperate regions in China is displayed by three different shadings.

Reference strains

The standard strains H99 (=CBS8710) (serotype A, MATα), KN99a (serotype A, MATa), JEC21 (=CBS10513) (serotype D, MATα), JEC20 (=CBS10511) (serotype D, MATa), WM276 (=CBS10510) (serotype B, MATα), E566 (serotype B, MATa), NIH312 (serotype C, MATα) and B4546 (serotype C, MATa) were used as control strains to determine the serotype and mating types; the reference strains representing each of the eight major molecular types were as follows, WM148 (=CBS10085) (VNI), WM626 (=CBS10084) (VNII), WM628 (=CBS10080) (VNIII), WM629 (=CBS10079) (VNIV), WM179 (=CBS10078) (VGI), WM178 (=CBS10082) (VGII), WM161 (=CBS10081) (VGIII), WM779 (=CBS10101) (VGIV); other reference strains involved in this study are listed in Tables S2 and S3.

DNA manipulation and genotyping

High molecular weight DNA was extracted as described previously (Meyer et al., 2003). Molecular types were determined by PCR fingerprinting with single primer of the minisatellite-specific core sequence of the wild-type phage M13 (5′-GAGGGTGGCGGTTCT-3′), and amplifications were performed as originally described by Meyer (2003). PCR was performed in a Bio-Rad thermal cycler (BIO-RAD, Hercules, CA). Amplification products were separated by electrophoresis in 1.4% agarose gels in 1 × Tris-Acetate-EDTA (TAE) buffer at 80 V for 12 cm, and visualized under UV light. The molecular types (VNI–VNIV and VGI–VGIV) were assigned by comparison to the major bands of the reference strains representing eight molecular types loaded on gel. The reference strains were used as internal controls to confirm reproducibility. All visible bands were included in the analysis, independent of their intensity.

Determination of serotype and mating type by PCR

Serotype and mating type were established by PCR with primers specific to the STE20α A (JOHE2472–JOHE2578), STE20α D (JOHE3069–JOHE3070), STE20a A (JOHE5169–JOHE5170), STE20a D (STE20a.D.F–STE20a.D.R), STE12α (JOHE1671–JOHE1672), STE20a (JOHE9421–JOHE9422), SOD1 B/C (JOHE7773–JOHE7775), GPA1 D (JOHE2596–JOHE3240), GPA1 A (JOHE2596–JOHE3241), CLA4 D (JOHE3065–JOHE3066), and CLA4 A (JOHE3066–JOHE3236) genes, as described previously (Lengeler et al., 2001; Fraser et al., 2003; Keller et al., 2003; Barreto de Oliveira et al., 2004; D'Souza et al., 2004). All PCR reactions were carried out in duplicate. The PCR products were electrophoresed on 1% or 1.5% agarose gels in 1 × TAE buffer at 100 V for 30 min and then visualized under UV light.

Mating assays

To test the mating type and mating ability of strains of C. gattii from China, serotype D fertile tester strains JEC20 (MATa) and JEC21 (MATα), serotype C fertile tester strains B4546 (MATa) and NIH312 (MATα), and their crg1Δ mutant derivatives JF109 (MATa) and JF101 (MATα) were used in mating assays. The procedure was conducted as described by Campbell (2005). Mating ability was assessed by observation of filament formation around the colony, and confirmed by observation of filaments, basidia and basidiospore formation under light microscopy.

Multigene sequencing

Fragments of four unlinked loci, including IGS1, PLB1, GEF1 and GPD1 genes, were sequenced and analyzed. Primer pairs designed to amplify and sequence the DNA fragments were as follows: for IGS1, IGS1F (5′-CTACGATCCACTGAGGCTAAGC-3′) and IGS1R (5′-AGCGACATCGAGACTGTATGC-3′); for PLB1, PLB1F (5′-ATTGAGCTTCAGGCGGAGAG-3′) and PLB1R (5′-AGCAACAACAGAGGGAAGGC-3′); for GEF1, GEF1CXYF (5′-GCATAGACTTTGCTTCTTCTGGTAG-3′) and GEF1R (5′-GGCTTCCATTGCT GATGAGTC-3′); and for GPD1, JOHE14968 (5′-CCACCGAACCCTTCTAGGATA-3′) and JOHE14969 (5′-CTTCTTGGCACCTCCCTTGAG-3′) (Fraser et al., 2005). PCR reaction procedure included denaturation at 94 °C for 5 min, followed by 35 cycles of 94 °C for 30 s, 57 °C for 30 s, and 72 °C for 1 min for all primers listed above. PCR products were run on 1% agarose gel and purified using DNA purification kit (TaKaRa, Dalian, CN), and sequenced using an ABI 3730 sequencer with the Big Dye Terminator cycle sequencing kit (Applied Biosystems, Forester City, CA).

