Introduction

Cryptococcosis, a life-threatening fungal infection, is mainly caused by Cryptococcus neoformans sand C. gattii species complexes affecting both immunocompromised and immunocompetent hosts. The mode of infection is through inhalation of basidiospores from environment. Clinical manifestations include pulmonary infection, chronic meningitis or dissemination depending upon the host immune status. Both species comprise genetically diverse subgroups that differ in ecology and epidemiology [1, 2]. C. neoformans includes 2 varieties with 3 serotypes: C. neoformans var. grubii (serotype A), C. neoformans var. neoformans (serotype D), and a hybrid (serotype AD) and C. gattii species complex including serotypes B and C [3]. Recently, seven species have been recognized in C. neoformans/C. gattii species complex viz. C. neoformans, C. deneoformans, C. gattii, C. bacillisporus, C. deuterogattii, C. tetragattii, C. decagattii [4, 5]. With worldwide distribution, C. neoformans var. grubii (serotype A) is responsible for approximately 95% of cryptococcal infections and 98% of infections among HIV infected populations [6]. Until recently, C. gattii species complex (serotypes B and C) were considered restricted to tropical and subtropical regions, with Eucalyptus trees as an ecological niche. However, the environmental shifting of this species had been highlighted by a recent outbreak in the Pacific Northwest of North America [7, 8]. Additionally, it was also thought to have a predilection for infecting apparent immunocompetent hosts which had been challenged by African studies where most of the patients infected were immunosuppressed individuals [9]. Diagnosis of cryptococcosis is primarily based on direct demonstration of the capsule in India ink mount, isolation on culture and detection of capsular antigen by latex agglutination test (LAT) or lateral flow assay (LFA) in cerebrospinal fluid (CSF) and serum.

Several molecular typing methods like M13-PCR fingerprinting, restriction fragment length polymorphism (RFLP), random amplification of polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP), multilocus microsatellite typing (MLMT) and multilocus sequence typing (MLST) have been used to study the molecular epidemiology of both C. neoformans sand C. gattii species complex [10]. Based on M13-PCR fingerprinting, these two species complexes are further divided into eight major molecular types: VNI and VNII (serotype A; C. neoformans var. grubii), VNIII (hybrid serotype AD), VNIV (serotype D; C. neoformans var. neoformans), VGI, VGII, VGIII VGIV and VGV (serotypes B and C; C. gattii species complex) [11,12,13,14,15,16]. VNI and VGI are the most prevalent genotype for C. neoformans var. grubii and C. gattii species complex, respectively [17,18,19], and VNB, a genotype in C. neoformans var. grubii , was discovered in Botswana [10]

According to 2018 World Health Organization (WHO) guidelines, for treating cryptococcal meningitis among people living with HIV, the preferred regime for induction therapy is a combination of amphotericin B deoxycholate with flucytosine followed by fluconazole and for the consolidated/maintenance phase, fluconazole is recommended [20]. Despite the introduction of HAART (Highly Active Anti Retroviral Therapy), cryptococcosis remains one of the common opportunistic infections among people living with HIV [21,22,23,24]. The previous (2002–2007) data published from our institute showed the cryptococcal distribution of 60% in HIV positive and 40% in HIV negative individuals [25]. Extensive studies on environmental isolates had been conducted in India [26,27,28,29,30,31,32]. However, data on epidemiological typing of clinical isolates from India is still limited. Therefore, this study was planned to provide more insight into the molecular epidemiology of clinically isolated Cryptococcus and their susceptibility profiles in the post-HAART era.

