Two new root endophyte and nematode cyst parasite species of the widely distributed genus Laburnicola

Fungal root endophytes, including the common form group of dark septate endophytes (DSEs), represent different taxonomic groups with potentially diverse life strategies. During surveys of DSE communities and of nematode cysts colonizing fungi, isolates representing Laburnicola (Didymosphaeriaceae, Pleosporales) lineages were discovered. Here we carried out a comprehensive study of the phylogenetic relationships and taxonomy of fungi collected from plant roots in Hungary, Mongolia, and Kazakhstan and from eggs of the cereal cyst nematode Heterodera filipjevi in Turkey. In addition to the study of the morphology and culture characteristics of the strains, four loci (internal transcribed spacer, partial large and small subunit regions of nuclear ribosomal DNA and partial translation elongation factor 1-alpha) were used to infer the molecular phylogenetic relationships of the strains within Laburnicola. The isolates were found to represent two distinct lineages, which are described here as novel species, Laburnicola nematophila and L. radiciphila. The interaction of the strains with plants and nematodes was examined using in vitro bioassays, which revealed endophytic interactions with the plant roots and parasitic interactions with the nematode eggs. Analyses of similar ITS sequences found in public databases revealed that members of the genus Laburnicola are widely distributed characteristic members of the plant microbiome, and they are reported as parasites of plant parasitic cyst nematodes here for the first time.


Introduction
Healthy roots of terrestrial plants host a plethora of various microorganisms, including fungal endophytes, which colonize plant tissues during some period of their life cycle, yet cause no symptoms of tissue damage to their hosts (Petrini 1991;Saikkonen et al. 1998;Schulz and Boyle 2005). A common form group of these root-associated non-mycorrhizal fungi is the so-called dark septate endophytes (DSEs) that are characterized by mainly melanized septate hyphae and microsclerotia (Jumpponen and Trappe 1998;Sieber and Grünig 2013). These fungi are common members of the root microbiome, and seem to be especially frequent in certain environments, e.g., in grasslands. Their role in the ecosystem and their effects on host plants are still enigmatic (Mandyam and Jumpponen 2005;Newsham 2011;Porras-Alfaro and Bayman 2011;Mayerhofer et al. 2013;Sieber and Grünig 2013). Recent findings of The genus Laburnicola in Didymosphaeriaceae was originally described with four species: Laburnicola centaurea, L. dactylidis, L. hawksworthii, and L. muriformis (Wanasinghe et al. 2016). Those four species were found on the stems of different plants in Italy and they were morphologically characterized by their sexual morphs (Wanasinghe et al. 2016). Later, L. halophila was described from a halophytic plant as a DSE forming only thalloconidia and no a s c o m a t a ( Y u a n e t a l . 2 0 2 0 ) . A s i x t h s p e c i e s , L. zaaminensis, was described and morphologically characterized by its asexual morph (Htet et al. 2021). These six Laburnicola species formed a monophyletic, well-supported clade in multi-locus phylogenetic analyses (Htet et al. 2021).
During investigations of root endophytes of grasslands of the Eurasian steppe belt in Hungary, Mongolia, and Kazakhstan (Knapp et al. 2012(Knapp et al. , 2019Akhmetova et al. 2021Akhmetova et al. , 2022, we have regularly isolated fungi phylogenetically representing the genus Laburnicola. Ever since the first isolate was found on Festuca vaginata in a semiarid sandy grassland of Hungary ('group-11' sensu Knapp et al. 2012), many additional isolates representing Laburnicola lineages from healthy roots of mainly gramineous plants were isolated (Knapp et al. 2019;Akhmetova et al. 2022). Also, numerous sequences can be found in public databases, which probably represent distinct Laburnicola lineages. Further strains with high sequence similarity to Laburnicola sequences were also isolated from eggs of the cereal cyst nematode (CCN) Heterodera filipjevi from the Central Anatolian Plateau of Turkey. Based on the isolation sources, these fungi resembled the isolation of Polyphilus sieberi, which was also identified independently from the roots of several plant species, nematode eggs and cysts and the ascomata of a truffle .
