Distribution of Aspergillus species and prevalence of azole resistance in clinical and environmental samples from a Spanish hospital during a three‐year study period

Surveillance studies are crucial for updating trends in Aspergillus species and antifungal susceptibility information.


| INTRODUC TI ON
Regarding opportunistic fungal pathogens, Aspergillus species stand out as major agents, causing a broad spectrum of clinical manifestation named aspergillosis. 1Aspergillus fumigatus is the most frequently isolated species from this genus and is the cause of, among other pathologies, invasive aspergillosis (IA), a critical clinical manifestation of aspergillosis associated with high mortality rates in immunocompromised hosts. 2,3Currently, triazole drugs are the antifungal of choice for prophylaxis and first-line treatment of Aspergillus infections. 4,5wever, the therapeutical options against A. fumigatus infections are being reduced as the reports of azole-resistant A. fumigatus strains have increased globally during the last decades. 5,6The development of azole resistance in A. fumigatus is caused by selective pressure associated with the employment of azole drugs and has been classically described by two different routes: a medical route that can occur inside the host, in patients that have been treated with long-term azole therapy; and another route related to the environment, where the acquisition of azole resistance happens in the agricultural scenario, due to the use of demethylation inhibitor fungicides (DMIs) to protect crops against fungal plant pathogens. 7,8Although there are many DMIs used, they share a similar chemical structure to clinical triazoles, thus generating cross-resistance between both antifungal classes. 9,10Regardless of the development route of azole resistance or the underlying azole mechanism, azole resistance is deeply associated with treatment failure. 5,11e 14α sterol demethylase (Cyp51) is the target of triazole drugs.The mode of action of these antifungals is based on the inhibition of the Cyp51 activity, an enzyme that plays a crucial role in the ergosterol biosynthesis pathway, encoded by the gene cyp51A and its homologue cyp51B. 12[15] Nowadays, two similar reference methods for antifungal susceptibility testing (AFST) are used globally, EUCAST (European Committee on Antibiotic Susceptibility Testing, https:// www.eucast.org/ ) and CLSI (Clinical and Laboratory Standards Institute, https:// clsi.org/ ).
Although these methods are validated and standardized, they present some limitations, such as being available only in specialized centres and being slow methodologies, taking at least 5-7 days to be completed.A rapid detection of azole resistance is crucial for increasing the possibilities of therapeutical success and the recovery of the patients.New screening methods, such as four-well azole agar plates, are very useful for detecting azole resistance in A. fumigatus, [16][17][18] being easier, simpler and faster to perform, and they are not restricted to specialized centres, improving the chances of successful clinical outcomes.Although results require confirmation through microdilution susceptibility testing and cyp51A sequencing, this method has been recommended by the EUCAST for screening procedures. 19An expanded version of this method was described and validated by our group, 20 with two types of four-well agar plates, one supplemented with clinical azoles with antifungal concentrations adapted to the last EUCAST breakpoints against A. fumigatus, 21 and four-well agar plates containing DMI antifungals.This method could be easily applied to surveillance studies due to the advantages mentioned before.
Moreover, in the last decade, the ECDC (European Center for Disease Prevention and Control) has recommended epidemiological surveillance studies to update locally A. fumigatus azole susceptibility information. 22[31] In addition to azole resistance surveillance studies, A. fumigatus genotyping is a useful methodology to determine the population structure of this species and to study the epidemiological association between environmental and clinical strains.3][34] While STRAf assay had a higher discriminatory power (D = 0.9993) compared to the TRESPERG typing method (D = 0.9972), the latter can be readily integrated in any clinical microbiology laboratory since it does not demand specialized equipment or trained staff.Both of them are used to determine and analyze genetic distances and have proven to be powerful instruments for A. fumigatus molecular typing. 34 this study, we aim to determine the distribution of Aspergillus species and prevalence of azole resistance using a 3-year prospective collection of clinical and environmental strains from the Severo Ochoa Universitary Hospital in Madrid.Also, we analyze the genetic relatedness of A. fumigatus strains isolated from clinical samples and those that coexist in the hospital environment.