Materials and Methods

Clinical Isolates

A prospective study for 3 years (2012–2015) was conducted in which 86 Cryptococcus isolates from 77 patients were obtained. Prospectively, patients were enrolled from the following hospitals: All India Institute of Medical Sciences (AIIMS), New Delhi (n = 35), Guru Teg Bahadur Hospital (GTB), New Delhi, (n = 20), Jawaharlal Institute of Postgraduate Medical Education & Research (JIPMER), Puducherry (n = 11), Institute of Human Behaviour & Allied Sciences (IHBAS), New Delhi,(n = 8), Maharaja Krishna Chandra Gajapati Medical College & Hospital (MKCG), Brahmapur, Odisha (n = 5), North Eastern Indira Gandhi Regional Institute of Health and Medical Sciences (NEIGRIHMS), Shillong, Meghalaya (n = 4), Sri Ramachandra Medical College and Research Institute (SRMC & RI), Chennai, Tamil Nadu (n = 1), Naga Hospital Authority Kohima (NHAK), Kohima, Nagaland (n = 1), Jubilee Mission Medical College Hospital and Research Institute (JMMCRI), Thrissur, Kerala (n = 1). Also, to observe any differences in the sequence types and susceptibility patterns previously collected 122 Cryptococcus isolates from 86 patients (AIIMS:99 isolates from 63 patients and IHBAS: 23 isolates from 23 patients) during 2005–2011 were also included. Therefore, 208 isolates so obtained from 163 patients were included in this study. Cryptococcus was mainly isolated from CSF (196/208, 94.23%). The remaining 12 isolates were from scalp abscess (3/208, 1.44%), sputum (3/208, 1.44%) and urine (3/208, 1.44%), whereas 2 were from blood and one from bronchoalveolar lavage (BAL) (S1 Table). This study was approved by the institute’s ethics committee i.e. All India Institute of Medical Sciences, New Delhi, India (Ref no. IEC/NP-364/2012 and RP-15/2012). The study was conducted using clinical samples obtained from patients for routine microbiological investigations. We were not directly involved with patient enrollment. The treating physician took verbal consent at the time of patient admission to the hospital.

Reference Strains

A set of standard laboratory reference strains provided by Dr. Wieland Meyer, The West mead Institute for Medical Research, Australia representing each of the 8 major molecular types were used for molecular typing: WM148 (VNI), WM626 (VNII), WM628 (VNIII), WM 629 (VNIV), WM179 (VGI), WM178 (VGII), WM175 (VGIII) and WM779 (VGIV).

Phenotypic Identification

Following growth on brain heart infusion agar, phenotypic identification of the isolates was made by standard mycological methods including urea hydrolyzing test, sensitivity to actidione, phenol oxidase production on birdseed agar [33,34,35]. Canavanine–Glycine–Bromothymol blue (CGB) agar was used to distinguish C. gattii/neoformans species complexes [36]

Molecular Characterization

DNA Extraction

DNA was isolated using the protocol as described earlier with slight modifications [22]. The cells were treated in pH 5.0 lysing buffer (1.0 M sorbitol–0.1 M sodium citrate) containing 10 mg/ml lysing enzyme (from Trichoderma harzianum; Sigma Aldrich, St. Louis, MO) for 3 h at 30 °C to generate protoplasts. After that, genomic DNA was extracted using the Qiagen DNeasy Mini Kit (Qiagen, Valencia, CA, USA), as per the manufacturer’s instructions.

MLST Analysis

Six genetic loci (CAP59, GPD1, LAC1, PLB1, SOD1, and URA5) and non-coding IGS1 region were undertaken for MLST analysis as per ISHAM consensus. PCR in 50 µL reaction volume was performed for each of the seven MLST loci using the primers described earlier with slight modifications in protocols [10, 14, 37,38,39]. The PCR products’ sequencing was outsourced to XCelris (XCelris Lab. Ltd., Ahmedabad, India). Sequences were edited using DNA BASER (v 4.36.0) and then aligned in MEGA 6.0 (www.megasoftware.net). The allele profile for each isolate was generated by assigning allele type (AT) to each of the seven loci. All sequence types (ST) were then described according to the ISHAM MLST scheme for C. neoformans/C. gattii species complex (https://mlst.mycologylab.org/.) Data from https://mlst.mycologylab.org were used for the global comparison.

Phylogenetic Relationship

Phylogenetic analysis was performed using the maximum likelihood method with 1000 bootstrap replicates implemented in MEGA v6.00. To represent comparison between original sources and allelic profiles of C. neoformans isolates, a minimum spanning tree was generated by Phyloviz v1.0 using goeBURST algorithm (https://goeburst.phyloviz.net/). For comparison with global isolates, data of commonly reported C. neoformans STs from multiple countries were retrieved from the ISHAM-MLST database (https://mlst.mycologylab.org). A total of 113 different STs (85 VNI, 17 VNB, 10 VNII, and 1 VNIV) were included for the same.

