Antifungal Resistance and Genotyping of Clinical Candida parapsilosis Complex in Japan

Non-albicans Candida infections have recently gained worldwide attention due to their intrinsic resistance to different antifungal agents and the limited therapeutic options for treating them. Although the Candida parapsilosis complex is reported to be the second or third most prevalent Candida spp., little information is available on the prevalence of antifungal resistance along with genotyping of the C. parapsilosis complex. In this study, we aimed to evaluate the prevalence of antifungal resistance, the genetic basis of such resistance, and the genotyping of C. parapsilosis complex isolates that were recovered from hospitalized patients in Japan from 2005 to 2019. Our results indicated that, with the exception of one single C. metapsilosis isolate that was dose-dependently susceptible to fluconazole, all other isolates were susceptible or showed wild phenotypes to all tested antifungals, including azoles, echinocandins, amphotericin B, and flucytosine. Molecular analyses for azole and echinocandin resistance via evaluating ERG11 mutation and FKS1 hotspot one (HS1) and hotspot two (HS2) mutations, respectively, confirmed the phenotypic results. Genotyping of our isolates confirmed that they belong to 53 different but closely related genotypes, with a similarity percentage of up to 90%. Our results are of significant concern, since understanding the genetic basis of echinocandin resistance in the C. parapsilosis complex as well their genotyping is essential for directing targeted therapy, identifying probable infection sources, and developing strategies for overcoming epidemic spread.


Introduction
Pathogenic fungi have become more prevalent in recent decades, posing a rising threat to public health, especially considering the scarcity of antifungal medications available to treat invasive infections, as well as the emergence of antifungal resistance [1].According to a recent estimate, fungal infection affects over a billion people and kills more than 1.5 million per year, which is similar to the outcomes of tuberculosis and is more than three-fold greater than the rates caused by malaria [2].The vast majority of annual deaths due to fungal infection are initially attributed to Candida and Aspergillus infections, which cause a high economic burden for the health care system [3,4].Among Candida infections, a recent concern has been directed to non-albicans Candida infections, owing to their intrinsically decreased susceptibility to commonly used antifungal drugs together with their increasing infection rates, and the development of their resistance to echinocandins and azole derivates [3][4][5][6][7].
C. parapsilosis is reported to be the second or third most prevalent Candida spp. in certain geographical regions, including Japan [5,8].For instance, C. parapsilosis is the second major cause of candidemia in Japan [8], Spain [9], and Iran [10].Furthermore, candidemia associated with C. parapsilosis has increased two-fold between 2008 and 2011 in North America, was responsible for 10 to 20% of all candidemia cases, and was associated with a wide range of clinical manifestations, including meningitis, endocarditis, vulvovaginitis, ocular infections, and urinary tract infections [11].The problems of C. parapsilosis infections are complicated by their increased MIC valuesto the first-line antifungal therapy (echinocandin), as compared to C. albicans or C. glabrata, with differences at the species level [12].Furthermore, recent reports indicate the emergence of fluconazole-resistant C. parapsilosis isolates, which has been associated with invasive infections [10,13].Based on genomic analysis, the C. parapsilosis complex consists of three genetically distinct species: C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis, which are phenotypically indiscernible from one another [10][11][12][13].
C. parapsilosis complex echinocandin resistance is exclusively attributed to the active mutations of FKS1 gene hotspot regions (HS1, HS2) that encode the 1,3-β-D-glucan synthase complex enzyme [10,13].FKS1 hotspot mutations have been confirmed as a predisposing factor of therapeutic failure in candidemic patients and are basically related to prior echinocandin therapy [14].Regarding azole resistance, two major mechanisms were reported in C. parapsilosis: (i) reduced azole accumulation caused by overexpression of the CDR1, CDR2, and MDR1 genes, causing active efflux of drugs; and (ii) an active mutation in the drug target, the ERG11 gene, which is associated with alterations in target protein structures, reductions in drug binding affinity, and a subsequently increased azole resistance [10][11][12][13].
To date, the prevalence of antifungal resistance, genetic mechanisms associated with resistance, and C. parapsilosis genotyping have never been tested in Japan.As far as we are aware, this is the first study to evaluate the epidemiology of antifungal resistance and genotyping of the C. parapsilosis complex recovered from clinical settings in Japan.

