Why is fungi erroneous

Many developments in this context work towards automated pipelines which rely principally on sequence similarity assessment based on the idea of a universal fungal barcoding marker, such as the ITS Majaneva et al. A major advantage of such curated databases is that curation, annotation and expansion of the database is being performed by the research community Abarenkov et al.

The ITS oligonucleotide hallmark approach attempted to refine DNA barcoding and its use in formalized interactive identification tools, by using a combination of short, species-specific sequence patterns motifs, anchors rather than overall sequence similarity Druzhinina et al. This approach should be revisited as an integrated tool as it allows adjustment to situations where more than one DNA barcode is needed, and for genome-wide studies through which diagnostic short sequences may subsequently be identified.

A growing number of studies is challenging the utility of ITS for delimiting, recognizing and identifying fungal species in certain lineages O'Donnell and Cigelnik ; Nilsson et al. A minor problem is that ITS may not amplify in all fungi Kijpornyongpan and Aime , but sequencing success is better than with many other markers Schoch et al.

More important caveats include lack of resolution and the potential presence of non-homologous ITS copies in the genome. It has been demonstrated that ITS does not provide sufficient resolution among closely related species of indoor and food-borne molds e. Occasional cases in fungal groups where ITS otherwise provides sufficient resolution, such as the subcosmopolitan and threatened macrolichens, Sticta fuliginosa and S.

In recently analyzed barcode datasets Vu et al. Meyer et al. Many plant-parasitic lineages in Dothideomycetes and Sordariomycetes cannot be resolved to species level using ITS Damm et al. On the other hand, for lichen-formers in Dothideomycetes , such as the genus Strigula , ITS provides a high level of resolution Jiang et al. A possible correlation between intragenomic variability of ITS and fungal life strategies should be explored further; the observed patterns indicate that fungal lineages exhibiting life strategies such as highly specific parasitism may undergo fast and complex speciation not immediately reflected in the ITS.

On the other hand, economically and medically important fungi are also more densely sampled, allowing for a more fine-grained taxonomy reflecting minor but important differences between individual strains.

As a result, the same underlying phylogenetic structure may translate into different taxonomic solutions, usually depending on the need. The level of precision to be achieved by DNA barcoding should therefore be dictated through context, regardless of how that precision is taxonomically formalized. In several fungal groups, ITS can only provide an initial approximation within a given clade, usually to a species complex, but cannot discriminate to the level of species.

More troubling than insufficient resolution is evidence of intragenomic variation of the ribosomal DNA rDNA cistron, including the ITS region, particularly when producing non-homologous discrete ITS variants, as this may result in conflicting molecular identifications.

Intragenomic ITS variation is well-documented for bacteria, plants and animals e. There is also growing evidence in certain fungal lineages Smith et al. Intragenomic ITS variation may largely stem from three processes: 1 stochastic point mutations resulting from DNA replication errors during cell division, 2 recombination through hybridization and introgression e. Neither hybridization and introgression nor gene duplication are unique to the ITS, but the specific challenge of utilizing ITS is its presence in multiple copies in the genome, as part of 18S-ITSS tandem repeats located on several chromosomes.

Intragenomic variation in point mutations is an obligate consequence of this, because DNA polymerases introduce stochastic errors during DNA replication. Under laboratory conditions, error rates of Taq polymerase vary between 0.

With an average number of copies in the fungal genome Lofgren et al. Such variation should not result in problems in ITS barcoding approaches, as it is substantially below even narrow identity thresholds. In contrast, processes such as hybridization and introgression or gene duplication introduce discrete ITS variants into the genome, which will result in serious identification errors if not properly recognized.

Intragenomic ITS variation is commonly misinterpreted, and its correct understanding is crucial for assessing potential problems. For instance, in the smut fungus Ceraceosorus Kijpornyongpan and Aime , intragenomic variation was found to be both stochastic and phylogenetically structured, affecting 25 and 15 out of sites, respectively.

While in the above study, the total number of stochastically varying sites 25 was high, individual sequences varied in up to four sites only, resulting in pairwise similarity of over The 15 sites with phylogenetically structured variation resulted in the formation of three clades Kijpornyongpan and Aime While these distinctive clades appear to represent non-homologous, discrete ITS copies, they may also be highly specific for this taxon and hence could be used for identification purpures.

