Identification of Microbial Community in Otomycosis by Metagenomic Next Generation Sequencing (mNGS): Potential Implication of Treatment with Terbinafine

The present study was designed to identify the microbial community as well as to analyze its diversity by means of metagenomic Next Generation Sequencing (mNGS) in 17 patients with otomycosis treated with terbinafine in the Department of Otolaryngology of Shandong Provincial Hospital from June 2021 to June 2022, so as to evaluate the relationship between microbial community and terbinafine resistance. Those 17 patients were divided into two groups, i.e., Terbinafine Effective Group (TEG, n = 14 cases) and Terbinafine Resistance Group (TRG, n = 3 cases) according to the therapy effect, whose microbial community of secretion of external auditory canal was identified using mNGS. We found that the sequence of bacteria was significantly more than that of fungi and, whereas, the difference between the two groups of bacteria was not significant. There were significant differences in fungal community between the two groups. Aspergillus was the main pathogenic fungus of TEG patients while Malassezia was a dominant fungus in TRG patients. In conclusion, the results from this work indicate that Aspergillus terreusis is the main pathogenic fungus in this cohort of otomycosis patients and MNGS sequencing can offer comprehensive information about the microbial community of otomycosis. The fungus community dominated by Malassezia is more likely to be resistant to terbinafine, which provides certain guidance for clinical treatment of otomycosis with terbinafine.


Introduction
In spite of the earlier emergence of fungal pathogens, for more than a century, it has been documented that the most common infections are caused by bacteria, viruses, and other pathogens [1].However, the prevalence of serious diseases resulted from fungi has increased in recent decades due to the increasing number of immunocompromised individuals [2].Meanwhile, in severely immunocompromised patients, some fungus can disseminate into the bloodstream and colonize internal organs, resulting inlifethreatening systemic infections [3,4].
Fungal communities in otomycosis may vary in different areas.According to the latest literature, A. niger, A. terreus, A. tubingensis, and A. awamori were the most frequent species in the Aspergillus genera, while C. albicans was by far the most common yeast among Candida species [9][10][11][12][13].
Predisposing factors include residing in tropical and humid climates, repeated swimming, the insertion of foreign bodies, use of hearing aids, the presence of cerumen, lack of hygiene, the use of long-term antibiotic or steroid therapy, repeated cleaning of the EAC with swabs, genetic factors, seborrheic dermatitis, diabetes, and immune defects, all of which benefit the germination of the spores and conidia of the prevalent fungi [14][15][16].
Treatment mostly requires the use of topical antifungals such as bifonazole, clotrimazole, ciclopiroxolamine, econazole, ketoconazole, miconazole, nystatin, terbinafine, and tolnaftate for at least 3 or 4 weeks [7,17,18].In clinical practice, early identification of fungal pathogens is critical for the diagnosis and treatment of otomycosis.The traditional methods such as culturebased methods can only identify the main pathogens in most cases [19], whereas, molecular microbial tools, like next-generation sequencing (NGS), have the advantage that they are culture-independent and more sensitive.The NGS uses an untargeted sequencing approach, which can identify and quantify bacteria and fungi present in a sample, including previously unknown microbes.Using an untargeted sequencing method, NGS can not only identify and quantify common bacteria and fungi, but identify previously unknown microorganisms as well.As a kind of NGS, mNGS (also termed high-throughput sequencing technology) can sequence the sample microbial genes, which can realize the sequencing of all microbial genomes, assemble and obtain the microbial genome information, and carry out the annotation and difference comparison of microbial potential functions.
The fungal pathogens of otomycosis have been reported by researchers from most countries [20][21][22].But as yet, there have been only few studies on the coexistence and interaction of bacteria and fungi in otomycosis [20].Bacteria may exist in the EAC of patients with otomycosis and cause otalgia [21].Assuming that bacteria cause otalgia, it is necessary to evaluate the diversity of bacterial and fungal communities in the EAC of patients with otomycosis.With the widespread application of antifungal drugs to treat otomycosis, more and more studies of antifungal drug resistance have been reported.Antifungal resistance is emerging as a public health challenge that needs to be addressed concurrently with antimicrobial and antiviral resistance.Currently, people in the United States are heavily infected with a super fungus called Candida auris, which has resistance to multiple antifungal drugs and causes about half of infected people to die within 3 months [23].In this study, 17 cases of otomycosis treated with terbinafine in Jinan, China, from June 2021 to June 2022 were analyzed retrospectively, and their microbial composition and possible interaction were studied by mNGS, with emphasis given on comparing the therapeutic effect of terbinafine, the relationship between different microbial communities and drug resistance to terbinafine.

