Molecular identification of Aspergillus fumigatus complex from lung transplant recipients using multilocus sequencing analysis (MLSA)

Abstract

BACKGROUND

Aspergillus infection causes significant morbidity and mortality among lung transplant recipients (LTRs). It is primarily caused by Aspergillus fumigatus. Other closely related species belonging to the section Fumigati have also been found. These cryptic species are often misidentified as A. fumigatus. Thus, we used multilocus sequencing analysis (MLSA) of the calmodulin, β-tubulin, and hydrophobin gene sequences to identify these species and to determine the frequency with which they occur among LTRs.

METHODS

A total of 81 A. fumigatus isolates were initially isolated from bronchoalveolar lavage fluid or sputum specimens collected from lung transplant patients. These isolates were then sub-cultured and genotyped using MLSA. Of these isolates, 53, 17, and 11 were isolated from double LTRs, single LTRs, and pre-LTRs, respectively.

RESULTS

All isolates (100%) carried DNA sequences identical to those of A. fumigatus reference strains and thus clustered in the same clade with A. fumigatus. Analysis of the MLSA data revealed that A. fumigatus species were the only species recovered in this population of LTRs. The MLSA results were consistent with those routinely obtained by conventional mycological procedures in the microbiology laboratory.

CONCLUSIONS

A. fumigatus appears to be the primary causative agent of colonization or invasive aspergillosis among LTRs. No cryptic species were identified.

Introduction

Aspergillus infection in lung transplant recipients (LTRs) has two clinically distinct presentations: Aspergillus colonization and invasive disease. In turn, invasive disease is divided into airway disease (tracheobronchitis or bronchial anastomotic diseases) or parenchymal disease (pulmonary or disseminated diseases) (1,2). Aspergillus colonization is an independent risk factor for the development of invasive disease and is managed in a similar fashion as invasive disease (2). Aspergillus colonization has been reported in as many as 14% of LTRs, and invasive disease has been reported to occur in as many as 8% (3,4).The mortality related to invasive aspergillosis (IA) in solid organ transplant recipients is 22% (5). Various antifungal treatment strategies, including universal prophylaxis, targeted prophylaxis, and pre-emptive therapy, are used for the prevention of invasive disease (2). Use of universal antifungal prophylaxis in LTRs may be associated with the development of resistant Aspergillus or infection with other moulds (6,7). Aspergillus fumigatus is the primary etiologic agent of IA. Other closely related species belonging to the section Fumigati, such as Aspergillus lentulus, Aspergillus viridinutans, and Aspergillus udagawae (Neosartorya udagawae), have occasionally been reported to be the causative agents of IA (1). However, the actual prevalence of these species may be underestimated as a result of their misidentification by conventional diagnostic approaches. Conventional procedures for the identification of Aspergillus spp are based on their morphological characteristics and microscopic features. However, the use of morphological characteristics to differentiate between species of A. fumigatus and others in the section Fumigati require an extremely high level of expertise and long-term observation to accurately distinguish between them. In addition, the species included in the section Fumigati are not morphologically uniform, and thus the use of morphological characteristics alone makes it difficult or impossible to distinguish among them. This has frequently led to species misidentification (8). Members in the section Fumigati have been reported to be more resistant to antifungal agents than A. fumigatus (9–11). Misidentification may therefore lead to adverse therapeutic consequences as a result of resistance to azoles and other antifungal agents (8,9).

The limited pathogenic role of Fumigati cryptic species (other than A. fumigatus) may be due to lower thermotolerance and decreased production of mycotoxins, such as gliotoxin (8,12). Although resistance to triazoles is rare among A. fumigatus, these cryptic species commonly exhibit reduced susceptibility to antifungal agents (10,11,13).

Therapeutic management of IA can be challenging after solid organ transplantation in the clinical setting, and this difficulty is related to the lack of accurate and reliable identification of A. fumigatus from those closely related species (8,11,14). Multilocus sequencing analysis (MLSA) may provide a useful technique for the accurate identification of A. fumigatus and closely related species isolated from lung transplant patients, which could in turn inform more appropriate empiric treatment decisions. Prevalence data on the frequency of cryptic species among LTRs are lacking. In the present study, we aimed to distinguish between A. fumigatus and other cryptic species isolated from clinical specimens collected from a population of LTRs managed with a pre-emptive antifungal strategy and to categorize them into their respective species using MLSA.