Another factor concerning the impact of intragenomic variation in the ITS is the sequencing technique. In genomes dominated by one functional copy, Sanger sequencing will mask variation in spurious background signal and provide clean sequences. If several frequent haplotypes with point mutations exist, variants may appear as ambiguous base calls in specific positions with Sanger sequencing. On the other hand, discrete variants originating from hybridization or gene duplication will produce largely unresolved sequence chromatograms, requiring cloning or other techniques.

One example is the nectar yeasts Metschnikowiaceae , which display high intragenomic rDNA variation Heeger et al. Similar considerations apply to other groups, such as arbuscular mycorrhizal fungi Lekberg et al. Therefore, metabarcoding data have to be interpreted with great care and multiple alignment-based approaches should be employed to identify and resolve potential issues see below.

The availability of well-documented reference data is of particular importance to properly assess ITS variants stemming from intragenomic variation. If ITS pseudogenes have been identified for a fungal lineage e. Li et al. Given that the average number of ITS copies in the fungal genome is around Lofgren et al.

While it is unclear whether the necessary high-molecular weight DNA can be obtained, since commonly used extraction techniques require a mechanical disruption of fungal cells, successful rDNA tandem repeat sequencing using a combination of PacBio and Oxford Nanopore sequencing has been performed in fungi Wurzbacher et al.

Another caveat of the ITS is interspecific and intragenomic length heterogeneity. In some groups, such as ascomycetous yeasts, the full length ITS1, 5. In most fungi, the length of the ITS is more uniform, but even minor variation may result in regions with low alignment confidence. Strategies to avoid this would be primer-based filtering or, as outlined above, anchoring with nuSSU or nuLSU flanking regions via long-fragment reads. While single-copy protein-coding markers proposed as secondary DNA barcodes in fungi do not exhibit the problems associated with multiple copies, phenomena such as paralogs may apply to them as well, such as in COX1 , RPB2 , and TUB2 Gilmore et al.

Regardless of the marker, the quality of reference data is of utmost importance, particularly in environmental metabarcoding. While it may not work for all fungi at the desired level of precision, ITS remains the first choice for fungal identifications at a broad level.

It is not only easily amplified with some exceptions; e. Kijpornyongpan and Aime , but it also is the most frequently sequenced fungal marker both in specimen-based and metabarcoding approaches, making it unchallenged as a reference compared to any other marker. Even if secondary barcode markers are increasingly employed, they only represent a small fraction of available sequence data compared to ITS.

The application of ITS is thus comparable to a first diagnosis across all fungi. Depending on the results, secondary DNA barcodes may be required to obtain the desired resolution. Unfortunately, in some common and diverse fungal genera, such as Fusarium and Trichoderma , due to lack of resolution, some taxonomists have stopped sequencing the ITS.

This practice is not recommended, as it excludes these taxa from being detected in metabarcoding surveys. Even if not necessarily providing enough resolution, ITS should be sequenced for each fungal lineage in addition to other markers, in order to provide a broad reference database that offers a compromise between coverage and precision.

Metabarcoding studies would then employ ITS as default marker and additionally one or several secondary barcodes e. These approaches have become popular as they are easily integrated into automated pipelines and allow the analysis of extremely large data sets in a short time and with little manual work involved Majaneva et al.

In contrast to multiple alignment-based phylogenetic approaches, pairwise similarity may wrongly assess positional variation and hence not accurately reflect taxonomic entities or phylogenetic relationships. For instance, a position with a varying indel comprising either [AG], [A] or [G], in a multiple alignment will align all [A] with either [A] or a gap, but not with [G], whereas pairwise alignment will interpret a single [A] and [G] as a substitution.

This issue may appear minor but can cause dramatic effects in OTU clustering, especially when such variation is caused by sequencing errors e. As a consequence, OTUs derived from clustering are different in number and composition when compared to actual phylogenetic entities Porter and Golding ; Powell et al.

Huse et al. Increased accuracy while not compromising in computational speed can also be achieved by hc-OTU clustering through homopolymer compaction Park et al. Clustering approaches require predefined similarity thresholds, but such fixed thresholds do not exist when it comes to the delimitation of species. In phylogenetic treatments based on ITS, sister species can differ in as few as three bases around Indeed, in certain groups of fungi, such as Hypocreales Fusarium , Gibberella , Trichoderma , species hypotheses delimited at Varying optimal thresholds have been determined for different lineages based on two large barcode datasets Vu et al.

