Introduction

Fungi are eukaryotic, heterotrophic organisms that grow as elongated, polarized cells (hyphae) or budding cells called yeasts, which reproduction occurs through meiotic and/or mitotic processes (Naranjo-Ortiz et al. 2019). The lifestyle of these microorganisms evolved independently in the Tree of Life (Hawksworth and Lucking 2018). Thus, this realm comprises one of the greatest species biodiversity among eukaryotic organisms (Willis 2018). Due to the complexity of their morphological structures and lifestyle that often depends on other organisms, the knowledge on distribution and diversity of fungi, particularly endophytic fungi, remains mostly unknown.

Endophytic fungi or fungal endophytes are a complex group of polyphyletic organisms (Chi et al. 2019). The constituent taxa are able to colonize intra/extracellularly the plant tissues (e.g. stem, flowers, leaves, fruits, roots) (Li et al. 2016a, b) during part or their entire life cycle without causing an apparent manifestation of disease in the tissue colonized (Bacon and White 2000).

In general, the methods employed to study endophytes are relatively simple, but involve complex steps that must be strictly followed to avoid misinterpretation (Fig. 1). After the plant material collection, the first step is the superficial disinfection of the sampled tissue to eliminate the epiphytic microbiota (Li et al. 2016a, b; Yao et al. 2017), followed by culture-dependent (Szűcs et al. 2018; Yao et al. 2019; Du et al. 2020; Fan et al. 2020) and/or independent techniques (Yao et al. 2019).

Fig. 1
figure 1

Flowchart for studying the diversity of endophytic fungi and their potential for producing bioactive compounds

Over the decades, the isolation of endophytic fungi was based on culture-dependent methodologies (Szűcs et al. 2018; Hamzah et al. 2018; Gong et al. 2019). For this, the plant samples are fragmented and distributed onto culture medium, incubated under controlled conditions and, after the fungal growth, the colonies are purified and the isolate identified using morphological characteristics or molecular analyses (Ibrahim et al. 2017; Pietro-Souza et al. 2017; Yao et al. 2017).

It is well known that culture-based methods do not reflect the real diversity of fungi in a niche (Fan et al. 2020; Lücking et al. 2020), since some of these microorganisms cannot be cultivated under laboratory conditions (Fan et al. 2020; Wu et al. 2019). In addition, identification of isolates using morphological features can provide misidentification, since fungi of different species have similar type of structure (Lücking et al. 2020), such as hyphae consisted of one or more cells surrounded by a tubular cell wall. Furthermore, some of these methods require modifications, including enriched-culture medium with a certain substrate to increase the fungal species diversity (Mourad et al. 2018), coupled with the use of molecular methods for identification of isolates (Noriler et al. 2018).

Despite advances in methodologies to study endophytic fungi, several barriers come up at all stages of commonly used protocols, requiring well-established strategies to avoid misinterpretation. Here, we discuss some methods mostly used to study endophytic fungi, emphasizing the main limitations derived from the different approaches, with alternatives to circumvent and/or minimize the existing barriers.

Collection and sampling of plant material

The first step to study endophytic microorganisms is the collection of plant material. However, there is no universal consensus on the adequate number of individuals and/or samples per individual that should be considered. It is strongly recommended a representative sampling according to the objectives proposed by the study. Considering the geographic factor, Yao et al. (2019) accounted eight individuals per plant species sampled and recommended a minimum distance of at least 50 m between each plant species in studies on diversity of endophytic fungi of different tropical mangroves. Wang et al. (2019) aimed to describe the three-years-old mycobiota of Zanthoxylum bungeanum and for these ten specimens were sampled excluding the spatial criteria. Fan et al. (2020) studied the endophytic mycobiome of the cultivated Huperzia serrata and three specimens were randomly chosen. Therefore, the number of plant individuals sampled depends on the objective of the research, but the quantity and quality of sampling plus biotic and abiotic factors that can interfere with the endophytic mycobiota must be considered (Fig. 2).

Fig. 2
figure 2

Biotic and abiotic factors influencing the assembly of the endophytic mycobiota. The community of endophytic fungi within the same plant species is highly dynamic and the richness and diversity of taxa can be influenced and determined by several biotic and abiotic factors. Image generated based on findings by David et al. (2016), Fernandes et al. (2018), Chi et al. (2019), Yao et al. (2019) and Wu et al. (2020)

Interestingly, most fungal genera and/or species are found in specific host plant species probably because the phytochemistry and micro/macronutrient contents present in the leaf tissue that acts by selecting the fungal species for colonization (Arnold et al. 2003).

Regarding sampling by the individual, a number of studies (Li et al. 2016a, b; Yao et al. 2017, 2019; Szűcs et al. 2018; Du et al. 2020; Fan et al. 2020) suggest > 5 samples per individual, but extraordinary cases should be analyzed. Biotic and abiotic factors that can cause interference on the endophytic fungi community harbored by the sampled individual should be accounted. In some cases, it is necessary to have a composite and representative sample (Fig. 3).

