Biology Methods and Protocols, 2020, 1–6 doi: 10.1093/biomethods/bpaa026 Methods Manuscript METHODS MANUSCRIPT Rapid and real-time identification of fungi up to species level with long amplicon nanopore sequencing from clinical samples Sara D’Andreano1,2, Anna Cuscó2 and Olga Francino 1, * 1 SVGM, Servei Veterinari de Genètica Molecular, Universitat Autònoma de Barcelona, 08193, Bellaterra, Barcelona, Spain and and 2Vetgenomics, Edifici EUREKA, Parc de Recerca de la UAB, Campus UAB, 08193, Bellaterra, Barcelona, Spain *Correspondence address. SVGM, Servei Veterinari de Genètica Molecular, Universitat Autònoma de Barcelona, 08193, Bellaterra, Barcelona, Spain. E-mail: olga.francino@uab.cat Abstract The availability of long-read technologies, like Oxford Nanopore Technologies, provides the opportunity to sequence longer fragments of the fungal ribosomal operon, up to 6 Kb (18S-ITS1-5.8S-ITS2-28S) and to improve the taxonomy assignment of the communities up to species level and in real-time. We assess the applicability for taxonomic assignment of amplicons targeting a 3.5 Kb region (V3 18S-ITS1-5.8S-ITS2-28S D2) and a 6 Kb region (V1 18S-ITS1-5.8S-ITS2-28S D12) with the What’s in my pot (WIMP) classifier. We used the ZymoBIOMICSTM mock community and different microbiological fungal cultures as positive controls. Long amplicon sequencing correctly identified Saccharomyces cerevisiae and Cryptococcus neoformans from the mock community and Malassezia pachydermatis, Microsporum canis and Aspergillus fumigatus from the microbiological cultures. Besides, we identified Rhodotorula graminis in a culture mislabelled as Candida spp. We applied the same approach to external otitis in dogs. Malassezia was the dominant fungal genus in dogs’ ear skin, whereas Ma. pachydermatis was the main species in the healthy sample. Conversely, we identified a higher representation of Ma. globosa and Ma. sympodialis in otitis affected samples. We demonstrate the suitability of long ribosomal amplicons to characterize the fungal community of complex samples, either healthy or with clinical signs of infection. Keywords: fungi; nanopore; ribosomal operon; Malassezia; otitis; dog Introduction [6, 17]. On the contrary, problems of unknown branches of fun- gal phylogenies still occur, due to considerable gaps in genetic Fungi are a vast kingdom of organisms with a range between 1.5 knowledge and to old species description [22]. and 6 million species [7], but only a modest part, around 140.000 Sequence-based methods allow to classify the fungi king- species, is phenotypical and genetically described [8, 27]. dom better. Still, the choice of the methodology used to study Usually, fungi have been identified by morphology on pure cul- the mycobiome, or even the intrinsic characteristics of a specific tures in agar medium. The main problem is that many species fungus, can impact the data generated and the results reached are difficult to isolate and culture, and even to classify [1, 24]. [24]. Public online databases for fungal identification are Both clinical approaches and research applications need noteworthy but still limited due to new updates for the best no- taxonomic classification to assign taxa to their functional traits menclature and identification of fungi species [16]. A large Received: 3 July 2020; Revised: 2 December 2020; Accepted: 18 December 2020; Editorial Decision: 15 December 2020; C The Author(s) 2020. Published by Oxford University Press. V This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 1 2 | S. D’Andreano et al. Figure 1: Fungal ribosomal operon: two hypervariable internal transcribed spacers regions (ITS1 and ITS2, marked in orange) and three conserved ones (18S, 5.8S and 28S rRNA, marked in grey) that contain variables domains, nine for the 18S and twelve for the 28S rRNA genes. Primers set used for the amplification of the ITS region (3.5 Kb) are shown in orange in the upper part of the operon, and the ones for amplification of the full operon (6 Kb) in grey in the lower part [20, 25]. number of online fungal databases are available for mycology regions of LSU subunit, fragments obtained from the amplifica- research, and they are growing thanks to the dedication of the tion can be assigned up to the species