If the marker of choice lacks resolution, then even the highest similarity threshold will not yield reliable OTU estimates. This latter threshold does reflect empirically derived estimates e. The potential underestimation of species richness using fixed pairwise similarity thresholds is counterbalanced by the overestimation of taxonomic units through OTU clustering bias.

As a result, a proportion of OTUs may not be real taxonomic entities, whereas a proportion of real taxonomic entities may be missed. Arbitrary variation of predefined thresholds, e. Similarity assessment through pairwise alignment also poses limitations for BLAST-based identifications of individual amplicon variant metabarcoding reads Callahan et al.

While amplicon variant BLAST mapping avoids potential bias of OTU clustering, it also relies on pairwise alignment scores, particularly max score, query cover, e value and percentage identity. Max score, the sum of match rewards and mismatch and gap penalties, depends on query and reference sequence length: shorter matches with higher identity may receive a lower score and not be immediately visible as best hits.

The e value, the number of expected hits of similar score that could be found by chance, is computed from max score and results in the same sorting of matches but depends on query sequence length and reference database size and hence is not comparable across databases.

Both max score and e value are also affected by the structure of reference sequences, such as partial ITS sequences that include long portions of the conserved nuSSU or nuLSU or are dominated by the 5. BLAST identity considers individual indels as mismatches and hence results in lower similarity values than the other two approaches for a given sequence pair. It is also more sensitive to homopolymer-based sequencing errors in the query reads and affected by improper trimming of low-quality terminal portions of reference sequences Nilsson et al.

As a result, sequences retrieved as best hits in BLAST searches are not necessarily most closely related e. Thiery et al. The latter is not possible for metabarcoding studies, as BLAST results cannot be inspected individually. However, automated verification can be achieved through phylogeny-based analysis of metabarcoding reads that compute statistical support values for alternative placements.

Read placement into a reference tree is a promising approach that increases accuracy and precision in metabarcoding studies compared to OTU clustering and BLAST-based amplicon variant read mapping Stark et al. The method, also dubbed phylogenetic binning, relies on three components: 1 a reference tree for a set of taxa which can be derived through phylogenetic analysis of existing data; 2 a fixed alignment of reference sequences corresponding to the metabarcoding marker e.

ITS for the taxa included in the reference tree; 3 a set of query reads from a metabarcoding study corresponding to the same barcoding marker. In a first step, the query reads are automatically aligned to the fixed reference alignment Berger et al. In a second step, each query sequence is individually placed into the reference tree based on its alignment by invoking the Evolutionary Placement Algorithm EPA; Stamatakis et al.

In addition to a maximum likelihood or maximum parsimony approach offered by the EPA, read placement can also be performed in a Bayesian framework using pplacer Matsen et al. Mirarab et al. Phylogenetic binning placed each query sequence at the most closely matching node under an evolutionary model: if the query sequence matches a terminal, it will cluster with that terminal; alternatively, it attaches to an internal node representing a higher taxonomic level, an approach that conceptually corresponds to the LCA.

While the Bayesian framework in pplacer offers direct assessment of statistical confidence, the EPA allows the computing of bootstrap support values for potential alternative read placements. Optionally, prior to invoking the EPA, the phylogenetic pattern of the metabarcoding marker over the fixed reference alignment can be analyzed using a maximum parsimony or maximum likelihood approach in order to compute a weight vector.

In doing so, potential homoplasy through saturation in highly variable regions of the metabarcoding marker can be assessed to improve the subsequent placement of query sequences into the reference tree. Therefore, the reference tree should be inferred based on markers that do not include the metabarcoding marker, to avoid circular conclusions.

Apart from bootstrapping and Bayesian posterior probabilities offering automated verification, phylogenetic binning has further, important advantages over OTU clustering and BLAST mapping.

Point variation in query reads, whether representing sequencing errors or real variation, does not prevent their accurate placement into a reference tree Berger et al. The absence of close relatives in a reference tree is immediately discernible by placement of a query read at a deeper node, a more accurate approach than LCA, as it avoids the ambiguity of low similarity values in the latter.

