Introduction

Aspergillus niger and its related species are grouped in Aspergillus section Nigri, commonly known as black aspergilli. These species are unevenly distributed globally and are often isolated from clinical samples1. Black aspergilli infections are the third most common cause of Aspergillus-associated infections, leading to conditions such as otomycosis, onychomycosis, and pulmonary aspergillosis2,3.

Triazoles comprise the first-line treatment for aspergillosis however, long-term therapy and widespread use of azole-based pesticides in agriculture have raised concerns because of an increase in resistance to the medical triazoles in various Aspergillus species4. The acquired resistance is not only because of the fungicidal effect of azoles but also caused by azole exposure in clinical and environmental settings5,6,7,8. Multi-azole resistance has been reported in patients and from the environment across Europe, China Japan, The Middle East, and India5,9,10,11,12. Several authors have reported azole-resistant Aspergillus isolates, which correlates with the poor therapeutic outcome of azole, thereby limiting the treatment options6,7,13,14. Previous studies have reported that the antifungal drug susceptibility of itraconazole (ITC) against clinical and environmental isolates of A. niger and A. tubingensis showed higher minimum inhibitory concentration (MIC)15,16,17,18.

The azole drugs act via non-competitive binding to the Cyp51 enzyme, a sterol 14α-demethylase, of the ergosterol biosynthetic pathway. The azole inhibits ergosterol synthesis and disrupts cell membrane19. The mechanism of azole resistance in Aspergillus fumigatus has been extensively studied. Several mutations in the cyp51A gene and overexpression of this gene have been reported in azole-resistant A. fumigatus strains20. However, azole resistance mechanisms have not been extensively investigated in the A. niger complex. The molecular mechanism of azole resistance in section Nigri was first reported by Howard et al.15. Since then, a few reports have elaborated on the resistance mechanisms against A. niger complex via mutation analysis of the cyp51A gene or its expression. Pérez-Cantero et al.21 reported that cyp51B gene expression in the Aspergillus section Nigri was not inducible after azole exposure. However, the underlying azole resistance mechanisms in Aspergillus section Nigri have not been fully explored.

Identification of a relatively large number of clinical azole-resistant A. fumigatus isolates lacking the cyp51A mutations and comparative genomics studies in yeasts prompted the investigations on alternative mechanisms of azole resistance. This led to the discovery of the role of efflux pumps in azole resistance. Efflux pumps are categorized into two main classes: the major-facilitator superfamily (MFS) proteins, encoded by 278 genes, and ATP-binding cassette (ABC) proteins, encoded by 49 genes22. Tobin et al.23 identified two ABC transporter proteins; MDR1 and MDR2 in A. fumigatus based on cloning and sequence homology. Further, overexpression of MDR3 and MDR4 in ITC-resistant A. fumigatus strains was reported24,25. In a recent study, overexpression of efflux pump genes has been reported in azole resistant A. niger isolates26. Overexpression of efflux pump genes has also been observed in response to amphotericin B in A. fumigatus. This report suggested that the fungus adapts by overexpressing genes and proteins involved in drug efflux, representing a mechanism to develop resistance and survive against antifungal drugs27. Moreover, overexpression of efflux pumps genes like abcC in azole-resistant A. fumigatus isolates has been highlighted in other studies28,29.

Here, we aimed to determine the correlation between azole resistance and the cyp51A gene mutation in black Aspergillus isolates isolated from agricultural soil samples across India through antifungal susceptibility testing, gene sequencing, expression analysis of selected genes and protein homology modeling. Further, we attempted to elucidate the role of efflux transporter genes, mdr1, and mfs, in azole-resistant environmental isolates.

Results

Identification and antifungal susceptibility testing

We screened 161 black aspergilli isolates isolated from 174 soil samples across India followed by antifungal susceptibility analysis against ITC, voriconazole (VRC), and posaconazole (POS). Results revealed that out of 161 isolates, 20 were resistance to at least one azole drug. Resistant isolates were obtained from the samples collected from Haryana (7/46), Bihar (3/23), Punjab (4/12), Madhya Pradesh (2/10), Uttar Pradesh (1/10 and Assam (3/6) region (Fig. 1). Furthermore, azole susceptibility revealed that 19/161 black aspergilli isolates was above its epidemiological cutoff value (ECV) of > 2 μg/mL for ITC. Five isolates had MIC > 0.5 μg/mL for POS (Table 1; Fig. 2). Four black aspergilli isolates were cross-resistant to ITC and POS. For VRC, all isolates (except AG1 and AU3) showed MICs below the ECV values (2 μg/mL), and only two isolates were cross-resistant to ITC and VRC (AU3 and AG1). We inferred that VRC and POS were the most effective triazoles against black aspergilli isolates, and ITC was the least effective triazole.

