Development of an efficient gene-targeting system for elucidating infection mechanisms of the fungal pathogen Trichosporon asahii

Trichosporon asahii is a pathogenic fungus that causes severe, deep-seated fungal infections in neutropenic patients. Elucidating the infection mechanisms of T. asahii based on genetic studies requires a specific gene-targeting system. Here, we established an efficient gene-targeting system in a highly pathogenic T. asahii strain identified using the silkworm infection model. By comparing the pathogenicity of T. asahii clinical isolates in a silkworm infection model, T. asahii MPU129 was identified as a highly pathogenic strain. Using an Agrobacterium tumefaciens-mediated gene transfer system, we obtained a T. asahii MPU129 mutant lacking the ku70 gene, which encodes the Ku70 protein involved in the non-homologous end-joining repair of DNA double-strand breaks. The ku70 gene-deficient mutant showed higher gene-targeting efficiency than the wild-type strain for constructing a mutant lacking the cnb1 gene, which encodes the beta-subunit of calcineurin. The cnb1 gene-deficient mutant showed reduced pathogenicity against silkworms compared with the parental strain. These results suggest that an efficient gene-targeting system in a highly pathogenic T. asahii strain is a useful tool for elucidating the molecular mechanisms of T. asahii infection.

Trichosporon asahii is a basidiomycete yeast that is widely distributed in the environment and is often isolated from human blood, sputum, skin, feces, and urine [1][2][3][4][5][6] . T. asahii causes severe, deep-seated fungal infections in neutropenic patients [7][8][9] . Deep mycoses caused by T. asahii has a twofold higher mortality rate than those caused by Candida albicans (80% vs 40%) 10 . Since T. asahii is resistant to echinocandin antifungals, patients treated with micafungin are susceptible to the development of severe infections 11 . T. asahii strains resistant to amphotericin B and azole antifungals such as fluconazole have also been isolated from patients 12,13 . Moreover, T. asahii forms a biofilm, a three-dimensional structure comprising microbe aggregates and extracellular matrix, on catheter surfaces in patients 14 . The T. asahii cells within biofilms are resistant to antifungal drugs 13 . T. asahii has morphological forms: yeast form, hyphae (filament form) and arthroconidia (chains of cells and asexual spores) 4 . Furthermore, arthroconidia of T. asahii may play a key role in biofilm formation by promoting cellular adhesion 15 . T. asahii is therefore a highly problematic clinical pathogen 9 . Since the technology to construct gene-deficient mutants of T. asahii has not been established, it has not been possible to study pathogenicity and drug resistance in T. asahii using a gene-deficient mutant.
In general, mammals such as mice are used as experimental models in studies of infectious diseases 16 . The use of mammalian animals in infection experiments requires specialized experimental facilities, and the large number of animals required for these studies is a severe limitation due to ethical issues regarding animal welfare 17 . T. asahii infection experiments are not easy to perform in mice because immunosuppressive drugs must be administered 18,19 . To address these issues, we established a silkworm infection model for elucidating the mechanisms of T. asahii infection 20 . Compared with mammals such as mice, the use of invertebrate silkworms is advantageous because they are less costly to house and easier to rear in large numbers in simple facilities, and www.nature.com/scientificreports/ fewer ethical problems are associated with their use. Therefore, the use of silkworms as an experimental animal enhances the feasibility of performing large-scale, in vivo screening using a large number of individuals 16 . Novel virulence genes in the pathogenic bacterium Staphylococcus aureus were identified using a silkworm infection model and a library of gene-deficient strains 21,22 . Silkworm infection models have also been used to identify virulence genes of the pathogenic fungi C. albicans and Candida glabrata 23,24 . In Cryptococcus neoformans, a basidiomycete yeast like T. asahii, a strain that is highly pathogenic to mice is also highly pathogenic to silkworms 25 . Moreover, a C. neoformans strain lacking the gene encoding the calcineurin subunit, which contributes to the pathogenicity against mice, was less virulent against silkworms 25 . Since the silkworm can be used to evaluate differences in the pathogenicity between strains of pathogenic fungi, the silkworm infection model with T. asahii may be useful for elucidating the infection mechanisms of T. asahii with the gene-deficient strains. We successfully established a T. asahii strain that expresses green fluorescent protein using an Agrobacterium tumefaciens-mediated gene transfer (ATMT) system 20 . A method for generating a gene-deficient strain of T. asahii, however, has not yet been established.
Homologous recombination (HR), a repair mechanism for DNA double-strand breaks (DSBs), is required to introduce mutations into a gene-targeting system using homologous DNA fragments 26,27 . Another repair mechanism is the non-homologous end joining (NHEJ) of DSBs 26 . These 2 main repair mechanisms affect gene-targeting efficiency by introducing homologous DNA fragments. NHEJ repair mediates the insertion of introduced homologous DNA fragments into genome sites that are different from the target region, thereby reducing the gene-targeting efficiency for generating a gene-deficient strain 26 . Therefore, gene-targeting efficiency can be increased by inhibiting NHEJ repair 26,28 . Ku70 and Ku80 proteins form heterodimers and are involved in the NHEJ repair for DSBs 29 . In several fungi, deletion of the genes encoding these proteins led to the increase of gene-targeting efficiency for generating gene-deficient strains 28,30,31 . In C. neoformans, gene-deficient strains could be generated in the ku80 gene-deficient strain, but not in the wild-type strain, by electroporation, a gene transfer method 32 . Therefore, strains with inhibited NHEJ repair due to disruption of the gene encoding Ku proteins are useful parental strains for promoting genetic studies.
In this study, we identified the T. asahii MPU129 strain, a clinical isolate that is highly pathogenic to silkworms, and generated a T. asahii MPU129 mutant deficient in the ku70 gene. Gene-targeting efficiency to obtain the ku70 gene-deficient strain was higher than that in the wild-type strain. Our findings suggest that a T. asahii strain showing high gene-targeting efficiency and the silkworm infection model are useful tools for studying infectious diseases as a preliminary step to conducting experiments in mice.