Here, we studied isolates obtained from both the roots of different plants and from the eggs of H. filipjevi. The results of the initial DNA barcoding using ITS sequences showed that those isolates represented the genus Laburnicola, and the lineages they formed were distinct from the known species of the genus. To determine their phylogenetic relationships and clarify their taxonomy, our aims were to (i) carry out multi-locus phylogenetic analyses and morphological studies and describe the potential novel species and to (ii) conduct in vitro resynthesis experiments with plants and nematodes to gain insights into their interaction to better understand their potential roles in the ecosystem.

Materials and methods
Sampling and isolation of fungal strains Nine fungal isolates originating from different hosts and geographic regions were examined in the present study (Table 1).
Five of these were isolated from healthy plant roots originating from Hungary, Mongolia, and Kazakhstan (isolation method as described in Knapp et al. 2012). Four isolates were obtained from symptomatic eggs of the plant parasitic nematode Heterodera filipjevi collected from wheat fields in the Central Anatolian Plateau of Turkey (isolation method as described in Ashrafi et al. 2017). The holotype specimens of the novel taxa were dried and deposited as metabolically inactive samples in the herbarium of the Hungarian Natural History Museum, Budapest, under the accession numbers 111910BP and 111911BP. Ex-type and other cultures were deposited in the German Collection of Microorganisms and Cell Cultures GmbH (DSMZ, Braunschweig, Germany) under the accession numbers DSM 112861-DSM 112868. Nomenclatural novelties and descriptions were registered at MycoBank (www. MycoBank.org, Crous et al. 2004).

Fungal growth, morphology, and sporulation
Four replicates of each of the nine isolates were sub-cultured onto potato dextrose agar (PDA) and corn meal agar (CMA) media (4-mm-diameter fungal plugs onto 9-cm Petri dishes). Growth rate and colony characteristics were recorded after cultures were grown for 3 weeks at temperatures from 5 to 35°C at 5°C intervals in the dark. The isolates that survived growing at 35°C were additionally incubated at 37 and 40°C for 4 weeks. To induce sporulation, the isolates were also cultured on autoclaved pine needles and stinging nettle stems laid on water agar (WA) media in Petri dishes (9-cm diameter) at 22°C for 6 months and on WA media supplemented with minced vegetables (carrot, turnip, celery, and kohlrabi) at a pH of 3.5, respectively, and checked for sporulation regularly.

Resynthesis experiments
Four inoculation experiments were run with each isolate to gain information on their symbiotic nature. We tested the interaction of the strains with plants in two experiments: (i) in vitro inoculation experiments in Petri dishes on MS (Murashige and Skoog Basal Salt Mixture, M5524, Sigma-Aldrich) media were run with leek (Allium porrum), a general host in DSE resynthesis experiments (see Mandyam et al. 2010;Knapp et al. 2012Knapp et al. , 2019. Leek seedlings were placed onto MS media and inoculated with three 5-mm fungal agar plugs (see Knapp et al. 2012) and were then grown in a 14 h light (24°C): 10 h dark (22°C) cycle and harvested 8 weeks post-inoculation. (ii) A pot experiment was set up using 1.5 dl pots containing twice autoclaved sand and zeolite (2:1) and wheat (Triticum aestivum) as gramineous host. Wheat seedlings were inoculated with five 5-mm fungal culture plugs and grown for 8 weeks in room temperature under a 14 h light: 10 h dark cycle. Five replicates of each fungal isolate were used in both experiments. After the plants were harvested, the media/soil was carefully removed from the roots. In case of the pot experiments with wheat, after separation from roots, shoots were dried at 50°C until constant weight and dry biomass was measured. To test the effect of inoculation, one-way analysis of variance (ANOVA) was applied and Tukey's test was used for post hoc analysis to identify the differences in shoot dry biomass among plants inoculated by different isolates using the PAST v4 software (Hammer et al. 2001). Roots of leek and wheat were studied microscopically. The cleared roots were stained with aniline blue following the protocol described in Knapp et al. (2012).