Figure 1
figure 1

Aspergillus section Nigri isolates (n = 161) identified from soil samples collected from different states of India. The blue bars represent the number of susceptible isolates, and the yellow bars represent the number of azole-resistant Aspergillus section Nigri isolates.

Table 1 In-vitro azole susceptibility profile of 161 Aspergillus section Nigri isolates.
Figure 2
figure 2

Distribution of Aspergillus section Nigri isolates based on the MIC values (n = 161). Aspergillus section Nigri was classified as resistant (R), when MIC was above the following ECV values: itraconazole ≤ 2 μg/mL, voriconazole ≤ 2 μg/mL, and posaconazole ≤ 0.5 μg/mL.

Molecular identification of black Aspergillus isolates

Molecular identification was performed for 20 azole-resistant black aspergilli isolates. Additionally, to provide a comparative context, we included 10 azole-susceptible black aspergilli isolates from the same geographical region as the resistant isolates. Using 18S internal transcribed spacer (ITS) and β-tubulin gene sequencing a total of 21 A. niger (13 resistant and 8 susceptible) and 9 A. tubingensis (7 resistant and 2 susceptible) isolates from the 30 Aspergillus section Nigri isolates were identified. All tested ITS and β-tubulin gene sequences displayed 99% to 100% nucleotide identity with the sequences available in NCBI database. The GenBank accession numbers generated for the submitted ITS sequences were OK342205, MZ305300, MZ305301, MZ305302 MZ305303, MZ292199, MZ292200, MZ292198, MZ292197, MZ292196, MW332573, MW282896, OQ938544, OQ938545, OQ938546, OQ935460, OQ935461, OQ935462, OQ935463, OQ935464, OQ935465, OQ935466, OQ935467, OR575635, OR575636, OR575637, OR575638, OR575639, OR575640 and for β-tubulin gene were OQ948174, OQ948175, OQ948176, OQ948177, OQ948178, OQ948179, OQ948180, OQ948181, OQ948182, OQ948183, OQ948184, OQ948185, OQ948186, OQ948187, OQ948188, OQ948189, OQ948189, OQ948190, OQ948191, OQ948192, OQ948193, OQ948194,OQ948195, OR584016, OR584017, OR584018, OR584019, OR584020, OR584021, OR584022, OR584023.

Mutation analysis of the cyp51A gene

The cyp51A gene was sequenced and analyzed for all 20 resistant (13 A. niger and 7 A. tubingensis) and 10 susceptible (8 A. niger and 2 A. tubingensis) Aspergillus section Nigri isolates. The sequences of the cyp51A gene of A. niger and A. tubingensis resistant isolates were aligned with the wild-type strains NT_166526 and JF450924.1, respectively (Supplementary Figs. S1, S2). We identified 15 non-synonymous mutations among the Aspergillus section Nigri isolates. Table 2 shows the amino acid alterations in the cyp51A gene. Among these, several mutations were found in both susceptible and resistant isolates. In A. niger, the amino acid change Q228R was observed in both resistant and susceptible isolates. However, the mutation S346R in combination with Q228R was identified in various isolates, with different susceptibility to azole drugs. This included one azole-susceptible isolate (PU2), two isolates resistant to either ITC or POS (RK5 and MP6), two isolates displaying cross-resistance to ITC and VRC (AU3 and AG1), and two isolates with cross resistance to ITC and POS (AU4 and MP8). Further, the amino acid change V383L, R501Q, I244F, E254D, D253Y, T267Y, Y268H, E278Q and L303M were found only in three A. niger resistant isolates, that exhibited resistance to only ITC (MD2, PR4, PI5).In the case of A. tubingensis, amino acid substitution T321A in combination with V377I was observed in three ITC resistant isolates (FA9, FA4, and FA6). Additionally, one of the identified ITC resistant isolate (FA10) exhibited this substitution along with K477N mutation. Mutation T321A was also observed independently in ITC-resistant isolate (AH6), an isolate exhibiting cross resistant to both ITC and POS (PR2), and a susceptible isolate (AH1). Furthermore, a new amino acid substitution, V329I was detected in one ITC-resistant (PR9) and in one susceptible (UT1) A. tubingensis isolates.