Fig. 3
figure 3

Leaf sampling strategy to study endophytic fungi. a The plant microbiome is subject to a mosaic of resource availability and environmental conditions that tend to vary on a microenvironmental scale, and are influenced by biotic and abiotic factors. Therefore, these factors should be considered during the sampling process. The red square refers to the specimen to be sampled; the blue and red arrows indicate the influence of biotic factors (inter- and intra-species interactions) that, for example, result in a lower incidence of sunlight in a specific portion of the tree crown to be sampled; the yellow arrows indicate the influence of abiotic factors (e.g. incidence of sunlight, availability of resources). b Experimental design of selection of individuals to be sampled to obtain a representative sample. c Leaf tissue sampling scheme of the lower, middle and upper leaves. In d, the top view of the sampled specimen is shown, and the black arrows indicate that the sampling should be carried out in different regions. In e, a tissue fragmentation scheme is presented to obtain a representative sample for the study of endophytic fungi. Image designed from the collection strategy by Wu et al. (2020) and Kluting et al. (2019), Denney et al. (2020), and Hurtado-McCormick et al. (2021) on population variations in microenvironments. Created with BioRender.com

Environmental conditions are responsible for species distribution not only in a macroscale, but also in a microscale, and even within a local site as shown in Fig. 3A. These variations within microenvironments expose natural population such as fungi to mosaics of resource availability and different abiotic (e.g. temperature, humidity, light intensity, availability of nutrients) and biotic conditions (e.g. interaction with nodes of the same species and other species) (Kluting et al. 2019; Denney et al. 2020; Hurtado-McCormick et al. 2019). Therefore, these factors must be taken into account during sampling in endophytic fungi studies.

After collecting the plant material, the samples should be placed in a sterile plastic container or bag, transferred to a cool box at a controlled temperature, and preferably processed within 24 h after sampling (Lundberg et al. 2012; Sharma et al. 2016; Li et al. 2016a, b; Ibrahim et al. 2017, 2021; Pietro-Souza et al. 2017; Yao et al. 2017, 2019; Szűcs et al. 2018; Hamzah et al. 2018; Gong et al. 2019; Dhayanithy et al. 2019; Thi et al. 2019; Jayatilake and Munasinghe 2020; Du et al. 2020; Fan et al. 2020; Chowdappa et al. 2020) under aseptic conditions.

Surface disinfection of plant material

Since the discovery of taxol in an endophytic fungus associated with Taxus spp. (Stierle et al. 1993), endophytic fungi have been drawing the attention of the scientific community either by the species diversity and/or the capacity to synthesize bioactive molecules. Nevertheless, several methodological obstacles appear when studying endophytes, including the complete removal of microorganisms that make up the epiphytic microbiota. Effective methods of surface disinfection to remove the epiphytic microbiota are mandatory first steps for studying endophytic microorganisms, which includes chemical or physical procedures.

The chemical sterilization of the sample surface is the most used method for removing microorganisms from the rhizoplane and many protocols are available (Table 1). In most cases, they consist of three basic steps: (1) submersion of the tissue in ethanol; (2) immersion of the tissue in the main sterilizing agent, and (3) successive washing with distilled water previously autoclaved. The main sterilizing agents used in rhizoplane disinfection are sodium hypochlorite (NaOCl) (Yao et al. 2019) and mercury chloride (HgCl2) (Du et al. 2020). Other sterilizing agents can also be used such as hydrogen peroxide (H2O2), paraquat, or 1% peroxyacetic acid (CH3COOOH) in 30% ethanol (Sieber, 2002). According to some authors (Píetro-Souza et al. 2017; Ibrahim et al. 2021), the samples should be washed in running tap water for some times to remove debris, dust, other particles, and main epiphytes (e.g. epiphytic bacteria, yeasts and filamentous fungi) prior to the superficial disinfection of the material to enhance the process.

Table 1 Botanical species, tissue type, methods used for surface disinfection, and culture medium used to study endophytic fungi

The physical surface disinfection can be mediated by sonication or ultraviolet (UV) light. In the case of sonication, the plant tissue should be previously washed in tap water and placed in a sonicator or sonication bath containing buffer solution where the sound waves will be propagated, promoting the removal and lysis of microbial cells from the sample surface (Guzmán et al. 2020). Although this method of superficial disinfection is useful for removing epiphytes present in the sample surface (Lundberg et al. 2012), the sound frequency used, application time and tissue characteristics may interfere in the method efficiency. Surface disinfection mediated by UV light also requires that the sample be washed repeatedly in tap water beforehand to remove as much dirt as possible from the plant tissue prior to expose to UV rays for 20 min (Sarsaiya et al. 2020). In general, UV light kills cells by damaging their DNA. However, even the highest-energy UV rays do not have enough energy to penetrate deeply into the plant tissue, so their main effect occurs on the plant tissue surface. Another physical method used to remove epiphytes is plant tissue buckling (Saldierna Guzmán et al. 2020), but this methodology is not widely used due to the various biases associated.

The choice of the disinfection protocol is a critical step to study endophytic microorganisms and the selected methodology should be able to completely remove the epiphytic microbiota without interfering with the endophytic microbiota. Therefore, the chosen method should considered the characteristics of the analyzed tissue/plant species. Burgdorf et al. (2014), for instance, evaluated the effectiveness of two surface disinfection methods, a chemical and a physical methods, on leaves of Triticum aestivum. For this, the leaves were artificially sprayed with Saccharomyces cerevisiae and then individually submitted to both treatments for the removal of epiphytes. The physical treatment was mediated by sonication for 5 min in a Biosonic sonication bath (Colténe/Whaledent, Altstätten, Switzerland), and successive washings with sterile distilled water. For the chemical treatment, the samples were consecutively immersed in 95% ethanol for 5 s, 0.5% sodium hypochlorite for 2 min, 70% ethanol for 2 min and successive washings with sterile distilled water. The rinse of the last wash from both procedures was seeded on Potato Dextrose Agar (PDA) medium and analyzed molecularly by PCR (Polymerase Chain Reaction). As result, the authors demonstrated that the physical treatment was the most consistent and effective method for the removal of microorganisms from the surface of host plants compared to the chemical treatment, and more reproducible.