Two other experimental set-ups were established to investigate the capabilities of isolates in the colonization of nematode cysts and eggs. First, the capacity of the fungal isolates to invade nematode eggs was tested using a slide-culture technique as described in detail in Ashrafi et al. (2017) with the modification that the eggs of the beet cyst nematode (BCN), Heterodera schachtii, were used. Second, to evaluate the ability of the isolates to colonize cysts and nematode eggs in planta, i.e., directly at the roots of host plant, a pathogenicity test was conducted in Petri dishes under sterile conditions. Using a modified protocol of Bohlmann and Wieczorek (2015), the BCN was propagated on roots of oilseed radish (Raphanus sativus) growing in 15-cm-diam. agar plates. A few weeks later, upon formation of nematode females on the growing roots, plates were inoculated with the fungal isolates of interest by placing a small agar plug of the fungus in the vicinity of the roots. Four to 8 weeks later, plates and roots were regularly monitored for pathogenicity of the fungi towards growing nematode females and newly formed cysts.
Wheat and leek root samples were examined and photographed using Nikon Eclipse 80i (Tokyo, Japan) microscope equipped with a Spot 7.4 Slider camera (Diagnostic Instruments, Inc.), and for samples including nematode infected roots, cysts, and eggs, Tagarno Prestige digital camera microscope (Horsens, Denmark), an Olympus SZX 12 dissecting microscope (Tokyo, Japan), and a Zeiss Axioskop 2 plus compound microscope (Göttingen, Germany) were applied using Nomarski differential interference contrast (DIC). The latter two were both equipped with a Jenoptik ProgRes (Jena, Germany) digital camera. Samples were photographed in water.

DNA extraction and amplification
Genomic DNA was extracted from fungal mycelia using the DNeasy PlantMini Kit (Qiagen, Hilden, Germany), following the manufacturer's instructions. Four loci were amplified and sequenced for all isolates. Three of these loci are portions of the nuclear ribosomal DNA repeat region: i.e., the internal transcribed spacers 1 and 2 including the 5.8S rDNA (ITS), partial 18S rDNA (small subunit, SSU), and partial 28S rDNA (large subunit, LSU). In addition, a part of the translation    Mongolia. Nalaikh District: from healthy roots of host near Kherlenbayan-Ulaan, in natural steppe zone (N47°43′ 47″, E107°13′ 30″); e ITS internal transcribed spacer region of nrDNA, LSU partial 28S large subunit of the nrDNA, SSU partial 18S small subunit of the nrDNA, TEF1 translation elongation factor 1-alpha gene, RPB1 partial RNA polymerase II largest subunit gene, RPB2 partial RNA polymerase II second largest subunit gene, TUB partial beta-tubulin gene. In case of the L. radiciphila isolate TU32 the ITS, LSU and TEF1 sequences were also deposited previously by Knapp et al. (2019) under the GenBank accession numbers MN537695, MN515273, and MN535270, respectively T Type material. Holotype materials of L. nematophila (111911BP) and L. radiciphila (111910BP) are deposited in the herbarium of the Hungarian Natural History Museum, Budapest, under the barcodes HNHM-MYC-024419 and HNHM-MYC-024418, respectively elongation factor 1-alpha gene (TEF1) was analyzed. In the case of isolates from Kazakhstan, the partial beta-tubulin gene (TUB) and partial RNA Polymerase II largest subunit gene (RPB1) were also sequenced. The following primers were used for amplification and sequencing: for ITS, ITS1F/ITS4 (White et al. 1990;Gardes and Bruns 1993); for SSU, NS1/ NS4 (White et al. 1990); for LSU, LR0R/LR5 (Rehner and Samuels 1994;Vilgalys and Hester 1990); for TEF1, EF1-983/EF1-2218R (Rehner and Buckley 2005); for TUB, Bt2a/ Bt2b (Glass and Donaldson 1995); and for RPB1, RPB1-Af/ RPB1-Cr (Stiller and