Results
Comparison of the pathogenicity of T. asahii strains using a silkworm infection model. Highly pathogenic strains are useful for understanding the molecular mechanisms of pathogens because several pathogenic strains obtain virulence genes by horizontal gene transfer and gene mutation 33,34 . First, we identified T. asahii strains that are highly pathogenic to silkworms. Using a silkworm infection model, we determined the median lethal dose (LD 50 values) on the basis of curves drawn by a simple logistic regression model (Fig. 1). The LD 50 values of the 17 clinical isolates were 9.3 × 10 3 -2.0 × 10 6 cells/larva and the LD 50 value of the MPU129 strain was the lowest, more than tenfold lower than that of the JCM2466 strain (Fig. 1g). The result suggests that the pathogenicity of the MPU129 strain against silkworms is highest among these T. asahii strains.
Generation of the ku70 gene-deficient mutant in the T. asahii MPU129 strain. We next obtained a ku70 gene-deficient mutant of the MPU129 strain using the ATMT system. The targeting plasmid, pAg1-5′UTR (ku70)-nptII-3′UTR (ku70), contained the nptII gene that leads to resistance against G418, an aminoglycoside used as a selective agent for eukaryotic cells (Fig. 2a). Colonies on Sabouraud dextrose agar containing G418 were obtained using the ATMT system (Fig. 2b). In the genome of the 414th candidate colony, polymerase chain reaction (PCR) amplification revealed DNA fragments of the predicted size ( Fig. 2c,d). The results suggest that the ku70 gene-deficient mutant in the T. asahii MPU129 strain was obtained using the ATMT system.
Effect of ku70 gene deficiency on growth and pathogenicity of the T. asahii MPU129 strain. We investigated whether the deficiency of ku70 gene in the T. asahii MPU129 strain affected its growth on nutrient media and its pathogenicity in silkworms. The growth of the ku70 gene-deficient mutant in RPMI-1640 or Sabouraud liquid medium was similar to that of wild-type at either 27 °C or 37 °C (Fig. 3a-d). Moreover, microscopic analysis did not reveal significant differences in the morphology (Fig. 3e). Furthermore, the time required for the ku70 gene-deficient mutant to kill all the silkworms was similar to that of the wild-type strain ( Fig. 3f-h). These results demonstrated that ku70 gene deficiency in the T. asahii MPU129 strain did not significantly affect its growth on nutrient media or its pathogenicity to silkworms.
Increased gene-targeting efficiency in the ku70 gene-deficient T. asahii mutant. We examined whether gene-targeting efficiency was increased in the ku70 gene-deficient mutant by determining the ratio of the strain lacking the cnb1 gene, which encodes the β-subunit of calcineurin. Since gene-targeting efficiency in ku80 gene-deficient mutant of C. neoformans was tested by electroporation, a faster and simpler gene transfer method 32 , we also used electroporation to investigate T. asahii. A DNA fragment, 5'UTR (cnb1)-NAT1-3'UTR (cnb1), was introduced to delete the cnb1 gene by electroporation. Nourseothricin-resistant strains were obtained, and each colony was confirmed by PCR to be deficient in the cnb1 gene ( Fig. 4a-d). Of the 21 nourseothricinresistant colonies obtained by introducing the 5'UTR (cnb1)-NAT1-3'UTR (cnb1) into the ku70 gene-deficient mutant, 4 were deficient for the cnb1 gene (Table 1). On the other hand, none of the 120 nourseothricin-resistant colonies obtained by introducing the 5'UTR (cnb1)-NAT1-3′UTR (cnb1) into the wild-type was deficient for the www.nature.com/scientificreports/ cnb1 gene (Table 1). These results suggest that the deficiency of the ku70 gene in the MPU129 strain increases the gene-targeting efficiency for generating a gene-deficient mutant by electroporation.
Attenuated pathogenicity of the cnb1 gene-deficient mutant against silkworms. In C. neoformans, the pathogenicity of the cnb1 gene-deficient mutant against silkworms was reduced 25 . We examined whether the cnb1 gene-deficient mutants of T. asahii had reduced pathogenicity against silkworms. The survival time of silkworms injected with the cnb1 gene-deficient mutants was longer than that of the parental strain (Fig. 5a). The LD 50 values of the cnb1 gene-deficient mutants were 89-fold higher than that of the parent strain (Fig. 5b). The result suggests that pathogenicity against silkworms was reduced by cnb1 gene deficiency in T. asahii.