Guzmán et al. (2020) evaluated the effectiveness of four surface disinfection methods in different tissues of two different plant species, namely Pinus contorta and Populus fremontii. The evaluated protocols consisted of three chemical methods using different sterilizing agents individually (100% ethanol for 1 min; 8.25% sodium hypochlorite for 5 min, and 30% hydrogen peroxide for 1 min) and one physical method consisted of sonication at 40 kHz for 10 min. At the completeness of each procedure, the samples were successively washed with sterile distilled water and the final rinse was recovered. PCR amplification of the final rinses from samples sterilized using the chemical methods gave negative results for all the treatments employed. Moreover, after superficial disinfection, seeding by printing the analyzed tissues in Lysogeny Broth (LB) culture medium, primarily used for bacterial growth, were also negative for all treatments. This showed the efficiency of the chemical treatments employed in removing the epiphytic microbiota. However, scanning electron microscopy (SEM) analysis demonstrated the presence of hyphae, yeast, and bacteria on the leaf surface of both species (Pinus contorta and Populus fremontii) when sonication (physical method) was employed. In addition, the leaves of P. fremontii were not damaged by chemical agents, differently of treatments with hydrogen peroxide and sodium hypochlorite in P. contorta leaves, which damage could interfere with the endophytic mycobiome. These results highlight the fact that different plant species require different methodologies to efficiently remove the epiphytic microbiota.

Some authors suggest other protocols of superficial disinfection in addition to those commonly used. For example, Waheeda and Shyam (2017) proposed a disinfection method that includes washing in 3.15% calcium hypochlorite for 10 min, followed by washing in 10% sodium hydrogen carbonate for 15 min, and a final washing in sodium azide 1% for 2 min.

Similar to plant tissue, fungal species are also not equally sensitive to sterilizing agents, and this fact should be considered when surface disinfection methods were chosen. According to Reissinger et al. (2001), ethanol and hypochlorite as well as the exposure time commonly used in surface disinfection protocols are ineffective for the removal and unfeasibility of Chaetomium globosum ascospores present in Hordeum vulgare leaves. Similarly, Holdenrieder (1989) tested a three-step protocol [(i) immersion in ethanol; (ii) immersion in hypochlorite, and (iii) washing with sterile distilled water] for superficial disinfection of Picea abies roots artificially contaminated with conidia of Penicillium sp. or Cylindrocarpon destructans. The authors demonstrated that the spores of C. destructans were eliminated after 60 min, while those of Penicillium remained viable for more than two hours. Penicillium conidia were only exterminated when hydrogen peroxide was used without prior submersion in ethanol.

The effectiveness of disinfection methods should be essentially evaluated prior when studying endophytic microorganisms, including cheaper or more expensive approaches. This step should be performed properly to validate the adopted procedure. Some authors used the sowing especially in PDA culture medium of the water recovered from the last rinse to validate the effectiveness of disinfection methods (Yao et al. 2017; Szűcs et al. 2018; Fan et al. 2020; Ibrahim et al. 2021). In this case, centrifugation should be used to concentrate the microorganisms present in the last rinse water, followed by the discard of supernatant, and sow of the centrifuged sediment in the culture medium (Sieber 2002). Another alternative is to print the tissue in the culture medium after surface disinfection (Szűcs et al. 2018; Dhayanithy et al. 2019; Jayatilake and Munasinghe 2020.

The effectiveness of the disinfection method can also be verified using DNA-based techniques such as PCR from the water recovered of the last rinse (Burgdorf et al. 2014; Guzmán et al. 2020). This technique is widely used to amplify a segment of DNA of interest. However, depending on the disinfection method employed, false-positive results may occur because the sterilizing agents can promote the cell death of epiphytic microorganisms without degradation of nucleic acids (Burgdorf et al. 2014). The combination of two or more techniques to assess the effectiveness of surface disinfection methods can ensure greater efficiency. The use of scanning electron microscopy (Guzmán et al. 2020) in association with seeding of water from the latter rinsing in culture medium may be an interesting approach.

Thus, the characteristics of the analyzed tissue, the composition of the epiphytic microbiota, the toxicity of the sterilizing agents, and methods used to validate the surface disinfection protocol should be taken into consideration in studies on fungal endophyte communities. Therefore, the performance of previous tests with different sterilization protocols, the previous description of the epiphytic microbiota, and combination of methodologies to validate the surface disinfection method employed are extremely necessary to avoid misinterpretation about the composition of the culturable endophytic mycobiota.

Culture medium for growth of endophytic fungi

There are many culture medium protocols available to recover endophytic fungi, with PDA as the most frequently used (Table 1). Other growth media such as Malt Extract Agar (Li et al. 2016a, b), Agar containing Murashige and Skoog (MS) vitamins and sucrose (Lundberg et al. 2012), Hagem Minimal Medium (Khan et al. 2016), Czapek medium (Chand et al. 2020), Tryptone Soybean Agar, Tryptone Bovine Extract Agar, and Luria–Bertani medium (Gong et al. 2019), which is also known as LB broth, Lysogeny broth or Luria broth, can be efficiently used to recover endophytic fungi from plant samples. Although several culture media are used for recovery and cultivation of endophytic fungi, they must have a slightly acidic pH range (5.8–6.0 pH) (Lundberg et al. 2012; Li et al. 2016a, b; An et al. 2020). Depending on the purpose of the study and the characteristics of the habitat (i.e. host plant species), the medium commonly used can be nutritionally supplemented and/or replaced by specific medium to better achieve the objectives of the study (Pietro-Souza et al. 2017; Hamzah et al. 2018; Hamzah et al. al. 2018).