Alleles of IGS1, PLB1 and GPD1 were sequenced for C. gattii isolates from China, and aligned with those published by Fraser (2005). GEF1 alleles were sequenced for all C. gattii isolates involved here, except for reference strains WM276, E566 and R265 (sequence data from GenBank).

PCR-restriction fragment length polymorphism (RFLP) analysis

The GEF1 gene, located within the MAT locus of C. neoformans (Fraser et al., 2004), was selected for RFLP analysis. The GEF1 sequences of reference strains representing distinct molecular and mating types, including strains WM276 (VGI MATα), E566 (VGI MATa) and R265 (=CBS10514) (VGII MATα), were obtained from GenBank (accession numbers AY710430, AY710429 and DQ096809), and those representative of the other molecular and mating types, including reference strains CBS1930 (VGII MATa), WM161 (VGIII MATα), B4546 (VGIII MATa) and WM779 (VGIV MATα), were generated in our study. Primers specific to the GEF1 alleles of C. gattii, GEF1F (5′-TCGCCCTTGTCTGACCTTG-3′) and GEF1R (5′-GGCTTCCATTGCTGATGAGTC-3′) were designed based on the consensus nucleotide sequences of the GEF1 gene of the reference strains listed above, and the specificity of this primer pair was taken into account based on GEF1 sequences of the C. neoformans type strains H99, 125.91 (=CBS10512), JEC21 and JEC20. PCR was performed in a final volume of 50 μL containing 50 ng DNA, 5 μL PCR buffer 1 × (10 mM Tris/HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl2), 0.2 mM each of dATP, dCTP, dGTP and dTTP, 40 ng primer each, 1.5 U Amplitaq. PCR was conducted in a Bio-rad thermal cycler at 94 °C for 5 min initial denaturation, followed by 32 cycles of 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 1.5 min with a final extension step at 72 °C for 6 min. PCR products were run on 1% agarose gel, and 15 μL PCR products were double digested with enzymes EcoT14I (10 U μL−1) and PstI (15 U μL−1) for 3 h and separated by 1.5% agarose gel electrophoresis at 100 V for 50 min, and visualized under UV light.

Phylogenetic analysis

Phylogenetic analysis of fragments of the GEF1 gene was performed using mega version 4.0 software (Tamura et al., 2007). A dendrogram was produced by neighbor-joining analysis using sequences alignment with the Kimura 2-parameter method. Gaps were treated as a complete deletion. Statistical support for each clade was assessed using bootstrap analysis with 500 replicates. Additionally, sequences from type strains of serotype A and D, namely H99, 125.91, JEC21 and JEC20, were used to root the tree.


Epidemiology of clinical cryptococcal isolates from China

Of 104 evaluable cases, 54 originated from males and 50 from females. Of 91 cases with data of age, the ages ranged from 4 to 84 years, and the numbers of the cases from each age group were as follows: six (<11 years), seven (11–20 years), 16 (21–30 years), 16 (31–40 years), 26 (41–50 years), nine (51–60 years), and 11 (>60 years). From cases of available HIV status, 70 out of 73 (95.9%) came from non-HIV-infected patients (Table S1).

All isolates from China were successfully genotyped by PCR fingerprinting with primer M13 (Fig. 2) and PCR analysis of serotype and mating type. Of the 115 clinical isolates tested, 106 were C. neoformans and nine were C. gattii. Of C. neoformans, 103 were Aα and belonged to molecular type VNI; two were AD hybrids (one A-Dα and the other AαD-); and one was serotype D, mating type a. As for geographic distribution, the AD hybrids and serotype D strain were isolated in East China, and the serotype A strains (VNI MATα) were recovered from all six regions. All C. gattii isolates were obtained from non-HIV-infected patients. Among these, eight belonged to the VGI molecular type and possessed the MATα mating type. All the VGI isolates were from East China, and seven were isolated from CSF and one from sputum. Only one C. gattii isolate sampled from CSF in South China belonged to the VGII molecular type and belonged to MATα. None of the other molecular types (VNII, VGIII and VGIV) were recovered in the present study (Tables S1 and S4).


