Meloidogyne enterolobii-induced Changes in Guava Root Exudates Are Associated With Root Rotting Caused by Neocosmospora falciformis

The procedures followed those outlined by Gomes et al. (2013). Briefly, seedlings of M. enterolobii-susceptible guava ‘Paluma’ were produced from stem cuttings and individually transferred to black plastic plant-growth tubes filled with 100 mL of autoclaved, washed riverbed sand. The base of each tube was closed with a plug of sterile cotton wool. The plants were kept in a growth chamber under a 12-hr photoperiod, with illumination of 35 μmol/(m²/s) provided by Grow-Lux (Sylvania, Lisboa, Portugal) light tubes. The day/night temperatures were set at 25 and 23 °C, respectively. The plants were watered with sterile, distilled water as needed. Biweekly, the plants were sprayed with foliar fertilizer Ouro Verde (Mato Verde Jardinagem, Bady Bassitt, Brazil), which provided nitrogen (urea), phosphorus (H₃PO₄), potassium (KCl), manganese, iron, boron, zinc, calcium, sulfur, and magnesium. For fertilizer application, the sand was covered with a plastic sheet to avoid getting drops on it.

Meloidogyne enterolobii was collected from a guava orchard (location coordinates: −21.649131; −41.042880) affected by guava decline. The nematode identity was ascertained by morphology and esterase phenotyping (Esbenshade and Triantaphyllou, 1985). The nematode was cultured in greenhouse-grown tomato plants. For inoculum preparation, tomato roots were cleaned with tap water and processed for nematode extraction through a modified version of the method detailed by Coolen and D’Herde (1972), i.e., without adding kaolin in the blender. The resulting suspension was passed through 60- and 500-mesh sieves (250- and 25-μm openings, respectively). The nematode eggs and second-stage juveniles (J₂) retained in the 500-mesh sieve were resuspended in water, and three 1-mL aliquots were counted on Peter’s slides under a stereomicroscope. The inoculum was calibrated to 200 eggs + J₂/mL of water. When the plants had rooted well and emitted two pairs of true leaves, 10 mL of inoculum (2000 eggs + J₂) were applied per plant, in two 3-cm-deep holes that were made in the sand around the plant. One hundred plants were inoculated (NP plants) and 100 plants received water only (NF plants).

When NP plants had developed abundant root galls, watering was suspended for 72 hr. The NP and NF plants were then watered with 50 mL of sterile, distilled water. This resulted in about 10 mL of root exudate per plant, which percolated through the cotton-wool plug. The NP and NF root exudates were separately pooled in sterile glass flasks wrapped in aluminum foil to minimize exposure to light. The root exudates were passed twice through a Whatman #1 paper filter (Whatman plc, Maidstone, UK) using a vacuum pump, and through a 0.22-μm-opening Teflon (Chemours, Wilmington, DE, USA) Millipore (MilliporeSigma, Burlington, MA, USA) filter membrane using the TTP vacuum filtration system (TTP, Switzerland). The exudates were collected into sterile, sealed TTP containers, frozen immediately, and then lyophilized in an L101 lyophilizer (Liotop, Brazil) within 2–3 days. Over about eight months, the 100 NP and 100 NF plants provided about 40 L of root exudates in each group. After lyophilization, this resulted in 23.1 g of NP and 16.4 g of NF dried matter.

Extracts were obtained with organic solvents of increasing polarity - hexane, ethyl acetate, methanol, and water. The NP dry matter (21.1 g) was immersed in 150 mL of hexane and agitated for 30 min at room temperature. The suspension was filtered through a cotton wool plug, and the solid residue underwent two more extractions with 150 mL of hexane each. The three liquid fractions were pooled, concentrated to dryness in a rotary evaporator, and lyophilized. The hexane-insoluble residue was then submitted to sequential extractions with ethyl acetate (3 × 150 mL), methanol (3 × 150 mL) and water (3 × 150 mL). The same extraction procedure was followed with the 14.6 g of NF dried matter, except that the volume of the solvents was 100 mL. The masses of the hexane, ethyl acetate, methanol, and water NP extracts (NP-01, NP-02, NP-03, and NP-04, respectively), and of the NF extracts (NF-01, NF-02, NF-03, and NF-04) are shown in Supplementary Table 1.