Hall 1997;Matheny et al. 2002). The RPB1 and TUB regions of some isolates were sequenced but were not used in the phylogenetic analyses, because of the lack of available sequences of these loci of related taxa ( T a b l e 1 ) . T h e s e q u e n c e s w e r e c o m p i l e d f r o m electrophoregrams using the Pregap4 and Gap4 software packages (Staden et al. 2000) and Sequencher 5.4 (GeneCodes Corporation, Ann Arbor, MI, USA) and deposited in GenBank (ON870557-ON870574, ON876670-ON876678, ON892832 -ON892840, ON892841-ON892843, Table 1). The obtained sequences were compared with the accessions in the National Center for Biotechnology Information database (NCBI, http:// www.ncbi.nlm.nih.gov/ Blast.cgi) using the nBLAST search (http://blast.ncbi.nlm. nih.gov/Blast.cgi) (Altschul et al. 1990).

Phylogenetic analyses
We aligned our sequences of the different loci with those from representative taxa in GenBank using the online version of MAFFT 7 (Katoh and Standley 2013) and the E-INS-i method. The alignments were examined and edited using MEGA 7 (Kumar et al. 2016). Two multi-locus datasets representing family and genus level were compiled for molecular phylogenetic analyses. For the family-level dataset (Supplementary  Table 1), we used ITS, LSU, SSU, and TEF1 sequences of the studied strains and representative sequences from species of the family Didymosphaeriaceae according to Samarakoon et al. (2020) and Yuan et al. (2020). In the second multi-locus analysis, we used ITS, LSU, SSU, and TEF1, as well as the indels coded from the ITS and SSU regions (Nagy et al. 2012) using a simple indel coding algorithm (Simmons et al. 2001;Young and Healy 2003) with the program FASTGAP (Borchsenius 2009). Therefore, in the second dataset, six partitions were set. An ITS-based single-locus phylogeny was also conducted using our and similar sequences from public databases. Similar sequences were gained by searches with the BLASTn algorithm (date of BLASTn analysis: 5th April 2022) and all the hits of ITS1 and ITS2 sequences of the nine studied isolates above 90% similarity were incorporated in the analyses. Bayesian inference (BI) analyses were performed with MRBAYES 3.

Colony morphology and sporulation
The nine Laburnicola isolates had variable colony morphology, growth characteristics, and color on different media and temperatures. Fungal growth and colony morphology were assessed on PDA and CMA culture media. Accordingly, colonies were moderately slow to slow-growing, and variable in color; exudates were present or absent (Fig. 1). Altogether four media (PDA, MEA, CMA, WA), three autoclaved plant parts, and numerous culture conditions were applied sensu (Knapp et al. 2015) to study culture characteristics and to potentially induce sporulation. Neither conidiomata and conidia nor ascomata, ascomata-like structures, or ascospores could be detected on any media used under any conditions during the study. No sporulation of the strains was observed in any of the media or conditions tested in the two different laboratories where the strains were isolated and maintained for years before this study. During the pot experiments, the isolate 20K3 formed globose ascomata-or conidiomata-like structures in three wheat plants ( Fig. 2A-C). These structures were sterile and likely immature and had variable sizes (25-90 μm in diam). Pigmented hyphae formed these structures on the surface or below the epidermal cells of the wheat roots. Dense intracellular and extracellular colonization by dark hyphae could be observed around the structures ( Fig. 2A-C). In the case of isolates 20K3 and KG133, also conidium-like, mostly ovoid or peanut-shaped small hyphal formations (4-6 × 7-9 μm) were occasionally observed on pigmented hyphae around the root surface (Fig. 2D, K).