Table 2 Amino acid substitutions observed in the cyp51A region of Aspergillus section Nigri isolates.

Cyp51A homology modeling and molecular docking

We constructed ten Aspergillus section Nigri Cyp51A homology models (A to L) to compare the amino acid profiles of azole-resistant isolates with those of wild-type isolates. Five mutation combinations observed in all A. niger isolates were: Q228R/V383L, Q228R/S346R, Q228R, Q228R/R501Q, and Q228R/I244F/D253Y/E254D/T267Y/Y268H/E278Q. Therefore, models A to E were constructed after incorporating these mutations in A. niger isolates (Fig. 3). Similarly, A. tubingensis models G to K were constructed for the mutations V329I/L492M, T321A, V377I/T321A, V329I, and K477N/T321A (Fig. 3), and two wild-type models F and L was constructed for A. niger and A. tubingensis, respectively. Table 3 showed the docking score of the different Cyp51A protein models of the aspergilli isolates with ITC and VRC. Compared to wild-type isolates, marked variations in the overall protein structure conferring resistance were not identified in models A to E, and G to K. The most negative docking score was obtained with model E for ITC, and docking scores were almost similar in all the models for VRC. H-bond interactions between the ligands (ITC and VRC) and Cyp51A protein structure were obtained in models A, D, H, J, K, and L. In model A, ITC interacted with CYS447, in model H, it interacted with TYR51, and in model K interaction was observed between ITC and TYR119, whereas, two H-bonds were formed between model J and ITC with TYR119 and SER360. VRC interacted with TYR119 via H-bond interactions in models A, H, I, J, K, and L, while, H-bond interactions were observed between CYS447 and VRC in models D and G.

Figure 3
figure 3

Ribbon representations of superimpositions of modeled (AE) and (GK) mutated Cyp51A protein structure with the wild-type protein of A. niger and A. tubingensis respectively. The green color structure represents the wild-type Cyp51A proteins. Models (A,D,E,G,I,K) correspond to structures associated with resistant isolates. Model (J) represents structure associated with susceptible isolate. In contrast, Models (B,C,H) represents structures associated with both susceptible and resistant isolates.

Table 3 Docking scores of the different Cyp51A protein models of the aspergilli isolates with itraconazole and voriconazole.

The obtained homology-modeled protein of each isolate was aligned individually with wild-type Cyp51A modeled protein. The differences in protein backbone structures are quantitated with the root mean square deviation (RMSD) of the similarity between two superimposed atomic coordinates of modelled proteins. The RMSD scores demonstrated high similarity among models A to E when compared to the susceptible A. niger model (model F), with all models exhibiting a difference of less than 0.1 Å. In the case of A. tubingensis, the RMSD scores also demonstrated substantial similarity among models G to K compared to the susceptible A. tubingensis model (model L), all models displaying a difference of less than 0.24 Å (Fig. 3). Table 4 depicted the RMSD values obtained (because of mutations) of the aligned protein models of the mutated sequences with wild-type Cyp51A of A. niger and A. tubingensis, respectively.

Table 4 RMSD values of the aligned Cyp51A models of the mutated isolates with wild-type Cyp51A of A. niger and A. tubingensis.

Gene expression analysis

Reverse transcription analysis from the RNA samples followed by qRT-PCR was performed to understand the expression modulation of transcripts encoding for cyp51A, cyp51B, mdr1, and mfs genes in the azole-resistant isolates of Aspergillus section Nigri. Figure 4 indicates the two-fold relative expression of crucial genes in Aspergillus section Nigri isolates compared to the susceptible isolates PI2 of A. niger and UT2 of A. tubingensis.