level [17]. experts’ team. Taxonomic revisions are still ongoing and the Here, we aim to investigate the applicability for taxonomic main databases for fungi classification are Species Fungorum assignment of the long-amplicon PCR approach to detect and (www.speciesfungorum.org), MycoBank (www.mycobank.org), identify the fungal microbiota present on complex microenvir- UNITE [13] and International Nucleotide Sequence Database onments at the species level. We use microbiological fungal Consortium (https://www.ncbi.nlm.nih.gov/taxonomy). cultures characterized phenotypically as positive controls. We One of the preferred markers for taxonomy assignment is then apply the same protocol to clinical samples of canine otitis the fungal ribosomal operon, which is almost 6.000 bp length. It as a complex microenvironment. contains three conserved units, the 18S rRNA gene (small subu- nit, SSU), 5.8S rRNA gene and 28S rRNA gene (large subunit, Materials and methods LSU), and two hypervariable units as internal transcribed Samples and DNA extraction spacers regions (ITS1 and ITS2). The ITS units flank the 5.8S RNA gene, and better represent the high variability among taxo- LETI laboratories (LETI Animal Health) kindly provided a total of nomic levels of fungi; showing a superior species discrimination eight microbiological fungal cultures grown at 28 C in Petri and PCR success rates [10]. The variable domains located at the dishes in Dermatophyte Test Medium (DTM agar) and conserved 18S (V1-V9) and 28S rRNA genes (D1-D12) (Fig. 1) are Sabouraud Chloramphenicol agar. Four of the cultures had been also worth considering to refine the taxonomy assignment. It is classified up to the genus level (Alternaria spp., Aspergillus spp., essential to recognize that the D1-D2 from the 28S rRNA gene Candida spp. and Malassezia spp.) and four other ones up to the domains are the ones that perform a higher level of taxonomic species level: three of Malassezia pachydermatis and one of assignment for fungi [20]. Microsporum canis. Also, fungal DNA of the ZymoBIOMICSTM Primer sets to amplify the fungal operon regions have been mock community (Zymo Research, Irvine, CA) containing described in different manuscripts, starting from 1990 until Saccharomyces cerevisiae and Cryptococcus neoformans was in- 2018 [9, 12, 21, 22, 25, 26]. These sets of primers target the appro- cluded in the study as a positive control. The DNA from all priate operon fragments to proceed with either short fragments, fungal samples was extracted by ZymoBIOMICSTM Miniprep kit by massively parallel sequencing (or second generation se- following the manufacturer’s instructions. quencing) as Ion Torrent or Illumina, or longer fragments with Four canine otitis samples were analysed as complex micro- single-molecule sequencing (or third generation sequencing) as bial microenvironments. Two of them were collected from a PacBIO or Oxford Nanopore Technologies. Petri dish, divided into two halves parts, to culture fungi Taxonomy with short reads is focused on ITS1 and ITS2 from both ears of a dog, one of the ears was healthy and the other regions, considered as the official barcoding markers for one showed clinical signs for external otitis (S02_healthy; species-level identification in fungi, due to their easy amplifica- S03_affected). The DNA was extracted by using ZymoBIOMICSTM tion, conserved primers sites, widespread use [19] and available Miniprep kit, as for the cultures. The other two otitis samples databases, such as UNITE or MycoBank. Usually, the ITS1 and (S01_affected; S04_affected) were collected by swabbing the inner pinna of the ear of two dogs using Sterile Catch-AllTM Sample ITS2 regions provide the taxonomy resolution within-genus and Collection Swabs (Epicentre Biotechnologies). DNA was extracted within-species level, but debates on which one provides the with QIAGEN-DNeasy PowerSoil Kit (Hilden, Germany). DNA best taxonomic skill are still under discussion [2]. Alternative quality control was checked by Nanodrop and QubitTM markers located in the small and large subunits of the rRNA Fluorometer (Life Technologies, Carlsbad, CA). genes can address the phylogenetic diversity, depending on the fungi species [2, 20]. Depending on the fungus species, different regions