In the case of a study focused on description of the diversity of easily cultivable endophytic fungi, a single culture medium such as PDA (Ibrahim et al. 2021) can be used. To increase the number of retrieved taxa, two or more different culture media can be employed in addition to PDA (Man et al. 2015; Gong et al. 2019). Supplementation of culture medium with plant material from the host species can also be used to improve the recovery of endophytic fungi (Mourad et al. 2018). In this case, the plant material should be properly processed and autoclaved in order to avoid contamination. The use of a water agar medium is also a great option to be considered in studies on diversity of fungal endophyte, since this medium is poor in nutrients and prevents fast-growing fungal species from standing out over slower growing ones (Stone et al. 2004).

Last but not least, the use of antibiotics to inhibit endophytic bacteria is essential in studies on diversity of cultivable fungal endophytes (Yao et al. 2017). The use of one or more broad-spectrum antibiotics such as sodium benzylpenicillin (Li et al. 2016a, b), chloramphenicol (Hamzah et al. 2018), streptomycin (Ibrahim et al. 2016), gentamicin (Ibrahim et al. 2021) and tetracycline (Chowdappa et al. 2020) is strongly recommended.

Cultivation conditions, isolation and culture purification

According to the objectives of the study, the cultivation conditions of endophytic fungi in artificial culture medium are diverse. The temperature commonly used for incubation of Petri dishes usually ranges from 25 ºC (Li et al. 2016a, b; Du et al. 2020) to 28 ºC (Yao et al. 2017), with period of incubation between 3 and 20 days (Li et al. 2016a, b; Yao et al. 2017; Du et al. 2020), which can be extended for up to 6 weeks if necessary (Yao et al. 2017). In the case of cultures that require a longer period of incubation or are kept under environmental conditions outside greenhouses or BOD incubators (Bio-Oxygen Demand), it is necessary to seal the plates with Parafilm to minimize risks of contamination and/or drying of the culture medium (Szűcs et al. 2018; Ababutain 2021). Cultivation can also be carried out with or without incidence of light (Duan et al. 2019; Agbessenou et al. 2020). In this sense, the endogenous circadian clocks present in both prokaryotes and eukaryotes provide the machinery by which they keep in synchrony with the external environment. In fungi, the clock has been shown to control daily rhythms (Dunlap and Loros 2017) in spore development and liberation, supporting their survival.

Plates seeded with fragmented plant tissue should be monitored daily. Fragmentation of plant tissue can be achieved with the aid of a scalpel, eyelet pliers or scissors, then printed on a culture medium and incubated. Although fragments with 5 mm × 5 mm are commonly used, other dimensions have also been employed (Table 1). As fungal growth frequently occurs from the edges of the inoculated plant tissue, the isolation is recommended by collecting the hyphae from the edges of the fungal colonies, followed by seeding in fresh culture medium with or without antibiotics (Li et al. 2016a, b; Ibrahim et al. 2017; Yao et al. 2017; Ababutain 2021). This step should be repeated several times until a monoculture of a endophytic fungus strain is reached with a uniform colony (Ibrahim et al. 2017). Then, the monosporic and/or hypha tip purification must be performed (Supaphon et al. 2013; Duan et al. 2019) for later deposit in a mycological culture collection and/or molecular identification of the isolates.

Morphological or morphotype grouping, and storage of endophytic fungi colonies

After the purification process, the fungal isolates are grouped based on their macro- and micromorphological characteristics (Dhayanithy et al. 2019; Jayatilake and Munasinghe 2020). For this, macroscopic vegetative characteristics that includes color, texture, topography, and diffuse pigmentation of structures, and color and topography of the colony’s dorsum, are considered, and also analyses of their microscopic reproductive structures, using microculture and/or sporulation methods (Dos Banhos et al. 2014). The macroscopic classification allows grouping the isolates into morphotypes, while the microscopic classification allows the morphological classification of the isolates. However, the arrangement in morphotypes and/or morphological features does not reflect the real phylogeny of the organisms, even more considering non-sporulating fungal endophyte species (Li et al. 2016a, b; Yao et al. 2017; Du et al. 2020) that requires robust tools such as molecular methods to achieve the taxonomic classification of the isolates.

To avoid losses of isolated colonies caused by contamination, re-isolation on PDA prepared inclined cryotubes and subsequent storage in a cold chamber at 4º C and freezer – 80º C is recommended (Li et al. 2016b; Wang et al. 2019). For long-term preservation of cultures, mycelia and spores should be transferred to 20% glycerol in ultrapure distilled water and stored at − 80° C (Wang et al. 2019). Another alternative for long-term preservation of monocultures is to transfer small pieces of pre-inoculated culture medium to sterile microtube containing 30% glycerol (v/v) and sterile rice medium, followed by incubation at 25 ºC. After a fungal growth is observed, the microtubes should be transferred to a cold room (4 °C), and storage in a freezer at − 20 ºC, respectively (Ibrahim et al. 2017).

Molecular identification of isolated endophytic fungi

Despite the development of different methods to promote fungal sporulation (Taylor et al. 1999; Sun et al. 2008a), more than 50% of endophytic fungal isolates do not sporulate in cultures (Mycelia sterilia) (Wang 2008; Wang and e Guo LD 2007; Sun et al. 2008b, 2011), which makes the conventional classification based on observation of the reproductive structures of fungi impossible. Furthermore, even for those isolates capable of sporulating, a polyphasic approach that includes molecular analysis plus observation of reproductive structures is strongly recommended to achieve a precise fungi identification (Sun and Guo 2012).

