In Vitro Assessment of Organic and Residual Fractions of Nematicidal Culture Filtrates from Thirteen Tropical Trichoderma Strains and Metabolic Profiles of Most-Active

The nematicidal properties of Trichoderma species have potential for developing safer biocontrol agents. In the present study, 13 native Trichoderma strains from T. citrinoviride, T. ghanense (2 strains), T. harzianum (4), T. koningiopsis, T. simmonsii, and T. virens (4) with nematicidal activity were selected and cultured in potato dextrose broth to obtain a culture filtrate (CF) for each. Each CF was partitioned with ethyl acetate to obtain organic (EA) and residual filtrate (RF) fractions, which were then tested on second-stage juveniles (J2s) of the nematodes Meloidogyne javanica and M. incognita in a microdilution assay. The most lethal strains were T. harzianum Th43-14, T. koningiopsis Th41-11, T. ghanense Th02-04, and T. virens Th32-09, which caused 51–100% mortality (%M) of J2s of both nematodes, mainly due to their RF fractions. Liquid chromatography–diode array detector-electrospray-high resolution mass spectrometry analysis of the most-active fractions revealed sesquiterpene and polyketide-like metabolites produced by the four active strains. These native Trichoderma strains have a high potential to develop safer natural products for the biocontrol of Meloidogyne species.


Introduction
Fungi belonging to the Trichoderma genus are cosmopolitan species, with 488 species identified [1]. Several of these species have been widely studied as biocontrol agents against phytopathogenic fungi [2] and nematodes [3,4], insects [5], and weeds [6], and as plant growth promoters [7,8]. The nematicidal potential of Trichoderma species is increasingly being harnessed to develop new and safer biocontrol agents against parasitic nematodes such as Globodera pallida, Heterodera avenae, Meloidogyne incognita, M. javanica, M. hapla, and Pratylenchus brachyurus [8,9]. In particular, Meloidogyne root-knot nematodes are considered the most harmful because they can affect a wide range of crops, causing production losses between 25% to 50% and millions of dollars. They thus continue to be controlled mainly with synthetic agrochemicals despite recognized problems for the environment and organisms [10][11][12] because of with lack of safer products and eco-friendly and holistic strategies. As harmful synthetic chemicals are withdrawn from the market, the search for alternatives such as crop rotation, resistant plant varieties, and biological control agents or their derivatives to control nematodes has intensified [4,11,13,14] Trichoderma species that have lethal effects against Meloidogyne species include T. harzianum, T. koningii, T. koningiopsis, T. longibrachiatum, T. citrinoviride and T. viride [15][16][17] against M. incognita; T. hamatum, T. harzianum, T. koningii, T. koningiopsis, and T. viridae against M. javanica [15,18,19]; T. asperellum, T. harzianum, T. viride and T. viride on M. hapla [9]; and T. harzianum against M. enterolobii [20].
The main mechanisms of action known for Trichoderma species are antibiosis, competition for space and nutrients, mycoparasitism, and induction of defense mechanisms. The antibiosis involves the production and secretion of secondary and primary metabolites that inhibit the growth and development of root-knot nematodes [19,21,22]. Regarding the production and secretion of metabolites, approximately 390 non-volatile metabolites have been identified from Trichoderma spp. [2,23], but only a few, such as gliotoxin, gliovirin, heptelidic acid, and viridin identified from T. virens [3,24], have been reported to have nematicidal activity. Therefore, a systematic bio-guided screening of Trichoderma species is a promising option to find novel nematicidal products.
During our ongoing bioprospecting studies in search of natural agrochemical products in the Yucatán península (Mexico), our group isolated 56 native Trichoderma species from soils. The foregoing is in response to the need to develop nematicides based on native species adapted to the areas where they are intended to be applied, with the knowledge of the active metabolites produced by fungal strains. These strains were tested for control of M. incognita, and 29 of these strains affected the viability of eggs and second-stage juveniles (J2s) in vitro and acted as plant growth promoters [25][26][27]. Moreover, these active Trichoderma strains decreased the severity of nematode damage on tomato and improved yields in the greenhouse [18,25].
In the present study, 13 Trichoderma strains from Yucatán were selected due to their activity against phytopathogenic fungi, nematodes, and as plant growth promoters. All selected Trichoderma strains were cultured in the liquid media potato dextrose broth. From each fungal strain, the culture filtrate, ethyl acetate fraction, and residual filtrate fraction were obtained and tested against the J2s of M. incognita and M. javanica. The chemical profiles of the most-active filtrates or fractions were analyzed by liquid chromatography-diode array detector-electrospray-high resolution mass spectrometry (LC-DAD-ESI-HRMS).