Materials and Methods

Fungal strains, growth, and DNA extraction

A total of 81 phenotypically identified A. fumigatus isolates were obtained from sputum or bronchoalveolar lavage (BAL) fluid of LTR patients collected for clinical diagnosis of Aspergillus colonization or IA at the University Health Network (UHN) in Toronto in 2018–2019. All isolates were identified using standard microbiological procedures, including microscopic and macroscopic morphological characteristics by the shared Sinai Health System/UHN Department of Microbiology. In addition to the A. fumigatus ATCC MYA-4609 strain, other well-characterized reference isolates of A. fumigatus (PHO4788), A. lentulus (PHO0940) and A. thermomutatus (PHO7412) obtained from the culture collection of the Public Health Laboratory of Ontario (PHOL) in Toronto, were included in this study as reference strains to ensure the accuracy of MLSA (Table 1).

Table 1:

Strains used in this study

Strain	GeneBank accession number
Aspergillus fumigatus isolate NRRL 4661	EF669911.1
A. fumigatus strain ATCC MYA3626	KT150504.1
A. fumigatus strain ATCC MYA4609	KT150505.1
A. fumigatus isolate M 56	KJ175558.1
A. fumigatus isolate AFB-623	KJ914711.1
A. fumigatus strain DBMCC204	KY421204.1
A. fumigatus strain CMXY6730	MG991519.1
A. fumigatus strain CMXY9825	MG991527.1
A. fumigatus strain KACC 41388	AY689331.1
A. fumigatus strain CMXY22843	MG991470.1
Neosartorya fischeri strain KACC 41664	AY870692.1
A. lentulus strain FH5	DQ094896.1
A. viridinutans strain KACC 41142	AY689362.1
Neosartorya udagawae strain KACC 41155	AY689372.1
A. novofumigatus strain IBT 16806	DQ094893.1
Neosartorya spinosus strain KACC 41119	AY689371.1
A. brevipes strain KACC 41136	AY689365.1
A. fumigatus isolate	PHOL reference
A. lentulus isolate	PHOL reference
Neosartorya pseudofischeri isolate	PHOL reference

PHOL = Public Health Ontario Laboratories

Each isolate was subcultured onto sabouraud dextrose agar (Oxoid Canada Ltd, Nepean, Ontario) and incubated for 3–5 days at 35°C. DNA was extracted from each isolate using the Fungi/Yeast Genomic DNA Isolation Kit (Norgen Biotek Corp, Thorold, Ontario), according to the manufacturer’s instructions.

PCR and phylogenetic analysis

MLSA, including polymerase chain reaction (PCR), DNA sequencing, and phylogenetic analysis, was carried out as described elsewhere (15). PCR amplified regions of the β-tubulin (benA), hydrophobin (rodA), and calmodulin (cmd) genes of A. fumigatus using the primers listed in Table 2 (16). PCR samples were prepared in a 25 µl reaction volume containing 12.5 μl of TopTaq Master Mix (Qiagen, Hilden, Germany), 1.5 μl of each fungal DNA extract, and 1 μl of forward and reverse oligonucleotide primers at 20 μM. The PCR assays were performed using an Applied Biosystems Thermo Fisher MiniAmp Thermal Cycler (Applied Biosystems, Foster City, California, USA), using the following cycling parameters: an initial 3 min denaturation step at 94°C, followed by 35 cycles of hybridization–elongation, including 30 s at 94°C, 30 s at 54°C, and 60 s at 72°C, with a final extension step of 10 min, and then cooling at 4°C. PCR products were visualized on a 1.5% agarose gel using gel electrophoresis method. Amplicons were purified using the QIAquick PCR purification kit (Qiagen, Hilden, Germany). Fifty ng of the amplicon was mixed with 0.7 µl of the forward and reverse primers in separate wells (same as the respective PCR primers) and then sent for Sanger sequencing by the Centre for Applied Genomics in Toronto.