Seeds of guava ‘Paluma’ were germinated in autoclaved sand. When the cotyledon leaves had spread, the seedlings were sequentially rinsed with sterile, distilled water, immersed in commercial bleach at 1% active chlorine for 1 min, and then rinsed with sterile, distilled water. The seedlings were then individually transferred to sterile 30-mL glass tubes filled with 20 g of autoclaved, washed riverbed sand (Fig. 1A). The seedlings were maintained in a growth chamber under a 12-h photoperiod, with an illumination of 35 μmol/(m²/s¹) provided by Grow-Lux light tubes. The day and night temperatures were set at 25 and 23 °C, respectively.

Figure 1:

Neocosmospora falciformis isolate UENF/CF 163 originated from an orchard (location coordinates: −21.649131; −41.042880) affected by guava decline (Gomes et al., 2011). This isolate was maintained in anhydrous silica gel. To keep it from losing its pathogenicity to guava, it was routinely revived in potato-dextrose-agar medium, inoculated in guava seedlings, re-isolated, and preserved.

For seedling inoculation, N. falciformis was cultured in 1.5%-agar-water medium. A 0.5-cm plug was collected from the edge of the colony and positioned in the seedling collar region. The inoculated seedlings were watered daily with 2-mL aqueous solutions prepared from the NP-01, NP-02, NP-03, NP-04, NF-01, NF-02, NF-03 or NF-04 extracts. These watering solutions were prepared so as to match the extract concentration in the root exudates collected from the guava plants, according to the formula C_F = (C_E × m_F)/m_E, in which C_F is the fraction (NP-01 to NP-04 or NF-01 to NF-04) concentration in the watering solution; C_E is the concentration of the exudate solution; m_F is the fraction mass; and m_E is the exudate mass from which the fraction was obtained. The blank controls consisted of fungus-inoculated and uninoculated seedlings watered with 2 mL of sterile, distilled water. Each of the 10 treatments had eight replicates, for a total of 80 seedlings. The seedlings were arranged in trays in the growth chamber in a completely randomized design, separated from each other by a few centimeters.

Fifteen days after fungus inoculation, the seedlings were carefully removed from the glass tubes and examined for collar and rootlet necrosis. In the treatments where the fungus caused necrosis (see Results), the tissues decayed quickly, so we measured the rootlets’ unrotten fresh weight (mg) and the shoots’ fresh weight (mg). This bioassay was repeated once.

Data from both bioassays were tested for homogeneity of variances (Cochran and Bartlett tests) and for normality of errors (Lilliefors test), at 5% probability. No data transformation was necessary. Data were submitted to ANOVA, with time being one of the factors. Since the F-test revealed no significance (p > 0.05) for the factor time, data from both assays were pooled. The treatment means were compared by the Tukey HSD test at 5% probability.

Only NP-03 and NP-04 enabled N. falcifomis pathogenicity to guava seedlings (see Results). Quantification of amino acids, soluble carbohydrates, and sucrose (mg/mL) in NP-03, NP-04, NF-03 and NF-04 were performed using the method developed by Passos (1996) and modified by Campos et al. (2012). Phenol quantification (mg/mL) was performed using a modified version of the method of the Association of Official Analytical Chemists (Anonymous, 1996), and alkaloids were quantified (mg/mL) using of Sreevidya and Mehrotra (2003)’s method. A detailed description of these methods is contained in the Supplementary Material and Methods. All quantifications were conducted twice.

For each chemical group or substance, data were tested for homogeneity of variances (Cochran and Bartlett tests) and for normality of errors (Lilliefors test), at 5% probability. No data transformation was necessary. Data were submitted to ANOVA, with time being one of the factors. Since the F-test revealed no significance (p > 0.05) for the factor time, data from both quantifications were pooled. For each chemical group/substance, the concentrations in the different extracts were compared by the Tukey test at 5% probability.

