Clonal Expansion of Environmental Triazole Resistant Aspergillus fumigatus in Iran

Azole-resistance in Aspergillus fumigatus is a worldwide medical concern complicating the management of aspergillosis (IA). Herein, we report the clonal spread of environmental triazole resistant A. fumigatus isolates in Iran. In this study, 63 A. fumigatus isolates were collected from 300 compost samples plated on Sabouraud dextrose agar supplemented with itraconazole (ITR) and voriconazole (VOR). Forty-four isolates had the TR34/L98H mutation and three isolates a TR46/Y121F/T289A resistance mechanism, while two isolates harbored a M172V substitution in cyp51A. Fourteen azole resistant isolates had no mutations in cyp51A. We found that 41 out of 44 A. fumigatus strains with the TR34/L98H mutation, isolated from compost in 13 different Iranian cities, shared the same allele across all nine examined microsatellite loci. Clonal expansion of triazole resistant A. fumigatus in this study emphasizes the importance of establishing antifungal resistance surveillance studies to monitor clinical Aspergillus isolates in Iran, as well as screening for azole resistance in environmental A. fumigatus isolates.


Introduction
Aspergillus fumigatus is the most common agent of various forms of aspergillosis, including allergic bronchopulmonary aspergillosis (ABPA), chronic pulmonary aspergillosis (CPA), aspergilloma, and invasive aspergillosis (IA) [1]. Voriconazole (VOR) is the recommended primary and most effective therapy in the management of aspergillosis [2]. However, azole resistant A. fumigatus isolates are increasingly found worldwide with major epidemiological and clinical implications [3,4]. Therapeutic failure caused by azole-resistant A. fumigatus is becoming a significant concern to clinicians who are caring for patients at high risk for IA [1,[4][5][6][7][8]. Azole resistance in A. fumigatus is mainly linked to cyp51A-mediated resistance mechanism, such as a 34-basepair (bp) sequence tandem repeat (TR 34 ) in the promoter region of the cyp51A gene, in combination with a L98H substitution and a 46 bp tandem repeat (TR 46 ) in the cyp51A promoter in combination with two amino acid changes (Y121F and T289A) in the CYP51A protein (TR 46 /Y121F/T289A) [9]. Isolates carrying these mutations exhibit a pan-azole resistant phenotype that can develop through long-term treatment with azole antifungals in the clinical setting or extensive exposure of the fungus to azole compounds in the environment [10,11]. Azole-resistant A. fumigatus with the TR 34 /L98H mutation isolated from environmental and clinical samples have been reported earlier in Iran, and we recently reported the occurrence of TR 46 /Y121F/T289A mutations in the cyp51A gene in A. fumigatus isolates from compost [12][13][14][15][16][17][18]. High concentrations of azole-resistant A. fumigatus spores are released during incomplete composting processes, especially when azole residues from agricultural waste are present [19]. Agricultural use of fungicides has driven the emergence and spread of azole-resistant A. fumigatus. The existence of an environmental route of azole resistance development involves serious risks for patients, as well, as they can become infected with azole resistant A. fumigatus strains before starting their treatment [12][13][14][15][16][17][18][19][20][21][22][23][24]. Notably, genetic exploration of azole resistant A. fumigatus strains indicates that isolates with the TR 34 /L98H allele are less genetically variable than susceptible isolates [12,23]. For instance, analysis of azole resistant A. fumigatus isolates in the Netherlands showed five distinct genotype groups in this country, while all the azole resistant isolates with the TR 34 /L98H mutation belonged to one group [5]. On the other hand, all clinical and environmental azole resistant A. fumigatus strains carrying TR 34 /L98H obtained from India were genetically identical [14]. These studies illustrate that A. fumigatus carrying this azole resistance mutation may preferentially spread clonal within a population. Major data gaps remain regarding the genotype distribution of azole resistance A. fumigatus in Iran. As ongoing reports indicate an expansion in the frequency of azole resistant A. fumigatus isolates worldwide, understanding the genetic structure of this potentially lethal fungus is critical. In this study, the genetic characterization of azole resistant A. fumigatus isolated from compost samples in Iran was explored.