Figure 4
figure 4

Relative expression of cyp51A ((a) and (b)), cyp51B ((c) and (d)), mdr1 ((e) and (f)), and mfs ((g) and (h)) genes in azole-resistant environmental isolates of A. niger and A. tubingensis, respectively. In A. niger isolates PI2, PI4, PU1, PU2, F2, TR1, PR5, J6 were the azole-susceptible while isolates RK5, MD2, PI3, PR3, PR4, F1, PI5, PU6, AU3, AU4, AG1, MP6, MP8 were azole-resistant. In A. tubingensis UT2 and AH1 were the susceptible isolates while FA9, FA10, FA4, FA6, PR9, PR2, AH6 were resistant isolates. Azole-susceptible isolates PI2 and UT2 were the reference isolates for A. niger and A. tubingensis, respectively. Actin is used as a housekeeping gene to normalize transcription level. Green-colored bars represent the upregulation of the gene, and red-colored bars represent the downregulation of the gene. ****p < 0.0001.

Results depicted a consistent upregulation (> 3 folds) of cyp51A gene in isolates exhibiting exclusive resistance to ITC (MD2, PU6, PI5, and PR4), with the exception of PR3 and F1 isolates. Additionally, a slight upregulation (1.5 folds) of this gene was observed in a ITC resistant isolate RK5. Upregulation was also noticed in two out of the three isolates (AU4 and PI3) that showed cross-resistance to ITC and POS. Notably, the cyp51A gene showed a 1.7 folds upregulation in a susceptible isolate (PR5). However, in isolates with cross-resistance to ITC and VRC or isolate resistant only to POS, no upregulation was observed. In case of A. tubingensis isolates, the cyp51A gene consistently showed upregulation in all the isolates, except the susceptible isolate AH1. The expression analysis of the cyp51B gene revealed downregulation in all A. niger isolates including susceptible and resistant isolates while the gene was > 2 folds upregulated in three azole-resistant (FA4, FA9 and FA10) A. tubingensis isolates. Relative expression of gene mdr1 showed over 2 folds increased expression in two ITC resistant isolates (PU6, and PI5) and one ITC resistant isolate (RK5) with a 1.5 folds upregulation. Additionally, there was a 4.19 folds upregulation in an isolate displaying cross-resistance to ITC and VRC (AU3), and 2 folds upregulation in an isolate with cross-resistance to ITC and POS (AU4). In the case of A. tubingensis, the gene was 2 folds upregulated only in two ITC resistant isolates (FA4 and FA9).Among the 13 A. niger isolates resistant to azoles, the expression of the mfs gene was found to be upregulated in five isolates resistant to ITC, three isolates displaying cross-resistance to ITC and POS, in one isolate resistant to POS, and one isolate demonstrating cross-resistance to ITC and VRC. Notably, the mfs gene exhibited high expression, with a 10 folds increase in isolate MD2, 6.19 folds in isolate F1, and a 5 folds upregulation in MP6 and MP8. Whereas, the mfs gene was downregulated in all A. tubingensis isolates.

It’s noteworthy that overexpression of both the mdr1 and mfs gene was not observed in any of the susceptible Aspergillus section Nigri isolates. The expression data were normalized by housekeeping gene actin.

Discussion

The members of Aspergillus section Nigri causes several human diseases such as keratitis and invasive aspergillosis30,31. However, the in vivo efficacy of antifungal therapy against the clinical isolates of black aspergilli is undetermined, and in-vitro data is limited. Several authors from different countries have highlighted the prevalence of azole resistance in A. fumigatus4,32,33. However, limited number of studies has reported in-vitro antifungal susceptibilities and resistance mechanisms in the black Aspergillus spp.

In our study, we isolated 161 black aspergilli from soil samples across diverse regions in India. Azole susceptibility assay identified 20 azole resistant isolates. Notably, the number of the resistant isolates was prominent in states known for extensive agricultural activities, including Haryana (7/46), Bihar (3/23), Punjab (4/12), Madhya Pradesh (2/10), Uttar Pradesh (1/10), and Assam (3/6). We hypothesized that this prevalence might be linked to the use of azole fungicides in these regions. This hypothesis aligns with similar observations in studies involving azole-resistant A. fumigatus isolates found in soils exposed to fungicides34,35. The absence of resistant isolates in other states like West Bengal despite being a major agriculture state may be due to the limited number of samples collected. Further studies with more extensive sampling are warranted. Azole susceptibility testing of these 161 isolates demonstrated that ITC resistance (19/161) was more common suggesting that ITC resistance may be more obvious among the species of black aspergilli. Similar result was observed in previous study conducted by Howard et al.15. In contrast, POS was the most effective azole against Aspergillus section Nigri isolates. Several authors have reported similar results17,32,36. Interestingly, cross-resistance for the three tested azole drugs were not obtained in our study.

