Hybrid genome sequence of Cryptococcus neoformans of Indian origin and comparative genome analysis

Abstract

Human opportunistic fungi C. neoformans are a global health concern as they cause severe pneumonia and cryptococcal meningitis in tropical and subtropical regions. Though the occurrence of infection has reduced from 2014 to 2022, the mortality rate remains high, and it is considered a neglected tropical disease. More studies are needed to elucidate the genomic architecture and functional traits required to underpin the disease management of Cryptococcus neoformans. This study showcases the first complete genome sequence of an Indian isolate of Cryptococcus, identified as subsp. grubii containing 4.8% of virulence genes in a total genome. A comparative virulome analysis revealed that study genomes Cn, H99 and JEC21 harbor crucial virulence genes associated with infection severity. While all study strains possess genes promoting antifungal resistance (AFR), most lack specific SNPs defining AFR. The whole genome Single Nucleotide Polymorphism (SNP) analysis categorized the strains as hyper- and non-hypermutators. The phylogenomic and pan-genome analysis revealed that the study strains share a core genome, highlighting evolutionary relatedness. The study shows gene expansion and contraction between the study strains. This pioneering study presents Cryptococcus’s first comprehensive genomic analysis from India, integrating genetic variations, virulome, resistome, and pan-genome dynamics. The study highlights the necessity of bioinformatic tools for efficient genome analysis and large-scale comparative genome research.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-25672-3.

Introduction

Cryptococcus sp. is an environmental fungus that majorly infects immunocompromised people and was categorized as a critical priority pathogen by the World Health Organization (WHO) in 2022 due to its high mortality rate, immunocompromised vulnerability, global health impact, growing antifungal resistance and limited treatment options¹.

Cryptococcal infections are more prevalent in tropical and subtropical regions². Sub-Saharan Africa has the highest incidence (6–10%) and mortality rates (24–47%), followed by Asia and the Pacific (1–3%)³. Environmental, ecological, socioeconomic, and healthcare factors contribute to these regional disparities. Globally, approximately 152,000 new cases and 112,000 deaths (19% mortality rate) occur annually, despite the Highly Active Antiretroviral Therapy (HAART) reducing the incidence; mortality rates remain alarmingly high³.

This ubiquitous pathogen frequently inhabits avian environments, marshy areas, and decaying wood^4,5. Infection occurs through the inhalation of spores/ infectious propagules, leading to pulmonary colonization and subsequent dissemination to the liver, kidney and bone marrow, finally disseminates to the brain via paracytosis, transcytosis and the Trojan horse mechanism^6,7. The Cryptococcus sp. exhibits neurotropism owing to the presence of proteinases and urease that allow it to penetrate the blood–brain barrier (degradation of Extracellular Matrix [ECM] components of the host) and survive⁸. In the recent decade, it’s been observed that even immune-competent individuals are acquiring the infection. The increased susceptibility is believed to be the organism’s enhanced pathogenicity, which may be attributed to the increased virulence, stress tolerance, or AFR⁶.

Taxonomically, the genus Cryptococcus is classified into two species (neoformans and gatti) and two subspecies (grubii and neoformans). Further, it is categorized into four serotypes and nine molecular types⁹. Particularly serotype A (C. neoformans var. grubii), comprising the molecular types VNI and VNII, is highly pathogenic and widespread in the Indian population¹⁰. The high pathogenicity of serotype A is due to the escalated virulence factors, environmental adaptability, and immune evasion capabilities^11,12.

In the earlier decades, Polymerase Chain Reaction- Restriction Fragment Length Polymorphism (PCR–RFLP) and PCR fingerprinting were used to study Cryptococcus molecular epidemiology, the genetic structure of fungi, and genome plasticity. Recent advancements in Whole Genome Sequencing (WGS) help understand Cryptococcus’s epidemiology and genetic structure¹³. Globally, the studies based on the WGS have demonstrated small genetic variability and a small number of chromosomal rearrangements across the molecular types and also the variations in gene content and structural changes due to translocations, aneuploidy and inversions were noted in C. neoformans. The WGS also helps reveal the hybrids between the Cryptococcus sp., complex¹³. In addition, WGS helps in identifying the microevolution signals of infecting strains, revealing the rapid adaptation of the organism through hypermutation to identify genes related to the adaptation of an organism to the stress condition that include virulence and pathogenicity features and monitoring of outbreaks^13,14.

Though the preliminary role of the stress tolerance genes was to overcome the environmental stressors (temperature, pH, reactive oxygen species, reactive nitrogen species and starvation), the recent insights of cryptococcal genomics reveal that stress-tolerant genes also play a pivotal role in modulating the virulence factors synthesis and promoting AFR¹⁵. In addition, the stress tolerance genes play a role in maintaining the cell wall/ cell membrane integrity and evasion of host immune response. Further, the autophagy genes (also stress-tolerant genes) in Cryptococcus sp. play roles in cell differentiation, cell growth, morphology and pathogenesis^16–18. Complete genome sequencing also revolutionized the understanding of AFR in C. neoformans. Though the inherent resistance (echinocandin resistance) is more common in Cryptococcus sp., the fungal genomics highlights the insights of acquired resistant mechanisms (target protein modification, over-expression of target proteins, and upregulation of multi-drug transporters) caused by genetic mutations^19–21. The mutations in non-primary target proteins also produce resistance to antifungal drugs^22,23.