There are many protocols used for molecular characterization of endophytic fungal isolates (Table 2). Briefly, this approach requires the following steps: (I) genomic DNA (gDNA) extraction; (II) PCR amplification of conserved DNA sequences (rDNA) using specific or universal oligonucleotide primers forward and reverse; (III) sequencing of PCR products; (IV) sequence processing and comparison with related sequences deposited in DNA databases; (V) data interpretation and phylogenetic tree reconstruction (Hamzah et al. 2018; Thi et al. 2019; Jayatilake and Munasinghe 2020). In addition, DNA cloning using cloning vectors can also be employed prior to sequencing (Atsatt and Whiteside 2014).

Table 2 Main steps and PCR protocols for molecular detection of endophytic fungi using culture-dependent methods

DNA extraction is a critical step to achieve the isolate characterization and should be carried out from monocultures well established previously (Li et al. 2016a, b; Wang et al. 2019). In general, it begins with the fungal cell lysis by maceration in liquid nitrogen (Li et al. 2016a, b) or cell lysis kits (Ababutain 2021). After cell disintegration, several homemade protocols such as the standard phenol–chloroform method (Ibrahim et al. 2017; Tang et al. 2020) and kits commercially available (Du et al. 2020; Ababutin 2021) can be employed. To avoid biases, it is strongly recommended the utilization of commercial kits according to the manufacturer’s instructions (Ababutin 2021). However, they can be adapted eventually to achieve better results (Du et al. 2020) when necessary. Subsequently, the extracted DNA should be purified, recovered and analyzed on 1% agarose gels after electrophoresis in 1 × TBE running buffer, for instance, and storage at − 20° C. The extracted DNA quality and integrity should be ensured prior to further analyses (Ibrahim et al. 2017; Tang et al. 2020).

PCR amplification of the extracted genomic DNA targeting specific DNA regions (Barth Reller et al. 2007) can be performed using different PCR conditions as shown in Table 2. The PCR technique is still widely used to study endophytic fungi from monocultures prior to sequencing, generating PCR products (i.e. amplicons) of target sequences. PCR reactions require PCR buffers, cofactors, deoxynucleotides (dNTPs—DNA building blocks), specific primers to the DNA target region (e.g. ITS and 18 rDNA), thermostable DNA polymerase, and the genomic DNA as template (Table 2). All components should be in a proper concentration following the manufacturer’s instructions. The DNA amplification is performed in a thermocycler through numerous cycles of temperature variation. Briefly, PCR thermal cycling conditions comprise three steps basically: double-stranded DNA denaturation, primer annealing, and extension/elongation (Chowdappa et al. 2020; Ibrahim et al. 2021) (Table 2). It is important to emphasize that the annealing temperature depends on the primers design (Sarsaiya et al. 2020), especially concerning the GC content. Finally, the amplicons obtained are purified and visualized on gel electrophoresis to confirm the PCR amplification, and then subjected to sequencing (Ibrahim et al. 2017; Tang et al. 2020; Du et al. 2020).

Despite advances in DNA sequencing technologies, sequencing of amplicons from fungal endophyte using Sanger sequencing is still the method largely used (Li et al. 2016a, b; Ibrahim et al. 2017; Yu et al. 2018; Du et al. 2020; An et al. 2020; Tang et al. 2020; Ababutain 2021). After sequencing completion, the nucleotide sequences obtained are treated in appropriate softwares for noise removal and quality scores (Phred > 20). Then, the newly digital sequences are submitted to DNA public databases, and compared with previous deposited sequences for identification (Ibrahim et al. 2017; Tang et al. 2020; Du et al. 2020). Pairwise comparisons are usually carried out using BLASTn tool (Basic Local Alignment Search Tool) from the NCBI (National Center for Biotechnology Information) (Li et al. 2016a, b; Ibrahim et al. 2017; Yu et al. 2018; Du et al. 2020; An et al. 2020; Tang et al. 2020). Sequence similarity of 94% allows classification at the genus level and > 97% at the species level for constitutive gene (Li et al. 2016a, b). Other bioinformatics tools and databases can also be employed for the taxonomic identification of endophytic fungi (Ababutin 2021). Nevertheless, fungal identification at the species level is hardly achieved using a single DNA region (Noriler et al. 2018) due to highly conserved sequences. Thus, analysis of DNA barcodes with hypervariable regions may improve the identification of endophytic fungi at the species level.

Finally, the taxonomic position of fungal isolates can be accessed by phylogenetic tree reconstructions. Phylogenetic analysis may extend the knowledge on geographic occurrence of the isolates and host plants especially in studies on diversity of endophytic fungi. Some tools to assess the phylogenetic relationships among endophytic fungi taxa can be found in Li et al. (2016a, b) and An et al. (2020).

