In total, 45 A. fumigatus strains from different clinical specimens were collected from May 2019 to April 2021. The patient information related to these strains in the study was showed in Supplementary Table 1. The AF293 strain was used as a control. All strains were stored in sterile ddH₂O at room temperature (Hartung de Capriles et al., 1989). Prior to analysis, the strains were cultured on PDA at 37°C for 3–7 days. Mouse alveolar macrophages (MH-S cells) were donated by Professor Jia Jing of the Affiliated Hospital of Southwest Medical University (Luzhou, China).

The AF1, AF2, AF4, AF5, AF8, and AF293 strains were cultured on PDA at 37°C for 5 days. Colonies were selected and stained with lactophenol cotton blue stain solution. Morphological characteristics of conidia, conidiophores, and mycelia were observed under a microscope.

Each strain was cultured in Roswell Park Memorial Institute (RPMI) 1640 liquid medium (Cytiva, Marlborough, MA, USA) at 37°C for 48 h. Afterward, the mycelia were collected. Genomic DNA was extracted using a commercial DNA extraction kit (Sangon Biotech Co. Ltd., Shanghai, China) in accordance with the manufacturer’s instructions. The internal transcribed spacer (ITS) sequences of the test strains were amplified by RT-qPCR with the primers listed in Supplementary Table 2. The Sanger sequencing results were compared with the Basic Local Alignment Search Tool.¹ An evolutionary tree of the ITS sequences was generated with Molecular Evolutionary Genetics Analysis software.²

Conidia of A. fumigatus, cultured for 5 days, were eluted with 0.05% Tween 80 and diluted to 1 × 10⁶ cells/mL. Then, 3 μL of diluted conidia were spotted in the center of PDA plates (diameter, 9 cm) and incubated at 37°C for 36 and 120 h, respectively. Then, all conidia were eluted with 0.05% Tween 80 in sterile water and counted under a light microscope (magnification, 400 ×).

The biofilm formation assay was performed as described in a previous report (Zhang et al., 2022). Briefly, 2 mL of conidia (1.0 × 10⁵ cells/mL) were added to the wells of a 12-well cell culture plate, covered with round coverslips, and cultured at 37°C for 24 h. Then, the unattached spores and mycelia were washed three times with PBS. Afterward, 1 mL of calcofluor white stain (CFW) (Beijing Solarbio Science & Technology Co., Ltd.) and 20 μL of 10% KOH were added to each well and the plate was incubated at room temperature for 3–5 min. Biofilm formation was observed with a confocal laser scanning microscope (Olympus Corporation, Tokyo, Japan).

To further confirm the biofilm formation ability of the test strains, 2 mL of spore suspensions (1.0 × 10⁵ cells/mL) were added into the wells of 12-well cell culture plates and cultured at 37°C for 24 h. Afterward, the unattached spores and mycelia were washed three times with PBS. Then, 200 μL of a mixture of 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-[(phenylamino)-carbonyl]-2H-tetrazolium hydroxide sodium salt and menadione were added to each well and the plate was incubated at 37°C for 2 h in the dark. Finally, the supernatant was transferred to the wells of a new 96-well plate and absorbance was measured at 490 nm.

The phagocytosis assay was performed as previously described (Lim et al., 2022). Briefly, MH-S cells were cultured in complete Dulbecco’s modified Eagle’s medium (Thermo Fisher Scientific, Waltham, MA, USA) containing 10% fetal bovine serum and 1% penicillin-streptomycin. Then, 1 mL of the cultured MH-S cells (1 × 10⁵ cells/mL) was added to each well of a 12-well plate, covered with round coverslips, and cultured overnight at 37°C under an atmosphere of 5% CO₂/95% air. Afterward, 1 mL of conidia (1 × 10⁶ cells/mL) was added to each well and the plate was cultured for 2 h under an atmosphere of 5% CO₂/95% air. The 12-well plate was washed three times with PBS pre-cooled at 4°C. Then, the round coverslips were stained with 50 μg/mL of CFW and observed under a microscope. The phagocytosis percentage (Eq. 1) and phagocytosis index (Eq. 2) were calculated in reference to the microscope images and repeated independently three times. In total, 200 MH-S cells were randomly counted.

Phagocytosis rate = number of MH-S cells of phagocytizing conidia/200 × 100% (Eq. 1).