Further, the molecular identification of these 30 isolates revealed that 21 as A. niger and 9 isolates as A. tubingensis. Other Aspergillus section Nigri isolates, such as A. welwitschiae and A. brasiliensis, were not identified in our dataset, despite being present in previous studies21,37. This discrepancy may be attributed to variations in geographical distribution.

We also tried to correlate the relation between azole drug resistance and mutations in the cyp51A gene of Aspergillus section Nigri isolates in this study.

Sequence analysis of the cyp51A gene revealed that amino acid alterations, including Q228R, S346R, V329I, and T321A were present in susceptible and resistant Aspergillus section Nigri isolates. Several new mutations (R501Q, I244F, E254D, D253Y, T267Y, Y268H, E278Q, and L303M) have been observed in ITC resistant isolates only. However, it remains unclear whether these mutations are responsible for azole resistance in these isolates or not, Hence, Cyp51A protein modeling was conducted to further examine the role of these mutations in azole resistance. Molecular docking results revealed docking scores for the drugs, and the most negative docking score was obtained with model E (Mutations, Q228R/I244F/D253Y/E254/D/T267Y/Y268H/E278Q) for ITC. The docking score was almost similar in all the models for VRC. Docking results suggested no significant change in the binding site of the Cyp51A protein due to amino acid substitutions. The modeling results and similarity between the protein structures (RMSD score) suggested that cyp51A mutations in these isolates do not cause a marked change in the overall protein structure and do not directly interfere with their binding to azole drugs to confer resistance. Despite the in-silico findings, it is crucial to acknowledge the limitations of computational models. To validate the real impact of these amino acid substitutions, further in-vitro experiments employing the CRISPR/Cas9 system are essential. The approach has previously been employed to study the cyp51A gene in Aspergillus species38,39.

This study also demonstrated that mutations Q228R, S346R, and T321A were not specific to a particular azole drug. The presence of these mutations were observed in different isolates displaying varied resistance viz, ITC or POS, as well as in isolates exhibiting cross-resistance to ITC and VRC or ITC and POS. This diversity in mutation patterns suggested a lack of uniformity in the correlation between specific mutations and resistance profiles.

Similar to the study conducted by Howard et al.15, which also proposed that mutations in the cyp51A gene may not be crucial for azole resistance in section Nigri.

Azole resistance is mainly associated with acquiring genetic mutations or overexpression of the cyp51A gene and genes associated with efflux pump40. The overexpression of the cyp51A gene plays a crucial role in azole resistance in A. fumigatus; therefore, the expression of the cyp51A gene was also investigated in all 30 Aspergillus section Nigri isolates. Our findings revealed that the expression of cyp51A gene was significantly (p value ≤ 0.05) upregulated in seven azole-resistant A. niger isolates when compared to the susceptible isolate PI2. Similarly, the gene exhibited overexpression in all the resistant A. tubingensis isolates in comparison to the susceptible isolate UT2.

Surprisingly, even a susceptible A. niger isolate (PR5) carrying the Q228R mutation in the cyp51A displayed significant overexpression of this gene. This observation implies that the overexpression of the cyp51A gene may not always result in azole resistance in this fungal species. These results were consistent with the observations reported in previous studies21,37. Furthermore, we found that the cyp51B gene was downregulated in both the resistant and susceptible A. niger isolates in comparison to susceptible isolate PI2. Conversely, in three ITC-resistant isolates (FA4, FA9, and FA10), characterized by the presence of T321A in combination with V377I or K477N mutations, where the cyp51B gene exhibited upregulation compared to susceptible UT2 isolate. The baseline expression of cyp51A was more than that of cyp51B. However, a recent study on ergosterol quantification has revealed that both enzymes have a comparable impact on the total ergosterol content within the Aspergillus section Nigri cell41.