Liquid Culture of Trichoderma Strains
The strains were grown in potato dextrose broth (PDB), which was made by adding 200 g of potato fragmented in distilled water at the boiling point (1000 mL) for 15 min, then filtered and 20 g of dextrose (Difco, Baltimore, MD, USA) added. A volume of 200 mL of the medium was deposited in Roux bottles and sterilized in an autoclave at 121 °C, 15 lb pressure, for 15 min. For each Trichoderma strain, cultured in PDA (8 days), a mycelial disk (7 mm diameter) was added to the medium in each of three Roux bottles. PDB without a Trichoderma strain was used as a control (blank). Three replicates of these cultures

Liquid Culture of Trichoderma Strains
The strains were grown in potato dextrose broth (PDB), which was made by adding 200 g of potato fragmented in distilled water at the boiling point (1000 mL) for 15 min, then filtered and 20 g of dextrose (Difco, Baltimore, MD, USA) added. A volume of 200 mL of the medium was deposited in Roux bottles and sterilized in an autoclave at 121 • C, 15 lb pressure, for 15 min. For each Trichoderma strain, cultured in PDA (8 days), a mycelial disk (7 mm diameter) was added to the medium in each of three Roux bottles. PDB without a Trichoderma strain was used as a control (blank). Three replicates of these cultures were incubated at 25 ± 2 • C, 12/12 h light/dark for 31 days. The mycelium was then removed from the culture broth by filtration through a double layer of cheesecloth. Each culture filtrate (CF), designated as 100% concentration, was then diluted with distilled water to 50% and 25% concentrations. The pH of 5 mL of the 100% CF was measured, then stored at 4 ± 2 • C until organic extraction (1-3 days).

Preparation of Fractions from Culture Filtrates
Each CF was liquid-liquid extracted with ethyl acetate three times (2:1, 1:1, 1:1 v/v), obtaining an ethyl acetate (EA) and residual filtrate (RF) fractions. The EA fraction was dried over sodium sulfate (Merck, New Jersey, USA), and the solvent was vacuum-evaporated at 40 • C in a rotary evaporator (IKA model RV-10, Staufen, Germany). The residual solvent in the RFs was also removed by evaporation, and the residue was designated as 100% concentration. The PDB control was processed the same way. All EAs were stored at 4 • C, and the RF fractions were frozen.

Nematode Inoculum
The population of M. incognita was obtained from the Tecnológico Nacional de México/ campus Conkal, Yucatán, México (30 ± 2 • C, 90% relative humidity) and M. javanica from the Instituto de Ciencias Agrarias, CSIC in Madrid, Spain (25 ± 1 • C, 70% relative humidity). Both nematodes were maintained on tomato plants (variety Marmanded) growing in pots in a greenhouse. Egg masses were collected from infested tomato roots and incubated for 72 h in sterile distilled water at 25 ± 2 • C for M. javanica and 30 ± 2 • C for M. incognita. The hatched J2s of Meloidogyne were adjusted to a final concentration of 100 J2 nematodes/100 µL distilled water to test CFs and RFs fractions (aqueous samples) and to 100 J2 nematodes/95 µL of distilled water solution to test EA samples [29,30].