Examples of PCR-fingerprinting profiles produced with single primer M13 from Chinese cryptococcal isolates. The isolates representing each of the molecular types from China are shown in panel.

Mating capability

In mating assays, all eight C. gattii VGI isolates were unable to mate with either the serotype D or serotype C wild-type tester strains using incubation of up to 21 days, while 87.5% isolates exhibited a weak mating response when cocultured with the serotype C MATa crg1Δ mutant tester strains. The only VGII isolate exhibited a weak mating response in a cross with any one of the MATa tester strains, in which a few filaments, basidia and basidiospores developed (Fig. S1). Meanwhile, as a positive control, serotype D or serotype C tester strains exhibited a robust mating when cocultured with the corresponding tester strains of the opposite mating type.

MLST for C. gattii isolates from China

Partial sequences of IGS1 (666-bp), PLB1 (600-bp), GEF1 (608-bp), GPD1 (547-bp) alleles of C. gattii isolates from China were compared with those from GenBank. All sequences were aligned using clustalw2 software. Based on the four loci, seven out of eight VGI MATα isolates shared identical MLST profiles with strains from Australia and USA (i.e. WM179, WM276 and B5242). The other VGI MATα isolate possessed an identical MLST profile to that of an environmental isolate from Australia (E286). The MLST profile for the one VGII MATα isolate was identical to that of the Vancouver Island outbreak minor genotype strain R272 (Table 1).

View this table:

Multilocus sequence typing profiles of Cryptococcus gattii isolates from China compared with strains originating from other geographic areas

StrainOriginSourceMolecular genotypeMLST profilesReferences/sources
WM276AustraliaEnvironmentalVGI3551Fraser (2005); W. Meyer
WM179AustraliaClinicalVGI3551Fraser (2005); W. Meyer
B5242USAClinicalVGI3551Fraser (2005); W. Schell
E286AustraliaEnvironmentalVGI35141Fraser (2005); W. Meyer
XH3ChinaClinicalVGI3551This study
XH4ChinaClinicalVGI3551This study
XH5ChinaClinicalVGI3551This study
XH23ChinaClinicalVGI3551This study
XH24ChinaClinicalVGI35141This study
XH31ChinaClinicalVGI3551This study
XH32ChinaClinicalVGI3551This study
XH72ChinaClinicalVGI3551This study
R265Vancouver IslandClinicalVGII4117Fraser (2005); J. Kronstad
R272Vancouver IslandClinicalVGII10267Fraser (2005); J. Kronstad
WM178AustraliaClinicalVGII1614178Fraser (2005); W. Meyer
XH91ChinaClinicalVGII10267This study
  • * Identical number represents possession of the same allele; the sequences of IGS1, PLB1 and GPD1 loci for strains shown in boldface are deposited under accession numbers EU443515EU443523, EU443524EU443532, and EU670735EU670743 (the accession number for GEF1 locus is listed in Table S3).

Molecular and mating type analysis by PCR-RFLP

A c. 1555-bp fragment of the GEF1 gene was amplified from nine Chinese isolates and 32 reference strains of C. gattii using gattii-specific primers (i.e. GEF1F and GEF1R) (Fig. 3). The 22 reference strains representing distinct molecular and mating types of C. neoformans could not be amplified with this primer pair (Table S2; Fig. 3). RFLP analysis of the fragments amplified from the 41 C. gattii strains yielded seven profiles (i.e. C1–C7), which were in concordance with the molecular and mating type groups of C. gattii. RFLP profile C1 corresponded to VGI MATα, C2 to VGI MATa, C3 to VGII MATα, C4 to VGII MATa, C5 to VGIII MATα, C6 to VGIII MATa, and C7 to VGIV MATα (Fig. 4). Using this simple tool, the molecular and mating types of the 41 C. gattii strains were all successfully determined (Table S3).