Candida parapsilosis Complex Isolates
In this study, a total of 79 clinical C. parapsilosis complex isolates recovered from 76 patients were tested, including 65 C. parapsilosis isolates recovered from 63 patients, 9 C. metapsilosis isolates recovered from 9 patients, and 5 C. orthopsilosis isolates recovered from 4 patients (Table S1).The isolates were obtained from inpatients of different hospitals in 13 prefectures across Japan (Figure S1) during a 15-year period, from 2005 to 2019 (Table S1).All of the isolates were provided through the National BioResource Project (NBRP), Japan "http://www.nbrp.jp/(accessed on 19 December 2023)".The study's protocols and procedures were approved (approval number MMRC-REC 21-27) by the Ethical Committee of the Medical Mycology Research Center, Chiba University.Identification and confirmation of the isolates were performed via sequencing and analysis of the ITS1-5.8SrRNA-ITS2 DNA region, as previously described in [3,4,14].

Antifungal Susceptibility Testing
The antifungal susceptibility profiles of all of the isolates were determined by evaluating the minimum inhibitory concentrations (MICs) for the different antifungal agents fluconazole (FLC), voriconazole (VRC), itraconazole (ITC), and miconazole (MZ), as representatives of azoles, caspofungin (CAS) and micafungin (MFG), as representatives of echinocandins, and amphotericin B (AMB) and flucytosine (5FC), through broth microdilution assays according to CLSI document M27-Ed4, using Eiken dried yeast-like fungal DP plates EF-47 (Eiken Chemicals, Tokyo, Japan) [15].C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 were tested as quality control strains and the antifungal breakpoints were reported according to CLSI document M60 [16].Resistances to FLC, CAS, and MFG were reported when the MIC values were ≥8 µg/mL, and were reported for VRC when the MIC value was ≥1 µg/mL [16].The susceptibility profiles of ITC, AMB, and 5FC were recorded according to the epidemiological cutoff values (ECVs), and an isolate was reported as a non-wild type (non-WT) when the ECVs were >0.5, >2, and >0.5, respectively [17].On the other hand, there are no established breakpoints or ECVs for MZ [16,17].

Genomic DNA Extraction
The genomic DNA of all isolates was extracted as previously described for Candida spp., with minor modifications [3,14].Briefly, all isolates were grown on Sabouraud dextrose agar (SDA) for 24-48 h at 35 • C, followed by mixing and vigorous vertexing of 1 to 2 loopfuls of the yeast culture with 150 µL of lysis buffer consisting of 30 mM EDTA, 0.5% (w/v) sodium dodecyl sulfate, and 200 mM Tris-HCl (pH 8.0).After incubation at 100 • C for 20 min, the solution was mixed with 150 µL phenol-chloroform-isoamyl alcohol (25:24:1) and centrifugated at 13,000 rpm for 4 min.The clear supernatant was mixed with 300 µL of previously chilled 96% ethanol in a new Eppendorf tube.The solution was gently mixed and incubated in ice for 10-15 min, followed by centrifugation at 13,000 rpm for 15 min at 4 • C for DNA precipitation.After washing each DNA pellet with 500 µL of previously chilled 70% ethanol, the pellet was dried and suspended in 100-200 µL of sterile TE buffer or sterile distilled water, followed by preservation at −20 • C. Before the PCR experiments, the DNA template was prepared with a 10-fold dilution of DNA in sterile distilled water, and 1 µL of the resulting solution was used.