Read placement also allows the implementation of quantitative species delimitation methods to automatically assess taxonomic diversity, an approach already integrated into the phylogenetic binning approach Zhang et al. Broad reference trees can be assembled and centrally maintained to be used in analytical pipelines Tedersoo et al. Given the large amount of data to be analyzed, often encompassing hundreds of thousands of reads, environmental metabarcoding of fungi requires a trade-off between speed on one hand and accuracy and precision on the other see below.

Up to the recent past, OTU clustering was the only viable approach to achieve this goal. However, phylogenetic binning is now possible through massive parallel computing on large clusters Barbera et al. As is true for other organisms, fungal species are not only defined horizontally through phylogenetic and phenotypic coherence, but also vertically through time of origin and subsequent diversification.

Individually different evolutionary histories thus make it impossible to apply universal and unambiguous criteria for the delimitation, recognition, and identification of fungi. Best practice depends on each group, and residual ambiguity remains in many cases, also due to incompleteness of identification tools and reference data.

Full exploration of the various conceptual approaches to delimit fungal species, including reproductive biology, is currently only feasible for selected taxa including model organisms. Since generalizations from model studies are limited to close relatives or ecologically equivalent taxa, this approach should be expanded to cover selected species in all groups of fungi, representing the diversity of phenotypes, lineages, and nutritional strategies.

For broadly cataloguing fungal diversity, an integrative polyphasic taxonomic approach seems most effective, adjusted to the group under study and combining molecular and phenotype data. In many groups, single-marker DNA barcoding may suffice, whereas more complex taxa require a combination of primary and secondary barcoding markers or multi-marker approaches.

Phylogenomics may be employed to resolve particularly difficult species complexes, but this approach demands large computational and personal resources and is currently limited to exemplar case studies.

The phenotype remains an integrative component of fungal taxonomy, encompassing also data derived from cultures and other sources. Taxonomists will continue to describe new species in the absence of molecular data, in groups where this approach is justified. However, phenotypic data should be thoroughly analyzed before establishing new species by any method.

If the material would allow the generation of molecular data but the methodology to do so is not available, then collaboration to produce such data is recommended. In general, the goal remains to document all fungi with molecular data. Phenotypic data are of particular importance when assessing the status of phylogenetically distinct clades through integrative taxonomy. In such cases, quantitative analysis of structured phenotype matrices should be implemented to assess phenotypic variation in a phylogenetic context, which will then also allow the detection of reliable diagnostic characters.

On a molecular level, ITS remains the universal fungal barcode marker to initially identify phylogenetic lineages. It can thus be considered a first diagnosis. Where ITS does not suffice to discriminate between species, secondary barcoding markers or multi-locus approaches need to be employed to achieve the desired level of precision and accuracy. How individual markers resolve species is determined by context, and feasibility of particular markers should not be uncritically transferred from one taxonomic group to another but instead empirically explored for each taxon.

ITS will likely remain the marker of choice for fungal metabarcoding studies, although long-read approaches or the addition of secondary barcoding markers will improve accuracy and precision.

However, metabarcoding approaches should move away from OTU clustering and BLAST mapping exercises and instead implement phylogenetic methods, such as read placement phylogenetic binning. Current issues arising with DNA barcoding of fungi are not primarily due to conceptual limitations of the approach but due to shortcomings of reference databases, including incompleteness in terms of taxonomic coverage, lack of properly documented genetic diversity, and inaccuracy of sequence labels.

Major efforts must therefore be directed at further improving these resources, particularly the continued and critical revision of existing data to achieve high quality labels. Data sharing is not applicable to this article as no datasets were generated or analysed specifically for this purpose.

New Phytologist — Article PubMed Google Scholar. The Quarterly Review of Biology — Aime MC Intercompatibility tests and phylogenetic analysis in the Crepidotus sphaerula group complex: concordance between ICGs and nuclear rDNA sequences highlight phenotypic plasticity within Appalachian species. Google Scholar. Studies in Mycology — Mycologia — Fungal Biology — Journal of Molecular Biology Molecular Phylogenetics and Evolution — Molecular Ecology Resources e—e MycoKeys — Article Google Scholar.