Sample Preparation and Nematicidal Bioassay
Aqueous samples (100 µL) of a serial dilution (100, 50, or 25%) of either the CF or RF samples of a Trichoderma strain or a blank were deposited in wells of 96-well plates with U-bottom (BD Falcon, San Jose, CA, USA). The J2s suspended in distilled water (100 µL) that had been filtered through a 25 µm mesh screen were then transferred into each well. The negative controls consisted of CF or RF blanks, distilled water (DW), and 100 J2s. EA samples (5 µL) dissolved in DMSO:0.6%-Tween 20 (DT) were transferred to each well containing nematode suspension (95 µL), with a final concentration of 1 µg/µL. In this case, the negative control consisted of blank extracts, a mixture of water-DT 95:5 (WDT), and 100 J2s. Four replicates for each treatment were performed. The experimental plates were sealed with parafilm to prevent evaporation and maintained at the same conditions described above for egg masses in the dark [29,31].
After 72 h, immobile and rigid J2s that lacked intestinal contents were counted as dead, using a binocular microscope, and expressed as the percentage of juvenile mortality (%M). The nematicidal activity data were corrected using Schneider-Orelli's formula [32]. A completely randomized design was used, and means were compared using the Scott-Knott test (p ≤ 0.05) in the statistical package Infostat Ver. 2018 [33].

Liquid Chromatography-Diode Array Detector-Electrospray-High Resolution Mass Spectrometry
The active EA and RF fractions were freeze-dried (Labconco FreeZone 2.5, model 7670520, Houston, TX, USA) and dissolved to 1% w/v with methanol, then 3 µL was injected onto a C8 column (Zorbax SB, 2.1 × 30 mm) in an Agilent 1200 liquid chromatograph (LC, Santa Clara, CA, USA) coupled to an Agilent diode array detector and a Bruker Maxis HR-QTOF mass detector (HRMS Bruker GmbH, Bremen, Germany). The samples were separated at 40 • C with a flow of 300 µL/min. The mobile phase was a mixture of wateracetonitrile 90:10 v/v 0.01% trifluoroacetic acid and 1.3 mM ammonium formate (solvent A) and 10:90 v/v 0.01% trifluoroacetic acid and 1.3 mM ammonium formate (solvent B). The gradient was from 10% to 100% of solvent B in 6 min, maintained in 100% B for 2 min, then returned to 10% B for 2 min [34]. Mass spectra (50 to 2000 m/z) in the positive mode were acquired, and components detected were compared against the MEDINA Microbial Dereplication Databases (with approximately 900 known bioactive molecules), the Chapman & Hall Dictionary of Natural Products (v25.1, CRC Press, Boca Raton, FL, USA), and a database available in the literature.
were acquired, and components detected were compared against the MEDINA Microbial Dereplication Databases (with approximately 900 known bioactive molecules), the Chapman & Hall Dictionary of Natural Products (v25.1, CRC Press, Boca Raton, FL, USA), and a database available in the literature.
In general, against both nematodes, the CFs and their RF and EA fractions from T. harzianum Th43-14 and T. virens Th27-08 were the most active. In addition, the RF fraction from CF of T. koningiopsis induced J2s mortality of both nematodes at the lowest dilution tested.

Identification of Components in Active Fractions from Trichoderma Strains by LC-DAD-ESI-HRMS
The results obtained from analyses of the liquid chromatograms and ultraviolet and high-resolution mass spectrometry data of the RF fraction from T. harzianum Th43-14, T. ghanense Th26-52, and T. virens Th27-08 and the EA fraction of T. koningiopsis are shown in Tables 3-6 No significant differences were observed between the chromatograms from the active strains T. ghanense Th02-04 and the T. virens Th32-09 and negative control unfermented culture medium.
The compounds were tentatively identified by taking into account the species of the producing strain, the UV spectrum, and the molecular formula assigned through analysis of the protonated and ammonium adducts of each molecule, helped in some cases by the presence of dimers and dehydration products, and through searches of in the internal MEDINA database of HRMS spectra, dictionary of natural products and other natural product databases.