Detection of ERG11 Mutations
PCR and DNA sequencing was performed to check for the presence of ERG11 mutations in all C. parapsilosis and C. orthopsilosis isolates.For C. parapsilosis ERG11 (CpERG11), NCBI accession number NW_023503279.1 for C. parapsilosis strain CDC317 was used for the design of primers and the ERG11 sequence of C. parapsilosis ATCC 22019 was used as reference.For PCR and DNA sequencing of ERG11, four newly designed primers were used, and they are listed in Table S2.For C. orthopsilosis ERG11 (CoERG11), two previously published primers were used for the PCR experiments (Table S2).Besides these primers, two other newly designed primers were used for the sequence of CoERG1 based on NCBI accession number MG601484.1 for C. orthopsilosis isolate Rome1 (Table S2).Unfortunately, the C. metapsilosis ERG11 (CmERG11) sequence is not available in the database, hence the CmERG11 sequence was not investigated in this study.

Detection of FKS1 (HS1 and HS2) Mutations
GenBank accession numbers EU221325.1,XM_003867859.1,and EU350514.1 for C. parapsilosis, C. orthopsilosis, and C. metapsilosis, respectively, were used as a reference and for the primer design of the FKS1 HS1 and HS2 regions.For the PCR reactions and DNA sequencing of both regions, four primers were designed and used for every species (Table S2 and Figures S2-S4).

Microsatellite Typing of C. parapsilosis Isolates
Genotyping of C. parapsilosis isolates was performed based on the microsatellite typing method using four loci designated as CP1, CP4, CP6, and B, composed of tandemly repetitive stretches of three nucleotides, which has previously been described to achieve a discriminatory power of 99.9% [18].For exact and accurate allele size determination, the forward primers were fluorescently labeled with VIC dye for CP1, PET dye for CP4 loci, and FAM dye for CP6 and B5 loci (Supplementary Table S2).The alleles were designated according to their sizes (in base pairs) by using GeneScan™ 500 ROX™ Size Standard (Applied Biosystems, Warrington, UK) in the 35-500 nucleotide range and examined with PeakScanner (Thermo Fisher Scientific, Waltham, MA, USA).Based on the allele sizes of the four diploid loci for each isolate, a dendrogram was constructed by using BioNumerics v7.6 software (Applied Maths Inc., Austin, TX, USA) and a clustering method using the unweighted pair group method with average linkage (UPGMA) settings, as described previously [3,6].

Data Availability
The C. parapsilosis ERG11 gene, FKS1 HS1 region, and FKS1 HS2 region sequences reported in this study have been deposited in GenBank under accession numbers OR536963 to OR537027, OR537028 to OR537092, and OR537093 to OR537157, respectively.The C. orthopsilosis ERG11 gene, FKS1 HS1 region, and FKS1 HS2 region sequences reported in this study have been deposited in GenBank under accession numbers OR537158 to OR537162, OR537163 to OR537167, and OR537168 to OR537172, respectively.

Mutations in the ERG11 Gene and FKS1 HS Regions
For C. parapsilosis, all isolates harbored ERG11 gene-synonymous mutations at T591C, and 33 isolates had missense mutations at R398I as compared to C. parapsilosis ATCC 22019 (Table S4).For C. orthopsilosis, four isolates harbored ERG11 gene-nonsynonymous mutations at Y13C and F420S, and one isolate harbored nonsynonymous mutations at Q211K, F420S, A421V, and V481I as compared to C. orthopsilosis isolate Rome1 (Table S4).However, none of the isolates with ERG11 missense mutations showed a higher MIC value for azoles.Furthermore, C. metapsilosis was not tested for the ERG11 sequence.Checking the HS1 and HS2 regions of FKS1 for all C. parapsilosis isolates, five C. metapsilosis isolates and all C. orthopsilosis isolates confirmed the absence of missense mutations.