Nucleotide Diversity

For C. neoformans, DNA polymorphisms including haplotype and nucleotide diversity, number of polymorphic and mutation sites were calculated using DnaSP v5.10 (https:// www.ub.edu/dnasp/).

Antifungal Susceptibility Testing

Antifungal susceptibility testing was performed using the broth microdilution assay, according to Clinical Laboratory Standards Institute (CLSI) approved standard M27-A3 guidelines suggested for yeasts [40]. Quality control isolates (Candida parapsilosis ATCC 22,019 and Pichia kudriavzevii ATCC 6258) were included. The antifungal drugs tested were: amphotericin B (AMB), flucytosine (5FC), fluconazole (FLU) and voriconazole (VOR) (Sigma Chemical Corporation, St. Louis, MO).

Statistical Analysis

Fisher’s exact test was used for statistical analysis. A p-value of ≤ 0.05 was considered statistically significant.

Results

Demographic Data and Molecular Types

Among 163 patients enrolled in the study, 126 were male and 37 were female. The HIV status was positive in 67 (67/163, 41.1%), negative in 90 (90/163, 55.21%) and status was unknown in 6 cases (6/163, 3.68%). Of the total of 208 isolates, 205 isolates were identified as C. neoformans belonging to VNI except one which belonged to VNII. Among the 3 C. gattii species complex, 2 were of molecular type VGI and one VGIV.

MLST Determination

A total of 10 sequence types (STs) were revealed following analysis of 208 Cryptococcus isolates by MLST. The aligned sequences were 4003 base pairs with 366 polymorphic sites. The 39 allele types (ATs) were obtained from all seven loci, one of which was novel to the Indian population (novel AT 5 of LAC1 gene belonging to ST 194). MLST analysis divided C. neoformans into seven STs: ST31 (n = 159), ST93 (n = 27), ST77 (n = 9), ST3 (n = 7), ST174 (n = 1) and ST194 (n = 1) of VNI and ST40 (n = 1) of VNII. In addition, three STs of C. gattii species complex from non-HIV population were identified; ST58 (n = 1) and ST215 (n = 1) of VGI and ST69 (n = 1) of VGIV (Fig. 1). Irrespective of the geographical distribution and HIV status, the most common ST type for C. neoformans was 31 (Table 1).

Fig. 1
figure 1

adapted from https://d-maps.com/m/asia/india/inde/inde15.pdf

Distribution of Cryptococcus sequence types (ST) in different parts of India. India map had been

Full size image
Table 1 Sequence types (STs) according to HIV status
Full size table

Phylogenetic Analysis of C. neoformans

Phylogenetic analysis of the global dataset using maximum likelihood method revealed 2 different clusters within the C. neoformans VNI population. VNB and VNII isolates clustered separately in different groups, VNIV was used as an out-group. Cluster I of the VNI population includes 23 different STs with the majority of European isolates. Cluster II is composed of 62 different STs with Asian isolates constituting the majority. Though isolates from different continents were dispersed in both the clusters, our isolates from the present study along with other Asian isolates fall under cluster II only (Fig. 2).

Fig. 2
figure 2

Phylogenetic analysis of C. neoformans using Maximum likelihood method. Each colored rectangle represents geographical origin of isolates. * indicates STs found in the present study. The numbers at each branch depict bootstrap values 50%, based on 1,000 replicates

Full size image

To understand the pattern of evolutionary descent among clusters of related genotypes, goeBURST analysis was performed. The goeBurst analysis differentiated our isolates into one main cluster and three singlet on STs VizST93, ST194 and ST40 (VNII) with ST174 as the founder one (Fig. 3a).