| Aspergillus spp. strains
A total of 335 Aspergillus spp.strains were analyzed: 283 clinical and 52 environmental isolates.All Aspergillus isolates were cultured using standardized mycological procedures and identified at the section or species level based on local routine procedures (i.e.phenotypic identification and/or sequencing).To extract Aspergillus DNA, conidia from every strain were cultured in liquid glucose-yeast extract-peptone (GYEP) medium (0.3% yeast extract, 1% peptone; Difco, Soria Melguizo, Madrid, Spain) containing 2% glucose (Sigma-Aldrich Química, Madrid, Spain) at 37°C for 24 h.After disrupting the mycelium mechanically through vortex-mixing with glass beads, the genomic DNA of the isolates was extracted using the phenol-chloroform method. 35All Aspergillus spp.strains identified from the Severo Ochoa Universitary Hospital were sent to the National Centre for Microbiology to screen for azole resistance and to genotype A. fumigatus strains.

| Environmental surveillance
Environmental air samples were obtained using an air sampler AESAP1075 (Sampl'air Lite, AES Laboratories).Two samples of 1 m 3 of air were captured per day of testing, one at the entrance of a hospital room and another at the centre of the room selected.

| Agar-based screening plates
A method consisting of two sets of four-well agar plates was used to screen for azole resistance in all strains of the study. 20The strains were considered resistant to every specific antifungal if the growth observed in the drug-containing wells was like that of the growth control.All isolates that showed growth in the agar plates were considered possible azole-resistant strains and were evaluated for AFST using the EUCAST.

| Microdilution antifungal drugs susceptibility testing
Antifungal susceptibility testing was performed following the EUCAST broth microdilution reference method 9.4. 36Antifungals used were the azoles ITC, VRC, POS and isavuconazole (ISV) (all from Sigma-Aldrich Química).MICs were performed at least twice for each isolate.Clinical breakpoints for interpreting AFST results established by EUCAST 37 were used for classifying the A. fumigatus strains as susceptible or resistant.Aspergillus isolates with an MIC above the usual epidemiological cut off values for at least one of the mould-active triazoles (VRC, POS and ITC) were submitted to sequencing of the entire cyp51A gene and promoter region for detection of mutations.

| PCR conditions for cyp51A amplification and sequencing
The full coding sequence of cyp51A including its promoter was amplified and sequenced, using primers and PCR conditions previously described. 12To exclude the possibility that any change identified in the sequences was due to PCR-induced errors, each isolate was independently analyzed twice.The amplified products were purified using Illustra ExoProStar 1-step (GE Healthcare Life Science, Buckinghamshire, UK) and both strands were sequenced with the Big Dye terminator cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) following manufacturer's instructions.All gene sequences were edited and assembled using Lasergene software package (DNAStar Inc., Madison, WI, USA).DNA sequences were compared against the cyp51A sequence of A. fumigatus reference strain CBS 144.89 (NCBI accession number AFUB_063960).The GenBank accession numbers for the cyp51A DNA sequences from both resistant strains are H100 PP392543 and H208 PP392544.

| Strains genotyping and genotypic diversity analysis
All A. fumigatus strains included in this study were genotyped following the previously described typing method TRESPERG. 34The combination of the genotypes obtained with each marker has a discriminatory value (D) of 0.9972 using the Simpson index. 38The genotypic diversity analysis was performed as described previously 33 and was represented graphically using a minimum spanning tree (MST) generated with the combination of TRESPERG typing data analyzed by BioNumerics (version 6.0.1)software (Applied Maths, Belgium).The date of isolation, source and genotype of all the strains of the study are displayed in Table S1.The GenBank accession numbers for all four TRESPERG loci have been added in Table S2.The year of isolation, geographical origin and genotypes of the A. fumigatus azole-resistant strains, all harbouring TR 34 /L98H mutation, isolated in Spain from 2012 to 2023, were included for comparison (Table S3).

| Agar-based screening plates
All strains tested grew in the control well without azole drug.Only two A. fumigatus strains grew in all wells supplemented with clinical azole drugs (ITZ, VCZ and POS) and in the agar wells supplemented with the DMIs, MET and EPZ, a pattern that we have previously reported as a possible underlying azole resistance mechanism Cyp51A dependant (TR 34 /L98H or TR 53 ).Although this screening method was explicitly designed to screen for azole resistance in A. fumigatus, we considered that it could be useful for the surveillance to use it with the rest of nonfumigatus Aspergillus species isolated in this study.

| Antifungal drugs susceptibility testing
Strains that were positive in the four-well screaning assay were subjected to azole drugs AFST (EUCAST).Two strains of A. fumigatus showed azole-resistant MICs, consisting of >8 mg/L to itraconazole, 4 mg/L to voriconazole, 0.5 mg/L to posaconazole and 8 mg/L to ISV.
The rest of the strains tested were all azole-susceptible.Thus, the prevalence of azole-resistant strains in this study was 0.6% (2 of 335 strains) and the prevalence of azole-resistant A. fumigatus was 1.15% (2 of 174 A. fumigatus strains).