Phagocytosis index = total number of phagocytic conidia/number of MH-S cells of phagocytizing conidia (Eq. 2).

G. mellonella larvae (n = 10/group) were injected with 10 μL of the test strain in suspension through the last pro-leg into the hemocoel using a 50-μL microinjector (Shanghai Gaoge Industry and Trade Co., Ltd., Shanghai, China) (Özkan and Coutts, 2015). Untreated, pierced only, and PBS-injected groups were used as controls. The experimental groups were injected with 10 μL of conidia (1 × 10⁷ cells/mL). After injection, each group of larvae was divided into two petri dishes and cultured at 37°C in the dark. The larvae were photographed and morphological changes were recorded every 24 h. Each experiment was repeated three times.

Different groups of infected larvae were sacrificed at 24 or 72 h after infection. The whole larvae were macerated in microtubes with 1 mL of ddH₂O. Afterward, 50 μL of the dilution were seeded on PDA medium and the plates were incubated at 37°C. Each experiment was repeated three times.

Drug sensitivity of the five strains was determined in accordance with Clinical and Laboratory Standards Institute (CLSI) document M38-A2—Reference Method for Broth Dilution Antifungal Susceptibility Testing of Filamentous Fungi (Wayne, 2008). The antifungal agents were used at the following concentrations: fluconazole (FLC), 64–0.0625 μg/mL; anidulafungin (ANI), amphotericin B (AmB), caspofungin (CAS), ISA, micafungin (MF), POS, and VRC, 32–0.03125 μg/mL. The conidia (1.0 × 10⁵ cells/mL) were dispensed into triplicate wells of 96-well microtiter plates and incubated at 37°C for 48 h. Conidia in drug-free wells were used as growth controls. The minimum inhibitory concentration (MIC) was defined as the lowest drug concentrations that caused complete visible inhibition of growth after incubation at 37°C for 48 h. The minimum effective concentration (MEC) was defined as the lowest drug concentration causing the mycelia to become shorter and thicker, as observed under a microscope, after incubation at 37°C for 48 h (Subcommittee on Antifungal Susceptibility Testing of the ESCMID European Committee for Antimicrobial Susceptibility Testing, 2008).

For the strip tests, the conidia (1.0 × 10⁶ cells/mL) were evenly coated on RPMI 1640 solid medium using cotton swabs. Once the plate was dry, an E-test strip (Liofilchem s.r.l., Roseto degli Abruzzi, Italy) was attached to the center of the medium with sterile tweezers. The results were read after incubation at 37°C for 24 or 48 h. The MIC was determined in reference to the scale at the intersection of the inhibition zone and the E-test strip. For both methods, strain AF293 was used for quality control. Each experiment was repeated three times.

Sanger sequencing of cyp51A, cyp51B, and hmg1, in addition to the promoter region of cyp51A, was conducted. The sequences were compared using DNAMAN software (Lynnon Biosoft, San Ramon, CA, USA) to identify short tandem repeats and amino acid mutations. The 3D structures of the Cyp51A proteins of the test strains were generated and analyzed with Discovery Studio software (BIOVIA, San Diego, CA, USA).

The ergosterol content was determined by HPLC (1260 infinity II; Agilent Technologies, Inc., Santa Clara, CA, USA) as described in a previous report (Zhang et al., 2022). Briefly, conidia (1 × 10⁶ cells/mL) were inoculated into 50 mL of RPMI 1640 medium and cultured at 37°C for 48 h at 200 rpm. The samples were exposed to ultrasonic waves at 40 kHz. Following the addition of methanol (w/w), the supernatant was collected by centrifugation as the sample to be tested. An ergosterol standard (Shanghai Macklin Biochemical Co., Ltd., Shanghai, China) was used as a reference and 100% methanol as a negative control. The concentration of ergosterol = (peak area of experimental group × concentration of ergosterol standard group)/peak area of ergosterol standard product. Each experiment was repeated three times.

Conidia (1.0 × 10⁶ cells/mL) of the test strain were inoculated into RPMI 1640 medium and cultured at 37°C for 16 h at 200 rpm, and then co-cultured with or without drugs for 0, 4, 8, or 12 h. The mycelia were collected and frozen in liquid nitrogen until assayed. Total RNA extraction, reverse transcription, and RT-qPCR reactions were performed using commercial kits (Takara Bio, Inc., Shiga, Japan) in accordance with the manufacturer’s instructions. Transcript levels were determined by a ABI 7500Fast Real-Time PCR Detection System (Thermo Fisher Scientific, Waltham, MA, USA) and measured with the 2^–ΔΔCT method against β-tubulin as an internal control (Vandesompele et al., 2002).