of the operon can be considered for taxonomy assign- MinION sequencing ment: the SSU and the LSU are used when taxonomy is investi- Two sets of primers were chosen (Table 1; Fig. 1): the first set gated up to family level, while lower taxonomy level analysis targeting the ribosomal operon from V3 region of 18S RNA gene requires the ITS regions. When primers sets include the D1-D2 to D3 region of 28S RNA gene (3.500 bp), and the second one Identification of fungi with long amplicon nanopore sequencing | 3 Table 1: Primers targeting the full ITS region (3.5 Kb) and the whole fungal operon (6 Kb) Name Sequence (50 -30 ) Target Amplicon Reference SSU515Fngs-F TTTCTGTTGGTGCTGATATTGCGCCAGCAACCGCGGTAA 18S-V3 3.5 Kb Tedersoo et al. [20] LR5-R ACTTGCCTGTCGCTCTATCTTCTCCTGAGGGAAACTTCG 28S-D3 3.5 Kb Tedersoo et al. [20] SR1R-Fw TTTCTGTTGGTGCTGATATTGCTACCTGGTTGATQCTGCCAGT 18S-V1 6 Kb Vilgalys Lab [25] LR12-R ACTTGCCTGTCGCTCTATCTTCGACTTAGAGGCGTTCAG 28S-D12 6 Kb Vilgalys Lab [25] The Nanopore Universal Tag in bold type. targeting the complete ribosomal operon from V1 region of 18S Data analysis RNA gene to D12 region of 28S RNA gene (6000 bp). The pri- For further in-depth analyses, the fast5 files with the sequenc- mers, both forward and reverse, included the Nanopore ing reads were basecalled and demultiplexed by Albacore v2.3.3 Universal Tags (Oxford Nanopore Technologies Ltd, UK). for the 3.5 Kb amplicon or guppy 2.3.5 for the 6 Kb amplicon. Two PCRs were performed: the first for the amplification of Barcodes and adapters were removed by using Porechop the target and the second one to add the specific barcode to (https://github.com/rrwick/Porechop). Taxonomy was assigned each sample. DNA initial concentration was of 5 ng DNA per with the cloud-based analysis WIMP application from the sample, in 50 ll of PCR final volume: 15 ll of DNA plus 35 ll of R EPI2ME platform (Oxford Nanopore Technologies Ltd, UK), PCR mix, which contained 10 ll of PhusionV High Fidelity Buffer which is based on Centrifuge (https://ccb.jhu.edu/software/cen (5), 5 ll of dNTPs (2 mM), 0.5 lM of primer forward and reverse, R trifuge/manual.shtml). and 0.02 U/ll of PhusionV Hot Start II Taq Polymerase (Thermo The fastq files output of each fungal amplicon with the Fisher Scientific GmbH, Dreiech, Germany). PCR profile included length of 3.5 Kb and 6 Kb are loaded on Zenodo (http://doi.org/ an initial denaturation of 30 s at 98 C, followed by 25 cycles of 10.5281/zenodo.3662300) and ENA (PRJEB41658). 10 s at 98 C, 30 s at 62 C, 80 s at 72 C and a final extension of 10 min at 72 C. Amplicons obtained were purified with Agencourt AMPure XP beads (Beckman CoulterTM A63880, Results Thermo Fisher Scientific GmbH, Dreiech, Germany), at 0.4 ratio We aim to develop a long-amplicon PCR approach to detect and for the fungal amplicon; then, they were quantified by QubitTM identify fungal microbiota present on complex microenviron- fluorometer (Life Technologies, Carlsbad, CA). ments, and to apply it to clinical samples (canine otitis). As posi- Following the PCR Barcoding kit protocol (SQK-PBK004; tive controls, we chose microbiological fungal cultures and Oxford Nanopore Technologies Ltd, UK), 0.5 nM per each sample fungal strains from a mock community. Some of the cultures was required for the second PCR to add barcodes of PCR barcod- were previously classified by classical microbiology up to the ing kit (EXP-PBC001). The final volume of second PCR is 100 ll, genus level as Alternaria spp., Aspergillus spp., Candida spp. and containing 20 ll of DNA template from the previous PCR at Malassezia spp. Others were classified up to the species level as 0.5 nM, 2 ll of specific barcode and 78 ll of mixture that include R Ma. pachydermatis and Mi. canis at LETI laboratories (LETI Animal 20 ll of 5 PhusionV High Fidelity Buffer, 10 ll of dNTPs (2 mM) R Health). The ZymoBIOMICSTM mock community fungal strains and 2 U/ll of PhusionV Hot Start II Taq Polymerase. PCR profile are S. cerevisiae and C. neoformans. included an initial denaturation of 30 s at 98 C, 15 cycles of 10 s at 98 C, 30 s at 62 C, 80 s at 72 C and final step of 10 min at 72 C. Identification of microbiological cultures and mock The amplicon obtained were purified again with Agencourt AMPure XP beads, at 0.4 ratio and quantified by QubitTM fluo- community with long amplicons rometer (Life Technologies, Carlsbad, CA). All samples were amplified for both amplicon sizes, 3.5 Kb and We proceeded then to the Library preparation with the 6 Kb. In the 3.5 Kb amplicons, we included those domains that Ligation Sequencing kit (SQK-LSK109, Oxford Nanopore better help in the taxonomic classification of fungi, as shown in Technologies Ltd, UK). Barcoded samples (1.5 lg) were pooled in Fig. 2. 