NP-03 and NP-04 were the extracts that enabled N. falciformis pathogenicity, and the former had considerably more mass (Supplementary Table 1), and thus it was chosen for further characterizations. Adapting the method of Still et al. (1978), a NP-03 aliquot of 4.6 g underwent fractionation by elution through a 7 × 15-cm column of flash silica gel (Merck, 40-63 μm). Five eluents were consecutively used (700 mL each): ethyl acetate/methanol (9:1, v/v); methanol; water; water/acetic acid (99:1, v/v); and 0.1 mol/L aqueous hydrochloric acid. The five fractions obtained were concentrated to dryness in a rotary evaporator, and lyophilized. Their masses were 0.017 g (NP-03-F1), 3.782 g (NP-03-F2), 0.135 g (NP-03-F3), 0.254 g (NP-03-F4), and 0.266 g (NP-03-F5).

The seedlings were germinated, inoculated, and watered like in bioassay 1, with one adaptation: the guava seedlings were maintained in water in Eppendorf tubes (Figure 1B) (Eppendorf, Hamburg, Germany). This reduced the volume of the watering solution to 1 mL daily. Consequently, less of the fractions' mass was spent to prepare the solutions. To prepare the watering solutions from the NP-03 fractions, we applied the formula described before. The blank controls consisted of fungus-inoculated and uninoculated seedlings watered with 2 mL of sterile, distilled water. Eight replicates were used for each treatment, for a total of 56 seedlings. Evaluations were conducted as described before. This bioassay was repeated once, and the statistical analysis was conducted as previously described.

Only NP-03-F2 enabled N. falcifomis pathogenicity to guava seedlings (see Results). NP-03-F2 was analyzed by thin layer chromatography (TLC), using silica-coated aluminum plates with a 250-μm-thick layer of fluorescent indicator (Merck, Germany). Initially, various solvents and combinations thereof were used for vertical development. For visualization of the substances after development, the plates were dried (~70 °C) and observed under both daylight and 254-nm UV light, before and after derivatization, with either iodine vapor or ninhydrin at 0.002g/mL in ethanol. The best result was obtained using ethyl acetate/methanol (7:3) for development, which afforded a retention factor (Still et al., 1978) of around 0.23 for the major components.

A 1.52-g aliquot of NP-03-F2 was eluted through a 4-by-15-cm flash silica gel column, with ethyl acetate/methanol (7:3; 600 mL), ethyl acetate/methanol (3:7, 400 mL), and water (400 mL). Twenty-nine 40-mL fractions were collected, except for the eluent water, for which a single fraction was collected. Fractions that were similar according to TLC analysis were pooled, resulting in eight fractions that were coded NP-03-F2-1 through NP-03-F2-8. These fractions were concentrated to dryness in a rotary evaporator and lyophilized. Their masses are shown in Supplementary Table 2.

Fractions NP-03-F2-3, NP-03-F2-4, NP-03-F2-5 and NP-03-F2-6, which had the greatest masses, underwent this bioassay. The procedures for seedling germination, inoculation, preparation of the watering solutions, and watering were as in bioassay 2. The controls consisted of fungus-inoculated seedlings, watered with sterile, distilled water or fraction NP-03-F2. Eight replicates were used for each treatment, for a total of 48 seedlings. Evaluations were conducted as described before, but a less-severe decay of tissues allowed us also to measure the rootlets’ rotten fresh mass (mg). This bioassay was repeated once. The statistical analysis was conducted as described before.

Only NP-03-F2-3 and NP-03-F2-4 enabled N. falciformis pathogenicity in guava seedlings (see Results). These fractions were combined due to their similar chromatographic behavior and were thereafter referred to as NP-03-F2-3+4. A 371.8-mg aliquot of NP-03-F2-3+4 underwent consecutive extractions with diethyl ether (2 × 35 mL), ethanol (4 × 35 mL), and methanol (4 × 35 mL). The liquid fractions and the final residue were concentrated to dryness in a rotary evaporator and lyophilized. Their masses are in the Supplementary Table 3.