Discussion
In this study, we identified a T. asahii strain that is highly pathogenic against silkworms and established a platform for generating a gene-deficient mutant. The cnb1 gene-deficient mutant obtained using the technique showed decreased pathogenicity against silkworms. To our knowledge, this is the first report of a method for obtaining a gene-deficient mutant of T. asahii. Our results suggest that the calcineurin pathway is involved in the pathogenicity of T. asahii.
In the silkworm infection model with T. asahii, the MPU129 strain showed high pathogenicity among clinical isolates used in this study. We assumed that the MPU129 strain can adapt to the host environments and appropriately regulate the pathogenicity compared with other isolates. To reveal the relationship between clinical information and the pathogenicity in the silkworm infection model among the clinical isolates will be an important study. www.nature.com/scientificreports/ Silkworms are suitable experimental animals for performing large-scale in vivo evaluations because they are relatively inexpensive and few ethical issues are associated with their use. Therefore, silkworm infection models are suitable for quantitative evaluation of the pathogenicity of microorganisms based on the calculation of LD 50 values. Using the silkworm infection model and a gene-deficient mutant library of S. aureus, we previously identified the virulence genes that contribute to pathogenicity against mice 21 . The gene-targeting system in T. asahii developed in the present study will facilitate the construction of a gene-deficient mutant library of T. asahii. It is expected that a gene-deficient mutant library of T. asahii for application to silkworm infection models will help elucidate the molecular mechanisms of T. asahii infection. The contribution of candidate virulence genes to pathogenicity that is identified using the silkworm infection model should be confirmed by infection experiments in mice.
Gene-targeting efficiency by electroporation was higher in the ku70 gene-deficient mutant, while growth on nutrient media and pathogenicity to silkworms remained unaltered. Therefore, the ku70 gene-deficient mutant is useful as a parental strain for elucidating the infection mechanism of T. asahii based on genetic studies. In the pathogenic fungus Aspergillus fumigatus, the pathogenicity of the ku80 gene-deficient mutant did not differ from that of wild-type against mice 35 . We considered that NHEJ repair of DSBs might not be greatly involved in the pathogenicity of T. asahii or A. fumigatus. Although electroporation is a faster and simpler gene transfer method than the ATMT system for obtaining gene-deficient mutants, homologous recombination by electroporation occurs at low frequency 32,36 . When using electroporation, no cnb1 gene-deficient mutants were obtained in the wild-type, but a 19% ratio of mutants was obtained in the ku70 gene-deficient mutant. The result suggests that NHEJ of DSBs occurs at a high frequency in T. asahii. Therefore, we reasoned that the generation of target gene-deficient mutants by electroporation requires the ku70 gene-deficient mutant that lacks NHEJ repair activity. When we obtain fungal colonies grown on a drug-containing agar medium, both strains with mutations in the targeted gene region and strains with non-specific gene insertions caused by NHEJ were obtained. Therefore, NHEJ may contribute to obtaining the strains with a non-specific gene inserted mutants by selecting drug-resistant strains. It can be constructed the gene-deficient mutants in T. asahii within two weeks by using the TR129 ku70 gene-deficient mutant with electroporation method. Construction of a gene-deficient mutant library of T. asahii using the ku70 gene-deficient mutant as a parent strain is thus planned for future studies.
Although we tried to obtain a cnb1 gene-deficient mutant using the ku70 gene-deficient mutant as the parent strain with the ATMT system, we did not obtain a drug-resistant candidate. Optimization of the ATMT system using the ku70 gene-deficient mutant is needed to obtain target gene-deficient mutants. Moreover, CRISPR-CAS9 technology was applied for gene editing in fungi including C. neoformans 37,38 . The establishment of the CRISPR-CAS9 mediated gene-editing method for T. asahii will be a future subject. www.nature.com/scientificreports/ In C. neoformans, the calcineurin pathway is involved in capsule production and melanin synthesis, which are responsible for evading host immunity 39 . A calcineurin-deficient strain of C. neoformans showed decreased pathogenicity against mice and silkworms 25,40,41 . The present study also showed that the calcineurin pathway is involved in the pathogenicity of T. asahii against silkworms. The calcineurin in C. neoformans regulates gene expression via the dephosphorylation of the transcription factors 39 . Therefore, we assumed that the calcineurin in T. asahii also regulates the virulence-related gene expression. To reveal the role of calcineurin in the T. asahii pathogenicity will be an important subject. Further studies are needed to investigate the generation of a revertant strain obtained by reintroducing the cnb1 gene into the cnb1 gene-deficient mutant and to perform a detailed functional analysis using the cnb1 gene-deficient mutant and its revertant strain.
In conclusion, we established a simple method for generating a gene-deficient T. asahii strain that is highly pathogenic against silkworms. The ku70 gene-deficient mutant in the T. asahii MPU129 strain is useful as a parental strain for genetic studies and an important tool for studying infectious diseases of T. asahii.