Despite global genomic studies, India-specific data on cryptococcal genomics are scarce. This study demonstrates the first hybrid sequencing of a clinical Indian isolate of C. neoformans and unravels the genomic aspects. A comparative genomic study was made with the clinical isolates and an environmental isolate from the National Center for Biotechnology Information (NCBI) datasets to understand the evolutionary relationships, virulome, resistome and pan-genome.

Digital tools have played a censorious role in genomic research over the last decade. In particular, digital tools play a pivotal role in the comparative genomic analysis of complete genome sequences²⁴. There is a lack of tools/ databases to study the fungal virulome, resistome and pan-genome on a large scale. The available databases list virulence and resistance factors in Cryptococcus that are not strain-specific. This study aims to assess the virulome, resistome, and SNPs in a strain-specific manner and pangenome analyses. The study highlights the need for bioinformatic tools for large-scale evaluation of virulome, resistome and pan-genome analyses. Researchers who want to develop new drugs would be able to learn about potential drug targets.

Result

Sequencing and assembly

Illumina sequencing resulted in 7.35 million raw reads at 116.05X coverage and Nanopore sequencing had 0.609 million raw reads at 104.8X coverage (supplementary Table 1, sheet 1). The genome sequences of the strain Cn were deposited under BioProject accession number PRJNA1141203. The strain Cn was deposited in the Microbial Type Culture Collection (MTCC), Chandigarh and obtained the accession number MTCC15081. The assembled genome was ~ 19 MB and the statistics are provided in supplementary Table 2, sheet 2. Quality assessment of the assembled genome indicated 89.6% completeness.

Gene prediction and annotation

About 5673 genes were annotated using the Biology Research And Knowledge Engineering Research (BRAKER) tool. The Gene Ontology (GO) annotation categorizes the proteins as biological processes, cellular components and molecular functions (Fig. 1a). The functional annotation using Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways annotated 5963 genes which fall majorly into 6 categories, including Biological Reaction, Interaction and Terminology Encyclopedia (BRITE) hierarchy (3186 genes), metabolism (1294 genes), genetic information processing (831 genes), cellular processes (419 genes) and environmental information processing (97 genes). The genes (n = 136) of unknown functions are categorized separately. About 29 cellular pathways were detected from the assembled genome of Cn using the KEGG Automatic Annotation Server (KAAS) database. The top 12 functional pathways are depicted in Fig. 1b.

Fig. 1.

Functional annotation: (a) GO annotation by DIAMOND database: the Gene ontology of the study strain Cn and (b) Pathway analysis by KAAS database: Top 12 pathway functions of the study strain Cn.

Genome identity search of clinical Cn with the reference genome

The genome identity search using the Multiple Alignment of Conserved Genomic Sequence with Rearrangements (MAUVE) and D Genies indicated a 99.9% identity of Cn with the reference genome (H99) (Figs. 2a and b). The ITS region of Cn showed 100% coverage and identity with the CBS8710 strain, isolated from a Hodgkin’s lymphoma patient from North Carolina.

Fig. 2.

Genome identity search: (a) Dot plot comparison using D GENIES;and (b) Linear genome image using MAUVE and (c) Comparative Nucleotide identity search: The whole genome nucleotide similarity search between the 12 study strains was represented.

Comparative nucleotide analysis

The nucleotide identity reveals that the Indian study strain (Cn) shares a high genetic identity with clinical strains from, U.S. (H99, C23) 99.9% identity, Australia (KBCN0140, KBCN0142, SACN00B1) 99.9% identity, South Africa (Bt65, Bt81, Bt89, Bt133) 98% similarity (Fig. 2c). Additionally, Cn exhibits 97.7% identity with a US environmental strain (VNII), and 88.3% identity with a U.S. clinical strain (JEC21), which belongs to a different subspecies (C. neoformans var. neoformans).

Whole genome SNP analysis

In total, 26,224 SNPs were identified among the 12 strains. The S. African and U.S. strains were categorized as hypermutators as they have SNPs > 1000. The strains from India, the U.S. (C23), and Australia are non-hypermutators as they have SNPs < 1000 (Fig. 3).

Fig. 3.

Whole genome SNP analysis: (a) Hypermutators (b) non-hypermutators. The hypermutators have < 1000SNPs, whereas non-hypermutators have > 1000SNPs.