Analysis of cyp51A
The two A. fumigatus azole-resistant strains (H-100 and H-208) were subjected to amplification and sequencing of the complete cyp51A gene.Sequence analysis revealed the same azole resistance mechanism in both strains, consisting of a 34-bp TR insertion in the promoter region of cyp51A together with a L98H substitution in the coding sequence of the gene (TR 34 /L98H).

| Genotypic variability in environmental and clinical samples
The 174 A. fumigatus strains were genotyped, although seven of them could not be amplified in one of the TRESPERG markers and were therefore excluded from the genotipic analysis.
The TRESPERG genotypes of these 167 strains can be found in Table S1.Out of the 167 strains, 31 were excluded from genotypic analysis because they share the same genotype as other strains from the same patient or were from the same day environmental search.Finally, 136 strains were included in the genotypic analysis.A total of 99 different genotypes were identified according to the TRESPERG typing assay.The TRESPERG results showed a very diverse population with 72.73% of the total genotypes being represented as a single genotype.The A. fumigatus clinical strains showed less diversity than the ones from environmental origin (Table 2).
The genotypic diversity of the A. fumigatus strains from clinical and environmental origin was graphically represented using a MST (Figure 1).
The strains were distributed in different clusters regardless of their origin, including strains from clinical and environmental origins in each cluster defined.Among the remaining 27 genotypes that were not unique, 15 genotypes were common among clinical A. fumigatus strains and two genotypes were shared among environmental A. fumigatus strains, some of them being isolated several times, even months apart.Ten of these 27 common genotypes TA B L E 1 Aspergillus species, number and percentages (%) of strains isolated by the 3 years study period.were shared between clinical and environmental strains (Table 3).
Also, some of these genotypes were common between clinical strains from diferent patients and environmental surveillance collected on different days.
None of the genotypes were coincident with any of the genotypes found in the azole-susceptible A. fumigatus strains of this study.The genotypic diversity of azole-susceptible and azole-resistant A. fumigatus strains from this study was evaluated using a collection of azole-resistant A. fumigatus strains, harbouring the TR 34 /L98H azole resistance mechanism, from different locations in Spain that were isolated between 2012 and 2023 (Table S3) and represented with a MST (Figure 2).The genotypes of the two azole-resistant strains from this study shared genotypes with azole-resistant A. fumigatus strains previously isolated in some locations in Spain.The azolesusceptible strains were widely distributed across the MST and all the azole-resistant strains were highly related and most of them grouped together in close clusters.