Isolate Collection
According to a previously described protocol, commercial and home-made compost samples from different region of Iran (located about 300 km apart) were collected. To recover A. fumigatus strains, 1 cm 2 of compost was dissolved in 5 mL sterile saline solution containing Tween 40 (0.05%), vortexed, and allowed to settle. For primary screening of azole-resistant A. fumigatus strains, 100 µL supernatant was plated on a Sabouraud dextrose agar plate (SDA; Difco, Franklin Lakes, NJ, USA), supplemented with 4 and 1 mg/L itraconazole and voriconazole, respectively, and incubated at 45 • C for 72 h in the dark [17]. Molecular identification of all A. fumigatus isolates that grew on the supplemented plate was performed with sequencing of the partial beta-tubulin gene as previously described [16].

In Vitro Antifungal Susceptibility Testing
Minimum inhibitory concentrations (MICs) were determined by broth microdilution susceptibility testing according to the methods in the Clinical and Laboratory Standards Institute (CLSI) reference standard (M38) [25]. For the preparation of the microdilution trays, itraconazole (Janssen, Beerse, Belgium) and voriconazole (Pfizer, Sandwich, UK) were obtained from the respective manufacturers as reagent-grade powders. All drugs were dissolved in 1% dimethyl sulfoxide (DMSO; Sigma, Zwijndrecht, the Netherlands) and were prepared at a final concentration of 0.031-16 mg/L. Paecilomyces variotii (ATCC 22319) and Candida parapsilosis (ATCC 22019) were used as quality controls [25].

Detection of Cyp51a Gene Mutations
All A. fumigatus isolates were subjected to a mixed-format real-time PCR assay specific for TR 34 /L98H and TR 46 /Y121F/T289A mutations of cyp51A gene leading to triazole resistance in A. fumigatus as described previously [26]. Those isolates with negative or inconclusive results in the real-time PCR assay, were further evaluated by sequencing the cyp51A gene as described previously [27].

Microsatellite Genotyping
Genotyping of all A. fumigatus isolates was performed with a panel of nine short tandem repeats (STRs) loci (namely short tandem repeats Aspergillus fumigatus (STRAf ) 2A, 2B, 2C, 3A, 3B, 3C, 4A, 4B, and 4C), as previously described [28]. Genotypes were considered identical when they showed the same alleles for all nine loci [29,30]. Finally, the genetic relatedness between Iranian isolates from compost and 633 resistant A. fumigatus strains with clinical or environmental sources collected during 2001-2019 from different countries (The Netherlands, India, United Kingdom, Tanzania, France, Colombia, Romania, Ireland, China, Kuwait, Germany, and Japan) and previous Iranian isolates in the database at the Center of Expertise in Mycology, Radboudumc/Canisius-Wilhelmina Ziekenhuis (CWZ), in Nijmegen, The Netherlands, already barcoded using a panel of nine short tandem repeat loci, were analysed using BioNumerics software v7.6.1 (Applied Maths, Saint-Martens-Latem, Belgium).

Triazole Resistant A. fumigatus with Mutation in cyp51A Gene
A total of 63 A. fumigatus colonies from 300 compost samples were obtained from SDA supplemented with itraconazole and voriconazole. Of these, 55 A. fumigatus isolates had high MICs of itraconazole (≥8 mg/L) and voriconazole (≥2 mg/L) by in vitro antifungal susceptibility testing. Exploring the mechanisms of resistance in these isolates by sequencing cyp51A and its promoter region showed that 44 isolates harbored the TR 34 /L98H mutation, three isolates the TR 46 /Y121F/T289A mutation and two isolates a M172V mutation. No mutations were found in 14 resistant isolates. Data of resistant isolates are summarized in Table 1. Details of isolates with the TR 46 /Y121F/T289A mutation have been previously described [17].

Microsatellite Typing Results and Evidence for Clonal Spread of a Single Triazole-Resistant A. fumigatus Genotype
Genotypic analysis identified that 41 A. fumigatus isolates with TR 34 /L98H shared the same allele across all nine examined microsatellite loci. These isolates came from compost in 13 different cities. The three remaining isolates with TR 34 /L98H exhibited three different genotypes. The two isolates with M172V differed by five microsatellite loci (2B, 2C, 3B, 3C, 4C). From the 14 azole resistant isolates with wild type cyp51A, which originated from 5 different cities, two isolates shared the same alleles across all nine microsatellite loci, while the 12 other isolates were genetically very diverse. A minimum spanning tree (MST) based on azole-resistant strains from various countries showed that the 41 Iranian A. fumigatus isolates with TR 34 /L98H formed a separate cluster (Figure 1).