To further clarify the azole resistance mechanism, conidia of strain AF5 prepared in RPMI 1640 medium were treated with or without VRC (MIC = 0.5) and incubated at 37°C for 8 h. Total RNA was collected and sequenced by Biomarker Technologies (Qingdao, China) with an Illumina RNA sequencing system (Illumina, Inc., San Diego, CA, USA). Then, quality control detection of the raw sequence data was performed on the Biomarker platform.³ Sequencing alignment was performed using the Hisat2 (version 2.2.1, and the reference genome is Aspergillus_fumigatus.GCF_000002655.1_ASM265v1.genome.fa). After that the samtools (version 1.9) was used to sort the sam files to bam files, and then matched Reads were assembled and quantified using the StringTie tool (version 1.3.4d). The tool for differentially expressed genes (DEGs) analysis is DESeq2 (version 1.6.3). Further gene function annotation was included COG (Cluster of Orthologous Groups of proteins), KO (Kyoto Encyclopedia of Genes and Genomes (KEGG) Ortholog database), and GO (Gene Ontology). The time-specific expression patterns of genes were determined by RT-qPCR analysis. Raw sequence data were deposited in the Genome Sequence Archive of the Beijing Institute of Genomics (Chinese Academy of Sciences, Beijing, China; accession number CRA011198).

The contents of NAD⁺ and NADH were measured using a commercial kit (Beyotime Institute of Biotechnology, Shanghai, China) as described in a previous report (Hsieh et al., 2020). Briefly, the conidia (1.0 × 10⁶ cells/mL) of strain AF5 were inoculated into RPMI 1640 medium and cultured at 37°C for 16 h, and then co-cultured with VRC (MIC = 0.5) for an additional 0, 4, 8, or 12 h. Afterward, the mycelia were collected and frozen in liquid nitrogen until assayed. Following the addition of working solution, the supernatant of each sample was collected and divided into two aliquots. One aliquot was directly transferred to the wells of a 96-well plate and the other was incubated at 60°C for 30 min to decompose NAD⁺ into NADH, and then transferred to the wells of a 96-well plate to determine the intracellular content of NADH.

Data analyses were conducted using GraphPad Prism software v9.0 (GraphPad Software, Inc., San Diego, CA, USA). Group comparisons were performed with the Student’s t-test or one-way analysis of variance. Mortality curves were generated by the Kaplan–Meier method and compared with the log-rank test. Each experiment was repeated three times. A probability (p) value < 0.05 was considered statistically significant.

The conidiophores of Aspergillus strains AF1, AF2, AF4, AF5, and AF8 were stained with lactophenol cotton blue stain solution and observed under a microscope (Figure 1A). Phylogenetic tree analysis of the ITS sequences revealed that all five test strains (AF1, AF2, AF4, AF5, and AF8) were clustered with strain AF293 (Figure 1B), confirming that all five test strains were A. fumigatus.

All five test strains were cultured on PDA medium for 5 days. As compared to strain AF293, there were no significant differences in the colony color and morphology of the five test strains (Figure 1C). As compared to strain AF293, the conidia yields of strains AF1 and AF2 were significantly increased by incubation for 36 and 120 h, while there was no significant difference among strains AF4, AF5, and AF8 (Figures 1D, E).

To explore the differences biofilm formation ability among test strains, as compared to strain AF293, strains AF1, AF2, AF4, AF5, and AF8 formed more and denser biofilms (Figure 2A), as confirmed by significantly increased optical density at 490 nm (Figure 2B).

To explore the interaction between the A. fumigatus conidia and MH-S cells, the conidia of the test strains were phagocytosed by MH-S cells after co-culture for 2 h, as observed with a fluorescence microscope (Figure 3A). As compared to strain AF293, phagocytosis of the conidia of strains AF2, AF4, AF5, and AF8 was significantly increased, while that of strain AF1 conidia was significantly decreased (Figure 3B). Further phagocytic index analysis, representing the average number of conidia phagocytosed per MH-S cell, showed that the phagocytosis index of MH-S cells to strain AF1 was significantly decreased, while that of strain AF4 was significantly increased (Figure 3C). These results indicate enhanced ability of strain AF1 to avoid phagocytosis, while the ability of the other strains to avoid phagocytosis was decreased.