Barcodes for identification of endophytic fungi and related primers

The nuclear ribosomal internal transcribed spacer (ITS) region is the most used DNA molecular marker employed to identify fungi (Schoch et al. 2012; Koljalg et al. 2013) (Fig. 4). Coupled with the 18S rRNA and 28S rRNA gene sequences, the ITS region is present in several copies in the fungal genome and other eukaryotes. Moreover, the ITS region is considered as a primary DNA barcode region for identification of fungi taxa (Kiss 2012). Barcode sequences are short universal and standardized DNA sequences easily PCR amplified (450–800 bp), and commonly found in a taxonomic group, but highly divergent at the species level, which enables accurate and rapid identification (Blackwell 2011). These sequences should be unique and stable within each species (Schoch et al. 2012). To be considered a good barcode, the interspecific variation must exceed the intraspecific variation to allow distribution of distances among different taxa (Hebert et al. 2003). As the ITS region is largely employed as a primary barcode region for fungal identification, several universal oligonucleotide primers have been designed (Lücking et al. 2020) (Table 3). The average size of amplicons generated by PCR amplification from this region either for studies using Sanger sequencing or high-performance sequencing generally ranges from 450 to 800 bp (Li et al. 2016a, b; Ibrahim et al. 2017; Yu et al. 2018; Du et al. 2020; An et al. 2020).

Fig. 4
figure 4

Ribosomal RNA nuclear genes and ITS regions. The ITS region includes the ITS1 and ITS2 sequences separated by the 5.8S rRNA gene, which are located between the 18S rRNA (SSU) and 28S rRNA (LSU) genes (Blackwell et al. 2011). The small arrows observed from the ITS region magnification demonstrate the annealing position of primers used for PCR amplification from this region (White et al. 1990; Gardes and Hester 1993; O'Donnell, 1993; Gerrits van den Ende and de Hoog, 1999; Beguin et al. 2012)

Table 3 Oligonucleotide primers for universal fungal barcodes to taxonomic identification

According to Schoch et al. (2012), the ITS region quite supplies the requirements for fungal species discrimination in the vast majority of cases. When associated with DNA sequencing for molecular identification of endophytic fungi, the ITS region is undoubtedly a good choice as a DNA barcode (Li et al. 2016a, b; Ibrahim et al. 2017; Yu et al. 2018; Du et al. 2020; An et al. 2020; Tang et al. 2020). However, this region may result in low resolution for fungal species-level differentiation (Noriler et al. 2018) particularly for the taxonomic classification of Aspergillus spp., Alternaria spp., Cladosporium spp., Colletotrichum spp., Fusarium spp. and Penicillium spp. (Lücking et al. 2020). Analyses of the ITS sequence accuracy showed that for 6 and 17% of yeast and filamentous fungi species, respectively, the classification at a species level was not possible from ITS as a barcode region (Vu et al. 2016, 2019). Furthermore, within the genome of a single fungal cell there may be numberless copies of ITS sequences (Kiss 2012), with intergenomic variability (Smith et al. 2007; Weitemier et al. 2015). Such variations can interfere with the PCR amplification (Xu, 2016) due to primer mismatching, jeopardizing the study. Therefore, in cases where precise identification is not possible through the ITS region, other DNA regions can be employed to allow a reliable identification (Lücking et al. 2020). In these situations, the use of secondary barcodes is strongly recommended to achieve a better discrimination among taxa. The D1/D2 regions, for instance, are well employed for species-level classification of yeasts (Scorzetti et al. 2002).

In addition, protein-coding genes employed as secondary barcodes are useful to identify fungal species. Some authors have reported their efficacy over the ITS region (Frisvad et al. 2004; Seifert 2009; Sarmiento-Ramirez et al. 2014). For the Ascomycota phylum, these genes are considered better than rRNA genes to solve close relationships at different taxonomic and phylogenetic levels (Schoch et al. 2009). Among the protein-coding genes, the β-tubulin II gene (TUB2), cytochrome c oxidase subunit I (COX1) and subunit II (COX2), DNA-directed RNA polymerase II largest subunit (RPB1) and second largest subunit (RPB2), translational elongation factor 1α (TEF1), DNA topoisomerase I (TOP1), and phosphoglycerate kinase (PGK) have been proposed (Lücking et al. 2020; Crous et al. 2020). To classify Penicillium spp. and Aspergillus spp. at the species level, the use of β-tubulin was reported by Frisvad et al. (2004) and Samson et al. (2014), respectively. Similarly, Short et al. (2013) employed the TEF1 sequence-based diversity for Fusarium spp.

The oligonucleotide primers available for PCR amplification of primary barcode regions should be specific and capable of amplifying a sufficient sequence length for further analyses (Lücking et al. 2020). The ITS1, ITS2, and ITS4 primers are commonly used to reach this goal because they produce amplicons of 400–800 bp (Li et al. 2016a, b; Ibrahim et al. 2017; Yu et al. 2018; An et al. 2020; Du et al. 2020; Tang et al. 2020). Other specific ITS primers were designed to avoid cross-reactions with animal and/or plant DNA such as ITS1F (Gardes and Hester 1993), IT2 (Beguin et al. 2012), NL4b (O'Donnell 1993), V9D, V9G and LS266 (Gerrits van den Ende and de Hoog 1999). The primers designed for secondary barcode regions should be universal and able to amplify a reliable region to allow the fungal strain identification when primary barcode regions possess insufficient resolution (Frisvad et al. 2004; Stielow et al. 2015). The EF1-1018F (Al33F)/EF1-1620R (Al33R) and EF1-1002F (Al34F)/EF1-1688R (Al34R) primer sets designed for TEF1α (Stielow et al. 2015) are good examples. Some primer sets designed for universal fungal barcodes to different loci are summarized in Table 3.

Secondary structure prediction of the ITS rRNA region and molecular phylogeny: an integrated approach for a better speciation of endophytic fungi

Novel species of non-sporulating fungi of the genus Muscodor with bio-fumigant and bio-preservative activities have been successfully identified using ITS sequences (Kapoor et al. 2018). For this, the secondary structural prediction of the ITS1, 5.8S, and ITS2 regions presented a vital accuracy for fungal speciation. In addition, the ITS hypervariable nature hinders the design of stable phylogenetic trees (Kapoor et al. 2018). Therefore, the ITS structural analysis along with the ITS primary sequence information have been provided new insights for identification of fungal species.