Metabolites from Residual Filtrate Fraction of Trichoderma harzianum Th43-14
The chromatogram of the RF fraction from T. harzianum Th43-14 showed eight main components ( Figure 3, Table 3). Five of these were tentatively identified as sesquiterpenes according to their UV and HRMS data. Component 5 (Rt = 2.86 min) presented in its HRMS an ammonium adduct (m/z 300.1807), a protonated molecular ion at m/z 283.1541, indicative of a molecular formula C 15  The largest peaks were from two unknown molecules (1 and 4), tatively assigned as 3,4,15-scirpenetriol. The HRMS of component 2 (Rt = 1.35 min) exhibited an ammonium adduct (m/z 316.1758), and a protonated molecular ion at m/z 299.1492 consistent with the molecular formula C15H22O6 (calc. C15H23O6 + , 299.1489) and was putatively identified as 3,7,8,15-scirpenetetrol. All compounds tentatively identified (2, 5, and 8) have the same structural skeleton.
The HRMS spectrum of the minor component 3 (Rt = 2.05 min) displayed several protonated fragments (m/z, 231.1383, 221.1544, 213.1272) and a protonated molecular ion at m/z 249.1488 in agreement with the molecular formula C15H22O6 (calc. for C15H23O6 + , 249.1489) and was putatively identified as illudin M. The HRMS spectrum of another minor component (6, Rt = 3.09 min) revealed a protonated molecular ion at m/z 249.1488 and was given a molecular formula C15H20O3 (calc. for C15H21O3 + , 249.1485) and tentatively identified as naematolin.
The largest peaks were from two unknown molecules (1 and 4),     Seven components were detected in the chromatogram of the residual fraction of T. ghanense Th26-52 ( Figure 4, Table 4). The most abundant component, 1 (Rt = 0.63 min) dis-played in its HRMS spectrum an ammonium adduct (m/z 278.1233) and a protonated molecular ion at m/z 261.0966, indicative of a molecular formula of C 11  Seven components were detected in the chromatogram of the residual fraction of T. ghanense Th26-52 ( Figure 4, Table 4). The most abundant component, 1 (Rt = 0.63 min) displayed in its HRMS spectrum an ammonium adduct (m/z 278.1233) and a protonated molecular ion at m/z 261.0966, indicative of a molecular formula of C11H16O7 (calc. C11H16O7 + , 261. 0973); this polar metabolite could not be identified. The HRMS of component 4 (Rt = 2.16 min) with a protonated molecular ion at m/z 169.0493 has a molecular formula C8H8O4 (calc. C8H9O4 + , 169.0500) and was putatively identified as atrichodermone D. The analysis of the HRMS data of the minor component 7 (Rt = 2.89 min) displayed an ammonium adduct (m/z 212.1640), and a molecular protonated ion at m/z 195.1379 supporting the molecular formula C12H18O2 (calc. C12H19O2 + , 195.1384) and tentatively assignment as the unsaturated lactone 6-heptyl-2H-pyron-2one. The other four unknown components included small metabolites with the molecular formulae of C8H9NO3 (Rt = 0.86 min), C8H10O3 (Rt = 1.22 min), C11H12O5 Rt = 2.40 min), and C14H18O3 (Rt = 2.64 min) according to their UV and HRMS data ( Table 4).

Metabolites from Ethyl Acetate Fraction of Trichoderma koningiopsis Th41-11
The chromatogram of the ethyl acetate extract from T. koningiopsis Th41-11 ( Figure 5) showed three components, which tentatively were assigned as koninginin isomers.

Metabolites from Ethyl Acetate Fraction of Trichoderma koningiopsis Th41-11
The chromatogram of the ethyl acetate extract from T. koningiopsis Th41-11 ( Figure 5) showed three components, which tentatively were assigned as koninginin isomers. The     The chromatogram of RF fraction of T. virens Th27-08 displayed four components not present in the blank sample ( Figure 6, Table 6). Component 3, eluting at Rt of 1.00 min exhibited an ammonium adduct (m/z 242.1020), dehydration fragments (m/z 207.0652, 191.1424), and a protonated ion at m/z 225.0757 indicative of a molecular formula of C 11 H 12 O 5 (calc. for C 11 H 13 O 5 + , 225.0762). Based on HRMS and UV data, compound 1 was tentatively identified as sepedonin. The major (1) and two minor components (2 and 4) were not identified, and molecular formulae were assigned as C 11  191.1424), and a protonated ion at m/z 225.0757 indicative of a molecular formula of C11H12O5 (calc. for C11H13O5 + , 225.0762). Based on HRMS and UV data, compound 1 was tentatively identified as sepedonin. The major (1) and two minor components (2 and 4) were not identified, and molecular formulae were assigned as C11H10O4 (Rt = 0.78 and 1.57 min) and C16H20O4 (Rt = 3.35) based on their protonated HRMS ion [M+H] + and additionally supported by its ammonium and dimer adducts.