Mycoscience — Coleoptera : Scolytinae on woody hosts. Experimental Mycology — Biological Reviews — Applications in Plant Sciences BMC Microbiology Ecology and Evolution — Systematic Biology — Ecological Informatics — Springer International Publishing, Cham, pp — Chapter Google Scholar.

Applied Microbiology and Biotechnology — Environmental Microbiology — Methods in Ecology and Evolution — Berger SA, Krompass D, Stamatakis A Performance, accuracy, and web server for evolutionary placement of short sequence reads under maximum likelihood.

Trends in Ecology and Evolution — Molecular Ecology Resources — Blackwell M The fungi: 1, 2, American Journal of Botany — Blackwell M, Vega FE Lives within lives: hidden fungal biodiversity and the importance of conservation. Fungal Ecology — Bonants P, Edema M, Robert V Q-bank, a database with information for identification of plant quarantine plant pest and diseases.

EPPO Bulletin — Yale University Press, New Haven. Journal of Clinical Microbiology — Fungal Diversity — Springer, Cham. BMC Bioinformatics Nature Methods — Microbiol Resource Announcements 8 29 :e—e Microbial Ecology — BMC Genomics Mutation Research — International Journal of Systematic and Evolutionary Microbiology — Frontiers in Microbiology IMA Fungus — ZooKeys — Fungal Genetics and Biology — Crawford SD Lichens used in traditional medicine.

Springer, Cham, pp 31— Journal of Microbiologcal Methods — Annual Review of Phytopathology — Mycological Research — Bioinformatics — Bioinformatics btaa Some fungi, such as lichens , even look like plants see Figure below. Both fungi and moss are growing on this tree. Can you tell them apart? Today, fungi are no longer classified as plants.

We now know that they have unique physical, chemical, and genetic traits that set them apart from plants and other eukaryotes. For example, the cell walls of fungi are made of chitin , not cellulose. Also, fungi absorb nutrients from other organisms, whereas plants make their own food. These are just a few of the reasons fungi are now placed in their own kingdom. Classification of fungi below the level of the kingdom is controversial.

Building the bioeconomy will include significant efforts in enzyme discovery to facilitate the diversification of substrates to be upgraded and of products to be developed from the biorefinery. Such discovery efforts will simultaneously result in massive accumulation of underutilized genome sequencing data Murphy et al. The vision is straightforward: if we become better at predicting functions from sequences, we could significantly shorten and sharpen the enzyme discovery process.

We could go directly from sequence to the subgroup of enzyme genes we would like to screen to identify the one with the highest potential for that specific biomass conversion process.

This list can be transformed into one providing an overview of all the functions illustrated as a list of EC numbers found in the secretome of that ecological niche. For each function, the protein families that have those functions, often more than one type of protein family for each EC function, can be specified. Another highly promising field for new microbial and fungal products is the use of inocula for strengthening crop plants, making the plants more robust to abiotic stress and more efficient with regard to water and nutrient utilization.

Since the days of the green revolution, plant breeding has taken place almost in isolation except for breeding for increased tolerance against certain plant diseases and pests. Now, advances in the field of fungal and microbial products for agriculture is making it possible for plant breeders, mycologists, and microbiologists, to work together to find the combination of plants and other organisms which will provide farmers and the world with more robust and resilient agriculture.

At a time when climate change challenges agriculture in many parts of the world, and where water and nutrition are at a minimum in many places, the combinations of fungi, microbes, and plants can provide opportunities for significant progress in global food, feed, and biomass production. It will be interesting to follow developments in this area. Again, a new field is emerging that builds on a platform of knowledge generated through the combined efforts of public and private mycological research, where fungi are seen as having the potential to contribute significantly towards a more sustainable world.

This is therefore a highly interesting area for contributions from mycologists specialized in endophytes, mycorrhizas, Penicillium, Aspergillus, Trichoderma, Fusarium , soil fungi, and consortia where bacteria and fungi work together to produce efficient and optimized systems. Next in demand are regimes of enzymes active under low temperature conditions and enzymes active and stable at high temperatures for decomposition of both plant and animal derived biomasses.

Here, mycologists specialized in extremophiles can contribute significantly Zajc et al. Could intensified studies of the archaean splicing mechanisms and protein expression lead to a breakthrough in the development of fungal production hosts for heat stable secreted enzymes from these microorganisms?