Discussion
In this study, about 70% A. fumigatus isolates from compost samples grew on SDA supplemented with azoles and had the TR 34 /L98H mutation in the cyp51A gene. Indeed, the high rate of resistance to azole drugs due to the TR 34 /L98H mutation in A. fumigatus in Iran outperforms previous studies done during 2013-2016. The prevalence of clinical or environmental azole-resistant A. fumigatus isolates harboring this mutation was much lower in a previous episode and has been estimated between 3.2-6.6% [16,18,31]. Concurrent genetic studies of worldwide A. fumigatus isolates harboring the TR 34 /L98H resistance mechanism also suggested clonal expansion from a common resistant ancestor [32,33]. In the current study the azole resistant A. fumigatus population with TR 34 /L98H was grouped into four microsatellite genotypes, in which the genotype with STRAf profile: 2A:22, 2B:10, 2C:9, 3A:9, 3B:9, 3C:23, 4A:8, 4B:10, 4C:8 included 41 (93%) identical isolates, showing clonal expansion across different geographic locations. Furthermore, MST showed Iranian A. fumigatus isolates harboring TR 34 /L98H were apart from isolates of other countries and previously recovered Iranian isolates. Similar to our finding, Chowdhary et al. described a clonal spread and emergence of environmental azole resistant A. fumigatus strains carrying the TR 34 /L98H mutation from different parts of India. All Indian azole resistant isolates shared the same multilocus microsatellite genotype not found in any other analyzed samples within India or from other Asian or European countries [14]. In agreement with our findings, there is strong evidence that azole-susceptible or cyp51A single point mutation resistance strains have a greater genetic diversity than isolates harboring TR 34 /L98H and TR 46 /Y121F/T289A mutations, since the expansion of latter strains at a local level is predominantly clonal [14,[34][35][36]. The dispersal of A. fumigatus with the TR 34 /L98H genotype supports the hypothesis that these strains have robust fitness in natural environments, with comparable or even higher fitness than that of wild-type strains [11]. Clonal spread of a single genotype in our study reinforced the hypothesis that geographic distances are not a barrier for the global spread from its centers of origin and their ability to cover thousands of miles by producing a large number of airborne spores or by anthropogenic means [14,31,[37][38][39]. The widespread application of azole fungicides in Iran could have contributed to the spread of azole resistant A. fumigatus in environment niches, such as compost. To mitigate spread of azole resistant A. fumigatus in environment, changing of practices to prevent fungal diseases in plants on the fields is necessary. Procedures, such as prudent and restricted use of fungicides, controlling doses, and periods of fungicide application could be helpful. In cases where resistance to fungicides is observed, either the dosage can be increased or alternative fungicides can be used. In addition, environmental surveillance studies aimed to collect precise information of azole resistance monitoring to investigate the size and impact of this emerging problem is necessary [40].
Interestingly, we found that a sizable number of isolates (8 out of 54 resistant isolates) with azole MICs ≥16 mg/L exhibited no mutations in cyp51A. Other mechanisms of resistance, such as increased production of drug target Cyp51A protein, multidrug efflux pumps, or other proposed but not yet fully characterized mechanisms of resistance, such as amino acid substitutions in 3-hydroxy-3-methylglutaryl-CoA, stress response, and biofilm formation, can contribute to azole resistance in these isolates [32]. The limitation of our study was the absence of STRAf profiles of TR 34 /L98H A. fumigatus from neighbor countries of Iran, such as Pakistan or Turkey, for comparison with Iranian isolates [41,42]. In addition, the absence of clinical A. fumigatus was another drawback of our study. As most clinical microbiology laboratories in Iran do not routinely perform antifungal susceptibility testing of Aspergillus, the prevalence of azole resistance and mechanism of resistance in clinical A. fumigatus isolates in Iran is unknown [17].

Conclusions
Clonal spread of triazole resistant A. fumigatus isolated from compost, which is used widely in gardens and indoor plants in Iran, is concerning. This study highlights the importance of antifungal resistance surveillance studies of clinical and environmental Aspergillus isolates in Iran.