Phylogenetic analysis using MLST

The midpoint of the tree was divided into two major clades. Clade 1: Clinical strain JEC21 (subsp. neoformans) from the U.S. Clade 2: Clinical and environmental strains belonging to subspecies grubii. Further, clade 2 was divided into Subclade 1: Environmental strain VNII from the U.S.; Subclade 2: Clinical strains of South Africa, grouped into 2 clusters and Subclade 3: Clinical strains from India, Australia and the U.S. (C23, H99), grouped into 5 clusters (Fig. 4). It was observed that the S. African isolates were grouped as a separate clade from the USA, India, and Australian strains. Also, the phylogenetic tree separated the clinical and environmental isolates and the subspecies grubii and neoformans.

Fig. 4.

Phylogenetic tree based on MLST genes: MLST phylogeny was constructed using the maximum likelihood method with MegaX. 1000 bootstrap replicates were calculated to assess robustness. The midpoint of the phylogenetic tree was divided into two major clades by the subspecies grubii (red line) and neoformans (yellow line). The environmental strain was represented by the bird (Cockatoo excrement) and the clinical strains were represented by brain, lungs and human (CSF, bronchial wash and unknown clinical source). The magenta colour represents the Botswana (South Africa) strains, the blue colour represents the TamilNadu (India) strain, the orange colour represents the Melbourne (Australia) strains, the red colour represents the U.S. strains.

Virulome and comparative virulome analysis

About 274 Virulence Related Genes (VRGs) were retrieved from the literature and the Database of Fungal Virulence Factors (DFVF) database. Based on the functionality, we categorized the VRGs into 15 domains and the genes belonging to each domain were illustrated in Fig. 5a. Most genes involved in virulome analysis are multifaceted, either directly or indirectly involved in more than one virulence trait.

Fig. 5.

(a) VRGs present in the study strain Cn: The chart represents the (n = 266) VRGs of Cn that were categorized into stress tolerance, cell wall-associated, cell membrane-associated, capsule-associated, morphogenesis associated, cell adhesion, immune evasion, iron acquisition, protease, nutrient acquisition, copper homeostasis, host cell membrane disruption and melanin associated genes; (b) Comparative virulome analysis: The graph represents the comparative study of the major VRGs in the (n = 12) study strains. The colored region represents the presence of genes.

In the comparative virulome analysis, H99 has the highest VRGs with 99.27% (n = 272) and the lowest in Bt133 with 82.84% (n = 227). However, both are clinical isolates. The Cn strain has 97.27% (n = 266) of VRGs. Being an environmental isolate, VNII has 83.94% (n = 230) VRGs. Further, the major VRGs were compared among the 12 strains, revealing that strains Cn, H99 and JEC21 have all major VRGs (n = 24) and could be more virulent than other strains (Fig. 5b; Table 1). Moreover, Cn and H99-infected animal models prove them highly virulent²⁵.

Table 1.

The major VRGs of C. neoformans study strains.

Virulence genes	Protein name	Functions
CAP 10	Capsular associated protein 10	Capsule synthesis and maintanence
CAP 59	Capsular associated protein 59	Capsule synthesis and maintanence, Biofilm production, GXM transport
CAP 60	Capsular associated protein 60	Capsule synthesis and maintanence
CAP 64	Capsular associated protein 64	Capsule synthesis and maintanence
LAC1	Laccase1	Production of melanin; regulates biofilm formation; Inhibits host cell phagocytosis
LAC2	Laccase2	Production of melanin; inhibits host cell phagocytosis
STU1	protein STU1	Stress responses; morphogenesis; capsule regulation
Hsp90	heat shock protein 90	Regulates heat shock, oxidative stress and various physiological/ environmental stress including stress of antifungal agents; regulates capsule formation, melanin production; Regulates host immune evasion
URE1	urease1	Facilitates nitrogen source utilization; alkalinization of the environment; calcium carbonate production; production of biofilm
PLB1	phospholipase B1	Immune evasion; fungal cell adhesion to host cells; nutrient acquisition; regulates titan cell formation
YME1	ATP dependent metalloprotease	Catalyzes mechanism that involves a metal and related to virulence (stress response )
RIM13	Zinc metalloprotease	Catalyzes mechanism that involves a metal and related to virulence (stress response )
APR1	Aspartyl protease	Involved in host immune evasion and food acquisition
PIM1	Protease	Host immune evasion by proteolysis
SPP1	S2P endopeptidase	Host immune evasion by proteolysis
RBD1	Serine peptidase	Fungal cell adhesion to host cell; immune evasion; nutrient acquisition
APP2	Aminopeptidase 2	Proteolysis; nutrient acquisition; capsule formation; immune evasion
CFT1	Cryptococcus Fe Transporter	iron acquisation
SIT1	Siderophore-iron transporter Str1	Iron acquisition
CIR1	GATA type zinc finger protein asd-4	Regulates iron homeostasis, capsule formation, melanin production, stress response, and host immune evasion and pathogenesis
RIM101	pH-response transcription factor pacC/RIM101	Regulates genes encoding cell wall-modifying enzymes such as chitinases, glucanases, and cell wall proteins; capsule production
ALG3	Dol-P-Man:Man(5)GlcNAc(2)-PP-Dol alpha-1,3-mannosyltransferase	Maintains cell wall integrity; Fungal cell adhesion to host cell; Immune evasion; Nutrient acquisition
SOD2	Mn superoxide dismutase	Oxidative stress tolerance (Catalyzes dismutation of toxic superoxide, converting superoxide to hydrogen peroxide and oxygen and related to virulence)
SOD1	Copper zinc superoxide dismutase	Oxidative stress tolerance (Catalyzes dismutation of toxic superoxide, converting superoxide to hydrogen peroxide and oxygen and related to virulence)

Resistome and comparative resistome analysis

Around 78 genes were predicted as AFR-promoting genes (Fig. 6a).