47 ll of nuclease-free water and the library was prepared follow- Both amplicons correctly detected and identified the ing the manufacturer conditions. ZymoBIOMICSTM mock community fungal strains (S. cerevisiae, With a final step of Agencourt AMPure XP beads 0.4, the C. neoformans), and Ma. pachydermatis and Mi. canis from microbi- DNA library was cleaned and ready to be loaded into the flow ological cultures. Looking in detail, S. cerevisiae, C. neoformans cell. We used two SpotON Flow Cells (FLO-MIN106; Oxford and Ma. pachydermatis were detected up to 100% by both 3.5 Kb Nanopore Technologies Ltd, UK) for three MinION runs, primed and 6 Kb long fragments, while 6 Kb amplicon better detected with a mixture of sequencing buffer and Flush buffer according Mi. canis. to the manufacturer’s instructions. A quality control of se- Both fragments identified Aspergillus genus as the main one quencing pores was done before each run. Libraries were mixed found in the culture, but looking at the species level, Aspergillus with Sequencing Buffer and Loading Beads in a final volume of fumigatus was the most abundant. Alternaria spp. and Candida 75 ll. The final mix was added, by dropping, in the SpotON sam- spp. showed different results from what we expected (Fig. 3). ple port. For the Alternaria culture, both amplicons size detected Sequencing runs were performed using the MinKNOWN 2.2 Metarhizium brunneum as the main fungus found in these cul- v18.07.2 and the MinKNOWN v18.12.9 (Oxford Nanopore tures. Conversely, Alternaria alternata was the species found in Technologies Ltd, UK). Nanopore sequencing from Oxford really low relative abundance (0.2% for 3.5 Kb and 15.6% for 6 Nanopore Technologies includes real-time analysis with the Kb). No similarities between these fungi were found: they be- EPI2ME platform [What’s in my pot, (WIMP)], allowing the identi- long to different order, Pleosporales and Hypocreales. Alternaria fication of the fungal species few minutes after the run started. is ubiquitous filamentous fungi, which are found in soil air and 4 | S. D’Andreano et al. Figure 2: ZymoBIOMICSTM mock community (S. cerevisiae and C. neoformans) and microbiological cultures of fungi after taxonomical classification of the 3.5 Kb and 6 Kb ribosomal amplicons. Figure 3: Fungal microbiological cultures showed unexpected results in the taxonomical classification after sequencing. Few reads from the Alternaria culture belonged to Alternaria spp., and it was classified at species level as A. alternata, but the most abundant fungus found was Me. brunneum. No reads from the Candida culture were classified as Candida spp. because of the presence of R. graminis. human/animal skin [15]; Metarhizium is commonly found as a Two of the samples correspond to the same dog, one from a parasite of insects and symbiont of plants [18, 23]. healthy ear (S02) and the other one (S03) with clinical signs Looking at the nomenclature of this fungus, Me. brunneum compatible with otitis externa, and Ma. pachydermatis is the belonged to Me. anisopliae strain [23, 29], but no correlation with main fungal species detected in both ears. The other two Al. alternata was found. samples (S01 and S04) came from the ear with otitis externa of For the Candida culture, the colonies of this fungus in Petri two dogs. In that case, other Malassezia species were detected dish were red/orange. Sequences revealed the presence of together with Ma. pachydermatis, such as Ma. globosa and Rhodotorula graminis, and only a few reads were classified as Ma. sympodialis (Table 2). Candida spp. Rhodotorula is a carotenoid biosynthetic yeast, part of the Basidiomycota phylum, easily identifiable by distinctive Discussion yellow, orange or red colonies [28]. This yeast produces three major carotenoids: b-carotene, torulene and torularhodin, and Our first approach with Oxford Nanopore Technologies se- is commonly associated with plants (http://www.antimicrobe. quencing was aimed to understand if long amplicons are suit- org/f16.asp#t1). In this case, we could confirm the presence of able markers to analyse the mycobiome in dog skin, and which two yeast species in the microbiological culture. size could be the best in the analysis of mycobiome. The micro- biological cultures were essential for the study as positive Canine Otitis controls because their genome sequences were used to validate Conscious that no differences were found in Ma. pachydermatis the correct detection of fungi in WIMP. analysis using both fragments sizes (Fig. 2), we sequenced mi- Primers used to amplify the fungal ribosomal operon crobiological cultures of Ma. pachydermatis as positive controls domains should be chosen depending on the fungus, but no and four complex samples with 3.5 Kb amplicon size. We run standard markers are defined yet. The longest amplicons WIMP for fungal communities’ detection: the positive controls should be considered to describe the communities at lower tax- were identified as Ma. pachydermatis, while the complex otitis onomy classification [22, 27]. Malassezia spp., S. cerevisiae, C. neo- samples showed other Malassezia species (Table 2). The reads formans, Mi. canis and Aspergillus spp. were correctly detected corresponding to Malassezia out of the total reads were 15 726 and identified from the microbiological cultures. However, the out of 15 940 for S02; 22 666 out of 23 067 for S03; 7188 out of 7842 microbiological cultures corresponding to Alternaria spp. and for S04 and 15 243 out of 17 150 for S01. Candida spp. were misidentified as per classical microbiology, Identification of fungi with long amplicon nanopore sequencing | 5 Table 2: Relative abundance of Malassezia species found in pure Ma. pachydermatis culture, and in four complex samples belonged to three dif- ferent dogs affected by otitis Sample Malassezia pachydermatis (%) Malassezia globosa (%) Malassezia sympodialis (%) Malassezia spp. (%) Others (%) M01 98.9 0.2 0.2 0.2 0.5 M02 99.0 0.1 0.2 0.2 0.5 M03 98.9 0.2 0.3 0.2 0.4 S02_healthy 98.2 0.1 0.2 0.2 1.3 S03_affected 95.1 1.1 1.7 0.4 1.7 S04_affected 78.8 4.0 5.0 3.9 8.3 S01_affected 73.7 4.7 5.1 5.3 11.2 Samples S02 and S03 belong to the same dog, while S04 and S01 belong to two different dogs. and other fungi were detected. It is noteworthy that the samples specific database of fungal ribosomal operons for the molecular plated came from dog skin, which is prone to environmental identification of fungi, such as UNITE (https://unite.ut.ee), as contamination, as has been previously described in skin micro- there are many more ribosomal operons sequenced than com- biome of healthy dogs [4, 5]. plete fungal genomes. However, UNITE targets only the ITS re- Few of the reads from the Alternaria culture were classified gion as the formal fungal barcode. The long-read sequencing as A. alternata with the 6 Kb amplicon, while most classified as either from fungal isolates or from microbiome samples will im- Me. brunneum. Discovered in Spain and used as an herbicide prove the database, which could be used with WIMP for the tax- against fly Bactrocera oleae [29], this fungus belongs to the same onomic assignment of fungi. phylum of Ascomycota, but it differs at lower taxonomy levels. This study is a first approach for the applicability of ribo- The Candida microbiological culture was misclassified, even somal long amplicons to identify some of the most common when showing an orange colour, caused by R. graminis. fungi in the skin of dogs. We evaluate the applicability of long In this study, the proper positive controls were those from reads to the fungal ribosomal operon as a whole, taking into ac- the ZymoBIOMICSTM mock community (S. cerevisiae and C. neo- count not only ITS1 and ITS2 as the most common markers for formans). The microbiological cultures were grown from dogs’ fungal taxonomy, but adding the variable regions of 18S and 28S skin samples and donated after morphological classification by rRNA genes in the 6 Kb amplicon. We used positive controls microbiologists as potential positive