Results of reference strains after PCR amplification of the GEF1 fragments with gattii-specific primers GEF1F and GEF1R. A c. 1555-bp fragment was amplified from strains representing seven groups of distinct molecular and mating types in Cryptococcus gattii, whereas Cryptococcus neoformans strains representing distinct molecular and mating types could not be amplified by this primer pair.


RFLP profiles of reference and Chinese strains of Cryptococcus gattii after double digestion of 1555 bp fragments obtained from PCR of GEF1 gene (EcoT14I and PstI). Seven profiles (C1–C7) which corresponded to groups of distinct molecular and mating types of C. gattii were established.

Phylogenetic analysis

By alignment of the 608-bp fragments of the GEF1 gene of the 41 C. gattii strains, 89–91% sequence similarity was found to be present in opposite mating types. Within each mating type, strains belonging to different molecular types showed 92–97% sequence similarity, but strains within the same molecular types exhibited 99–100% sequence identity.

In a neighbor-joining dendrogram, the 41 C. gattii isolates were grouped into two major clades according to the bipolar mating system, in which MATα and MATa clusters were strongly supported by bootstrap values of 97% and 100%, respectively. In each of these two clades, the isolates clustered according to the molecular genotypes with bootstrap values of 100% for all molecular genotypic clusters. There were 15 alleles found among 41 C. gattii isolates, including 12 alleles belonging to the MATα locus and three alleles of the MATa locus (Fig. 5).


A neighbor-joining tree for 41 Cryptococcus gattii strains by analysis of 608-bp fragments of the GEF1 gene. Cryptococcus neoformans reference strains of H99 (Aα), 125.91 (Aa), JEC21 (Dα), and JEC20 (Da) are used as outgroup. Bootstrap values (500 replicates) are indicated for the main branches. Strains of C. gattii group with the mating types, then cluster according to the molecular types. Chinese isolates are shown in boldface.


Unlike other studies (Casali et al., 2003; Meyer et al., 2003; Jain et al., 2005; Viviani et al., 2006), in which the male-to-female gender ratio was over 2.9 : 1, our data set showed no prominent gender bias in the incidence of cryptococcosis in China. The distribution among distinct age groups was found to be more equal than in Europe (Viviani et al., 2006).

Similar to observations in other areas (Chaturvedi et al., 2005; Jain et al., 2005; Viviani et al., 2006), MATα strains are found to be predominant. Among the MATα strains, C. neoformans var. grubii (VNI) revealed to be the most commonly involved agent responsible for cryptococosis, which is concordant with results obtained from other geographic areas (Casali et al., 2003; Meyer et al., 2003; Jain et al., 2005). This wide distribution may be related to specific dispersal traits and relatively higher virulence (Casadevall & Perfect et al., 1998). Moreover, one serotype D isolate and two AD hybrids were observed, and only the MATα allele was found in both AD hybrids. This has been reported in other areas as well (Cogliati et al., 2006).

Here, we determined eight VGI MATα strains and one VGII MATα strain from China by PCR fingerprinting and PCR amplification of mating type alleles, and PCR-RFLP analysis of the GEF1 gene. The latter method, which is based on the difference in sequence variation of the GEF1 gene among groups of distinct molecular and mating types, also successfully determined the molecular and mating types of another 32 reference strains of C. gattii, indicating the potential for simultaneous genotyping and identification of mating type in this primary pathogen. In addition, phylogenetic analysis showed that the GEF1 gene exhibited similar genetic traits as the CAP1 gene (Kidd et al., 2005).

Molecular types VGI and VGII have been reported to represent the majority of clinical C. gattii isolates in Asia, whereas VGIII and VGIV strains have been isolated more rarely (Jain et al., 2005; Tay et al., 2006), which may be related to the geographic location and climate factors (Granados & Castañeda et al., 2006). Other surveys indicated that the VGI and VGII molecular types are more widely distributed than was previously thought (Kidd et al., 2004, 2007; Fraser et al., 2005). In several studies, MLST revealed the presence of some subtypes in each of the molecular types of C. gattii, suggesting the presence of random genetic drift, strain dispersal or recombination. Identical MLST profiles had been found among strains from distinct areas, and epidemiological links were established (Fraser et al., 2005; Kidd et al., 2005; Bovers, 2008b).