The G. mellonella larvae infection models were subsequently opted to explore variations in virulence among test strains. G. mellonella larvae infected with A. fumigatus conidia begin to melanize and die after 24 h. However, there was no significant correlation between melanization and strains (Supplementary Figure 1). As compared to strain AF293, the larvae survival rate was significantly decreased in the AF1, AF2, AF4, and AF8 model groups at 120 h after inoculation with 10⁷ cells/mL (Figure 4A). The fungal burden was significantly decreased in the AF1 and AF4 groups at 24 h (Figure 4B). At 72 h, the fungal burden was significantly decreased in the AF1 and AF5 groups, while significantly increased in the AF2 and AF8 groups (Figure 4C).

The MIC₉₀ and MEC of azoles, AmB, and polyenes against the test strains are shown in Table 1. All test strains were resistant to FLC, as determined by the broth microdilution method, while strains AF1 and AF2 were resistant to ITR and ISA, and sensitive to AmB, ANI, CAS, MF, POS, and VRC. Meanwhile, strains AF4, AF5, and AF8 were resistant to VRC and ISA, and sensitive to AmB, ANI, CAS, MF, ITR, and POS. In addition, the E-test results were consistent with the results of the broth microdilution method (Supplementary Figure 2).

Sanger sequencing and comparison analysis showed that strains AF1 and AF2 carried the cyp51A mutations TR34/L98H/S297T/F495I, strains AF4 and AF8 carried the cyp51A mutations TR46/Y121F/T289A, and strain AF5 carried no mutation (data not shown). Further analysis with Discovery Studio software showed that there was no difference in the 3D structure of Cyp51A between strains AF293 and AF5. Moreover, although there were small spatial differences in the branch chain of Cyp51A among the other four strains, the spatial structure of the main chain conformation was consistent with that of strain AF293 (Supplementary Figure 3). HPLC analysis found that there was no significant difference in ergosterol content (data not shown). In addition, Sanger sequencing and comparison analysis showed that none of the five test strains carried a cyp51B mutation, while strains AF1 and AF2 carried the hmg1 mutation S541G (data not shown).

Differences in the expression levels of genes associated with ergosterol synthesis, including cyp51A, cyp51B, erg1, erg4, erg24, and hmg1, were detected by RT-qPCR analysis. As compared to strain AF293, cyp51A expression was significantly up-regulated, while the expression levels of erg1, erg24, and hmg1 genes were significantly down-regulated in strain AF1 (Figure 5A). Treatment of strain AF1 with ITR (32 μg/mL for 8 h) significantly up-regulated the expression levels of cyp51A, cyp51B, erg1, erg24, and hmg1 (Figure 5B). As compared to strain AF293, cyp51A and erg24 were significantly up-regulated, while erg4 and hmg1 were significantly down-regulated in strain AF2 (Figure 5C). Treatment of strain AF2 with ITR (32 μg/mL for 8 h) significantly up-regulated hmg1 and significantly down-regulated cyp51A and cyp51B (Figure 5D). As compared to strain AF293, cyp51A, cyp51B, erg1, and erg24 were significantly up-regulated, while hmg1 was significantly down-regulated in strain AF4 (Figure 5E). Treatment of strain AF4 with VRC (16 μg/mL for 8 h) significantly up-regulated cyp51A and hmg1, and significantly down-regulated erg1 (Figure 5F). As compared to strain AF293, cyp51A, cyp51B, erg4, and erg24 were significantly up-regulated, while hmg1 was significantly down-regulated in strain AF8 (Figure 5G). Treatment of strain AF8 with VRC (16 μg/mL for 8 h) significantly up-regulated cyp51A, cyp51B, erg1, erg24, and hmg1 (Figure 5H). These results indicate that the resistance mechanism of strains AF1, AF2, AF4, and AF8 is related to cyp51 mutation or overexpression.