Silkworm infection experiments. Silkworm infection experiments
To prepare competent cells for electroporation, T. asahii MPU129 strain was spread on a Sabouraud dextrose agar plate and cultured at 27 °C for 3 days. T. asahii cells on the agar were suspended by physiologic saline solution (2 ml), and the suspension was transferred to a 1.5-ml tube. The fungal cells were collected by centrifugation at 8000 rpm for 3 min (TOMY-MX100, TOMY Digital Biology Co. Ltd, Tokyo, Japan) and suspended by adding 1 ml of ice-cold water and centrifuged at 8000 rpm for 3 min. This washing process was repeated 4 times. The washed cells were suspended by adding 1 ml of 1.2 M sorbitol solution and centrifuged at 8000 rpm for 3 min. The obtained fungal cells were suspended with 0.2 ml of 1.2 M sorbitol solution as competent cells. The PCRamplified 5′-UTR (cnb1) -NAT1-3′-UTR (cnb1) fragment (180 ng/2 µl) was added to the T. asahii competent cells Table 2. Primers used in this study.

Primers
Nucleic acid sequence pAg1-ku70(3′UTR)-nptII-ku70(5′UTR) for cloning www.nature.com/scientificreports/ (40 µl) and placed on ice for 15 min. The suspension was added to a 0.2-cm gap cuvette (Bio-Rad Laboratories, Inc.) and electroporated (Time constant protocol: 1800 V, 5 ms) using a Gene Pulser Xcell (Bio-Rad Laboratories, Inc.). The cells were suspended by adding 500 µl YPD containing 0.6 M sorbitol and incubated at 27 °C for 3 h. After incubation, the cells were collected by centrifugation at 10,000 rpm for 5 min and suspended in 100 µl of physiologic saline solution and applied to Sabouraud dextrose agar containing nourseothricin (300 µg/ml). The cells were incubated at 27 °C for 3 days and the growing colonies were isolated as cnb1 gene-deficient strain candidates. Introduction of the mutation into the genome of the candidate strains was confirmed by PCR using the primers shown in Table 2 and the extracted genome as a template DNA.

Statistical analysis.
All experiments were performed at least twice and the representative results were shown. The significance of differences between groups in silkworm infection experiments was calculated by the log-rank test based on the curves by the Kaplan-Meier method using Prism 9.1.2.

Data availability
The datasets generated during the current study are available from the corresponding author on reasonable request.