Table 2:

Sequences of primers used in the MLSA assays (16)

Primer	Target	Sequence	Reference	Amplicon size
BT2A	β-tubulin	5’-AATTGGTGCCGCTTTCTGG-3’	(16)	471 bp
BT2B	β-tubulin	5’-AGTTGTCGGGACGGAATAG-3’	(16)
RODA1	Hydrophobin	5’-GCTGGCAATGGTGTTGGCAA-3’	(16)	461 bp
RODA2	Hydrophobin	5’-AGGGCAATGCAAGGAAGACC-3’	(16)
CMD5	Calmodulin	5’-CCGAGTACAAGGAGGCCTTC-3’	(16)	461 bp
CMD6	Calmodulin	5’-CCGATAGAGGTCATAACGTGG-3’	(16)

MLSA = Multilocus sequencing analysis

According to both the neighbour-joining (NJ) and maximum-parsimony (MP) algorithms, phylogenetic relationships among all sequenced strains were established by MEGA software version 6 (17), using 1,000 bootstrap analysis and 70% or more as the cutoff value for generation of the phylogenetic tree for each locus. Bootstrap values were generated from 1,000 pseudoreplicates.

The obtained sequences were viewed, analyzed, and trimmed when needed using CLC Genomics Workbench version 8.53 (Qiagen, Germantown, Maryland, USA). Each set of homologous gene sequences for each gene was aligned using MUSCLE algorithms within the MEGA application. Sequences of other closely related isolates in the Aspergillus section Fumigati were obtained from the GenBank database and included in the phylogenetic analyses (Table 1). Also, the gene sequences obtained from the PHOL reference and ATCC strains used in this study were included in the analysis. Aspergillus clavatus, a member of section Clavati, was chosen as an outgroup to ensure the specificity of the MLSA. It was previously shown to be distinct from the Aspergillus section Fumigati by morphologic and phylogenetic analyses (15). The DNA sequence datasets are deposited in GenBank with accession numbers MW221835–MW221915.

Results

Clinical isolates

Aspergillus clinical isolates recovered in this study were all cultured from either BAL fluid or sputum. On the basis of the association of each isolate with its corresponding specimen type–patient, they were categorized as double LTRs (53 isolates), single LTRs (17 isolates), or pre-transplant recipients (11 isolates).

The baseline characteristics of the pre- and post-LTR patients are provided in Table 3. The ratio of men to women was similar (41 men, 40 women).

Table 3:

Baseline characteristics of lung transplant recipients

Characteristic	Type of transplant			Total (N = 81)
	Single lung (n = 17)	Double lung (n = 53)	Pre-transplant (n = 11)
Sex
Female	7	29	4	40
Male	10	24	7	41
Underlying disease
Hypersensitivity pneumonia	4	1	0	5
Pulmonary fibrosis or emphysema	5	5	0	10
Interstitial lung disease	4	11	0	15
Cystic fibrosis	0	11	2	13
Idiopathic pulmonary fibrosis	4	6	1	11
COPD		12	4	16
Others	0	7	4	11
Final diagnosis
Aspergillus colonization	4	35	6	45
Probable IPA	9	15	1	25
Pre-transplant colonization	0	5	1	6
Aspergilloma	4	1	0	5
CMV donor status
Positive	10	31	0	41
Negative	7	14	0	21
N/A	0	8	11	19
CMV recipient status
Positive	15	34	0	49
Negative	2	15	0	17
N/A	0	5	10	15

COPD = Chronic obstructive pulmonary disease; IPA = Invasive pulmonary aspergillosis; CMV = Cytomegalovirus; N/A = Not available

Phylogenetic analyses

Phylogenetic analyses of 1,000 replicates was used to estimate the inferred tree branches’ reliability and robustness of MLSA (18,19). Phylogenetic analyses of each locus generated trees with similar topology (Figures 1–3).