Resynthesis experiments
During the root colonization tests established with leek in Petri dishes and wheat in pots, hyphae of all nine Laburnicola isolates were detected on the surface of the root. Although intraradical colonization of the roots of both plants was rarely detected, at some parts, mainly in the leek roots, we could observe extra-and intracellular hyphae and hyphal structures such as microsclerotia in the case of the isolates 20K3 and KG133 ( Fig. 2E-J). In the roots of wheat, the hyphae were mostly pigmented and could not be stained, while in leek, the hyphae were rarely pigmented and could be stained blue (Fig. 2). Although most strains colonized the roots only sporadically at some regions, isolate 20K3 showed extensive colonization in each inoculation with both plant species. In addition, it formed the abovementioned globose ascomataor conidiomata-like structures at certain parts of the root of wheat ( Fig. 2A-C). The inoculation by the distinct isolates affected the shoot biomass of wheat in the pot culture experiment. The effect of the isolate 20K3 to the wheat biomass compared to F6B1, TU32, 20AD, and KG133 differed; however, the differences of the inoculated plants compared against the control plants were not significant (Fig. S1).
The BCN Heterodera schachtii was used as a model to test the pathogenicity of the fungal isolates 20AD, 20K1, 20K2, 20K3, KG133, and F6B1 towards cysts and eggs of the nematode. The in planta colonization experiment showed that these strains couldto different extentinfect females and newly formed cysts developing at the roots of the host plant. In comparison to healthy females and cysts (Fig. 3A, B), infected ones were discolored (Fig. 3). Among the strains examined, the fungal strains 20AD, 20K1, 20K2, and 20K3 could colonize more nematode females and cysts ( Fig. 3C-R), while the in vitro plants inoculated with the strains KG133 contained fewer symptomatic cysts (Fig. 3S). The plant samples inoculated with the strain F6B1 contained also very few infected cysts (Fig. 3V) indicating that the latter isolates (KG133 and F6B1) could rarely infect the nematode cysts. All examined fungi rendered the initially healthy females and cysts highly pigmented from dark brownish to black. The fungal isolates could infect the nematodes during deployment of females or formation of cysts. Infected eggs collected from infected cysts became also highly pigmented by fungal development, with colors ranging from orangish to dark brown (Fig. 3).
The slide culture studies showed that strain 20AD could colonize the nematode eggs by simple hyphal penetration; no specialized colonizing structure (e.g., appressorium) was observed. The fungus colonized both un-embryonic and D-F: Colonies of L. radiciphila isolated from healthy plant roots, F6B1 (D), KG280 (E), and TU32 (F). All isolates were grown for 21 days except the isolate 20K3, which was grown for one month. Scale bars = 2 cm embryonic eggs. Infected eggs displayed no discoloration and were mainly colonized by hyaline hyphae. Inside the colonized eggs, the mycelium developed by formation of moniliform, thick-walled hyphal cells filled with guttules (oil-like droplets) (Fig. 3Y-AC). Using the slide culture methodology, no other strain could colonize the nematode eggs.
We found approximately 300 ITS sequences with at least 90% sequence similarity to those of the nine isolates studied here and to known Laburnicola species deposited in GenBank (Fig. 6). Using a 508-character long ITS dataset with 302 sequences, the most similar sequences deposited to GenBank as either uncultured or unidentified cultures formed a highly-supported clade (ML-BS = 95, B-PP = 1) with the ITS sequences of the known and our Laburnicola sequences. Besides the clades of the six described and the two new Laburnicola species, sequences deposited in GenBank formed several well-supported distinct lineages, probably representing 15-20 undescribed species. These Laburnicola sequences were obtained from environmental DNA sampling of bulk soil and the rhizosphere as well as isolates collected as endophytes from roots of various plant species, of which grasses and halophytes were the common hosts (Fig. 6). The sequences originated from different countries of Asia, Europe, and North America. None of these sequences originated from the southern hemisphere. Further similar sequences used in this phylogeny represented mainly Alloconiothyrium, Kalmusia, Microdiplodia, and Paraconiothyrium species.
