Fungal DNA barcoding is the process of identifying species of the biological kingdom Fungi through the PCR amplification and sequencing of specific DNA sequences and their comparison with sequences deposited in a DNA barcode database such as the ISHAM reference database, or the Barcode of Life Data System (BOLD). In this attempt, DNA barcoding relies on universal genes that are ideally present in all fungi with the same degree of sequence variation. The interspecific variation, i.e., the variation between species, in the chosen DNA barcode gene should exceed the intraspecific (within-species) variation.
A fundamental problem in fungal systematics is the existence of teleomorphic and anamorphic stages in their life cycles. These morphs usually differ drastically in their phenotypic appearance, preventing a straightforward association of the asexual anamorph with the sexual teleomorph. Moreover, fungal species can comprise multiple strains that can vary in their morphology or in traits such as carbon- and nitrogen utilisation, which has often led to their description as different species, eventually producing long lists of synonyms. Fungal DNA barcoding can help to identify and associate anamorphic and teleomorphic stages of fungi, and through that to reduce the confusing multitude of fungus names. For this reason, mycologists were among the first to spearhead the investigation of species discrimination by means of DNA sequences, at least 10 years earlier than the DNA barcoding proposal for animals by Paul D. N. Hebert and colleagues in 2003, who popularised the term "DNA barcoding".
The success of identification of fungi by means of DNA barcode sequences stands and falls with the quantitative (completeness) and qualitative (level of identification) aspect of the reference database. Without a database covering a broad taxonomic range of fungi, many identification queries will not result in a satisfyingly close match. Likewise, without a substantial curatorial effort to maintain the records at a high taxonomic level of identification, queries – even when they might have a close or exact match in the reference database – will not be informative if the closest match is only identified to phylum or class level.
Another crucial prerequisite for DNA barcoding is the ability to unambiguously trace the provenance of DNA barcode data back to the originally sampled specimen, the so-called voucher specimen. This is common practice in biology along with the description of new taxa, where the voucher specimens, on which the taxonomic description is based, become the type specimens. When the identity of a certain taxon (or a genetic sequence in the case of DNA barcoding) is in doubt, the original specimen can be re-examined to review and ideally solve the issue. Voucher specimens should be clearly labelled as such, including a permanent voucher identifier that unambiguously connects the specimen with the DNA barcode data derived from it. Furthermore, these voucher specimens should be deposited in publicly accessible repositories like scientific collections or herbaria to preserve them for future reference and to facilitate research involving the deposited specimens.
Barcode DNA markers Internal Transcribed Spacer (ITS) – the primary fungal barcode
In fungi, the Internal transcribed spacer (ITS) is a roughly 600 base pairs long region in the ribosomal tandem repeat gene cluster of the nuclear genome. The region is flanked by the DNA sequences for the ribosomal small subunit (SSU) or 18S subunit at the 5' end, and by the large subunit (LSU) or 28S subunit at the 3' end. The Internal Transcribed Spacer itself consists of two parts, ITS1 and ITS2, which are separated from each other by the 5.8S subunit nested between them. Like the flanking 18S and 28S subunits, the 5.8S subunit contains a highly conserved DNA sequence, as they code for structural parts of the ribosome, which is a key component in intracellular protein synthesis.
Due to several advantages of ITS (see below) and a comprehensive amount of sequence data accumulated in the 1990s and early 2000s, Begerow et al. (2010) and Schoch et al. (2012) proposed the ITS region as primary DNA barcode region for the genetic identification of fungi.
UNITE is an open ITS barcoding database for fungi and all other eukaryotes.
Primers The conserved flanking regions of 18S and 28S serve as anchor points for the primers used for PCR amplification of the ITS region. Moreover, the conserved nested 5.8S region allows for the construction of "internal" primers, i.e., primers attaching to complementary sequences within the ITS region. White et al. (1990) proposed such internal primers, named ITS2 and ITS3, along with the flanking primers ITS1 and ITS4 in the 18S and the 28S subunit, respectively. Due to their almost universal applicability to ITS sequencing in fungi, these primers are still in wide use today. Optimised primers specifically for ITS sequencing in Dikarya (comprising Basidiomycota and Ascomycota) have been proposed by Toju et al. (2012).
For the majority of fungi, the ITS primers proposed by White et al. (1990) have become the standard primers used for PCR amplification (with the most common pairing being ITS1 + ITS4). These primers are as follows:
Forward primers: ITS1: ITS2: ITS5:
Reverse primers: ITS3: ITS4:
Advantages and shortcomings A major advantage of using the ITS region as molecular marker and fungal DNA barcode is that the entire ribosomal gene cluster is arranged in tandem repeats, i.e., in multiple copies. This allows for its PCR amplification and Sanger sequencing even from small material samples (given the DNA is not fragmented due to age or other degenerative influences). Hence, a high PCR success rate is usually observed when amplifying ITS. However, this success rate varies greatly among fungal groups, from 65% in non-Dikarya (including the now paraphyletic Mucoromycotina, the Chytridiomycota and the Blastocladiomycota) to 100% in Saccharomycotina and Basidiomycota (with the exception of very low success in Pucciniomycotina). Furthermore, the choice of primers for ITS amplification can introduce biases towards certain taxonomic fungus groups. For example, the "universal" ITS primers fail to amplify about 10% of the tested fungal specimens.