The fraction with the greatest mass—NP-03-F2-3+4-ET—was eluted through a flash silica gel column with ethyl acetate/methanol (8:2, 400 mL; 6:4, 400 mL; 1:1, 200 mL). Fifty 20-mL fractions were collected, followed by a single 300-mL methanol fraction. The fractions were analyzed by TLC using silica-coated aluminum plates impregnated with fluorescent indicator F254. For visualization of the substances after development, the plates were dried (~70 °C) and observed under daylight and 254-nm UV light, before and after exposure to iodine vapor. The plates were then dipped in a solution of 0.07 g/mL of phosphomolybdic acid in ethanol, and heated. The fractions were pooled according to the TLC analysis, concentrated to dryness in a rotary evaporator, and lyophilized. Their masses are enumerated in Supplementary Table 4.

Fractions NP-03-F2-3+4-ET-02 and NP-03-F2-3+4-ET-05 had pure substances according to TLC analysis. Therefore, they were analyzed by HPLC in a Shimadzu (Kyoto, Japan) apparatus equipped with a DAD SPD-M20A-type UV-visible detector, two pumps LC-6AD, Rheodyne 7725i-type manual injector (all also Shimadzu, Kyoto, Japan), and a Luna 5μm column (phenyl-hexyl, 250 by 4.60 mm) (Phenomenex, Torrance, CA, USA). The eluent was ultrapure water. The fractions were also submitted to elemental analysis in an Truspec Micro Element Analyzer (LECO, USA). Approximately 2 mg of each fraction was incinerated at 1075 °C in a quartz tube, and the resulting gases were analyzed with infrared and thermal detectors for carbon and nitrogen quantification, respectively.

The fractions were also submitted to analysis by EDXRF using a Shimadzu EDX-720. This analysis was carried out in the qualitative/quantitative mode under reduced pressure, for detection of chemical elements from ¹¹Na to ²³⁸U. To obtain ¹H and ¹³C NMR spectra, the fractions were dissolved in hexadeuterated dimethylsulfoxide (DMSO-d₆), and analyzed in an Avance III (Bruker, Carteret, NJ, USA) NMR spectrometer (¹H: 600.13 MHz; ¹³C: 150.9 MHz). To obtain the mass spectra, the fractions were dissolved in methanol, and injected into a Daltonics microTOF-Q apparatus (Bruker, Carteret, NJ, USA) equipped with an ESI-type ionizer.

Data from mass spectrometry, elemental analysis, NMR and EDXRF indicated 1,5-dinitrobiuret as the substance present in the fraction NP-03-F2-3+4-ET-05 (see Results). Therefore, conformational searches for all possible tautomers of 1,5-dinitrobiuret were undertaken, taking into account the geometric arrangements around the generated double bonds (Supplementary Fig. 1). For this, we used the computer programs Chemsketch 12.01 (Advanced Chemistry Development, Toronto, Canada); OpenBabel 2.3.2 (O’Boyle et al., 2011); Open3Dalign 2.282 (Tosco et al., 2011); Mopac 2016 (http://openmopac.net/, J.J.P. Stewart, Stewart Computational Chemistry, Colorado Springs, CO, USA), with the conductor-like screening model COSMO (Klamt and Shüürmann, 1993); and NWChem 6.8.1 (Aprà et al., 2020). A detailed description of the computational calculations is contained in the Supplementary Material and Methods.