Fig. 6.

(a) AFR genes present in the study strain Cn: The chart represents the (n = 78) AFRs of the Cn that were categorized into transporter, structural, DNA mismatch, nucleic acid synthesis, stress response and transcription factors; (b) Comparative resistome analysis: The graph represents the comparative study of the major AFR genes present in the (n = 12) study strains. The colored region represents the presence of genes and (c) Comparative analysis of SNPs in the AFR genes: The graph represents the comparative study of the SNPs present in the AFR genes of the (n = 12) study strains. The colored region represents the presence of genes.

Among the 12 strains, the Cn was found to harbor abundant AFR genes across certain antifungals. Azole resistance was contributed by ERG11, ERG24, ERG1, SRE1, AFR1, AFR2, PMR5, TUB1, SEY1, YOP1, TUB1 and YAP1 genes. Echinocandin resistance was contributed by FKS1, CRM1, and Hsp90 genes. The gene for P4 ATPase and CRM1 confers intrinsic resistance to echinocandins and the resistance to pyrimidine analogs was conferred by the genes FCY2, TMP1 and FUR1. Finally, allylamine resistance is conferred by ERG1. The Cn (n = 78; 100%) has the highest AFR genes. Despite being an environmental strain, the VNII has n = 53 AFR genes (68.83%).

Among 20 major AFR genes, 13 were commonly shared among the 12 genomes. The genes ERG11, ERG24, SEY1 and ERG1, belonging to sterol biosynthesis, provide resistance by alteration of target protein structure, overexpression of target protein and increased activation of efflux pump. The gene for P4-ATPase and the gene CRM1 provide inherent resistance by regulating the intracellular calcium concentration. The gene FKS1 (β-glucan synthase; Cell wall synthesis) and Hsp90 (Heat shock protein 90; Cell wall regulation) provide resistance by alteration, underexpression of the target protein, and stress regulation. The genes TMP1 (Thymidylate synthase 1) and FUR1 (Uracil phosphoribosyltransferase 1) from nucleotide biosynthesis provide resistance by alteration and overexpression of target proteins. The gene PMR5 (ABC transporter) is the only common transporter protein that provides resistance to antifungals (Fig. 6b).

Among the 78 AFR genes, SNPs were detected in ERG11 (lanosterol 14-α demethylase), FKS1 (β-glucan synthase), UXS1 (UDP-glucuronate decarboxylase 1), FUR1 (Uracil phosphoribosyltransferase), TUB1 (β tubulin) and FCY2 (Cytosine permease) genes. Among them, ERG11 was noted to have the highest number of mutations, followed by UXS1, FKS1, FUR1 and FCY2 (Fig. 6c; Table 2). In the FKS1 gene, serine at the 643rd position was mutated to proline. This resulted in the conformational change in the lipid-enriched molecules, as it has a role in anchoring the lipid molecules, and the amino acid serine at the 643rd position is involved in the binding site. In the ERG11 gene, phenylalanine at 126th position has been changed to serine and glycine at the 464th position has been mutated to valine. Neither of these residues was conferred to the active site binding pocket. In the TUB1, histidine at the 6th position was changed to glutamine, glutamic acid at 98th position was mutated to glycine and phenylalanine at 200th position was mutated to asparagine. Glycine at the 190th position was mutated to glutamic acid in the ABC transporter gene FUR1. In all the study genomes, though the mutation was observed, none of them were conferring to the active site binding pocket, which is also concurrent with our in vitro susceptibility test, where the strain Cn showed sensitivity to amphotericin B and heteroresistance to fluconazole²⁶.

Table 2.

The SNPs present in AFR genes of C. neoformans study strains.

Genes	Position of Residue in protein	Mutation detected differently than proposed in journals
ERG11(9)
Y145F	active site residue	–
G484S	active site residue	–
S460T	not an active site residue	–
A457V	not an active site residue	–
I99V	not an active site residue	–
Y132F	active site residue	–
K143R	not an active site residue	–
F126L	not an active site residue	F126S
G464S	not an active site residue	G464V
FKS1 (1)
S643P	not an active site residue	–
TUB1 (3)
H6Q	not an active site residue	H6R
E98	not an active site residue	E98G
F200	not an active site residue	F200N
UXS1 (2)
D306G	not an active site residue	–
Y217C	not an active site residue	–
FUR1 (3)
G190D	not an active site residue	–
G87D	not an active site residue	–
G190E	not an active site residue	–
FCY2 (1)
A157V	not an active site residue	–