Discussion
The results of this study complement our previous discoveries on native Trichoderma strains with nematicidal properties against two Meloidogyne species as part of our continuing work to find and develop safer biocontrol agents. Herein, we demonstrated that 92% of the screened tropical Trichoderma strains (all 13 except T. simmonsii) are mortal to J2s of Meloidogyne javanica and confirmed that the CFs and the EA or RF fractions of all strains were highly mortal to M. incognita. T. koningiopsis UFSMQ40 [12], T. harzianum Th6, JX1614550 [18,19,35], and Trichoderma sp. EF1671 [5] have been shown to have nematicidal

Discussion
The results of this study complement our previous discoveries on native Trichoderma strains with nematicidal properties against two Meloidogyne species as part of our continuing work to find and develop safer biocontrol agents. Herein, we demonstrated that 92% of the screened tropical Trichoderma strains (all 13 except T. simmonsii) are mortal to J2s of Meloidogyne javanica and confirmed that the CFs and the EA or RF fractions of all strains were highly mortal to M. incognita. T. koningiopsis UFSMQ40 [12], T. harzianum Th6, JX1614550 [18,19,35], and Trichoderma sp. EF1671 [5] have been shown to have nematicidal effects against J2s of M. javanica, but our report is the first to demonstrate the nematicidal activity of T. citrinoviride, T. ghanense, and T. virens against M. javanica.
The mortality data revealed that M. javanica was less sensitive than M. incognita to the CFs and EA fraction from Trichoderma strains. This differential sensitivity behavior of both Meloidogyne species against extracts, compounds, or fungal strains has been previously reported. For example, fungus Arthobortys sp. MVD18 caused less mortality against M. javanica (92.9%) than against M. incognita (99.0%) after a 48 h exposure. However, two Trichoderma sp. strains (KAV2 and KAV3) were more active against M. javanica than M. incognita [36]. Acetic acid and hexanoic acid yielded an EC 50 of 162.4 and 339.3 µg/mL, respectively, against M. javanica but EC 50 of 38.3 and 41.1 µg/mL against M. incognita, respectively, after 1 day [37,38]. Sensitivity differences between different species and even different populations of the same species have been attributed to the habitat and environmental conditions to which the organisms are exposed. Other saprophytic fungi have been reported to have low nematicidal effects against J2s of M. javanica; for example, a CF of Arthrobotrys oligospora, A. conoides, and Hypocrea lixii (sexual state of T. harzianum) at 100% concentration [16,39] caused from 16.14 to 64.5% mortality. An EA extract from Trichoderma sp. EFI 671, however, was not active [5].
The present study is also the first nematicidal bio-guided fractionation of the CFs from Trichoderma species and screening the organic and residual aqueous fractions against J2s. After the fractionation, 69% of the RF fractions from the non-nematicidal CFs had a higher mortal effect on M. javanica. This effect could be attributed to antagonistic action between metabolites and the concentration of the metabolites in the RFs of the strains. By contrast, the nematicidal effect of the CF from T. harzianum strain Th20-07 was not confirmed, suggesting that enzymes were mainly involved in the nematicidal activity of strain Th20-07 and were subsequently denatured by the solvent during the fractionation. The antibiosis produced by chitinase and protease action from Trichoderma species has been widely described. For example, an enzymatic filtrate obtained from T. koningiopsis UFSMQ40 is mortal to J2s of M. incognita (90.4% mortality) and M. javanica (63.2%) after 24 h of exposure [15]. Chitinase (51.42 U/mL) and protease (4.27 U/mL) from T. harzianum ITCC 6888 caused high mortality (93%) against M. incognita [40].
On the other hand, in our study, the highest lethal activity against both Meloidogyne species was found for T. koningiopsis Th41-11, T. harzianum Th43-14, T. virens Th27-08, and the two T. ghanense strains tested. The most investigated of these species has been T. harzianum for its properties as a biological control agent and its metabolites [2,4,14]. Other in vitro studies of CFs from T. harzianum strains grown in PDB have shown a significant mortal effect on the J2s of M. incognita. For example, CFs from T. harzianum strains ThU, JX1614550 and Th.6 caused a mortality of 33, 64.5 and 75%, respectively, at 100% concentrations after 72 h [19,41,42].