The nine isolates studied in detail in this study formed two distinct clades supported by multi-locus phylogeny and morphological and functional studies. These clades are interpreted to represent two novel Laburnicola species we formally describe here.
Typification  Notes: Symptomatic nematode cysts show dark brown discoloration; infected eggs become pigmented from orangish to dark brownish by fungal development. Strains were obtained from surface-sterilized nematode eggs and plant roots, and maintained on general fungal culture media. The growth rates of strains are variable from moderately (isolates 20AD, KG133) to very slow growing (isolates 20K1, 20K2, 20K3). Isolates of Laburnicola Description: Colonies on PDA are moderately growing, at 25°C reaching 37 mm diam (21 days); optimum temperature for growth 25°C; at 5°C 1 mm (21 days), at 10°C 5 mm (21 days), at 15°C 14 mm (21 days), at 20°C 25 mm (21 days), at 30°C 31 mm (21 days). On other examined culture media (CMA) at 25°C, reaching 60 mm diam (21 days); no growth observed at 35°C. Colonies on PDA elevated centrally, surface velvet, creamy in the central part to pale brown towards the margin, radially striate. Margin even and flattened. Reverse dark brown, pale brown staining around the margin. Fungal growth rate and colony morph vary among conspecific isolates (20K1, 20K2, and 20K3), colonies very slow growing, on PDA at 25°C reaching 11-mm diam after 21 days; colony surface slightly elevated at the center, smooth, olivaceous brown, reverse dark olivaceous.
Sexual morph is unknown. The conidia observed in the case of 20K3 and KG133 are produced solitarily from pigmented hyphae attached to the root surface of Allium porrum and Triticum aestivum. Production of pigmented sterile/immature ascomata-or conidiomata-like structures formed by 20K3 on the surface of T. aestivum could be observed.
Laburnicola radiciphila D.G. Knapp,Ashrafi,Akhmetova,Maier & Kovács,Figs. 1,2,3,4,5,6 Etymology. Referring to the association (the Latin word philia meaning brotherly love, like, having an affinity for something) with roots (radix in Latin).  Notes: Isolates of Laburnicola radiciphila were obtained from surface-sterilized healthy roots of different grasses and are considered dark septate endophytes. The strains colonize experimentally the wheat and leek roots only sporadically at some regions and can barely colonize plant parasitic cyst nematodes Heterodera filipjevi and H. schachtii.
Sexual and asexual morph is unknown. No spore and conidia production could be observed.

Discussion
Here we introduce and formally describe two novel Laburnicola species, L. radiciphila and L. nematophila isolated from roots of gramineous plants and from the eggs of cyst nematodes and wheat roots. The genus Laburnicola (Didymosphaeriaceae, Pleosporales) was erected by Wanasinghe et al. (2016) who originally described four species, L. centaurea, L. dactylidis, L. hawksworthii, and L. muriformis based on multi-locus phylogeny and morphological characters of mainly the ascomata. All those species were considered to be saprotrophic fungi collected from herbaceous stems, hanging branches, and dead branches, stems and wood. No asexual morph was recorded. The obpyriform, immersed ascomata, with the peridium fused to the plant host tissues containing long pedicellate asci and ellipsoidal to fusoid ascospores were defined as the main characteristics of these species. Yuan et al. (2020) described L. rhizohalophila, whole genome of which has been sequenced recently (He and Yuan 2021;Yuan et al. 2021). This species originated from the healthy roots of the halophyte plant Suaeda salsa (Amaranthaceae) collected in China and produced no sexual reproductive structures, but large amounts of mainly peanut-shaped thalloconidia with occasional pigmentation. Such asexual propagules were described neither in other Laburnicola species nor in other members of the family before. The L. rhizohalophila isolates displayed considerable phenotypic and physiological variation. In resynthesis experiments, they successfully