controls. We checked them from the ZymoBIOMICSTM mock community, and from microbi- with the long amplicon approach to reach the species level ological cultures that cannot always be considered pure cul- (when and if possible). The Candida and Alternaria colonies were tures. We demonstrate the suitability of this approach to originally misclassified, confirming that morphology only is not characterize the fungal community of otitis samples, either enough for fungal taxonomic classification. healthy samples or samples with clinical signs of infection. It is Finally, we investigate the possibility of reaching species out of the scope of this study the comparison with other non- level in complex samples from the skin of dogs affected by oti- ribosomal loci commonly used for taxonomic assignment when tis, finding that Malassezia was the most abundant genus. The ITS is not able to resolve to species level. Further steps are classification at the species level was performed to investigate needed to evaluate the ribosomal long amplicon as a potential possible changes between health status and diseased one. universal barcode, without discarding the need to rely on non- Malassezia pachydermatis has been reported as the most ribosomal loci for sequence-based identification of some fungal abundant species in the ear canal of healthy dogs [11]. WIMP taxa. correctly identified all the Malassezia samples, and we were able On the contrary, nanopore sequencing is a technology that to identify Malassezia at the species level from four complex ca- evolves fast and both the library preparation and the bioinfor- nine otitis samples. Two of the samples corresponded to the matics tools improve every few months for providing better same dog, one from a healthy ear (S02) and one with clinical results in real time. The next steps will lead to simplify the li- signs (S03) that were compatible with otitis externa. Malassezia brary preparation with the Rapid Barcoding kit from ONT and pachydermatis is the main fungal species detected (98% of the the analysis of complex samples from different origins to detect reads for the healthy ear – S02 – and 95% for the sample with the causal agent of the disease in a clinical metagenomics clinical signs compatible with otitis externa – S03). The other approach. two samples (S01 and S04) came from the ear with otitis externa of two other dogs. In those cases, other Malassezia species were Data Availability detected together with Ma. pachydermatis, such as Ma. globosa and Ma. sympodialis. The fastq files output of each fungal amplicon with the length The results agree with previous studies on Malassezia spp., of 3.5 Kb and 6 Kb are loaded on Zenodo (http://doi.org/10.5281/ describing it as a commensal microorganism in human and ani- zenodo.3662300). mal skin that may become pathogenic [3, 14]. It is worthy to note that WIMP taxonomic assignation is Funding based on Centrifuge, which could misclassify some closely- related species due to the absence of complete reference This study was funded by a grant awarded by the genomes in the database used. However, Malassezia is a genus Generalitat de Catalunya (Industrial Doctorate Program, with reference genomes for each one of its species, and such a grant 2015DI044) and by Vetgenomics. misclassification is not observed for the positive control or the two other samples. The procedure will benefit by using a Conflict of interest: AC and OF work for Vetgenomics, SL. 6 | S. D’Andreano et al. References 15. Pastor FJ, Guarro J. Alternaria infections: laboratory diagnosis and relevant clinical features. Clin Microbiol Infect 2008;14: 1. Arbefeville S, Harris A, Ferrieri P. Comparison of sequencing 734–46. the D2 region of the large subunit ribosomal RNA gene R 16. Prakash P, Halliday C, Irinyi L et al. Online databases for tax- (MicroSEQV) versus the internal transcribed spacer (ITS) onomy and identification of pathogenic fungi and proposal regions using two public databases for identification of com- for a cloud-based dynamic data network platform. J Clin mon and uncommon clinically relevant fungal species. J Microbiol Methods 2017;140:40–6. Microbiol 2017;55:1011–24. 