In Candida albicans and Candida glabrata, overexpression of efflux pumps, ATP-binding cassette transporters, and transporters of the major facilitator superfamily has been extensively studied42. Other studies have highlighted the significance of efflux pump genes overexpression in azole resistant A. fumigatus as well28,29. However, to the best of our knowledge, the expression of efflux pump genes in Aspergillus section Nigri isolates in India has not been investigated. In a recent study overexpression of these genes in Aspergillus section Nigri isolates has been reported, suggesting a possible drug resistance mechanism26. Therefore, we analyzed the expression of efflux pumps genes. The relative expression of the MDR efflux pump gene mdr1 was upregulated in five A. niger resistant isolates (AU3, AU4, PU6, RK5, and PI5) and two A. tubingensis resistant isolates (FA4 and FA9). The overexpression of mdr1 gene was observed along with the overexpression of cyp51A and mfs gene in 3 isolates (AU4, PU6, and RK5) carrying Q228R mutation alone or in combination with S346R. 2 folds upregulation of the gene was also observed in PI5 isolate carrying mutation R501Q in combination with Q228R. Conversely in isolate AU3, carrying similar mutations (Q228R, S346R) showed 4 folds overexpression only for mdr1 gene. Additionally, two A. tubingensis resistant isolates displayed elevated expression of the mdr1 gene, along with overexpression of cyp51A and cyp51B. The diverse gene expression pattern even among isolate with similar mutation highlights the complex azole resistance mechanisms in these fungal isolates.

The mfs gene, was upregulated in 10 out of 13 resistant A. niger isolates, and the gene was downregulated in all A. tubingensis isolates. Remarkably, among these three A. niger isolates (AG1, MP6, and MP8) displaying diverse susceptibility patterns, a common mutation profile of Q228R in combination with S346R was associated solely with the upregulation of the mfs gene in these isolates. The mfs gene exhibited high expression, with a 10 folds increase in isolate MD2, 6.19 folds in isolate F1, and a 5 folds upregulation in MP6 and MP8.

However, it’s important to note that efflux pumps genes mdr1 and mfs was not upregulated in all the susceptible Aspergillus section Nigri isolates. These findings suggest a possible association between the overexpression of these genes and drug resistance in Aspergillus section Nigri isolates. Previous studies have also investigated the overexpression of the mdr1 gene in azole-resistant Aspergillus flavus isolates lacking mutations in the cyp51A region43. Additionally, reports in A. fumigatus have indicated the upregulation of the mfs gene without cyp51A mutation44. Furthermore, the overexpression of mdr1 and mfs genes has also been reported in other studies in A. flavus45,46.

Although we identified mutations in the cyp51A gene of Aspergillus section Nigri isolates but our analysis using Cyp51A homology modeling did not show significant changes in protein structure due to these mutations. Therefore, we analyzed the role of efflux pumps in conferring resistance and found that 92% of A. niger resistant isolates exhibited overexpression of efflux pumps genes either mdr1 or mfs. The overexpression of these genes may cause azole resistance in Aspergillus section Nigri isolates. However, in the context of A. tubingensis isolates, we noted overexpression of the efflux pump gene (mdr1) in only two of the isolates. Interestingly, one of the A. niger isolate (PR3) exhibited resistance to the drug ITC. However, the gene expression data for selected genes (cyp51A, cyp51B, mdr1, and mfs) did not show significant overexpression in comparison to sensitive isolate. This observation suggests the presence of an alternative mechanisms contributing to azole resistance. The potential mechanisms may involve the upregulation of ABC transporter genes such as cdr1B and mdr4, which could result in reduced drug concentrations with in these fungal isolates. To comprehensively understand the reasons behind this susceptibility profile and to explore potential resistance mechanisms further experiments will be required.

The excessive use of triazoles in agriculture leads to their accumulation in the environment leading to the development of azole resistance. These resistant strains may infect the immunocompromised individuals and are subsequently detected in clinical settings. Our previous research, as well as other studies, have reported instances of azole resistance in A. fumigatus in the environment due to the use of azole fungicides33,47. In the present study, the identification of azole-resistant Aspergillus section Nigri isolates from the same geographical area and the overexpression of efflux pump genes in these isolates without exposure to azole drugs, has prompted us to hypothesize. We suggests that the emergence of resistance in Aspergillus section Nigri might be linked to the environmental application of azole fungicides, potentially triggering a stress response leading to upregulation of efflux pump genes.