Fig. 3
figure 3

a Minimum spanning tree showing C. neoformans isolates from different parts of India using goeBURST algorithm. Each colour of the circle represents a different part of India with sequence type. b Minimum spanning tree showing C. neoformans isolates from different continents using goeBURST algorithm. Each colour of the circle represents different continents with sequence type

Full size image

Eight groups were formed when our isolates were compared with the global C. neoformans MLST dataset using goeBURST analysis. Groups were defined according to a single locus variant (SLV), and all the STs present in one group should be SLV for any one of the ST in that group. The founder ST of each group has been shown in figure as black circle. Most of the Asian isolates were found in group 1 and group 4. Isolates in the present study were grouped with other Asian isolates in group 1. Group 2 consists of the majority of STs originated from Europe. All the VNII isolates formed a separate group (Group 3). Group 5 contained four STs of VNI from all the continents with the majority from Europe. Most of the African VNI sequence types were clustered together in groups 6 and 8. Group 7 was formed by 2 VNII isolates from Europe. Most of the VNB isolates and few VNI, VNII and VNIV from different parts of the world were scattered as singletons (STs without group) (Fig. 3b).

Phylogenetic Analysis of C. gattii Species Complex

Phylogenetic analysis using Maximum-likelihood among the global C. gattii species complex revealed ST215 and ST58 from this study along with isolates from China clustering together in VGI. VGIV included ST69 from this study and ST243 previously reported from India (Fig. 4).

Fig. 4
figure 4

Phylogenetic analysis of C. gattii from India and different countries of world using Maximum likelihood method. * Indicates C. gattii isolates found in the present study

Full size image

Association Between Sequence Types, HIV Status and Geographical Origin

No significant correlation between STs and HIV status (p value = 0.249), STs and geographical origin (p value = 0.087) could be established.

Nucleotide Diversity for C. neoformans

Among seven loci analyzed (CAP59, GPD1, IGS1, LAC1, SOD1, PLB1 and URA5), the highest nucleotide diversity (p) of 0.26685 was observed in IGS1, followed by GPD1 (p = 0.00123) and LAC1 (p = 0.00071). The haplotypes number (alleles) varies for different loci: 1 for CAP59, PLB1, SOD1 and URA1, 2 for LAC1 and 3 for GPD1 and IGS1. The haplotypic diversity ranged from 0.33 for LAC1 to 0.733 for IGS1 (Table 2).

Table 2 Genetic variability of C. neoformans in the present study
Full size table

Antifungal Susceptibility Testing

In vitro antifungal susceptibility testing of 208 isolates was performed for 4 drugs, namely AMB,

5FC, FLU and VOR. All of the isolates, irrespective of the species, showed low MICs for all the antifungals tested (Table 3).

Table 3 In vitro antifungal susceptibility results according to the sequence types
Full size table

Discussion

In the present study, 98.5% of the isolates were C. neoformans with molecular types VNI, VNII and 1.4% were C. gattii species complex with molecular types VGI, VGIV. With worldwide distribution, VNI had been reported as the most common molecular type reported for C. neoformans [18]. Molecular type VNII had been reported from varied continents including North America, South America, Europe, Africa and Asian countries like Thailand, India and Japan [10, 18, 19, 41,42,43]. It also accounted for 1%, 4.5% and 13.7% of the isolates from Taiwan, Malaysia and China, respectively [44,45,46]. Another frequent molecular type VNB, mainly considered endemic in Botswana, had now been reported from Rwanda, RD Congo, South Africa, Italy, Portugal, Brazil, Columbia but never from Asian countries [10, 13, 47]. The molecular epidemiology of 205 isolates in the present study showed the same paradigm with 204 isolates (99.5%) belonging to molecular type VNI and one isolate (0.5%) to VNII [18].

MLST of C. neoformans revealed a total of seven STs, of which one ST was not previously reported from our country, India. In the present study the predominant STs identified were ST31 followed by ST93, accounting for 77.5% and 13.1%, respectively. These 2 STs have previously been reported from India, as well as from other Asian countries like China, Japan, Korea, Qatar, Kuwait, Thailand and Indonesia, and also from Africa [18, 19, 47,48,49,50,51,52]. ST93 was reported as the most common ST type accounting for 63.6% among 143 clinical and environmental Brazilian isolates [53]. ST93 has recently been documented as a predominant ST (76.7%) in clinical and environmental samples from India [54]. ST77 has mostly been reported from clinical samples in India, along with few cases from Thailand, Uganda, Brazil and France [18, 19, 53, 55]. The remaining few genotypes in our study ST174 have also been reported previously from India and Kuwait, and ST3 from Thailand isolated from clinical isolates [18, 19]. ST194 has not been reported previously from our country India but from another Asian country China [18]. The predominant STs reported from Asian countries like China, Hong Kong, Japan and Thailand include ST4, ST5 and ST6; however, none of them were identified in our study [18, 48, 51]

