Cysts of H. filipjevi were collected from experimental wheat fields of CIMMYT in the Central Anatolian Plateau of Turkey in 2013. The fields located in Yozgat (39°08'N, 34°10'E; altitude 985 m.a.s.l) and Haymana (39°26'N, 39°29'E, altitude 1260 m.a.s.l) were naturally nematode infested. The samples including soil and roots were collected at random from the rhizosphere of wheat plants at the end of the growing season. Cysts were extracted from the collected samples using the modified flotation decanting method (Coyne et al. 2007). From the extracted suspensions, cysts were manually collected under a dissecting microscope and stored in 1.5 ml microtubes at 4 ˚C either in dry condition or in sterile tap water until further use. For taxonomic and phylogenetic inferences, additional fungal strains were obtained from the Westerdijk Fungal Biodiversity Institute (formerly CBS-KNAW, Utrecht, Netherlands).

The field-collected cysts of H. filipjevi were scrutinised by using a dissecting microscope to separate symptomatic cysts showing defined discolourations or bearing discernible hyphae, from healthy-looking (i.e. homogeneously brown) or empty cysts. Symptomatic cysts were selected, surface-sterilised in 5% sodium hypochlorite (NaOCl), and dissected to collect their egg contents. Only the nematode eggs showing symptoms of fungal infection were processed for fungal isolation and culture-dependent species identification. A portion of the fungal infected eggs were additionally used for culture-independent identification. The methods applied here, have been described in greater detail in Ashrafi et al. (2017).

Growth rates were determined at various temperatures from 15 to 35 °C at 5 °C intervals in the dark or in ambient conditions by placing agar disks (5 mm diam.), excised from the margin of a young potato dextrose agar (PDA) culture onto five replicate plates of PDA, cornmeal agar (CMA), oatmeal agar (OA; 30 g oatmeal, 18 g agar-agar, 1L deionised water), synthetic nutrient-poor agar (SNA; Nirenberg (1976)), and malt extract agar (MEA). The colony diameter was measured weekly for a 3 week period. Colour changes of fungal structures formed in culture were checked using 3% potassium hydroxide (KOH) watery solution.

The antagonistic potential of the below described species and M. bulbillosum, respectively, was assessed towards H. filipjevi in vitro as previously described (Ashrafi et al. 2017). Briefly, healthy cysts and eggs were surface-sterilised and placed either on or at the margin of the growing mycelium of one-month-old PDA or 2% water agar (WA) cultures of the two fungal species. To document the process of colonisation of eggs of H. filipjevi by the new fungal species, a slide culture technique was also performed using PDA 1/3 strength (compare Ashrafi et al. (2017)).

Nematode eggs and fungal structures were examined and photographed by a Zeiss Axioskop 2 plus compound microscope and an Olympus SZX 12 stereo microscope equipped with a Jenoptik ProgRes® digital camera. Images were recorded using CapturePro 2.8 software (Jenoptic, Jena, Germany). Nematode eggs colonised by fungi, and fungal structures were mounted in water or lactic acid and photographed. Cysts were photographed in water in a square cavity dish (40×40×16 mm). To illustrate different stages of fungal development and fungal colonisation of nematode eggs, slide cultures were prepared (Gams et al. 1998) and then photographed. Nomarski Differential Interference Contrast (DIC) optic was used for observation and measurements. All measurements were taken in water, and are given as x1–x2 (x3 ± SD), with x1 = minimum value observed, x2 = maximum value observed, x3 = average, and standard deviation (SD), followed by the number of measurements (n).

Scanning electron microscopy was performed on a Quanta 250 scanning electron microscope (FEI Deutschland GmbH, Frankfurt, Germany). Fungal structures of interest were obtained from a one-month-old OA culture grown at 23 °C in the dark and directly analysed using environmental scanning electron microscopy (ESEM). For the experiment, pressures between 410 and 490 Pa at 4 °C were employed. For cooling the sample chamber was equipped with a Peltier stage. Fungal mycelia with abundant conidia were placed on non-conductive double-sided adhesive discs on a flat specimen stub and positioned on the Peltier stage for cooling. Images were taken at acceleration voltage of 12.5 kV. Scanning speed was 60 µsec. For imaging of beam sensitive fungal structures, the scanning modus was changed to 3 µsec with 20-fold line integration. Images were adjusted in brightness and contrast using Adobe Photoshop software CS 5.1.