As ITS sequence is a popular primary fungal barcode marker, the nuclear ribosomal internal transcribed spacer 1 (ITS1) region has used in studies on anaerobic fungi (phylum Neocallimastigomycota) in particular. Due to problems involving the reconstruction of stable phylogenetic trees, Koetschan et al. (2014) proposed a common core secondary structure of ITS1 from anaerobic fungi employing a Hidden Markov Model-based ITS1 sequence annotation and a helix-wise folding approach. The authors integrated the additional structural information into phylogenetic analyses and presented an automated sequence-structure-based taxonomy of ITS1 to anaerobic fungi. According to the authors, this methodology is possible to be applied to other fungal groups, probably including endophytic fungi.

In addition, exploring the secondary structure of the nuclear ribosomal internal transcribed spacer 2 (ITS2) has been a promising approach in fungal species delimitation. Sundaresan et al. (2019) investigated the presence of insertion/deletion (INDELs) and substitutions in fungal delineation at species level by the analysis of ITS2 molecular morphometrics effectiveness in species delimitation of Ascomycota. As result, the authors reported that the ITS2 molecular morphometric analysis is an efficient third dimensional study of fungal species boundaries, which may help to alleviate the challenges in molecular characterization.

An endophytic fungus with considerable amount of antimicrobial activity isolated from healthy leaf tissues of Houttuynia cordata Thunb., an ethnomedicinal plant of North East India, was previously identified as Colletotrichum sp. and then characterized by its genomic ITS rDNA and ITS2 sequences (Talukdar et al. 2021). According to phylogenetic analyses, the isolate was clustered with Colletotrichum coccodes. However, to confirm the endophytic lifestyle of the isolate, the ITS2 RNA secondary structure study was undertaken, and showed differences in the folding pattern as well as in motif structures compared to those of pathogenic C. coccodes, since species of Colletotrichum are also reported as plant pathogens. The consensus ITS2 secondary structure of the pathogenic C. coccodes consisted of a conserved core bulge of which radiating three major helixes (H1, H2, and H3), while the ITS2 secondary structure of the endophytic C. coccodes as well as for the author’s isolate comprised one incomplete (H1) and four complete helixes (H2, H3, H4, and H5) (Fig. 5).

Fig. 5
figure 5

ITS2 secondary structures of C. coccodes with different life styles showing helixes emerging from a central bulge, A pathogenic, B endophytic and C author’s isolate (MN128230.1). Motifs in the structures are indicated by arrow. Retrieved from Talukdar et al. (2021)

Methods for studying the chemical diversity of endophytic fungi

Endophytic fungi are known to synthesize a wide range of bioactive molecules with the most diverse applications (Toghueo 2019; Sagita et al. 2021; Zheng et al. 2021). Currently, there are many methods available for evaluation, characterization and isolation of compounds produced by endophytic fungi. Thus, the choice of the best methodology depends on the objectives proposed by the study. To study secondary metabolites, for instance, purified isolates are usually subcultured in liquid medium and incubated for a time course of 7–14 days at 25–28 ºC, under constant agitation (160–180 rpm/min) (Li et al. 2016a, b; An et al. 2020; Tang et al. 2020; Wei et al. 2020). Then, the fermented broth is treated to cause cell lysis, filtered and washed one or more times with different organic solvents, followed by vacuum volatilization or in a water bath to obtain the extracted metabolites, which are separated, identified and tested according to the study (Li et al. 2016a, b; Tang et al. 2020). Moreover, previous screening of fungal endophytes for metabolites production are recommended prior to chemical analysis (Luo et al. 2015; Sagita et al. 2021), including PCR detection of specific genes involved in the biosynthesis of metabolites.

Finally, the identification of fungal metabolites can be performed using different approaches. With demystification of omic technologies, metabolomics has been implemented to rapidly expand our understanding of cellular processes. Metabolomics is the large-scale study of end products of all cellular processes, providing a direct functional readout of cellular activity and physiological status at a set time. Thereby, high-performance gas or liquid chromatography (GC and LC, respectively) coupled with mass spectrometry (MS) and nuclear magnetic resonance (NMR) spectroscopy (chromatographic profile-guided experiment) are methods largely employed nowadays to analyze bioactive compounds produced by endophytic fungi (Silva et al. 2018; Teimoori-Boghsani et al. 2020; Wei et al. 2020; Sagita et al. 2021).

There is currently little information on endophytic fungi at the omics levels concerning their potential for production of secondary metabolites. Calcarisporium arbuscula, an endophytic fungus from the fruit bodies of Russulaceae, produces a variety of secondary metabolites with anti-cancer, anti-nematode and antibiotic activities (Cheng et al. 2020). The whole genome of C. arbuscula NRRL 3705 was sequenced using the PacBio Sequel platform and Illumina NovaSeq PE150, and transcriptomic survey was performed in order to determine its potential for biosynthesis of metabolites. The genome component prediction included the prediction of coding genes, repetitive sequences and non-coding RNAs, and gene clusters for biosynthesis of metabolites by the web-based software antiSMASH database (antibiotics and Secondary Metabolite Analysis 4.0) through the genome mining approach. According to the authors, a comprehensive survey of the genome and transcriptome of this endophytic fungus will allow to understand its capacity to biosynthesize secondary metabolites as well as generate pathways for its exploitation.