Therefore, large-scale epidemiological studies are required to monitor the resistant fungal strains in the environment. Further, the antifungal susceptibility profiles of environmental or clinical fungal isolates should be accurately monitored.

Conclusion

The identified 15 amino acid substitutions in the cyp51A gene of Aspergillus section Nigri isolates in our study does not completely correlate with resistance data, which suggests, other genes such as mdr/ mfs needs to be investigated for mutational analysis. Advanced techniques such as CRISPR/Cas9 could be applied to study the genetic changes. The results in the study suggests that overexpression of mdr1 and mfs genes could potentially play a role in drug resistance in Aspergillus section Nigri. Studies on expression of the cyp51A, cyp51B, and multidrug efflux transporter at transcript and protein levels will help identify other genes and proteins involved in resistance. Further, genome-wide profiling may provide a better understanding of the mechanisms of azole resistance in Aspergillus section Nigri isolates.

Methods

Environmental sampling and isolation of black Aspergilli

A total of 174 agricultural soil samples were collected from 19 states across India including Assam (n = 12), Bihar (n = 21), Chhattisgarh (n = 2), Delhi (n = 12), Haryana (n = 23), Himachal Pradesh (n = 3), Jammu and Kashmir (n = 1), Karnataka (n = 4), Maharashtra (n = 3), Manipur (n = 3), Madhya Pradesh (n = 10), Orissa (n = 4), Punjab (n = 12), Rajasthan (n = 9), Sikkim (n = 1), Tripura (n = 1), Uttaranchal (n = 8), Uttar Pradesh (n = 31), and West Bengal (n = 14).

All samples were processed and inoculated on potato dextrose agar (PDA) plates as previously described by Sen et al.48. The plates were incubated at 28 ± 2 °C for 5 days. All black aspergilli isolates grown on PDA plates were identified based on their macroscopic and microscopic morphologies48. The isolates were further sub-cultured on PDA and stored at 4 °C till further use.

Antifungal susceptibility testing

The spores (conidia) of all the black aspergilli isolates were harvested in sterile phosphate buffered saline (1 × PBS) supplemented with 0.05% Tween 20; suspension was then adjusted to 1 × 104 conidia/mL in potato dextrose broth. In-vitro susceptibility testing of black aspergilli isolates was performed to determine the MICs of ITC, VRC and POS using CLSI M38- A2 broth microdilution method49 in a 96-well flat bottom polystyrene plate (Tarsons, India). The experiment was performed in triplicate for each isolate. The azole drug stocks were prepared in dimethyl sulfoxide. Two-fold dilutions were prepared in a 96-well microplate to obtain final concentrations ranging from 64–0.125 µg/mL for ITC, VRC, and POS. Conidial suspension (100 µL) was added to each well except negative control. The plates were incubated statically for 4 days at 28 ± 2 °C. The proposed ECVs for ITC, VRC, and POS were 2, 2 and 0.5 µg/mL, respectively50,51. These values were used to interpret the results. The MIC value of a drug was determined as the lowest concentration with no visible growth relative to the drug-free control.

Genomic DNA extraction

To identify the azole resistant black Aspergillus isolates and analyse their mutations related to azole susceptibility, we used the cetyl trimethyl ammonium bromide (CTAB) method52,53 to extract genomic DNA from 20 resistant black aspergilli isolates. Additionally, to establish a comparative context, we have selected an extra set of 10 azole susceptible black Aspergillus isolates from the same geographical region as the resistant isolates.

Molecular identification of black Aspergillus isolates

All the 30 isolates (20 resistant and 10 susceptible) were identified by the amplification and sequencing of full-length 18S ITS region, partial sequencing of β-tubulin genes using the ITS1 and ITS454 and Tub5 (5' TGACCCAGCAGATGTT 3') and Tub6 (5' GTTGTTGGGAATCCACTC 3')55 primers, respectively as previously described47.