Pan-genome analysis and phylogenetic analysis using orthologous genes

The pan-genome analysis aids in understanding the genome architecture of a species. The preliminary analyses showed the number of gene clusters, proteins and singleton genes in each genome (Figs. 7a, b, c, d and e; Table 3). The orthologous analysis generates two data sets- the upset chart (Supplementary file 1, 2 and 3), where the unique orthologous clusters and proteins present in the 12 genomes are independent of one another and the occurrence chart, where the unique orthologous clusters and proteins depend on each other. Both charts visualize the presence/ absence of orthologous gene clusters and proteins across the study genomes. The unique clusters and proteins of the study genomes derived from the upset chart are demonstrated in Table 4. The unique clusters and proteins of the study genomes derived from the occurrence chart are depicted in Table 5. From the occurrence chart of the gene cluster, it was illustrated that there were about n = 4676 clusters (73.6%) containing n = 57,595 proteins (77.3%) present in all the study genomes (Supplementary Figs. 1, 2 and 3). Additionally, 4304 single-copy orthologous clusters are identified among the 12 genomes. The single-copy clusters were categorized into biological process-associated, molecular function-associated and cellular component-associated gene clusters (Fig. 8; Table 6; Supplementary Table 3, sheet 3). The phylogenetic tree was constructed by the maximum likelihood method using the orthologous genes depicted that the midpoint was divided into two clades, where the JEC21 (subsp. neoformans) was in Clade 1 and the rest in Clade 2, occupied by the subspecies grubii. Clade 2 was further divided into 2 subclades, where subclade 1 has the environmental strain VNII and subclade 2 has clinical strains further categorized into 6 clusters. The gene expansion and contraction studies using the CAFE5 showed duplication (gene expansion) and deletion (gene contraction) of gene families. The highest contraction of gene families was reported in the VNII and the strains from Australia showed merely stable expansion and contraction (Fig. 7f).

Fig. 7.

The preliminary orthologous analysis (a) the number of gene clusters, (b) proteins, (c) singletons present in each study strain; (d) The overall orthologous clusters and (e) the overall orthologous proteins; (f) Phylogenetic tree based on orthologous genes: Orthologous gene-based phylogeny was constructed using the maximum likelihood method with CAFE5. The midpoint of the phylogenetic tree was divided into two major clades by the subspecies grubii (red line) and neoformans (yellow line). The tree picture represents the environmental strain and the human represents the clinical strain. The magenta colour represents the Botswana (South Africa) strains, the blue colour represents the TamilNadu (India) strain, the orange colour represents the Melbourne (Australia) strains, the violet colour represents the Tennessee (USA) strain and the red colour represents the California (USA) strains. In the pie chart near each strain, the blue area represents the contraction of gene families and the orange area represents the expansion of gene families.

Table 3.

The overall read statistics of orthologous gene.

All clusters	6353
Single copy clusters	4360
All proteins	74,110
All singletons	325
% of singletons	0.44%
Total	6353

Table 4.

The unique orthologous cluster and proteins present in the 12 genomes- Upset table where the unique clusters and protein count are independent.

Strain	Unique cluster	Unique protein
Cn	9	9
H99	89	89
JEC21	19	19
Bt65	4	4
Bt81	2	2
Bt89	1	1
Bt133	1	1
C23	88	88
VNII	6	6
KBCN0140	87	87
KBCN0142	89	89
SACN00B1	87	87
Total	482	482

Table 5.

The unique orthologous cluster and proteins present in the 12 genomes- Occurrence table where the unique clusters are correlated with the protein count.

S.no	Strain name	Unique clusters	Unique proteins
1	JEC21	19	108
2	Cn	9	27
3	Bt65	4	84
4	Bt81	2	9
5	VNII	6	17
6	Bt133	1	2
7	Bt89	1	2
Total		42	249

Fig. 8.

Single-copy gene clusters present in 12 genomes classified as (a) cellular component genes, (b) biological process gene and (c) molecular function genes.

Table 6.

The single copy number genes present across the 12 genomes.