| DISCUSS ION
The spectrum of pathologies caused by Aspergillus species is named aspergillosis with IA as one of the most critical diseases due to its high mortality rates among immunocompromised hosts. 1,39,40pergillus fumigatus is the most frequently isolated species among the Aspergillus genus in different parts of the world.23,24,26,28,31 Similarly, in our 3-year surveillance results, out of a total of 335 Aspergillus spp.isolates, more than half (51.94%) of the strains were identified as A. fumigatus.This result was similarly found in other Spanish surveillance studies, although the prevalence order of the rest of Aspergillus species differs.23,24 In our study, the number of A. fumigatus isolates was followed by A. niger, A. terreus and A. flavus.
Nowadays, the rise of A. fumigatus azole-resistant strains has become globally alarming, 7 representing a severe threat to a successful clinical outcome, because azole resistance is closely associated to treatment failure and a higher mortality rate. 5,11,41,42 response to this urgent issue, the ECDC has declared that epidemiological surveillance studies are a useful tool to provide local information regarding A. fumigatus azole susceptibility levels. 22In Spain, multiple studies have evaluated the distribution of Aspergillus species and the prevalence of azole resistance from clinical samples, but most of these studies were limited due to a lack of environmental samples, 23,24,27 although a few studies have included these type of isolates. 29,31The inclusion of environmental isolates has been reinforced by the finding of the hospital setting as a hypothetical source of dissemination of azole-resistant A. fumigatus. 43oth microdilution reference methodologies can constitute a considerable laborious and time-consuming way to perform surveillance studies.The employment of four-well screening methods is affordable and simple to perform in any mycology laboratory and can easily detect azole-resistant A. fumigatus strains. 19Moreover, they can presumably identify the resistance mechanism involved, 20 despite the fact that those strains considered as resistant have to be confirmed by AFST and cyp51A sequencing.The four-well agar expanded method 20 has been used in this study and it permits for screening the entire collection of strains included, and it detected two A. fumigatus azole-resistant strains from clinical and environmental origin.Both strains were pan-azole-resistant and harboured a TR 34 /L98H azole resistance mechanism, the most frequent azole resistance mechanism described in A. fumigatus. 7,13,14The preva- genotyping showed that the genotypes of the strains included in this study are very diverse, with 72.73% being a single genotype.
Although the diversity of the genotypes was high in both settings, we found that it was higher among strains with an environmental origin, in consonance with the findings in other studies. 49,51,52wever, this result should be taken cautiously because there is a considerable difference in the numbers of clinical/environmental strains included in the study.
According to the genotyping results, the two azole-resistant A. fumigatus strains had different genotypes and did not share their genotype with any azole-susceptible A. fumigatus strains found in this study.Furthermore, the azole-resistant A. fumigatus strain isolated from a clinical source comes from an azole-naïve patient, so the development of the azole resistance mechanism could not possibly happen inside the host since there was not selective pressure.
Recently, azole-resistant A. fumigatus isogenic strains have been found in a patient and in their bathroom, which suggests two hypotheses: that the environmental setting could be contaminated with azole-resistant A. fumigatus that could colonise the patient; or that the patient was the source of the environmental contamination. 434][55] A very interesting finding in this study is that several genotypes were shared between clinical and environmental strains, in alignment with the different hypotheses proposed before.
Moreover, if we had tracked the different locations of the environmental captures and the locations of the patients, we could have determined if patients hospitalized in different parts of the same hospital were infected with the same spore population as other studies have found. 49e two azole-resistant strains were genotypically different and with no genetic relation with the rest of the azole-susceptible strains included in the study.7][58] Although the reason why these strains that harbour TR 34 / L98H azole resistance mechanism are so genetically related remains unclear, a better adaptation to persist in the environment or a relation with A. fumigatus genetic instability have both been suggested. 59,60

| CON CLUS IONS
Resistance of the human pathogenic fungus A. fumigatus to azole drugs is rising.However, the link between patient infections and their potential acquisition from hospital environmental sources remains vague.In this work, we used two recent methodological techniques that for their simplicity allow for easy integration into any clinical microbiology laboratory, fulfilling all the needs of surveillance for azole resistance, combined with a suitable typing assay.In this study, we found that A. fumigatus genotypes were highly diverse in both settings, emphasizing the highly mixed nature of A. fumigatus populations.However, identical clonal genotypes were found to occur both in the clinical strains and in the Sabouraud/Gentamicin (28 μg/mL)/Chloramphenicol (240 μg/mL) agar plates, irradiated with the air sample, were sealed and incubated at 35°C for 5 days.The environmental surveillance was performed by the preventive medicine department at the hospital.The air sampling procedure was sometimes altered by repetitions due to internal cleaning protocols, COVID-19 impact and construction work inside the hospital.Due to this, the number of air samples was irregular during the environmental surveillance: 230 air samples in 2019, 222 air samples in 2020 and 317 air samples in 2021.

47 TA B L E 2 F I G U R E 1
Aspergillus fumigatus genotypes found in air and in clinical samples collected during the 3 years of the study.Minimum spanning tree showing the genotypic diversity Aspergillus fumigatus strains from clinical (in orange) and environmental (in blue) origin.Each circle shows a unique genotype, and its size represents the number of strains belonging to the same genotype.Connecting lines between circles show the similarity between genotypes: solid and bold (shaded in black) indicate only one marker difference, a solid line indicates differences in two markers, and dashed lines for differences in three or more markers.

F I G U R E 2
Minimum spanning tree showing the genotypic diversity of azole-susceptible (yellow) and azole-resistant (purple) A. fumigatus strains.Each circle shows a unique genotype, and its size shows the number of strains belonging to the same genotype.Connecting lines between circles show the similarity between genotypes: solid and bold (shaded in black) indicate only one marker difference, a solid line indicates differences in two markers, and dashed lines for differences in three or more markers.The two azole-resistant strains obtained in this work are indicated: H-100 and H-208.

Aspergillus species No. of strains isolated (%)
Abbreviations: C, Clinical origin, E, Environmental origin; Genotypes shared between clinical and environmental Aspergillus fumigatus strains.
Note: Date of isolation: yyyy/mm/dd.The shades are to make more relevant Aspergillus types that are identical between clinical and enviromental strains.