Sample Collection
The samples were collected from the EAC of 17 patients with otomycosis treated with terbinafine in the Department of Otolaryngology of Shandong Provincial Hospital from June 2021 to June 2022.Among them, 14 patients recovered with terbinafine treatment after 1 month, and 3 patients didn't recover with terbinafine treatment after 1 month.

Sampling Method
17 patients were all unilateral.The affected ear was washed with sterile physiological saline, and the washing solution was stored in a sterile sampling tube, and stored at -80 °C for DNA detection.

Inclusion Criteria
The patients were between 18 and 60 years old, all of whom were for the first time.The patients had at least one of the symptoms, such as pruritus, tinnitus, otalgia, otorrhea, aural fullness, and hyperacusis.Punctate, villous, or lumpy plaque could be seen in EAC under the otoscope (Fig. 1A-C).All fungal smears were positive, and fungal hyphae or spores could be seen (Fig. 1D-F).

Exclusion Criteria
The exclusion criteria were as follows: patients who had applied antibacterial drugs locally and systemically within 1 month; patients with serious systemic diseases and immune deficiency; a diagnosis was different from otomycosis; and pregnant and lactating women.

DNA Extraction
The QIAamp DNA Microbiome Kit (50) of Kaijie Company was used to extract DNA from each sample in strict accordance with the product operating instructions.The DNA degradation degree and potential pollution were monitored on the 1% agarose gel, and the purity and integrity of DNA were analyzed.The DNA concentration was accurately quantified with Qubit 4.0 fluorometer.

Construction of Standard Sequencing Library
The DNA samples were randomly broken into fragments with a length of about 350 bp using a Covaris ultrasonic crusher, and the library was constructed through terminal repair, A-tailed addition, splicing, purification, and PCR amplification.After the construction of the library, a Qubit 4.0 fluorometer was used for preliminary quantification, and the library was diluted to 2 lg/ll.Subsequently, the length of inserted fragments in the library was detected using an Aglent 2100 biological analyzer.

High-Throughput Sequencing
According to the standard scheme, the Illumina platform (PE150 sequencing method) was used for standardized sequencing to obtain the required DNA sequence.

Data Processing
FastQC (version v0.11.9) was used to perform base quality statistics on the original sequencing data, and R statistical software was used to visualize the results.Trimmatic (Version 0.39) was used to cut the joint sequence, double-ended low-quality sequence, and double-ended pairing sequence greater than 36 bp was reserved.BMTagger (version 3.102) was used to remove the host genome sequence.

Beta Diversity Analysis
Non-metric Multi-dimensional Scaling (NMDS) based on Bray-Curtis distance was used to compare the composition of microbial community among different samples, and the difference of microbial community structure between different samples was evaluated by the distance between points.

Statistical Analysis
All statistical analyses were performed using SPSS 23.0 (SPSS Inc., Chicago, IL, United States).Enumeration data were expressed as rate (%).

Raw Data and Quality Control
The bacterial and fungal DNA in the EAC secretion of patients with otomycosis was sequenced with high flux, and the sequences obtained from each sample ranged from 29,815,652 to 87,812,501.After quality control, the sequence number of fungi varied from 11,144 to 1,746,156, while the number of bacteria varied from 418,491 to 8,968,300.The sequence number of bacteria were significantly higher than that of fungi.