GO annotation number	Count of unique input accession	Function	% contribution
(a) Biological process cluster
GO:0008150	1762	Biological process	18.48
GO:0008152	1576	metabolic process	16.53
GO:0044237	1114	Cellular metabolic process	11.68
GO:0009987	1037	Cellular process	10.87
GO:0043170	968	Macromolecule metabolic process	10.15
GO:0044238	783	Primary metabolic process	8.21
GO:0006807	778	Nitrogen compound metabolic process	8.16
GO:0016070	563	RNA metabolic process	5.90
GO:0065007	496	Biological regulation	5.20
GO:0046483	459	Heterocycle metabolic process	4.81
Total	9536	–	–
(b) Molecular function cluster
Molecular function
GO:0016787	53	Hydrolase activity	16.11
GO:0016491	50	Oxidoreductase activity	15.20
GO:0016740	38	Transferase activity	11.55
GO:0043167	38	Ion binding	11.55
GO:0003676	35	Nucleic acid binding	10.64
GO:0003674	31	Molecular function	9.42
GO:0005215	26	Transporter activity	7.90
GO:0008233	26	Peptidase activity	7.90
GO:0000166	16	Nucleotide binding	4.86
GO:0048037	16	Cofactor binding	4.86
Total	329	–	–
(c) Cellular component cluster
Cellular component function
GO:0016020	36	Membrane	19.56
GO:0005575	31	Cellular compartment	16.85
GO:0044464	27	Call part	14.67
GO:0005634	24	Nucleus	13.04
GO:0005622	19	Intracellular	10.33
GO:0043226	14	Organelle	7.61
GO:0043229	14	Intracellular organelle	7.61
GO:0005739	10	Mitochondrion	5.43
GO:0005576	5	Extracellular region	2.71
GO:0005794	4	Golgi apparatus	2.17
Total	184	–	–

Discussion

This was the first study to report the sequenced genome of C. neoformans from India. In a comparative genomic analysis of 12 isolates, we deciphered the diversity of genomes, phylogenomics, virulome, resistome and pan-genome profiles of C. neoformans across various countries.

In the initial analysis of nucleotide identity search, the clinical strains of S. Africa (grubii) showed 98% similarity to the clinical strains (grubii) belonging to other regions. The potential reason is their molecular type variation and their hypermutation nature. The S. African strains belong to the VNB molecular type and the other clinical strains (grubii) belong to the VNI molecular type²⁷. Since the environmental strain, VNII is a member of serotype A but has a different molecular type (VNII), it exhibits 97% resemblance to the other clinical strains belonging to the VNI molecular type and serotype A. The clinical strain JEC21 belongs to the molecular type VNIV and serotype D, shows only 88% identity with other strains²⁸.

The ISHAM ITS barcoding helped to confirm the subspecies of the Cn to be grubii (100% identity), which is in accordance with the phylogenetic tree. The Cn strain’s subspecies identifications proved that it belongs to serotype A and molecular type VNI. Due to the lack of 100% matching sequences in the ISHAM database, we anticipate that the strain Cn belongs to a new ST type.

The phylogeny using MLST and orthologous genes supported the lineages and sub-lineages. These results are consistent with whole-genome SNP analysis and nucleotide similarity of the study genomes. The phylogenetic tree constructed using the MLST and orthologous genes is exact (same clade) on strains from Botswana (South Africa), showing convergence in their genome. Sub-Saharan Africa has been the hub for Cryptococcal infection and Botswana belongs to the southern part of Sub-Saharan Africa³. Genetic convergence is obvious in this part due to factors like strong selective pressures, high transmission rate, limited gene flow, and host adaptation. The strains of the U.S., Australia and India showed diversity in the clades.

The virulome analysis depicts the presence of multifaceted stress tolerance genes that contribute to virulence regulation. The ability of Cryptococcus to adapt to the host environment depends on how well it can maintain metabolic homeostasis under stress. All 12 isolates have most of the major VRGs (above 80%). The environmental factors, such as ecological niches, temperature and host immune responses, enhance the potential of Cryptococcal virulence. The avian and marsh environments are nutrient-rich, support growth and enhance the virulence potential due to the conserved cAMP/ PKA signaling pathway that plays a crucial role in nutrient sensing and the regulation of VRGs²⁹. The ability of Cryptococcus to thrive at 37 °C (human body temperature) is crucial for its pathogenicity. The interaction of Cryptococcus with the host immune system is critical for its virulence as it modulates the immune response pathways, particularly inhibiting the inflammasome pathway, which is essential for effective antifungal immunity³⁰. This immunomodulatory activity helps Cryptococcus survive and multiply within the macrophages, further enhancing the virulence. Cryptococcus’s ability to adapt to various ecological niches and capacity to modulate the host immune responses underscore the importance of environmental influences on its pathogenicity. The strains C23, Bt89, and Bt133 were found to lack the major VRG URE1, which serves to breach the BBB and host immune system. Except for Cn, H99 and JEC21, all the others lack the LAC2, contributing majorly to the melanin production, biofilm production and inhibition of phagocytic uptake. This is also concurrent with our previous study, in which the Cn was found to have a dominant capsule, melanin production and biofilm formation²⁶. Despite high nucleotide similarities (99.9%) between the same subspecies, the number of VRGs varies. The factors contributing to such discrepancy are gene variation and divergence, gene acquisition and loss, strain-specific adaptation, gene regulation and expression, genomic rearrangements, gene copy number variation, genome fragmentation, evolutionary pressures, and mobile genetic elements.

Though all the genomes possessed the AFR genes with certain mutations, most mutational sites don’t belong to the active site binding pocket. As additional supporting data, the in vitro investigation on the Cn and the H99 reference strain demonstrated that both strains were susceptible to amphotericin B (polyenes) and fluconazole (azole). With this knowledge, it was anticipated that a mutation outside the active site binding pocket might not cause AFR. This comparative genomic study on virulence and AFR gives insights into the potent targets for developing novel antifungal drugs.