Fungal genomic DNA was isolated from mycelia grown on PDA using a modified CTAB method, and from individual nematode eggs infected by fungi using the Qiagen DNeasy Plant Mini Kit (Qiagen, Hilden, Germany) as reported in Ashrafi et al. (2017).

For each specimen, four nuclear loci were amplified: The internal transcribed spacers including the 5.8S rDNA gene (ITS) using the primers ITS1F (Gardes and Bruns 1993) and ITS4 (White et al. 1990); the 5’ end of the ribosomal large subunit (LSU) DNA with the primers LROR (Rehner and Samuels 1994) and LR5 (Vilgalys and Hester 1990); partial RNA polymerase II largest subunit 1 (rpb1) using the primers cRPB1af and RPB1cr (Castlebury et al. 2004); and partial translation-elongation factor 1-α (TEF) using the primers EF1-983f and EF1-2218r (Castlebury et al. 2004). All PCR reactions were performed as described previously (Ashrafi et al. 2017) with the following thermal programmes: 95 °C (2 min) for initial denaturation followed by 40 cycles of denaturation at 95 °C (30 s), annealing at 52 °C (ITS), 51 °C (LSU), 54 °C (rpb1), and 60 °C (TEF) (40 s), extension at 72 °C (1 min for ITS, LSU and rpb1, and 1 min and 20 sec for TEF), and a final extension at 72 °C (10 min). Amplicons were purified using the DNA Clean & Concentrator^TM-5 kit (Zymo Research Corp., Irvine, California, USA) and sequenced by Eurofins Genomics GmbH, (Ebersberg, Germany) with the same primers as used for PCR amplification. Obtained sequences were assembled, edited and trimmed with Sequencher 5.4.1 (Gene Codes Corporation, Ann Arbor, Michigan, USA) and deposited in GenBank under the following accession numbers: MF681481–MF681514. The sequences generated were compared to sequences available in GenBank using a BLASTn search (http://blast.ncbi.nlm.nih.gov/Blast.cgi) (Altschul et al. 1990).

The newly generated sequences together with closely related sequences selected as revealed by BLASTn searches were used for phylogenetic analyses (Table 1). The sequences were aligned using the online version of Mafft v.7 (Katoh and Standley 2013). All sequences were aligned using the iterative refinement methods: Sequences of the rpb1 and TEF gene regions were aligned using the algorithms implemented in L–INS–i, while LSU and ITS were aligned applying the Q–INS–i algorithm. Only the start and end of the alignments were trimmed manually in Se-Al v2.0 (Rambaut 1996). The following phylogenetic analyses were applied: a Bayesian method of phylogenetic inference using Metropolis Coupled Monte Carlo Markov chains (MC³) as implemented in the computer program MrBayes v3.2 (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003). We used MrModeltest v2.2 (Nylander 2004) to determine the best fitting DNA substitution model for the Bayesian approach. Both the hierarchical likelihood ratio test (hLRT) and the Akaike Information Criterion (AIC) selected the general time reversible model of DNA substitution with gamma distributed substitution rates and invariate sites (GTR+I+G) as the best fitting model for all individual data sets and was implemented for the analyses accordingly. For the Bayesian analyses four incrementally heated simultaneous Monte Carlo Markov chains were run with 2.000.000 generations using random starting trees and flat prior distributions. Trees were sampled every 500 generations resulting in a total of 4001 sampled trees. A 50% majority rule consensus tree was computed only from trees of the plateau, and if, additionally, the split frequencies were below 0.01. Thus, 501 trees were discarded as “burnin”. Maximum likelihood (ML) analyses were performed using RAxML 7.2.8 (Silvestro and Michalak 2012; Stamatakis 2014) implemented in Geneious 8.1.2 applying the general time-reversible (GTR) substitution model with gamma model of rate heterogeneity and 1000 replicates of rapid bootstrapping. Neighbor-joining (NJ) analyses (Saitou and Nei 1987) was done in PAUP 4.0b10 in the batch file mode (Swofford 2002) applying the Kimura two-parameter model of DNA substitution (Kimura 1980) with a transition/transversion ratio of 2.0 to compute genetic distances. Support for internal nodes was estimated by 1000 bootstrap replicates (Felsenstein 1985). Two members of Bionectriaceae, Bionectria byssicola (Berk. & Broome) Schroers & Samuels and Ijuhya vitellina Ashrafi, W. Maier & Schroers, were selected as outgroup to root the trees. The phylogenetic trees were visualised using FigTree v. 1.4.2 (http://tree.bio.ed.ac.uk/software/figtree).