Cultivation-independent methods for the study of endophytic fungi

Recent advances on development of culture-independent methods have allowed the study of microbial communities in a better comprehensive way (Rojas et al. 2019; Ruiz Gómez et al. 2019), since these tools provide a closer detection of the real diversity of microorganisms from environmental samples directly. In addition, these approaches may lead to identification of novel taxa, which are uncultivable through conventional culture-dependent methods and artificial conditions (Forbes et al. 2017).

Cultivation-independent methods have been widely used to study the diversity of fungal communities in a variety of habitats, including soils (Pang et al. 2019), plants (Chi et al. 2019), and hospital environment (Donovan et al. 2018). These methodologies are considered extremely useful to deeply understand the genetic and functional diversity of fungal species (Staniek et al. 2009; Sagita et al. 2021; Chi et al. 2019).

Using culture-dependent in combination with culture-independent (DNA metabarcoding) methods, Chi et al. (2019) accessed the diversity of fungal taxa in Acanthus ilicifolius. Through the culture-dependent method, 203 fungal strains were isolated and classified into 30 species, while 111 operational taxonomic units (OTUs) were identified using DNA metabarcoding, which demonstrates the advantages of cultivation-independent methods compared to cultivation-dependent methods in environmental studies of microbial diversity.

Staniek et al. (2009) designed specific primers to study taxol-coding genes in endophytic fungi without the need of cultivation-based procedures mostly associated with methodological and observational biases. Among the culture-independent techniques employed to study endophytic microbial diversity are Denaturing Gradient Gel Electrophoresis—DGGE (Bogas et al. 2015) and DNA metabarcoding (Chi et al. 2018). However, these methodologies are based on PCR amplification of rDNA ITS1-5.8S-ITS2 sequences with specific primers (Chi et al. 2018) that may exclude non-target fungal DNA from environmental samples (O'Donnell 1993). For analysis of PCR products subjected to sequencing and precise species characterization, several pipelines were developed from ITS data, including Plutof, Clotu, PIPITS and CloVR-ITS (Abarenkov et al. 2010; Kumar et al. 2011; White et al. 2013; Gweon et al. 2015). Other bioinformatics pipelines for performing microbiome analysis that includes BioMaS, Mothur and Qiime have also been used to analyze fungal ITS amplicons (Schloss et al. 2009; Caporaso et al. 2010; Fosso et al. 2015) derived from next-generation sequencing (NGS).

Despite advances in culture-independent DNA-based techniques for studying microbial communities, these methods are still under improvement due to methodological limitations. DNA metabarcoding, for instance, usually generates short DNA fragments (~ 250 bp), making unfeasible the taxa classification at genus and/or species levels (Chi et al. 2019). Anyway, limitations should be considered when choosing an appropriate research methodology.

In the last 2 decades, omics-based approaches and systems biology have emerged, and revolutionary concepts on molecular biology studies are available. Omics involves high-throughput analyses on genomics, transcriptomics, proteomics, metabolomics, lipidomics, and others, which can be integrated through robust systems biology, bioinformatics, and computational tools to study mechanisms of interaction and biological functions of organisms. These approaches are also employed to collective analyses of environmental microbial communities (e.g. metagenomics, metatranscriptomics, and metaproteomics). Thus, NGS (next-generation sequencing) is a massively sequencing technology that has revolutionized the biological sciences and offered ultra-high throughput, scalability, and speed in biological analyses. This technology has been largely used to determine the order of nucleotide sequences in genomes or targeted regions of DNA or RNA, and become an everyday tool to address biological questions.

Zhuoyan et al. (2017) reported the application of high-throughput internal transcribed spacer rRNA metagenomics analysis in deciphering endophytic fungi diversity of Dendrobium Officinale. Metagenomic mining of functional genes related to glycyrrhizin synthesis from endophytes of licorice (Glycyrrhiza uralensis), a traditional Chinese medicine, was employed by Chen et al. (2021). Glycyrrhizin is the main active ingredient in licorice, a pentacyclic triterpenoid compound with many pharmaceutical functions (e.g. anti-inflammatory, antiviral, and liver protection), which has also been used in the clinical treatment of COVID-19. Illumina-based analysis yields new insights into the diversity and composition of endophytic fungi in cultivated Huperzia serrata (Fan et al. 2020). Ascomycota and Basidiomycota were the dominant phyla, and Cladosporium, Oidiodendron, Phyllosticta, Sebacina and Ilyonectria were the dominant genera according to the taxonomic classification. The heat map relative abundance at the genus level suggested that H. serrata had characteristic endophytic fungal microbiomes. In addition, fungal communities were tissue-type and tissue-site specific. According to the authors, this study provided new insights into the complex composition of endophytic fungi in H. serrata.

Final considerations

There are many methodological biases and limitations on endophytic fungi studies from field collection to data processing that jeopardizes information on fungal diversity. In most cases, the cultivation-depend methods are extremely imprecise to access the diversity of endophytic fungi (Taylor et al. 2016; Wu et al. 2019). However, many fungal taxonomists consider the cultivation-independent approaches taxonomically problematic because of the absence of isolated specimens (Lücking et al. 2020). Nevertheless, these techniques cover a high diversity of organisms, including uncultivable fungi by conventional culture-dependent methods and artificial conditions. Therefore, it is strongly recommended a polyphasic approach combining different methodologies to produce meaningful information on endophytic fungi. Finally, tips and possibilities to minimize some limitations on endophytic fungi studies are summarized in Table 4 as a contribution to improve future studies.

Table 4 Common errors in studies on endophytic fungi and tips to avoid misinterpretation