Approximately 70% of Cryptococcal genomes have conserved functional elements (genes involved in signaling and metabolism) that are pivotal for survival and enhanced pathogenicity. Investigation of the conserved functional elements helps to develop effective therapeutic strategies to combat this opportunistic pathogen.

The pan-genome analysis gives various insights into the orthologous genes and proteins. The number of gene clusters and proteins are not necessarily correlated. The KBCN0142 has the highest number of clusters (n = 5995) and the Bt81 has the highest number of proteins (n = 6366). This might be due to high redundancy, high similarity among the proteins, stringent clustering parameters, and the presence of paralogous genes. The singleton proteins in the genomes are essential as they unfold uniquely within the dataset and don’t have any closely related homologs, suggesting their evolutionary significance. The Cn has the highest number of singletons (n = 136). The orthologous analysis demonstrated the intersection of gene clusters and proteins; unique clusters and proteins of each genome. Though there were unique clusters and proteins present in all the study genomes, as represented by the upset charts of cluster count and protein, the occurrence chart, where the gene clusters are correlated with the protein, shows the absence of a unique cluster or protein in H99, C23, KBCN0140, KBCN0142 and SACN00B1. The single-copy clusters are indispensable in comparative genomics and evolutionary studies as they represent the conserved genes that reduce the noise from gene duplication or loss events in the compared species. These clusters were categorized based on functional nature as a biological process, molecular process and cellular component. Among them, the biological process-related GO cluster contributes to a maximum of 96.69% (n = 17,262), molecular function contributes to about 2.23% (n = 398) and finally, the cellular component contributes to about 1.08% (n = 193). This shows that the biological process-related genes are more conserved across species. Though the strains JEC21 and Cn belong to different subspecies, the pairwise heat map shows high similarity between them, which might be because both subspecies were descendants of the same ancestor. The phylogenetic tree constructed with the orthologous genes shows more evolutionary relatedness than the tree constructed with the MLST genes. The reason behind this is the comprehensive representation of the genome, robustness to homoplasy, reduced noise of gene duplication or gene loss and improved alignment quality. These factors reflect the true evolutionary relationships among the species. In this study, the phylogenetic tree constructed using the MLST genes and orthologous genes is similar. The gene expansion and contraction studies showed that there was expansion (additional gene copies have been gained in their lineage) or contraction (number of gene copies lost in the lineage) among the genomes of the strains being isolated from the same region. The strains from Australia alone showed mere stable gene expansion and contraction due to the presence of stable gene families (showing no significant changes in size due to duplication or loss events) or equal evolutionary changes (either the expansion/contraction occurring uniformly/ not occurring, across the species being analyzed leading to unique changes in a single lineage).

In conclusion, this was the first study to report the complete genome sequence of a Cryptococcus sp. from India. Our study revealed that phylogenetic tree construction using orthologous genes yields more accurate evolutionary relationships than MLST. Furthermore, the study identified that the clinical strains from S. Africa and the environmental strain from the U.S. are hypermutators. The virulome analysis identified that all major VRGs are present in Cn (India) and H99 (U.S.), consistent with in vivo virulence assays. Resistome analysis detected major AFR genes, but mutations were absent in the active site binding pocket, aligning with in vitro MIC data. Pan-genome analysis showed gene expansion and contraction across strains, showcasing genome plasticity as a critical factor for Cryptococcus in adaptability, virulence and resistance. Deciphering these mechanisms provides insights into this pathogen’s evolution and suggests potential targets for therapeutic interventions. In summary, the findings on virulome and resistome profiles shed light on the pathogenicity of different strains, which is crucial for understanding disease management. The study advocates the need for bioinformatic tools for resistome and virulome analysis in eukaryotes.

Materials and methods

Culture isolation and DNA extraction

The Cn strain was collected from Trichy Medical College, Tamil Nadu, India. The preliminary identification was performed by Indian ink (HiMedia) staining after growing the yeast colonies in potato dextrose agar (HiMedia). The purified yeast colonies were stored in 30% glycerol at -80℃.

Approximately 10⁷ log-phase cells were obtained by centrifugation at 7500 rpm for 10 min. The isopropanol precipitation and urea lysis buffer were used to extract the DNA from the pellet^31,32. The concentration and purity of genomic DNA were quantified using a Nanodrop Spectrophotometer (Thermo Scientific; 2000) and Qubit dsDNA HS assay kit. DNA integrity was observed on agarose gel electrophoresis.

Complete genome sequencing

Illumina sequencing and preprocessing

The sequencing library was generated using the QIASeq FX DNA Library Kit (2023)³³. It was quantified using a Qubit fluorometer (Thermo Fisher Scientific, MA, USA) and the Agilent 2200 Tape Station (Supplementary Fig. 4).

