ABSTRACT

   Dactylellina haptotyla is a typical nematode-trapping fungus that has
   garnered the attention of many scholars for its highly effective lethal
   potential for nematodes. Secondary metabolites play an important role
   in D. haptotyla-nematode interactions, but which metabolites perform
   which function remains unclear. We report the metabolic functions based
   on high-quality, chromosome-level genome assembly of wild D. haptotyla
   YMF1.03409. The results indicate that a large variety of secondary
   metabolites and their biosynthetic genes were significantly upregulated
   during the nematode-trapping stage. In parallel, we identified that
   2-furoic acid was specifically produced during nematode trapping by D.
   haptotyla YMF1.03409 and isolated it from fermentation production.
   2-Furoic acid demonstrated strong nematicidal activity with an LD[50]
   value of 55.05 µg/mL against Meloidogyne incognita at 48 h.
   Furthermore, the pot experiment showed that the number of galls of
   tomato root was significantly reduced in the experimental group treated
   with 2-furoic acid. The considerable increase in the 2-furoic acid
   content during the infection process and its virulent nematicidal
   activity revealed an essential synergistic effect during the process of
   nematode-trapping fungal infection.

IMPORTANCE

   Dactylellina haptotyla have significant application potential in
   nematode biocontrol. In this study, we determined the chromosome-level
   genome sequence of D. haptotyla YMF1.03409 by long-read sequencing
   technology. Comparative genomic analysis identified a series of
   pathogenesis-related genes and revealed significant gene family
   contraction events during the evolution of D. haptotyla YMF1.03409.
   Combining transcriptomic and metabolomic data as well as in vitro
   activity test results, a compound with important application potential
   in nematode biocontrol, 2-furoic acid, was identified. Our result
   expanded the genetic resource of D. haptotyla and identified a
   previously unreported nematicidal small molecule, which provides new
   options for the development of plant biocontrol agents.

   KEYWORDS: nematode-trapping fungi, Dactylellina haptotyla, infection,
   genomics, metabolomic, 2-furoic acid

INTRODUCTION

   Plant-parasitic nematodes are major agricultural pests that cause
   devastation in agriculture around the world. The annual global economic
   losses caused by plant-parasitic nematodes are estimated to be between
   $80 and $173 billion ([38]1, [39]2). Nematode-trapping fungi are
   natural nematode predators with enormous application potential in
   managing plant-parasitic nematode populations and lowering agricultural
   losses ([40]3, [41]4). Nematode-trapping fungi mainly live
   saprophytically in soil. The formation of a trap is a signature event
   of nematode-trapping fungi from the saprophytic to parasitic stages
   ([42]5, [43]6); it can be produced spontaneously or induced in the
   presence of nematodes or some metabolites ([44]7). As the most
   essential weapon for nematode-trapping fungi, the structure of the trap
   is diverse ([45]8), resulting in substantial variances in its predation
   process. Nematode predation by nematode-trapping fungi is a complex and
   delicate process ([46]7, [47]9). Despite the fact that several research
   have been conducted on this predating process, many crucial questions
   remain unanswered.

   Microbial metabolites are mediators of biological communication and
   competition between microorganisms and the environment ([48]10,
   [49]11). These metabolites, formed by various enzymatic mechanisms that
   have evolved over thousands of years, contribute to biomass production
   and environmental niche assurance ([50]12, [51]13). During interactions
   between pathogens (or parasites) and hosts, metabolites produced by the
   pathogens (or parasites) often play key roles in the infestation
   process ([52]14, [53]15). In 1995, Anderson et al. reported three new
   derivatives of oligosporon from nematode-trapping fungus Orbilia
   oligospora (Arthrobotrys oligospora), which showed toxic activity
   against intestinal parasitic nematodes ([54]16). Researchers have also
   obtained a series of oligosporons and arthrosporols from O. oligospora
   that possess the ability to modulate the differentiation of traps
   ([55]17 [56]– [57]19). In 2013, Wu et al. isolated paganins A and B
   from O. oligospora and found that the compounds could promote trap
   generation ([58]18). A volatile compound, methyl 3-methyl-2-butenoate,
   obtained from O. oligospora, can attract a variety of nematodes,
   including Caenorhabditis elegans ([59]20). Additionally, Arthrobotrys
   flagrans (Duddingtonia flagrans) produce a polyketone 6-methyl
   salicylic acid that regulates the spatiotemporal control of trap
   formation and acts as a chemoattractant to C. elegans ([60]21). The
   discovery of these metabolites and the resolution of their activity
   have improved our understanding of nematode-trapping fungi. In the
   1960s, Olthof and Estey tested the activity of filtrates of O.
   oligospora that had already infected nematodes and found that many
   worms became immobile and appeared dead ([61]22). Therefore, they
   speculated that nematode-trapping fungi can generate metabolites that
   play a synergistic role in the infection process.