MLST Genotyping, Phylogeny, and Population Genetics
The microsatellite typing method using four loci designated as CP1, CP4, B, and CP6 loci was performed.Since C. parapsilosis is a diploid species [18], one or two PCR fragments per locus were produced for each strain, and each fragment was allocated to an allele.When a strain produced two PCR products, it was classified as heterozygous, whereas strains that produced only one amplification product were categorized as homozygous.Our analysis of the 63 isolates showed that all microsatellite loci were exhibiting between 15 and 30 alleles and were from 16 to 32 different genotypes (Table 2).The size ranges (bp) of the CP1, CP4, B, and CP6 alleles were 216-269, 253-479, 116-197, and 213-328, respectively (Table 2).The microsatellite genotyping using a panel of four loci markers identified 53 different genotypes (Table S5, Figure 1), of which 50 were observed only once.Three genotypes, numbered one to three, were found multiple times, and they were identified from four, three, and two isolates, respectively, from nine different patients (Table S5).The remaining 50 genotypes involved only one patient each, with four isolates being isolated from two different patients (two isolates each).Using the clustering approach and BioNumerics software version 7.6, phylogenetic analyses of the isolates were carried out in order to ascertain the links between the identified genotypes.Our results confirmed a close relationship between all genotypes, with a similarity percentage of up to 90% (Figure 1).

Discussion
Recently, special attention has been paid to non-albicans Candida (NAC) species infections, with particular interest in the C. parapsilosis infection, owing to it being reported as a major cause of candidemia in different countries [8][9][10][11].The progressive increase in the rates of antifungal resistance in most candida infections, and in the C. parapsilosis complex in particular [13], along with the narrowing therapeutic options [7], emphasizes the importance of studying the prevalence of antifungal resistance, as well as their genotyping.
In accordance with prior publications regarding other Candida spp.[3,4,6], our findings showed that C. parapsilosis infections are typically seen in elderly individuals and patients with underlying illnesses.Furthermore, previous reports have confirmed that, throughout the world, Candida species continue to be the leading cause of opportunistic infections, primarily affecting patients over 65 years old [19].The propensity of C. parapsilosis to build a biofilm on catheters and other implanted devices makes it an exogenous pathogen that is primarily found on skin surfaces as opposed to mucosal surfaces.In nursing homes and hospitals, it is transmitted via hand contamination.The fact that elderly patients frequently get at-home health care with indwelling catheter use owing to various chronic conditions is consistent with the observation that most C. parapsilosis infections in our study are identified in elderly patients [19].However, contrary to the statewide data reported by Pfaller et al. that suggest C. parapsilosis, a member of the NAC species, is responsible for the majority of invasive candidiasis cases in children (of nine years old) and neonates in North America [20], only 17% of the patients in this study were children.Furthermore, C. parapsilosis was one of the major NAC species responsible for neonatal candidiasis in different countries including Canada, the UK, and Norway [5].
Our results confirm the absence of azole and echinocandin resistance among the tested C. parapsilosis complex isolates.The low worldwide level of azole and echinocandin resistance in C. parapsilosis has also recently been confirmed by different studies [5,[21][22][23][24][25].For instance, surveys of fluconazole and itraconazole resistance among isolate collections revealed resistance rates ranging from 0 to 4.6%, and from 1.5 to 4%, respectively [5].Furthermore, globally, the fluconazole resistance rate ranged between 2 and 5% among C. parapsilosis isolates [21,22], and fluconazole resistance was reported in 3.4% of 6023 examined isolates in a recent review [23].Notably, 33 C. parapsilosis azole-susceptible isolates had ERG11 missense mutations at R398I.Previous reports have confirmed the lesser role of R398I in azole resistance, as it was recently identified in fluconazole-susceptible C. parapsilosis isolates; and even when R398I was identified in resistant isolates, it was accompanied by other missense mutations such as Tac1 L877P, Tac1 L877P and Mrr1 P250S, Tac1 L877P and Mrr1 S1081P, or Tac1 L877P and Mrr1 P295R [24].Furthermore, our results confirm the absence of ERG11 Y132F variants in Japan.On the other hand, azole-resistant outbreaks of C. parapsilosis associated with the Y132F substitution have been recently identified in different countries including South Korea [24], China [25], Mexico [26], Turkey [27], and Brazil [28].However, conducting other large-scale nationwide studies is essential to monitor the prevalence of such important resistance mechanisms in Japan.
Also, in accordance with our findings regarding echinocandin resistance, in a prospectively collected series of C. parapsilosis isolates, only 0.6% were resistant to echinocandins [29], and a very recent study in China confirmed their very low level of resistance (0.03%) to echinocandins [30].Our results verify that the MIC geometric means of both examined echinocandins (CAS and MFG) do not significantly differ from one other.Other studies have verified that caspofungin outperforms both micafungin and anidulafungin in terms of in vitro activity against C. parapsilosis isolates [30], which is consistent with global surveillance program reports [22,29,30].The susceptibility of Candida species to echinocandins varies; among the three echinocandins, C. albicans, C. tropicalis, C. glabrata, and C. lusitaniae were generally most sensitive to micafungin, while C. krusei and C. pelliculasa were most vulnerable to anidulafungin [30].Our results showed a close relationship between the MIC results and the results from the genetic analysis, as all of the isolates showed an absence of FKS1-HS missense mutations, which are responsible for echinocan-din resistance.Our findings supporting the reliability of azole and echinocandin MIC values obtained via CLSI and EUCAST methods to evaluate the resistance in C. parapsilosis complex, which is unlike other Candida species such as C. glabrata [3] and C. krusei [4], especially regarding echinocandin resistance.With both C. glabrata and C. krusei, we have to depend on FKS1 HS mutation rather than MIC (especially CAS) results to determine echinocandin-resistant isolates.
Candida genotyping has a significant role in the detection of emerging clones and the identification of relationships between certain genotypes and virulence traits, mortality rates, and gene polymorphisms, along with in investigating the potential source of infection [3,4].Microsatellite genotyping, which has a greater discriminative strength than other techniques like DiversiLab typing, was the method we used in this investigation [12,18].Although microsatellite genotyping characterized that our isolates are classified into 53 different genotypes, phylogenetic analysis of the isolates confirmed the close relationship between all of the genotypes, with a similarity percentage up to 90%.As far as we know, this is the first report to confirm this close relationship between Japanese clinical C. parapsilosis isolates.The diversity of the genotypes detected in this study points to the possibility of numerous causes contributing to the occurrence of C. parapsilosis infections in Japan.In line with our findings, two recent studies in Brazil identified different C. parapsilosis genotypes among pediatric patients [31,32], but their results also confirmed whether these genotypes are phylogenetically related or not.Moreover, highly related genotypes have caused outbreaks of C. parapsilosis candidemia in neonatal intensive care units in the USA [33].Furthermore, other studies have also documented the occurrence of clonal complexes of closely related genotypes as a result of microevolution caused by the inherent instability of microsatellite loci [34].