Figure 1:

Phylogenetic tree inferred from neighbour-joining analysis of partial calmodulin gene sequences from all Aspergillus fumigatus isolates included in this study

The numbers next to the nodes represent bootstrap values of ≥70% (out of 1,000 bootstrap replications). The number of nucleotide changes between taxa is represented by branch length. The evolutionary history was inferred using both the neighbour-joining method and the maximum parsimony

Figure 3:

Phylogenetic tree inferred from neighbour-joining analysis of partial hydrophobin (rod A) gene sequences from all Aspergillus fumigatus complex isolates included in this study

The numbers next to the nodes represent bootstrap values of ≥70% (out of 1,000 bootstrap replications). The number of nucleotide changes between taxa is represented by branch length. The evolutionary history was inferred using both the neighbour-joining method and the maximum parsimony

Figure 2:

Phylogenetic tree inferred from neighbour-joining analysis of partial β-tubulin gene sequences from all Aspergillus fumigatus complex isolates included in this study

The numbers next to the nodes represent bootstrap values of ≥70% (out of 1,000 bootstrap replications). The number of nucleotide changes between taxa is represented by branch length. The evolutionary history was inferred using both the neighbour-joining method and the maximum parsimony

All 81 A. fumigatus isolates possessed the three gene sequences (calmodulin, β-tubulin, and hydrophobin), which aligned and clustered with those of Fumigati, including A. fumigatus reference strains as well as A. clavatus from GenBank (Figures 1–3). The primers used for the calmodulin gene amplified a total of 461 nucleotides. In comparison, the β-tubulin gene primers amplified a total of 471 nucleotides, and the hydrophobin gene primers amplified a total of 461 nucleotides.

Both NJ and MP tree topology revealed that the various A. fumigatus isolates clustered together as a monophyletic group with high bootstrap support (≥90%) that fell into the A. fumigatus clade (MP tree not shown) (Figure 1).

Discussion

A. fumigatus is responsible for the majority of cases of IA, particularly among immunocompromised hosts (1). Other species in the section Fumigati (A. lentulus, A. viridinutans, Aspergillus fischeri, Aspergillus pseudofischeri, and A. udagawae) have also been reported to occasionally cause IA (20–23). The prevalence of cryptic species in the setting of pre-emptive antifungal therapy in lung transplantation is not known. In our study, MLSA data analysis revealed that A. fumigatus were the only species recovered in our LTRs, and this correlated well with results obtained by conventional mycological procedures. Other Fumigati species appeared to be uncommon among our patient samples and were not detected in LTRs or pre-LTRs. To our knowledge, this is the first study to evaluate the Fumigati in LTRs using MLSA analyses.

Conventional diagnostic methods depend mainly on growth conditions and morphological features for proper identification of Aspergillus species. This requires a long incubation time for cultures and a high level of experience to identify correctly (24,25). Recent studies have highlighted the limitations of species identification based only on morphologic features because of morphological variability and differences in growth conditions. In addition, culture methods followed by morphologic identification have relatively low sensitivity, making it difficult or impossible to identify isolates to the species level (25–27). Accurate identification of Aspergillus species associated with IA is essential to choose appropriate early empiric antifungal therapy pending the results of in vitro susceptibility testing, which should improve patient outcomes (28,29).

Comparative DNA sequence-based identification formats appear to be promising in terms of speed, ease of use, and objectivity for Aspergillus species identification. Standard molecular methods use specific primers or probes that target conserved regions of the ribosomal gene, such as the internal transcribed spacer (ITS) regions (30). However, ITS regions do not have enough phylogenetic strength with strong bootstrap support to resolve Aspergillus species’ evolutionary relationship within the section Fumigati (25,31–33). Some have reported that comparative sequencing analysis of the partial sequences of β-tubulin, calmodulin, and hydrophobin genes is reliable and accurate for identifying Fumigati isolates to the species level (25,34).

In a study of 499 A. fumigatus clinical isolates cultured from a variety of sources, including sputum, bronchoscopy, and tissue biopsy specimens collected from multiple medical centres around the world and identified by the A. fumigatus–specific Luminex assay as A. fumigatus (35,36), 96 randomly selected isolates were subjected to PCR sequencing of the β-tubulin region. The comparative sequence analysis of the β-tubulin region revealed that all 96 isolates were A. fumigatus with 99%–100% concordance with the reference strain NRRL 6113 (GenBank accession no. EF669854). Interestingly, non–A. fumigatus species (other Fumigati species) were not recovered from this global collection. According to this study, one possible explanation for this discrepant finding could be that the non–A. fumigatus isolates have geographical specificity (36).