Improved expression of basidiomycete genes in ascomycetous expression hosts, such as yeasts, also builds on an improved understanding of the variations in protein expression mechanisms within the fungal kingdom. There are discoveries to be made of new kinds of antibiotics from fungal hotspots for antibiotics. Discovery of novel drug candidates with new modes of action and central nervous system active metabolites from fungi, which manipulate insects to adopt behavioral patterns that optimize the chances for dispersal of fungus spores formed on the insect after the insect dies e.

Maybe such research efforts can reveal new concepts for neurosignalling, induced in animals, based on stimuli given by fungal metabolites or proteins, but which may also be of relevance for increased understanding of the human central nervous system? This is important because there will be a steep increase in amounts of fungal biomass available for upgrade as the technology of biomass conversion in biorefineries develops.

In nature, complex patterns and mechanisms of collaboration are expected to occur especially between fungi and bacteria. This is a most interesting area for basic studies, and has the potential to reveal organized consortia we have not even dreamt about so far. A model to use as a starting point for further studies is the enrichment consortium of waste water treatment Nielsen et al. In the future, when complex organic substrates are to be broken down to complex products to solve challenging problems, fungal and microbial consortia could provide a short-cut to a solution.

Last, but most importantly, further studies are urgently needed to prevent, control and cure serious human mycoses caused by Coccidioides species Taylor The number of fatalities caused by this fungal disease, especially in Africa, approaches that caused by tuberculosis and malaria, but far less effort has been invested in preventing and curing human mycoses. Perhaps basic studies of the secretome including substances bound to the outer wall structures , and new molecular insights, instruments and technologies could bring us a step further in this very difficult significantly under prioritized area of health and pharmaceutical investment.

It is evident from the above descriptions that investment in strengthening mycology globally is worthwhile. The following efforts are urgently needed to enable this to occur:. National Center for Biotechnology Information , U. IMA Fungus. Published online Dec Lene Lange 1. Lene Lange 1 Aalborg University, A. Author information Article notes Copyright and License information Disclaimer. Received Sep 30; Accepted Dec 1. You are free to share - to copy, distribute and transmit the work, under the following conditions: Attribution: You must attribute the work in the manner specified by the author or licensor but not in any way that suggests that they endorse you or your use of the work.

Non-commercial: You may not use this work for commercial purposes. No derivative works: You may not alter, transform, or build upon this work.

Any of the above conditions can be waived if you get permission from the copyright holder. This article has been cited by other articles in PMC.

Abstract In the new bioeconomy, fungi play a very important role in addressing major global challenges, being instrumental for improved resource efficiency, making renewable substitutes for products from fossil resources, upgrading waste streams to valuable food and feed ingredients, counteracting life-style diseases and antibiotic resistance through strengthening the gut biota, making crop plants more robust to survive climate change conditions, and functioning as host organisms for production of new biological drugs.

It further provides a vision for how mycology can be strengthened: The time is right to make the world aware of the immense importance of fungi and mycology for sustainable global development, where land, water and biological materials are used in a more efficient and more sustainable manner.

Keywords: bioeconomy, fungal research, fungal products, resource efficiency, global solutions, funding opportunities. Open in a separate window.

The following efforts are urgently needed to enable this to occur: 1 Understanding the fungal kingdom, through phylogeny and phylogenomics: support for global participation and global coverage of strains used for resolving the Fungal Tree of Life.

Science : — ISME Journal 6 : — Applied and Environmental Microbiology 79 : — Nucleic Acids Research 37 : D—D Journal of Biotechnology : — Trends in Biotechnology 16 : — Summary Report. Current Opinion in Biotechnology 2 : — A guide to identification of the food and air-borne terverticillate penicillia and their mycotoxins. Studies in Mycology 1— Extremophiles 14 : — Mycology 2 : — BMC Genomics 14 : Eukaryotic Cell 4 : — Mycological Research : — Enzyme and Microbial Technology 67 : 47— Nature : — PhD thesis, Copenhagen University.

Annual Review Phytopathology 52 : — Database: bar doi Environmental Microbiology 14 : — Mycologia 98 : — Mycologia 87 : — BMC Genomics 14 : 1— Support Center Support Center. External link. Please review our privacy policy.