Conclusions
In conclusion, our findings confirm the absence of antifungal resistance among clinical and C. parapsilosis complex isolates recovered in Japan.Our phenotypic susceptibility results were supported by genetic examination, as all of the isolates showed the absence of the missense mutations responsible for azole and echinocandin resistance.For the first time, microsatellite genotyping and phylogenetic analysis has confirmed that different, closely related genotypes are responsible for C. parapsilosis infections in Japan.

Figure 1 .
Figure 1.UPGMA dendrogram showing the similarities among 61 C. parapsilosis isolates (a single isolate for each patient) based on the microsatellite typing method using four loci designated as CP1, CP4, B, and CP6.Two isolates (IFM65553 and IFM64439) were not tested due to failure in the analysis of the CP4 segment despite several trials.Abbreviations: M, male; F, female; UN, unknown.

Figure 1 .
Figure 1.UPGMA dendrogram showing the similarities among 61 C. parapsilosis isolates (a single isolate for each patient) based on the microsatellite typing method using four loci designated as CP1, CP4, B, and CP6.Two isolates (IFM65553 and IFM64439) were not tested due to failure in the analysis of the CP4 segment despite several trials.Abbreviations: M, male; F, female; UN, unknown.

Table 1 .
Summary of antifungal susceptibility profiling of C. parapsilosis complex isolates.

Table 2 .
Characteristics of microsatellite loci for C. parapsilosis isolates.