Another study screened a large number of clinical A. fumigatus species isolates collected as a part of a nationwide fungal surveillance study (TRANSNET) that were identified using a two-step molecular format that included comparative sequence analyses of the ITS and the β-tubulin regions. This study found that A. fumigatus was isolated from 94% of the cases studied, with only 6% represented by three other species within A. fumigatus (Fumigati species)—A. lentulus, A. udagawae, and A. pseudofischeri (37). Although A. lentulus appears to have a global distribution (25), the TRANSNET study found that all A. lentulus isolates were recovered from only one geographical area (Seattle, Washington), which was not represented in the study (37).

Phylogenetic studies integrate data of partial DNA sequences of various genes, such as β-tubulin, calmodulin, and hydrophobin, to distinguish the different species and subspecies within the section Fumigati (16,34,38). Currently, several studies have found that comparative sequence analyses of these genes is the most reliable method for proper species and subspecies identification of A. fumigatus (8,16,25,27,34,39). In this study, we identified all Fumigati clinical isolates to the species level using MLSA of calmodulin, β-tubulin, and hydrophobin gene sequences. Phylogenetic analysis was strengthened by using both MP and NJ analysis of these genes. Our study confirms that the identification of our isolates using phylogenetic sequence analysis was in concordance with conventional laboratory identification methods used in the clinical microbiology laboratory. We did not identify any cryptic Fumigati species in the absence of prior antifungal treatment.

Some species in the section Fumigati, such as A. lentulus, A. pseudofischeri, and A. udagawae, have been reported to be resistant in vitro to the azole antifungals, including itraconazole, miconazole, posaconazole, ravuconazole, and voriconazole (8). Therefore, accurate identification of A. fumigatus and Fumigati species using molecular methods is important for early clinical management of IA.

Our study has certain limitations that we would like to acknowledge. First, although we used consecutive isolates stored in the mycology lab, it is possible that some of the isolates could have been missed because of the lack of collection (in cases in which bronchoscopy was performed outside our hospital) or absence of growth in the subculture media. However, only a few isolates did not grow in the media, and the overwhelming majority of bronchoscopies were performed at our centre. Second, because we collected the data retrospectively, some of the cases may be misclassified. However, we maintain a robust lung transplant database, and categorization of the clinical syndrome was done by two clinicians independently and, in case of dispute, was resolved by the senior author (S Husain). Finally, in the absence of data regarding the prevalence of the cryptic species, it is difficult to ascertain whether we had an adequate sample size to report the true prevalence of the cryptic species. However, this is the most extensive series to date among LTRs, and we can say with reasonable certainty that the prevalence of the cryptic species is negligible.

On the basis of our preliminary data, routine use of MLSA testing is not warranted in LTRs without prior antifungal agent exposure. A. fumigatus appears to be the primary causative agent of IA among our LTRs in the setting of pre-emptive antifungal therapy. Other cryptic Fumigati species were not detected.

Conclusions

Clinical surveillance with more phylogenetic analysis of Aspergillus isolates from lung transplant patients may be useful in detecting the emergence of other Fumigati species over time. This may help guide the use of prophylactic, pre-emptive, and treatment courses of antifungal agents with these patients. In view of the lack of presence of cryptic species in our cohort, routine use of MLSA for identification of cryptic species may not be required.

Declarations:

All methods included in this study were performed according to the standards guidelines and referenced protocols.

Availability of Data and Materials:

All data generated or analyzed during this study are included in this published article (and its supplementary information files). The raw data analyzed during the current study are available from the corresponding author on reasonable request. The DNA sequence datasets generated in this study are deposited in GenBank with accession numbers MW221835-MW221915.

Ethics Approval:

The local Research Ethics Board of the University Health Network (UHN), University of Toronto, approved this study (REB no. 13-7093-CE). All samples were collected as fungal culture isolates from the microbiology laboratory in Mount Sinai Hospital/UHN, and permission was obtained from the Microbiology Department to use these isolates in the study. Clinical specimens of patients were not directly processed in this study.

Informed Consent:

N/A

Registry and the Registration No. of the Study/Trial:

N/A

Funding:

No funding was received for this work.

Disclosures:

The authors have nothing to disclose.

Peer Review:

This manuscript has been peer reviewed.

Animal Studies:

N/A