In the present study, the highest nucleotide diversity was demonstrated at IGS1 (0.26685) followed by GPD1 (0.00123) and LAC1 (0.00071) locus for C. neoformans. Similarly the highest diversity at IGS1 locus (0.0045) followed by LAC1 (0.0018) and GPD1 (0.0014) was reported by Khayhan K et al. on Asian isolates [18].

Indian C. neoformans along with other Asian isolates were found to be less diverse when compared with African, North and South American, and European population. Some isolates from Africa, North and South America and Europe shared the same Asian haplotypes. In contrast, the allele of very few Asian isolates displayed haplotypes similar to those seen in isolates of other continents [18].

A significant correlation between STs and HIV negative had been observed in various studies for C. neoformans [17, 18, 43, 56]. However, in contrast to the above statement, no such correlation between ST and HIV status could be established in the present study (p value > 0.5). Similar to the previous report, in the present study, no change in STs distribution over the years was observed [51].

Five molecular types were recognized within C. gattii species complex: VGI, VGII, VGIII, VGIV and VGV with VGI as the predominant one. VGII is mostly associated with outbreaks and has been reported from British Columbia, Western Australia, Brazil, Venezuela, Pacific Northwest of the USA and Asia [14, 22, 57,58,59]. VGIII has been reported in Mexico, Colombia and the USA [8, 60,61,62]. VGIV has been reported mainly from India, Africa and South America [14, 47, 63]. In our study, two C. gattii species complex isolates belong to VGI with ST58 and ST215 and one to VGIV with ST69. ST58 had previously been reported from Columbia [62]. However, to the best of our knowledge ST215 (VGI) and ST69 (VGIV) have never been reported from India till date.

In the present study, antifungal susceptibility testing was done on all isolates. The MIC50 and MIC90 values for all the antifungals tested are in accordance with previously published data [64, 65]. In the current study, the MIC50, 2 µg/ml and MIC90, 4 µg/ml for fluconazole were found similar to the published surveys from the USA and South Africa [64, 65]. These lower MICs for fluconazole [22, 66,67,68] and 5-flucytosine [29] observed are incongruent to the previous data from our country India. Higher MIC value than the epidemiological cut-offs reported earlier was found for one isolate to voriconazole (MIC 0.25 μg/ml) and for 2 isolates to amphotericin B (MIC 1 μg/ml) [69, 70].

In our study, no significant differences in susceptibility to the tested antifungal drugs were observed among isolates with different STs. This might be due to the high degree of clonality observed in this region. The limitations of the study included the lack of clinical isolates from the western part of the country and lack of environmental isolates, and due to the financial constraints we could not include AFLP in our methodology.

Conclusion

The present study for the first time from India provides detailed information on molecular epidemiology and antifungal susceptibility pattern of 208 clinical Cryptococcus strains isolated from different regions of the country. The majority of isolates identified were C. neoformans belonging to molecular type VNI with predominant STs 31 and 93. The phylogenetic analysis revealed our isolates were clustered along with other Asian isolates in a group. Worldwide, the highest diversity was seen in African isolates [18]. Only 3 isolates of C. gattii species complex were obtained, belonging to ST58 and ST215 of VGI and ST69 of VGIV. Low MIC to different antifungals tested was observed for all the included isolates. Hence, fluconazole still can be given both in prophylaxis and maintenance therapy in the Indian scenario. Interestingly, no association between prevalent STs and HIV status, geographical origin or MIC value could be established. Additionally, we noticed an increase in the isolation of Cryptococcus species from HIV negatives. The present study database including both molecular epidemiology and antifungal susceptibility testing can be further utilized in a better understanding of the pathogenesis and treatment modalities in the future.