Community Composition of Bacteria and Fungi
At the phylum level, the difference of bacteria between the two groups was not obvious (P = 0.78).Three categories of microorganisms, namely Firmicutes, Actinobacteria and Proteobacteria occupied an absolute dominant position, and their proportions in different samples ranged from 89.9 to 99.8% (Fig. 2A).
At the phylum level, the fungal community structure of TEG and TRG was significantly different.Ascomycota was the main fungi in TEG, accounting for more than 95% in 8 samples, while Basidiomycota was dominant in TRG, accounting for more than 79% in 3 samples (Fig. 2B).
At the genus level of bacteria, Staphylococcus and Bacillus_ A were the most common bacteria in the two groups (Fig. 3A).For fungi, Aspergillus of Ascomycetes was the main fungus in TEG, of which 7 cases accounted for more than 90%, while the fungi in TEG was mainly Malassezia, of which 2 samples accounted for more than 65% and 1 case accounted even for 97% (Fig. 3B).
At the bacterial species level, Staphylococcus aureus of Staphylococcus was dominant, while Staphylococcus epidermidis, Staphylococcus capitis and Bacillus_A bombysepticusin were also common bacteria (Fig. 4A).For fungi, the most common fungus in Fig. 1 A-C Punctate, villous, and lumpy plaque can be seen in EAC under the otoscope.D-F Scattered fungal hyphae and spores can be seen under microscope TEG was Aspergillus terreus (Fig. 4B), while the most common fungus in TRG was Malassezia restricta, of which one case accounted for more than 90%.
At the same time, in all samples, Staphylococcus aureus and Aspergillus terreus accounted for the largest proportion of bacteria and fungi respectively (Fig. 5A, B).

Analysis of Microbial Community Difference
Beta diversity analysis was performed in order to show the difference between the two groups.Non-Metric Multi-Dimensional Scaling (NMDS) based on species level showed that the bacteria in two groups were not completely separated, and there was no significant difference between samples (Fig. 6A), which indicated that there was no obvious difference in bacterial community structure in otomycosis, and bacteria were not the main factor causing fungal drug resistance.
For fungi, there were significant differences between the two groups.The sample points of TEG and TRG were gathered respectively and separated significantly, indicating that there were significant differences in the fungal community structure between the two groups (Fig. 6B).Samples of TRG were resistant to terbinafine because of their different fungal community structures.