   Dactylellina haptotyla is a representative species of nematode-trapping
   fungi. Its spores are poke-shaped; it can capture its prey and enter
   the parasitic stage by forming knobs ([62]Fig. 1). Comparative genomic
   analysis suggests that knob formation is an evolutionary tendency of
   morphological simplicity and efficiency enhancement ([63]23, [64]24).
   The high pathogenicity of D. haptotyla suggests that it may have huge
   application potential in the biological control of nematodes ([65]25,
   [66]26). In a previous study of D. haptotyla, Tunlid and his group
   provided several excellent results: after studying differentially
   expressed genes (DEGs) between traps and hyphae, as well as
   transcriptome and proteome data during D. haptotyla infection, they
   identified several extended-domain highly expressed families, including
   subtilisin, common in several fungal extracellular membrane proteins,
   the DUF3129 family (gas1), small secreted proteins, and cell surface
   adhesins. Most of them are known to be associated with the
   pathogenicity of pathogenic fungi ([67]5, [68]26, [69]27). The genomic
   mechanism of parasitic evolution of predatory nematode fungi was
   elucidated through comparative genomics and transcriptomics. These
   findings provide a wealth of data for further understanding the
   function of D. haptotyla infection-associated metabolites ([70]28).

Fig 1.

   [71]Fig 1
   [72]Open in a new tab

   The characterizations of D. haptotyla YMF1.03409. (A) The growth status
   of D. haptotyla YMF1.03409 on PDA. (B) and (C) The conidiophores of D.
   haptotyla YMF1.03409, which are marked by red arrows. (D) and (E) The
   knobs of D. haptotyla YMF1.03409, which are marked by blue arrows.

   Biosynthetic gene clusters (BGCs) can be predicted from microbial
   genome analysis (e.g., antiSMASH and PRISM) through the reporting of
   large-scale genomes and the outcomes of big data in natural product
   research. The quality of assembly and annotation has a considerable
   influence on the outcomes of genome-based research since assembled and
   annotated genome sequences frequently serve as the raw material for
   genome mining ([73]29). The integrity of gene content and continuity
   within the chosen genome sequence heavily influence our ability to
   comprehend the diversity of BGCs ([74]30). Long-read technology can
   obtain a more complete and continuous genome assembly ([75]31). In this
   study, we obtained high-quality genome assembly at the chromosome level
   of D. haptotyla YMF1.03409 by high-throughput sequencing using Hi-C
   technology and constructed the first metabolome landscape of this
   fungus. Furthermore, we discovered some unique antinematode molecular
   features of D. haptotyla YMF 1.03409, such as EVM08G003990.1, which is
   located in the shikimic acid pathway, and its expression level was
   significantly increased during reciprocal interaction. We also
   discovered that the small molecule compounds (such as 2-furoic acid)
   were also specifically present in the reciprocal metabolome samples;
   nematicidal activity experiments demonstrated the tremendous nematode
   lethality of 2-furoic acid. Additionally, by comparative genomic
   analysis, we found that D. haptotyla YMF1.03409 exhibited a significant
   gene family contraction, which may be because D. haptotyla YMF1.03409
   discarded several genes not associated with parasitism during
   evolution.

MATERIALS AND METHODS

Fungal, nematode materials, and growth conditions

   The strain used in this study was D. haptotyla YMF1.03409, which was
   maintained at the State Key Laboratory for Conservation and Utilization
   of Bio-Resources and Key Laboratory for Microbial Resources. D.
   haptotyla YMF1.03409 was routinely cultured on potato dextrose agar
   (PDA) plates. Mycelium samples for genome sequencing were provided in
   the supplemental material. The nematodes used in this study were
   Panagrellus redivivus, Meloidogyne incognita, and C. elegans. P.
   redivivus was routinely cultured in an oat medium at 25°C. M. incognita
   was obtained by picking nematode egg masses from the roots of infested
   plants that were then incubated at 25°C. C. elegans was routinely
   cultured in a nematode growth medium at 25°C, and Escherichia coli OP50
   was used as the food source.

   The methods of genome sequencing, assembly, gene and genomic component
   prediction, and comparative genomic analysis are provided in the
   supplemental material.