2. Blaalid R, Kumar S, Nilsson RH et al. ITS1 versus ITS2 as DNA 17. Raja HA, Miller AN, Pearce CJ et al. Fungal identification using metabarcodes for fungi. Mol Ecol Resour 2013;13:218–24. molecular tools: a primer for the natural products research 3. Cafarchia C, Gallo S, Capelli G et al. Occurrence and popula- community. J Nat Prod 2017;80:756–70. tion size of Malassezia spp. in the external ear canal of dogs 18. Samish M, Rot A, Ment D et al. Efficacy of the entomopatho- and cats both healthy and with otitis. Mycopathologia 2005; genic fungus Metarhizium brunneum in controlling the tick 160:143–9. Rhipicephalus annulatus under field conditions. Vet Parasitol 4. Cuscó A, Catozzi C, Vin ~ es J . et al Microbiota profiling with 2014;206:258–66. long amplicons using Nanopore sequencing: full-length 16S 19. Schoch CL, Seifert KA, Huhndorf S et al., Fungal Barcoding rRNA gene and the 16S-ITS-23S of the rrn operon. F1000Res Consortium. Nuclear ribosomal internal transcribed spacer 2019;7:1755. (ITS) region as a universal DNA barcode marker for Fungi. Proc 5. Cuscó A, Sánchez A, Altet L et al. Individual signatures define Natl Acad Sci USA 2012;109:6241–6. canine skin microbiota composition and variability. Front Vet 20. Tedersoo L, Anslan S, Bahram M et al. Shotgun metagenomes Sci 2017;4:1–12. doi: 10.3389/fvets.2017.00006. and multiple primer pair-barcode combinations of amplicons 6. Dayarathne M, Boonmee S, Braun U et al. Taxonomic utility of reveal biases in metabarcoding analyses of fungi. Mc 2015;10: old names in current fungal nomenclature: conflicts, confu- 1–43. sion & clarifications. Mycosphere 2016;7:1622–48. 21. Tedersoo L, Lindahl B. Fungal identification biases in micro- 7. Hawksworth DL, Lücking R. Fungal Diversity Revisited: 2. 2 to 3. biome projects. Environ Microbiol Rep 2016;8:774–9. 8 Million Species, 2017. doi:10.1128/microbiolspec.FUNK-0052- 22. Tedersoo L, Tooming-Klunderud A, Anslan S. PacBio meta- 2016.Correspondence barcoding of fungi and other eukaryotes: errors, biases and 8. Hibbett D, Abarenkov K, Ko ~ ljalg U et al. Sequence-based clas- perspectives. New Phytol 2018;217:1370–85. sification and identification of Fungi. Mycologia 2016;108: 23. Tiago P V, Oliveira N T D, Lima E Á D L A. Biological insect con- 1049–68. trol using Metarhizium anisopliae: morphological, molecular, 9. Ihrmark K, Bödeker ITM, Cruz-Martinez K et al. New primers and ecological aspects. Cienc Rural 2014;44:645–51. to amplify the fungal ITS2 region - evaluation by 454-se- 24. Usyk M, Zolnik CP, Patel H et al. Novel ITS1 fungal primers for quencing of artificial and natural communities. FEMS characterization of the mycobiome. mSphere 2017;2:1–11. Microbiol Ecol 2012;82:666–77. 25. Vilgalys Lab. Conserved Primer Sequences for PCR Amplification of 10. Kalan L, Grice EA. Fungi in the wound microbiome. Adv Fungal rDNA, Duke University, updated Feb 3, 1992, 1–5. Wound Care 2017;7:wound.2017.0756. doi:10.1089/wound. 26. White TJ, Bruns T, Lee SJWT et al. Amplification and direct se- 2017.0756. quencing of fungal ribosomal RNA genes for phylogenetics. 11. Korbelik J, Singh A, Rousseau J et al. Analysis of the otic myco- PCR Protoc 1990;18:315–22. biota in dogs with otitis externa compared to healthy individ- 27. Wurzbacher C, Larsson E, Bengtsson-Palme J et al. uals. Vet Dermatol 2018;29:417–e138. 12. Lee J, Lee S, Young JPW. Improved PCR primers for the detec- Introducing ribosomal tandem repeat barcoding for fungi. tion and identification of arbuscular mycorrhizal fungi. FEMS Mol Ecol Resour 2019. doi: 10.1111/1755-0998.12944. Microbiol Ecol 2008;65:339–49. 28. Yadav S, Manjunatha KH, Ramachandra B et al. 13. Nilsson R H, Larsson K-H, Taylor A F . et al The UNITE data- Characterization of pigment producing rhodotorula base for molecular identification of fungi: handling dark taxa from dairy environmental samples. Asian J Dairy Foods Res and parallel taxonomic classifications. Nucleic Acids 2014;33:1. Res 2019;47:D259–64. 29. Yousef M, Alba-Ramı́rez C, Garrido Jurado I et al. Metarhizium 14. Ngo J, Taminiau B, Fall PA et al. Ear canal microbiota - a com- brunneum (Ascomycota; Hypocreales) treatments targeting parison between healthy dogs and atopic dogs without clini- olive fly in the soil for sustainable crop production. Front Plant cal signs of otitis externa. Vet Dermatol 2018;29:425–e140. Sci 2018;9:1–11.