Changes to the expression levels of genes associated with drug efflux pumps (e.g., atrF, cdr1B, mdr1, mdr2, mdr4, mdrA, and mfsB) were further confirmed. As compared to strain AF293, mdr2, and mdrA were significantly up-regulated, while atrF, mdr1, and mfsB were significantly down-regulated in strain AF1 (Figure 6A). Treatment of strain AF1 with ITR (32 μg/mL for 8 h) significantly up-regulated cdr1B and mdr1 (Figure 6B). As compared to strain AF293, mdr4 was significantly up-regulated, while atrF, mdr1, and mfsB were significantly down-regulated in strain AF2 (Figure 6C). Treatment of strain AF2 with ITR (32 μg/mL for 8 h) significantly up-regulated atrF, mdr1, mdr4, and mdrA, and significantly down-regulated mfsB (Figure 6D). As compared to strain AF293, cdr1B, mdr2, mdr4, and mdrA were significantly up-regulated, while atrF, mdr1, and mfsB were significantly down-regulated in strain AF4 (Figure 6E). Treatment of strain AF4 with VRC (16 μg/mL for 8 h) significantly up-regulated atrF, mdr1, mdr4, and mfsB, and significantly down-regulated cdr1B and mdrA (Figure 6F). As compared to strain AF293, mdr1, mdr2, mdr4, and mdrA were significantly up-regulated, while atrF, cdr1B, and mfsB were significantly down-regulated in strain AF8 (Figure 6G). Treatment of strain AF8 with VRC (16 μg/mL for 8 h) significantly up-regulated cdr1B, mdr1, mdr2, mdr4, and mdrA, with no significant difference in expression of atrF and mfsB (Figure 6H). These results indicate that the azole resistance mechanisms of strains AF1, AF2, AF4, and AF8 are related to overexpression of genes associated with efflux pumps.

The results of previous experiments found that cyp51A, cyp51B, and hmg1 of strain AF5 carried no mutations, and there was no change in the ergosterol content. As compared to strain AF293, the expression levels of genes associated with ergosterol synthesis (i.e., cyp51A, cyp51B, erg1, erg4, and hmg1) and efflux pumps (i.e., cdr1B, mdr1, mdr4, and mdrA) were significantly up-regulated in strain AF5. After treatment with VRC (8 μg/mL for 8 h), the expression levels of genes associated with ergosterol synthesis (i.e., cyp51B, erg24, and hmg1) and efflux pumps (i.e., atrF, mdr1, mdrA, and mfsB) were significantly down-regulated in strain AF5 (Figures 7A, B). Transcriptomics sequencing was conducted to further explore the mechanism of VRC resistance of strain AF5. The results showed that among these DEGs, 73 genes were significantly up-regulated and 147 genes were significantly down-regulated (Figure 7C). DEGs were functionally grouped into GO classes comprising 40 functional categories (Supplementary Figure 4), and the DEGs were involved in biological processes including drug efflux, glucose metabolism, and ribosome metabolism (Supplementary Figure 5). Further analysis found that there are 32 DEGs of up-regulated genes were assigned to 20 KEGG enrichment pathways, the top one pathway is ABC efflux pumps (4.1%), and another interesting pathway is the glycolysis pathway (2.74%), whereas there are 144 DEGs of the down-regulated genes were assigned to 20 KEGG enrichment pathways, and the top one pathway is ribosome metabolic pathway (50.34%) (Figures 7D, E), which are consistent with the GO analysis.

To confirm the results of transcriptome analysis, RT-qPCR analysis of genes associated with efflux pumps and energy metabolism was performed. The results showed that the expression levels of genes associated with effluent pumps (i.e., abcA, abcC, atrA, atrF, mdr1, mdrA, and mfsB) were significantly up-regulated (Figure 7F), in addition to genes associated with glycolysis (i.e., hxkA and PK), the tricarboxylic acid cycle (i.e., IDH, DLST, and MDH), the electron transport chain (i.e., ND5, CBR, and COX5), the pentose phosphate pathway rate-inhibiting enzyme (i.e., PRPS1), the malic acid cycle (i.e., Pdc), and the cholesterol and fatty acid anabolic pathways (i.e., HMG-COA and ACC). Meanwhile, the expression levels of Sui (a key gene of ATP synthase) and ATPase-related genes were also significantly up-regulated (Figure 7G). Further biochemical analysis showed that the NAD⁺/NADH ratio was significantly increased in strain AF5 at 4, 8, and 12 h after administration of VRC (Figure 7H). These results suggest that strain AF5 responds to VRC by increasing energy production and overexpression of genes associated with efflux pumps.