Discussion
Early identification of pathogens is essential for the diagnosis and treatment of otomycosis.In general, detection of pathogens of otomycosis mainly depends on traditional culture-based methods and modern sequencing technology.However, culture-negative strains or resident funguses could be hardly detected by regular culture conditions [19].Some scholars had proved that the positive culture rate of the agents causing fungal infections was 50%, especially when Malassezia species or Mucorales were highly suspected [20].It had been reported that eight causative fungal genera were identified by ITS sequencing while five fungal genera were identified by culture in fungal keratitis [21].Moreover, Gu et al. [22] identified the Fig. 3 Relative abundance of bacterial (A) and fungal (B) communities at genus level in EAC of patients in TEG and TRG fungal community of otomycosis patients in Nanjing, Jiangsu Province, China, through ITS, and successfully detected some uncommon funguses such as Sagenomella and Cladosporium.Unfortunately, they overlooked that bacteria may be associated with otomycosis and play an important role in the development of the disease.Through ITS we can't simultaneously identify multiple pathogens, and the limitation of ITS is that it is not suitable for the marking of species within the genus due to the small differences in the interval because of the evolutionary order and variation of fungi.
It is known that mNGS is highly sensitive and informative, which possesses the ability to detect mixed infections of bacteria, fungi and virus.Therefore, it has the advantage of congenital thickening for Fig. 4 Relative abundance of bacterial (A) and fungal (B) communities at species level in EAC of patients in TEG and TRG diseases caused by multiple pathogens.There have been several reports on the use of mNGS for the detection of pathogenic microorganisms in infections of human body such as respiratory tract, blood and central nervous system [24][25][26].However, data on the application of mNGS in patients with otomycosis are scarce.In this work, fungi and bacterial communities in otomycosis were identified simultaneously by mNGS.The fungal community of otomycosis may vary in different analysis techniques, application of different antifungal drugs and different region.Most researchers from different countries indicated that A. niger served as the most common otomycosis agent [27][28][29].However, through mNGS, we found that the most common fungus in otomycosis was Aspergillus terreus, which was consistent with the study of Zhang et al. [11] in Hangzhou, China.In western China, Aspergillus tubingensis was the most common fungi, followed by Aspergillus fumigatus and Aspergillus terreus [30].Studies in Iran had found that Aspergillus flavus was the most common species of otomycosis, followed by Aspergillus tubingensis and Aspergillus niger [8], while Aspergillus awamori was considered to be the most common fungi causing otomycosis in studies in southern Hungary [31].In India, Aspergillus niger and Aspergillus fumigatus were dominant in otomycosis [16].Candida albicans had also been reported as a common pathogen of otomycosis, which mainly occurred in patients with abnormal immune function [32].To date, however, the bacterial community in otomycosis has been studied sparingly.In the present study, we found through mNGS that the most common bacteria in otomycosis patients was Staphylococcus aureus.At the same time, staphylococcus aureus is the main pathogen of bacterial external otitis, which can proliferate and infect the human body, leading to inflammation when human immunity is reduced and the environment of the ear canal changes [33].When fungi infect the EAC skin of patients with otomycosis, causing changes in the EAC environment and weakening immune system, Staphylococcus aureus may proliferate and infect the human body.
Currently, the treatment of otomycosis currently relies mainly on topical use of antifungal drugs.Terbinafine is more effective against partial Aspergillus species in vitro than itraconazole or amphotericin B [34].Meanwhile, terbinafine is considered to be an effective drug in the treatment of otomycosis [35].It has been reported that the cure rate of terbinafine in the treatment of otomycosis reaches 100% [36], while, in this study, 14 of the 17 patients treated with terbinafine have been cured, with a cure rate of 82.4%, which b Fig. 5 Heatmap of relative abundance of bacterial (A) and fungal (B) communities at species level in all patients Fig. 6 Non-Metric Multi-Dimensional Scaling (NMDS) based on species level in two groups.Each point in the diagram represents a sample, and samples from the same group are represented by the same color.For grouped samples, ellipses will be used to display the distinguishing areas of the sample group, and the distance between points indicates the degree of difference.Generally, when the Stress is less than 0.2, it indicates that NMDS analysis has certain reliability might be related to the increasingly severe fungal resistance at present.It is worth noting that Aspergillus is the main cause of cured cases in our study, being consistent with the above report [36].Some Malassezia strains have been reported to be resistant to terbinafine [37].We found that the fungal community dominated by some Malassezia is more likely resistant to terbinafine.In our work, the dominated fungi are Malassezia restricta and Malassezia globosa in TRG, while the dominated fungi are Aspergillus terreus, followed by Aspergillus sp.A31 in TEG.It is worth noting that, there is a case whose fungal community dominated by Malassezia slooffiae is sensitive to terbinafine in TEG, indicating that the resistance of Malassezia may be related to different strains.The relationship between different Malassezia strains and terbinafine resistance and the precise mechanism of resistance need to be further studied.
In summary, the present study identifies by means of mNGS that the most common fungal species among patients with otomycosis in the city of Jinan, Shandong Province, of the People's Republic of China is Aspergillus terreus, and the most common bacterial species is Staphylococcus aureus.Fungi dominated by Malassezia are more likely resistant to terbinafine than fungi dominated by Aspergillus.Findings from this work have important guiding significance for the clinical treatment of otomycosis with terbinafine.The relationship between different Malassezia strains and terbinafine resistance and the specific mechanism underpinning drug resistance need to be further researched.

Fig. 2
Fig. 2 Relative abundance of bacterial (A) and fungal (B) communities at phylum level in EAC of patients in TEG and TRG