Sequencing of cyp51A gene

Different primer sets were used for amplification of the cyp51A gene of twenty resistant and ten susceptible isolates (Table S1)21. Nucleotide sequencing was performed via Sanger’s sequencing using ABI 370 XL (Applied Biosystems). The sequence of the products was compared to the A. niger (Accession no. NT_166526) and A. tubingensis (Accession no. JF450924.1) cyp51A wild-type sequence using the NCBI alignment service, Align Sequence Nucleotide BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) and Clustal Omega tool (https://www.ebi.ac.uk/Tools/msa/clustalo/).

Homology modeling

Cyp51A protein structure of A. niger and A. tubingensis were not available on RCSB-PDB hence; FASTA sequence of A. niger (XP_001394224.1) and A. tubingensis (AEK81606.1) were retrieved from NCBI to construct the homology models of the wild type strains. Amino acid substitutions were incorporated in the reference protein sequences of A. niger and A. tubingensis to perform homology modeling. The X-ray crystal structures of protein 4UYM and 5FRB were used as reference for the model generation of A. niger and A. tubingensis based on the sequence similarity. The Schrödinger Maestro multiple sequence viewer (MSV) tool56 was used to construct a structure-based alignment of the mutated templates. The template structure for alignment was identified through searching the wild type sequences using the BLAST tool incorporated in Schrodinger Maestro multiple sequence viewer. The homology models were built with Prime in Schrödinger Suite (Schrödinger, LLC, New York, NY). Models were refined using minimization and loop refining tool of prime. The results of homology modeling were further validated using Ramachandran plot. Modeled structures were aligned and root mean square deviation (RMSD) was calculated using Maestro Schrödinger.

Molecular docking

The 3D structures of VRC (PubChem ID- 71,616) and ITC (PubChem ID- 55283) were retrieved from the PubChem (https://pubchem.ncbi.nlm.nih.gov/). The 3D structures of these drugs were prepared using Ligprep in Schrödinger Suite and possible states were generated at pH 7.0 ± 2.0 (LigPrep 2022). The homology models were prepared for docking using one-step protein preparation workflow in Schrödinger Maestro (Maestro 2022) by adding and refining missing hydrogen atoms. Molecular docking calculations were completed using Glide in Schrodinger docking suits (Glide 2021)57. Modeled proteins were prepared by restrained minimization using force field OPLS3e. Grids centers were determined from ligands of reference proteins. Receptor grid maps representing the shape and chemical properties of the binding site were generated using Schrödinger Glide. The binding site of modeled proteins was also confirmed by the Sitemap predicting possible binding pockets. The grid sites were created using Glide receptor grid generator with docking length of 20 Å. Docking was performed using the Schrödinger Virtual Screening Workflow tool using Glide Standard Precision (SP) with the OPLS3e force field58. One pose per ligand was kept for post-docking full force field minimization (optimization of ligand pose geometry followed by recalculation of interaction strength between ligand–protein using the scaled Coulomb-van der Waals term and the Glide score). Docking scores are reported in kcal/mol, the more negative the number, the better binding.

Quantification of gene expression by qRT-PCR

To assess the expression level of the cyp51A, cyp51B, mdr1, and mfs genes in thirty Aspergillus section Nigri isolates, quantitative real time reverse transcription PCR (qRT-PCR) was used. Further, to carry out qRT-PCR, RNA was extracted from the harvested mycelia using TRIzol reagent (Thermo Fisher)59,60. This experiment was carried out in the absence of azole drug exposure. The extracted RNA was then reverse transcribed into first-stand cDNA using the Hi-cDNA synthesis kit (HiMedia, India) by following the manufacturer’s recommendations. Real-time qPCR was performed using an ABI QuantStudio 3 (Applied Biosystems, Streetsville, Canada) as previously described in a study by Gupta et al.60. The gene expression was estimated using the 2−ΔΔCt method, with actin as the reference gene21.

Accession numbers XM_001398369.2 and XM_035497978.1 were used for the primer designing of mdr1 and mfs gene, respectively. The gene specific primers were designed using the Primer 3 software (http://primer3.ut.ee/). The primer sets used in this study are listed in Table S1.

Statistical analysis

One-way ANOVA test was used to compare relative gene expression levels. The experiment was conducted in biological and technical triplicate. The p value of ≤ 0.05 was considered significant. Statistical analysis was also performed using GraphPad Prism v8.0.2.263.