From the results of the present study, T. koningiopsis Th41-11 is another promising strain that produces nematicidal metabolites. The LC chromatogram of its EA fraction revealed small amounts of koninginins B, L, and T, tentatively identified based on UV, ESIHRMS data analyses, and comparisons with the literature. Koninginin B is a bicyclic polyketide that has also been reported from T. koningii [61], T. neokoningin [62], and T. applanatum [63]. Koninginin L and T are tricyclic polyketides with an oxygen bridge (C7 and C9) at C2 and an alcohol group at C4. Koninginins B, L, and T from T. koningiopsis QA3 and YIM PH30002 were recently reported to be weakly antibacterial [64][65][66].
Among the four T. virens strains studies, only Th27-08 had activity against both nematodes, with the RF fraction achieving the highest mortality. In the LC chromatogram of the RF fraction, we detected sepedonin, a tropolone first isolated from Sepedonium chrysospermum Fries (teleomorph Hyphomyces chrysospermus Tul.) and later from S. ampullosporum, S. chalcipori, S. microspermum, and S. chrysospermum and having antimicrobial activity against several bacterial and fungal pathogens [67]. The artifact anhydrosepedonin (C 11 H 10 O 4 ) is produced during the isolation process due to the instability of sepedonin [68,69]. We additionally detected three unknown metabolites in the RF fraction of T. virens Th27-08. Other metabolites previously reported from T. virens include cathequin, caffeic acid, ferulic acid, and 33 other non-volatile metabolites [2,55,70,71]. Our report of sepedonin is thus a new contribution to the chemical composition of this species.
Except for illudin M, the metabolites reported from our native Trichoderma species were not previously tested on nematodes [48]. The nematicidal efficacy of the metabolites tentatively identified from Trichoderma species herein studied are likely due to the alcohol or carboxylic acid groups in their structure. Ntalli et al. [38] reported that acetic acid and hexanoic acid, furfurol (syn. furfuryl alcohol), and furfural paralyzed J2s an EC 50/1 h of 1-100 µg/mL or less after 24 h, and the alcohols and aldehyde were more effective than the organic acids. They also demonstrated that acetic acid damages the cuticle and, the nuclei of pseudocoel cells and vacuolizes the cytoplasm of M. incognita, while (E)-2-decenal, and undecanone induced malformation of somatic muscles of the nematodes [72].
In general, few metabolites with nematicidal properties have been isolated from Trichoderma species; some of these are acetic acid, gliotoxin, trichorzianine, viridin [3], trichodermin [73], and cyclonerodiol [65,74]. Therefore, more studies should focus on bioassay-guided isolation and characterization of metabolites responsible for nematicidal activity in the fractions from Trichoderma strains. In addition, studies to optimize the production of the extracts with the most promising active compounds, evaluate their efficacy in greenhouses, and assess their toxicity on plants and beneficial organisms must be carried out before the compounds can be evaluated in the field and further developed as safe bionematicide products.

Conclusions
The present contribution enriches our knowledge of the nematicidal potential of 13 tropical Trichoderma species isolated from the soils of Yucatán state. The most effective species against M. incognita and M. javanica were T. ghanense strains Th02-04 and Th26-52, T. harzianum Th43-14, T. koningiopsis Th41-11, and T. virens Th27-08. The LC-DAD-ESIMS chemical profiles of the residual filtrate fractions and ethyl acetate fractions of culture filtrate from Trichoderma spp. revealed the presence of novel metabolites for the genus and others with molecular formulas not found in natural products databases. These results highlight the ability of Trichoderma strains to produce bioactive secondary metabolites that could be developed to manage M. incognita and M. javanica. However, more studies are needed to determine activity and potential phytotoxicity in plant applications and doses for effective, efficient biocontrol effect.