The tandem repeats of the ribosomal gene cluster cause the problem of significant intragenomic sequence heterogeneity observed among ITS copies of several fungal groups. In Sanger sequencing, this will cause ITS sequence reads of different lengths to superpose each other, potentially rendering the resulting chromatograph unreadable. Furthermore, because of the non-coding nature of the ITS region that can lead to a substantial amount of indels, it is impossible to consistently align ITS sequences from highly divergent species for further bigger-scale phylogenetic analyses. The degree of intragenomic sequence heterogeneity can be investigated in more detail through molecular cloning of the initially PCR-amplified ITS sequences, followed by sequencing of the clones. This procedure of initial PCR amplification, followed by cloning of the amplicons and finally sequencing of the cloned PCR products is the most common approach of obtaining ITS sequences for DNA metabarcoding of environmental samples, in which a multitude of different fungal species can be present simultaneously. However, this approach of sequencing after cloning was rarely done for the ITS sequences that make up the reference libraries used for DNA barcode-aided identification, thus potentially giving an underestimate of the existing ITS sequence variation in many samples.
The weighted arithmetic mean of the intraspecific (within-species) ITS variability among fungi is 2.51%. This variability, however, can range from 0% for example in Serpula lacrymans (n=93 samples) over 0.19% in Tuber melanosporum (n=179) up to 15.72% in Rhizoctonia solani (n=608), or even 24.75% in Pisolithus tinctorius (n=113). In cases of high intraspecific ITS variability, the application of a threshold of 3% sequence variability – a canonical upper value for intraspecific variation – will therefore lead to a higher estimate of operational taxonomic units (OTUs), i.e., putative species, than there actually are in a sample. In the case of medically relevant fungal species, a more strict threshold of 2.5% ITS variability allows only around 75% of all species to be accurately identified to the species level.
On the other hand, morphologically well-defined, but evolutionarily young species complexes or sibling species may only differ (if at all) in a few nucleotides of the ITS sequences. Solely relying on ITS barcode data for the identification of such species pairs or complexes may thus obscure the actual diversity and might lead to misidentification if not accompanied by the investigation of morphological and ecological features and/or comparison of additional diagnostic genetic markers. For some taxa, ITS (or its ITS2 part) is not variable enough as fungal DNA barcode, as for example has been shown in Aspergillus, Cladosporium, Fusarium and Penicillium. Efforts to define a universally applicable threshold value of ITS variability that demarcates intraspecific from interspecific (between-species) variability thus remain futile.
Nonetheless, the probability of correct species identification with the ITS region is high in the Dikarya, and especially so in Basidiomycota, where even the ITS1 part is often sufficient to identify the species. However, its discrimination power is partly superseded by that of the DNA-directed RNA polymerase II subunit RPB1 (see also below).
Due to the shortcomings of ITS-based primary fungal DNA barcoding, the necessity of establishing a second DNA barcode marker was expressed. Several attempts were made to establish other genetic markers that could serve as additional DNA barcodes, similar to the situation in plants, where the plastidial genes rbcL, matK and trnH-psbA, as well as the nuclear ITS are often used in combination for DNA barcoding.
Translational elongation factor 1α (TEF1α) – the secondary fungal barcode The translational elongation factor 1α is part of the eukaryotic elongation factor 1 complex, whose main function is to facilitate the elongation of the amino acid chain of a polypeptide during the translation process of gene expression.
Stielow et al. (2015) investigated the TEF1α gene, among a number of others, as potential genetic marker for fungal DNA barcoding. The TEF1α gene coding for the translational elongation factor 1α is generally considered to have a slow mutation rate, and it is therefore generally better suited for investigating older splits deeper in the phylogenetic history of an organism group. Despite this, the authors conclude that TEF1α is the most promising candidate for an additional DNA barcode marker in fungi as it also features sequence regions of higher mutation rates. Following this, a quality-controlled reference database was established and merged with the previously existing ISHAM-ITS database for fungal ITS DNA barcodes to form the ISHAM database.
TEF1α has been successfully used to identify a new species of Cantharellus from Texas and distinguish it from a morphologically similar species. In the genera Ochroconis and Verruconis (Sympoventuriaceae, Venturiales), however, the marker does not allow distinction of all species. TEF1α has also been used in phylogenetic analyses at the genus level, e.g. in the case of Cantharellus and the entomopathogenic Beauveria, and for the phylogenetics of early-diverging fungal lineages.
Primers