infected the host and formed microsclerotia-like structures in the cortical cells of roots of S. salsa. The majority of the isolates promoted the growth of host seedlings. Yuan et al. (2020) emphasized that Laburnicola could accommodate more DSE species, because they found many similar fungal sequences from other halophytes, and concluded that L. rhizohalophila may be a generalist, melanized endophyte in halophytic plants. Our results further support this hypothesis: using the analyses of GenBank sequences, we found that identical or highly similar sequences to L. rhizohalophila originated not only from Suaeda salsa, but also from Su. maritima, Su. japonica, Salicornia europaea, and Sa. patula in China, Korea, and Poland ( Fig. 6) (You et al. 2012;Maciá-Vicente et al. 2016;Furtado et al. 2019aFurtado et al. , 2019b. Both the cosmopolitan genus Suaeda and the mainly northern hemispheric genus Salicornia belong to the Amaranthaceae comprising highly salt-tolerant species that therefore occur at coastal regions, tidal wetlands, and mangroves, but also at salty inland habitats, all of which are extreme habitats for land plants (Piirainen et al. 2017). The sixth species, L. zaaminensis recently Fig. 6 Maximum likelihood (RAxML) tree based on ITS sequences of described Laburnicola species (in bold black) and similar sequences from GenBank (black normal font). Sequences obtained in this study are shown in bold red. After the accession number and sequence name, the isolation source and the country of origin of each sequence are shown in brackets. ML bootstrap support values (≥70) are shown before slashes or above branches; Bayesian posterior probabilities (≥0.90) are shown after slashes or below branches. Described Laburnicola species and closely related sequences are highlighted in blue and sequences representing further Laburnicola-related lineages are highlighted in green. Abbreviations: Cl., clone; RAF, root-associated fungus; Unc., uncultured. Spegazzinia tessarthra (SH287) served as outgroup. The scale bar indicates expected changes per site per branch described by Htet et al. (2021), was isolated from the dead stem of a wild rose plant in the subalpine region of Uzbekistan. The fungus had no sexual morphs but formed coelomycetous globose conidiomata.
In the present study, conidia were only formed in planta by the isolates 20K3 and KG133 of L. nematophila and emerged solitarily from pigmented hyphae attached to the root surface of Allium porrum and Triticum aestivum during the inoculation tests. The shape of these occasionally produced conidia resembles the peanut-or oval-shaped thalloconidia of L. rhizohalophila (Yuan et al. 2020); however, we did not observe conidial chains in the case of the L. nematophila isolates in our in vitro experiments. Ascomata were neither detected for L. rhizohalophila nor for the two Laburnicola species described here, similarly to other DSE fungi (Jumpponen and Trappe 1998;Sieber and Grünig 2013;Knapp et al. 2015;Vohník et al. 2019;Yuan et al. 2020;Zheng et al. 2020;Pintye and Knapp 2021;Romero-Jiménez et al. 2022). It was not possible to induce ascomata formation despite considerable efforts. This is in contrast to Knapp et al. (2015) where sporocarp-like structures and ascomata, respectively, could be induced in five out of six species in Darksidea (Knapp et al. 2015). The ascomata-or conidiomata-like structures produced by the L. nematophila isolate 20K3 were smaller (below 100 μm) than the 120-230-μm-diam. globoid conidiomata of L. zaaminensis (Htet et al. 2021) and the 150-250-μm-diam. globoid ascomata of L. dactylidis, the smallest ascomata in the genus (Wanasinghe et al. 2016). Since these structures were only produced occasionally, they might represent immature structures and could only be observed in L. nematophila; therefore, we applied here the unique fixed allele positions in the diagnosis and description of the two novel Laburnicola species similarly to other asexual, non-sporulating DSE species (Knapp et al. 2015;Ashrafi et al. 2018;Pintye and Knapp 2021).