RNA-seq and analysis

   The strains were cultured on PDA plates at 28°C for 10 days. We then
   collected the spores and incubated them in 1% sucrose plates at 28°C
   for 3 days until spore germination. For the treated group,
   approximately 150 nematodes (P. redivivus) were added to the plates,
   and we sampled two time points (24 and 48 h after treatment), labeled
   as PD24 and PD48. D. haptotyla YMF1.03409 and P. redivivus were grown
   in parallel under identical conditions and used as control groups,
   labeled as D24, D48, P24, and P48, respectively. The transcriptome
   analysis time point was chosen based on the infection status of D.
   haptotyla YMF1.03409. The beginning, middle, and end of the infestation
   were determined to be 12, 24, and 48 h following the addition of
   nematodes, respectively. The nematode infectious time was also adjusted
   as a covariate. The nematode mortality rate and quantity of adhesive
   knobs considerably increased over the course of the infection. Nearly
   all nematodes perished by the conclusion of the infestation (48 h after
   nematode inoculation), and knobs counts peaked ([76]Fig. S1 and S2).

   The samples were collected, and RNA was extracted using the TRIzol
   method. The RNA-seq analysis method is described in the supplemental
   material.

Metabolomic sample preparation from the infestation of P. redivivus by D.
haptotyla YMF1.03409

   Aliquots (600 µL) of conidial suspension were spread on 6 cm Petri
   dishes with water agar (WA) medium and incubated at 28°C for 3 days
   until conidia germination. Approximately 150 nematodes (P. redivivus)
   were added to the middle of the WA plates, and the fungus-nematode
   interaction samples were collected at 24 and 48 h (after the injection
   of P. redivivus) and marked as PD24 and PD48, respectively. The
   nematodes and fungi used for the control were prepared in the same way
   as described above and marked as P24, P48, D24, and D48, respectively.
   The interaction, nematode, and fungus samples were harvested separately
   and extracted with 200 mL of organic reagent mixture (ethyl
   acetate:methanol:acetic acid = 80:15:5, vol/vol/vol) for 3 days. The
   extract liquids were separated by filtration, and the filtrate was
   concentrated by drying under vacuum. Each dried sample was resuspended
   in 2 mL of chromatographic methanol, filtered through 0.22 µm Steritop
   units (Millipore), and stored at 4°C prior to liquid
   chromatography-mass spectrometry (LC-MS) analyses. Metabolomic data
   acquisition and statistical analysis are presented in the supplemental
   material.

Purification and characterization of 2-furoic acid from D. haptotyla
YMF1.03409

   After fermentation on rice medium [60 g of rice, 0.3 g of
   (NH[4])[2]SO[4] and pork liver, and 30 mL of H[2]O]. After separation
   and purification using various chromatography techniques, 14 compounds
   were ultimately isolated after the crude extract (35.64 g) was obtained
   using an organic solvent. The fermentation, extraction, isolation, and
   spectroscopic data are presented in the supplemental material.

Virulence assay

   To understand the virulence of 2-furoic acid, we conducted a
   nematicidal activity test. In particular, a certain concentration of
   2-furoic acid solution was prepared and added to 3.5 cm plates, 2 mL
   per plate. Subsequently, 100–150 nematodes (M. incognita, P. redivivus,
   and C. elegans) were added to the plates. The number of dead and live
   nematodes was recorded after 12, 24, and 48 h under a light microscope
   (Olympus). The assay was conducted in triplicate (an aqueous solution
   under the same conditions served as the control group).

Root-knot nematode infestation experiment

   Three treatment groups were used in this experiment: (i) positive
   control avermectin solution, (ii) negative control aqueous solution,
   and (iii) experimental group 2-furoic acid solution. The soil weighed
   approximately 80 g, with three tomato seedlings per pot. Before
   transplanting the tomato seedlings, the roots (substrate soil
   containing the roots) were soaked in each of the three groups of
   solutions at a concentration of 350 µg/mL for 10 min and then
   transplanted into pots. Three hundred milliliters of the corresponding
   treatment group were irrigated to the tomato roots of each experimental
   pot after transplanting and watered normally at a later stage. After
   two days, M. incognita was added, and the number of M. incognita was
   approximately 5,000 per pot. Normal watering was conducted during the
   later period. Tomato plants from each treatment group were removed
   after 20 and 35 days, and the roots were observed and counted for
   root-knot nematode parasitism after the soil was gently rinsed off from
   the roots.