Among the field-collected samples, a high proportion of cysts was found containing blackish bodies resembling microsclerotia-like structures upon microscopy (Fig. 1A). By dissecting the infected cysts, microsclerotia-like black bodies were found to be colonising the individual nematode eggs (Fig. 1B, C). In some infected eggs the developing juveniles were found to be entirely destroyed exhibiting an olivaceous brownish appearance (Fig. 1D, E). Eggs were colonised by one or occasionally two microsclerotia. When cultured on PDA, hyphae grew out of the microsclerotia of the infected eggs (Fig. 1F), and formed colonies at first white creamy, later becoming blackish dotted centrally with a general dark appearance due to the dense pigmentation (Fig. 1G).

Figure 1. 

Naturally infested cysts and eggs of Heterodera filipjevi with Monocillium gamsii, and pure cultures obtained from infected eggs. A Field collected symptomatic cysts bearing parasitised eggs. B–E Nematode eggs infected by M. gamsii B, D Nematode eggs containing microsclerotia of M. gamsii E An embryonated egg containing a second stage juvenile (J2) parasitised by M. gamsii (arrow points at nematode’s stylet) F A nematode egg containing microsclerotia, and hyphae growing out of it G–H colony of M. gamsii grown on PDAG colonies developing from three individually plated infected eggs H A 25-d-old culture grown at 25 °C in the dark I The surface of a five-month-old culture detailing the sclerotioid masses covering the colony surface. Single microsclerotia can be seen as little black dots at the margin of the culture. Scale bars: 800 µm (A); 30 µm (B–E); 50 µm (F); 2 cm (H); 5 mm (I).

The DNA sequences of four different gene regions obtained from the examined specimens of the here described nematode-parasitic fungus were either identical (in TEF and RPB1), or nearly identical (1 base pair (bp) substitution in LSU, and up to 2 bp substitutions in ITS). The most similar DNA sequences found in GenBank using BLASTn searches belonged to Hyaloseta nolinae, the sexual morph of Monocillium nolinae, and shared similarities of 96% in the ITS region, 99% in the LSU, and 89% in rpb1. A similar BLASTn search in MycoBank showed identities of the ITS sequence of 96.6% with M. bulbillosum, 93.9% with H. nolinae and 92.7% with Niesslia exosporioides, and of the LSU sequence of 99.5% with H. nolinae, and 99% with both M. bulbillosum and Niesslia exosporioides suggesting a close relationship with the representatives of the Niessliaceae. Fungal DNA could also be directly isolated and sequenced from individual eggs displaying the typical symptoms of fungal infection. These DNA sequences were identical to the sequences retrieved from pure cultures supporting the conspecificity of the symptom-causing structures within the egg with the isolated pure cultures derived from the eggs.

The final combined ITS, LSU, rpb1 and tef dataset comprised 11 strains representing 7 species with a total alignment length of 2949 bp (603 ITS, 797 LSU, 649 rpb1, 900 tef). The topologies of the phylogenetic trees were identical using Bayesian inference (Fig. 2), neighbor-joining or maximum likelihood (not shown). The four sequenced strains of the here described nematode egg-colonising fungus were recovered as a highly supported monophyletic group with a close sister group relationship to M. bulbillosum and with H. nolinae as the next-closest relative. In the second monophyletic clade of Niessliaceae, two strains of the type species of Niesslia, N. exilis, proved to be paraphyletic with respect to M. ligusticum (Fig. 2).

Figure 2. 

Bayesian inference of phylogenetic relationships using four strains of the here described nematode parasite and all Niessliaceae present in GenBank based on an alignment of ITS, LSU, rpb1, and tef sequences using GTR+I+G as nucleotide substitution model. Depicted is a 50% majority rule consensus tree derived from 3500 trees from the stationary phase of a Monte Carlo Markov Chain. A posteriori probability (BIpp) values greater than 0.95, and bootstrap values of neighbor-joining (NJBT) and maximum likelihood (MLBT) analyses greater than 0.7 are given above branches (BIpp/NJBT/MLBT). Two representatives of the Bionectriaceae, Bionectria byssicola and Ijuhya vitellina, were used to root the tree.