MLST analysis based on IGS1, PLB1, GEF1 and GPD1 loci showed that seven out of eight VGI isolates from China shared identical MLST profiles with reference strains WM179, WM276 and B5242, which all belonged to a MLST genotype based on eight MLST markers used in a previous study. This MLST genotype was shown to be the most common representative in the VGI group and has a wide distribution (Fraser et al., 2005). The other VGI isolate from China possessed an identical MLST profile to that of strain E286, which represents a MLST genotype commonly found in the Australian environment. Interestingly, strains affiliated to these two MLST genotypes were found to be sterile (Fraser et al., 2003; Campbell et al., 2005), suggesting asexual reproduction. Although most VGI isolates from China displayed a weak mating response to the crg1Δ mutant, all of the strains did not mate with the wild-type tester strains used in our mating assays. The VGII isolate from China had an identical MLST profile to that of the Vancouver Island outbreak minor genotype strain R272. Similarly, a VGII isolate from Thailand was reported to share identical alleles with strain R272 at four loci (Kidd et al., 2005). In our study, the patient with cryptococcosis caused by the VGII isolate, had never gone abroad before diagnosis, suggesting that the VGII isolate was acquired within China. The minor outbreak genotype was found in Australia, Vancouver Island and Caribbean Islands (Fraser et al., 2005). The existence of the genotype, such as the representative strain R272 from Vancouver Island, was suggested to be associated with the import of eucalypts from Australia (Kidd et al., 2004; Fraser et al., 2005). It remains to be elucidated if this is true for China as well.

In general, our data suggests that VNI MATα strains (89.6%) are the most common agent for cryptococcosis in China. In evaluable cases, 58 out of 70 (82.9%) non-HIV-infected cases are due to this subtype. All C. gattii isolates are obtained from non-HIV-infected individuals. The most common (77.8%) agent responsible for C. gattii infection in China are related to strains of a widely distributed MLST genotype, and the only VGII isolate from China resembled the Vancouver Island outbreak minor genotype. However, more sampling among clinical and environmental sources, in combination with molecular analyses, are needed to further elucidate the epidemiology of C. gattii and C. neoformans in this part of the world, and to assess relationships with global distribution trends.

Supporting Information

Table S1. Details of origin and molecular characteristics of clinical cryptococcal isolates from China.

Table S2. Cryptococcus neoformans reference strains used in PCR assays to verify the specificity of gattii-specific primers GEF1F-GEF1R.

Table S3. Origin, molecular type, mating type, RFLP profile, and GEF1 partial sequence of Cryptococcus gattii strains involved in this study.

Table S4. PCR analysis of the mating type and serotype of cryptococcal isolates from China.

Fig. S1. Mating morphology of VGII and representative VGI isolates from China when crossed with tester strain JEC20 (JEC21 acted as a positive control).


We thank Joseph Heitman, Sheryl Frank, Anna Floyd (Duke University Medical Center, Durham, USA), Wieland Meyer (University of Sydney, Sydney, Australia), C. De Vroey (Prince Leopold Institut of Tropical Medicine, Belgium), Maria Anna Viviani (Istituto di Igiene e Medicina Preventiva, Milano, Italy) for reference strains. We thank Qiangqiang Zhang (Huashan Hospital, Fudan University, Shanghai), Weida Liu (Institute of Dermatology, Chinese Academy of Medical Sciences, Nanjing), Liyan Xi (The Second Affiliated Hospital of Sun Yat-sen University, Guangzhou), Yihui Yao (Zhongshan Hospital, Xiamen University, Xiamen), Yingchun Xu (Peking Union Medical College Hospital, Beijing), Lanfeng Lu (Zhongshan People's Hospital, Zhongshan), Hongbing Li (The First Affiliated Hospital, Kunming Medical College, Kunming), Bo Ling (Xinhua Hospital, Shanghai Jiaotong University School of Medicine, Shanghai), Shubing Li (The Children's Hospital of Hunan Province, Changsha), Yuechen Zheng (Union Hospital, Tongji Medical College, Huazhong University of Sciences & Technology, Wuhan), Yun Pan (Central Hospital of Jinhua, Jinhua), and Li Wan (The Fourth Hospital of Hebei Medical University, Shijiazhuang) for clinical isolates from China.

This work was supported by key subject fund from Xinhua hospital, Shanghai Jiaotong University School of Medicine.


  • Editor: Teun Boekhout


View Abstract