RESULTS

Chromosome level genome of D. haptotyla YMF 1.03409

   We extracted DNA from the D. haptotyla YMF1.03409 strain, which was
   isolated from China, for sequencing and genome assembly. A total of
   6.99 Gbp clean sequencing data were obtained using nanopore sequencing
   technology from Oxford Nanopore Technologies. Nanopore long reads were
   then assembled into 11 contigs with an N50 length of 3,741,584 bp. We
   then conducted Illumina short-read sequencing with an average depth of
   approximately 123× to polish the contigs generated by long-read
   sequencing. The short-read remapping ratio was 98.68% and covered 99.7%
   of the contigs, which demonstrated that our primary assembly was of
   high quality. The contigs were then anchored to the chromosome model
   using Hi-C data (7.41 Gbp clean data; [77]Fig. 2B). The mapping rate
   was 93.58%, and 80.16% of the Hi-C reads were uniquely mapped to the
   primary contigs. The mapped reads were used to cluster, orientate, and
   order the contigs to the chromosomes. Finally, we successfully
   constructed a chromosome-level assembly of D. haptotyla YMF1.03409 with
   a size of 39.55 Mbp ([78]Fig. 2A; [79]Table S1, GenBank assembly
   accession: [80]JAQJAX000000000).

Fig 2.

   [81]Fig 2
   [82]Open in a new tab

   The new chromosome genome assembly of D. haptotyla YMF1.03409. (A)
   Circos plot of D. haptotyla YMF1.03409. The outer to inner circles
   represent the mapping identity with MHA_v2, repetitive region density,
   and gene density, respectively. (B) The Hi-C heatmap of D. haptotyla
   YMF1.03409 indicates the satisfactory quality of the assembly. (C) The
   dotplot for genome-wide alignment between MHA_v2 and D. haptotyla
   YMF1.03409.

   The assembly of D. haptotyla YMF1.03409 showed higher sequence identity
   with the previously published D. haptotyla CBS 200.50 genome assembly
   MHA_v2 (GenBank assembly accession: GCA_000441935.1) ([83]28), which
   was constructed by 454 pyrosequencing method ([84]Fig. 2C). The
   one-to-one comparative analysis indicated an improvement in the
   chromosome-level assembly ([85]Table 1). The D. haptotyla YMF1.03409
   assembly has a longer total length (39.55 Mb vs 39.53 Mb) and a higher
   level of completeness (98.0% vs 97.6%). The identified genes were more
   than those in the previous D. haptotyla CBS 200.50 genome assembly
   MHA_v2 (11,073 vs 10,959). The number of transposable elements
   identified was also higher than that in the previous version (408 vs
   314, [86]Table S2).

TABLE 1.

   Statistics of D. haptotyla CBS 200.50 assembly and D. haptotyla
   YMF1.03409 assembly
   D. haptotyla CBS 200.50 D. haptotyla YMF1.03409
   Sequencing platform 454 pyrosequencing technology Nanopore + NGS + Hi-C
   Assembly size (bp) 39,531,920 39,557,028
   Depth 28× 176.60 (SMRT[87] ^a ) + 123.21× (NGS[88] ^b )
   GC content (%) 45.30 45.44
   Protein-coding genes 10,959 11,073
   Number of scaffolds 1,279 10
   BUSCO completeness (%) 97.6 98.0
   Longest sequence length (bp) 937,461 7,783,607
   [89]Open in a new tab
   ^^a

   SMRT, single molecule real-time sequencing.
   ^^b

   NGS, next-generation sequencing.

Pathogenic genes widely existed in D. haptotyla YMF1.03409 genome

   To provide a general view of potentially pathogenic genes in the
   nematode-trapping fungi, the basic local alignment search tool was used
   on the entire D. haptotyla YMF1.03409 genome against the pathogen-host
   interaction database (PHI-base). Among the 2,727 genes matching a
   PHI-base entry, 47% of the mutant phenotype of those genes was
   annotated as “reduced virulence” or “loss of pathogenicity.” A large
   number of virulence genes showed the capacity of D. haptotyla
   YMF1.03409 to control nematodes ([90]Table S3).

   The D. haptotyla YMF1.03409 genome contains abundant genes related to
   the fungal lifestyle. Carbohydrate-active enzymes (CAZYs) are involved
   in nutrient sensing and acquisition and in determining the saprophytic
   ability of fungi ([91]32). Information about CAZYs could provide
   insights into the species’ biology and pathogenicity. In order to gain
   further insight into the pathogenic potential of D. haptotyla
   YMF1.03409, based on the references to other existing literature