The libraries were sequenced on Illumina NovaSeq 6000 (Illumina, San Diego, USA) using 150 bp paired-end chemistry. The data obtained from the sequencing run was de-multiplexed using Bcl2fastq2 software v2.20. Based on the distinct dual barcode sequences, FastQ files were generated. The sequencing quality was assessed using FastQC3 v0.11.8 software and Trimgalore—v0.4.0.

Nanopore sequencing and preprocessing of reads

The genomic DNA was end-polished and A-tailed (NEBnext ultra II end repair kit, New England Biolabs, MA, USA). A blunt TA ligase master mix (M0367L) barcodes the end-prepared samples. NEB Quick T4 DNA Ligase ligated the sequencing adapters (SQK-NBD114) onto double-stranded DNA fragments. The prepared library was purified using AMPure XP beads. Qubit quantified the prepared library.

The libraries were sequenced on the PromethION system (PromethION P24 and Data Acquisition Unit, ONT, Oxford, UK) using a PromethION flow cell (FLO-PRO114M). Porechop-v0.2.4 was used to process the Nanopore raw reads.

Assembly and annotation

The MaSuRCA5-v3.4.2 assembler was used to assemble high-quality processed reads by de novo assembly³⁴. The assembled genome was further improved by gap closure using LRGapcloser6-v3.0 and scaffold construction using the LRScaf-v1.1.12³⁵. The assembled genome was validated using BUSCO-v5.4.2³⁶. Further, the annotation was executed by Braker-v2.1.4³⁷. The predicted proteins were aligned with reference using DIAMOND-v0.8.29³⁸. The KAAS database was used to retrieve information for functional annotation and pathways analysis³⁹. The collected assembled genomes from NCBI datasets were annotated by AUGUSTUS v3.4.0⁴⁰.

Comparative genomics of assembled and reference genome

The assembled genome was compared with the reference genome using the Average Nucleotide Identity (ANI) calculator, D-Genies and Mauve- v2.4.0⁴¹.

Retrieval of global genome data

The assembled genome of 11 isolates of C. neoformans was collected from the NCBI datasets. The metadata of the isolates was provided in supplementary Table 4, sheet 4. Due to the lack of fungal databases (virulome and resistome), we have limited our study to 12 isolates (n = 11 isolates- NCBI datasets and n = 1 sequenced genome).

Nucleotide diversity analysis

The study genomes were compared (nucleotide identity) by EZBioCloud, a web-based tool⁴².

SNP analysis

The SNP analysis was performed on the Galaxy v0.7.19 platform. The assembled reads were mapped to the reference H99, using ‘Map with BWA-MEM- Galaxy v0.7.19’. The SNPs were detected using ‘Freebayes- Galaxy v1.3.10 + galaxy0’. The ‘VCFfilter- Galaxy v1.0.0_rc3 + galaxy3’ removes the variants with low-quality scores, high error rates/ ambiguous calls. The SnpSift Extract Fields- Galaxy v4.3 + t.galaxy0 were used to convert the VCF file into a tabular file⁴³.

Phylogenetic analysis

MLST genes (CAP59, GPD1, IGS1, SOD1, LAC1, URA5 and PLB1) of the study genomes were concatenated and the phylogenetic tree was constructed by the maximum likelihood method using MegaX v11.0.13⁴⁴. Tree editing and annotation were performed using the interactive Tree of Life (iTOL v7.2)⁴⁵.

VRG identification and comparative study

The VRGs (n = 274) of C. neoformans were collected from the literature and the DFVF^{15,16,46–53}. The listed genes were retrieved from Uniprot and their presence in the annotated genomes was identified using BLAST. The sequence with 99 to 100% identity was considered for the presence of VRGs. The comprehensive result of the comparative virulome analysis was consolidated using iTOL.

AFR identification and comparative analysis of mutation profiles

The AFR genes of C. neoformans were collected following the same protocol used for VRG identification. Further, the mutations responsible for the development of AFR provided in the AFRbase database were compared with the annotated genomes of the strains using CLUSTAL Omega. The results of the comparative AFR and mutation profiles were consolidated using iTOL.

Pan-genome analysis

The pan-genome analysis was performed with Orthovenn3⁵⁴. Orthologous genes were identified using the OrthoFinder algorithm in combination with the Markov clustering method. The phylogenetic tree was constructed using FastTree2. The gene family contraction and expansion analysis were performed using CAFE5.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

JS performed data curation, genome analysis, methodology, validation, visualization and writing- the original draft; JR performed the roles of conceptualization, methodology, resources (including funding), and supervision, review and editing of the original draft.

Funding

The author(s) reported no funding associated with the work featured in this article.

Data availability

Sequence data of our lab isolate used in this study have been deposited in the NCBI under the BioProject accession number PRJNA1141203. Also, the lab isolate was deposited in the Microbial Type Culture Collection (MTCC) and obtained the accession number MTCC15081. Apart from our lab isolate we used 11 other Genomic sequence data that were assessed from NCBI datasets.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

This study did not involve experiments on humans or the use of human tissue samples. All analyses were performed using a previously isolated Cryptococcus neoformans strain. No new biological samples were obtained from human participants.