Laburnicola species were reported from various plant species of several plant families: L. centaureae was obtained from Centaurea sp. (Asteraceae), L. dactylidis from Dactylis sp. (Poaceae), L. hawksworthii from Laburnum anagyroides (Fabaceae), the generic type species L. muriformis also from La. anagyroides (Fabaceae), L. halophila from Suaeda salsa (Amaranthaceae) and L. zaaminensis from Rosa sp. (Rosaceae) (Wanasinghe et al. 2016;Yuan et al. 2020;Htet et al. 2021). Isolates from plant tissues representing the two Laburnicola species of this study were collected from roots of Festuca vaginata, Triticum aestivum, and Stipa krylovii (Poaceae). Analyses of similar ITS sequences from public databases also showed the association of these species to grasses (Fig. 6). A major clade (ML-BS = 71, B-PP = 1) accommodated L. radiciphila which nested (84/0.98) within three major clades comprising numerous (63) environmental ITS sequences and also isolates from roots of a dominant grass species of North American prairie ecosystems, Bouteloua gracilis (Porras-Alfaro et al. 2008;Khidir et al. 2010). Interestingly, several Laburnicola-related sequences found by BLAST derived from sequences of grassland/sedgeland soils from the Tibetan Plateau (e.g., MF971581, MF971582, MF971712, MF971713, MF972001; Yang et al. 2017) (Fig.  6). Three sequences (MH300035, MH300042, MH300044) forming the same and a closely related clade with L. nematophila were collected also from roots of Kobresia (Cyperaceae), which is a common species in the Tibetan Plateau . Therefore, we could assume the association of L. nematophila and L. radiciphila to roots of plant species in grass-and sedge-dominated environments.
Both Laburnicola nematophila and L. radiciphila show intraspecific variability of the color and morphology of the isolates. While the strains 20AD and KG133 formed similar colony morphologies and had a similar growth rate, the strains 20K1, 20K2, and 20K3 formed nearly identical colonies, which were different in color and growth rate from the abovementioned strains 20AD and KG133. These strains were originally isolated from wheat (KG133) or from the cereal cyst nematode H. filipjevi (20AD, 20K1, 20K2, 20K3) collected from wheat fields in semi-arid regions and could colonize the nematode eggs in the resynthesis experiments. To our knowledge, this is the first report on nematode parasitism within Didymosphaeriaceae. We did not observe differences in the colonization process of nematode eggs among the evaluated isolates in our in vitro experiments. Fungal penetration into the eggs and nematode body cavities was caused by (hyaline) hyphae without formation of any specialized structure, followed by digestion of egg content and resulting in proliferation of pigmented hyphae. The nematode isolated strains studied here could similarly colonize the cysts of two plant parasitic nematode species, H. filipjevi and H. schachtii, and parasitize their egg content. The observed parasitic interaction of L. nematophila with the eggs of H. filipjevii resembles the infection process in the previously studied DSE P. sieberi ) and the pleosporalean strain DSM 106825 (Helaly et al. 2018), suggesting that these DSEs might have a bifunctional lifestyle as root endophytes and nematode parasites. Similarly to L. nematophila, the helotialean DSE described by Ashrafi et al. (2018), Polyphilus sieberi, is also a nematode-parasitic fungus colonizing roots and nematodes and was found in the same environments. In addition, sequences representing Polyphilus in GenBank also originated from mainly plant-associated fungi, soil samples and one (HQ446082) from stromata of the entomopathogenic fungus Ophiocordyceps sinensis ). Here we also found a sequence (HQ446062) similar to Laburnicola species from O. sinensis (Zhang et al. 2010). These findings are in accordance with our previous hypothesis  that nematode cysts or truffle ascomata may serve as microenvironments in arid environments, where root-, rhizosphere-, and soil-associated fungi may have a better chance of survival. Therefore, we expect further fungal lineages that display a bipartite lifestyle as root colonizers and nematode parasites, especially in ecosystems with strong abiotic stresses.