After the conformational searches described above, 1,5-dinitrobiuret and its tautomers, biuret and its tautomers, and different protonation states of N-formyl-D-aspartic acid (Supplementary Figs. 2, 3, and 4, respectively), were inserted into the binding sites of biuret hydrolases (Esquirol et al., 2018) obtained from the RCSB Protein Data Bank (https://www.rcsb.org/), using the computer programs Lovoalign 21.027 (Martínez et al., 2007) and UCSF Chimera 1.15 (Pettersen et al., 2004). The resulting complexes were submitted to an optimization step with Chimera 1.15 and Antechamber (Wang et al., 2006), which employed the General Amber Force Field (GAFF) (Wang et al., 2004). Cyscore 2.0.3 (Cao and Li, 2014) was then used to calculate the affinities of the substances to the enzymes. A detailed description of these procedures is available in the Supplementary Material and Methods (Supplementary Figures 5–7). The resulting values underwent normality (Shapiro-Wilk) and homoscedasticity (Bartlett) tests at 5% probability (Snedecor and Cochran, 1989). Data was submitted to ANOVA, and means were compared using the Scott-Knott test at 5% probability (Scott and Knott, 1974), employing the R software.

Virtually all masses in the NP and NF root exudates were soluble in water or methanol (Supplementary Table 1). Among all extracts, only NP-03 and NP-04 enabled N. falciformis pathogenicity to guava seedlings (DF = 9; F = 6.4721; p < 0.05) (Table 1, Fig. 1C).

The concentration of amino acids, carbohydrates and sucrose was significantly (p < 0.05) higher in NP than in NF extracts (DF = 3.2, 3.2 and 3.2, respectively; F = 1971010.67, 88610.84 and 61468163.7, respectively) (Fig. 2). No clear distinction (p > 0.05) occurred between NP and NF extracts in phenol concentration (DF = 3.2; F = 1139.67). Regarding alkaloids, NF extracts had slightly higher (p < 0.05) concentration than NP extracts (DF = 3.2; F = 6085.62).

Figure 2:

Among the NP-03 fractions, only NP-03-F2 enabled fungal pathogenicity to guava seedlings (DF = 6; F = 5.4567; p < 0.05) (Table 2; Figs. 1D,E). Among the NP-03-F2 fractions, only NP-03-F2-3 and NP-03-F2-4 enabled fungal pathogenicity (DF = 5; F = 3.9741; p < 0.05) (Table 3, Fig. 1F).

Solvent extractions of NP-03-F2-3+4 resulted in an ethanol-soluble fraction with the greatest mass, which was coded NP-03-F2-3+4-ET (Supplementary Table 3). Fractionation of NP-03-F2-3+4-ET resulted in seven fractions that appeared to be composed of different amounts of two substances. TLC analysis suggested that fractions NP-03-F2-3+4-ET-02 and NP-03-F2-3+4-ET-05 were two pure substances (Supplementary Table 4). HPLC analyses confirmed that fraction NP-03-F2-3+4-ET-05 corresponded to a pure substance, while NP-03-F2-3+4-ET-02 seemed to correspond to another substance that had been contaminated by the substance present in the fraction NP-03-F2-3+4-ET-05 (Supplementary Fig. 8).

The elemental analysis of NP-03-F2-3+4-ET-05 revealed a C:N ratio of 1.93:5.16, which can be rounded up to 2:5. For NP-03-F2-3+4-ET-02, the ratio was 1.93:8.61, which can be rounded to 2:9. The EDXRF analysis revealed that the substances present in these fractions could only be composed of chemical elements with atomic numbers lower than that of sodium. For these substances, the ¹H NMR spectra presented broad signals centered at approximately 7.40 ppm. In the ¹³C NMR spectra, a single signal occurred at 159.5 ppm.

An analysis by mass spectrometry of fraction NP-03-F2-3+4-ET-05 showed two main signals in the negative mode, with respective m/z (mass/charge) of 147.0171 and 210.0119 u. For NP-03-F2-3+4-ET-02, a signal with m/z of 147.0142 u was revealed in the negative mode, while in the positive mode the main signal had m/z of 441.0401 u.

Among all the possible conformations and tautomers for 1,5-dinitrobiuret (Supplementary Fig. 1), one corresponded to approximately 89.9% of the total according to the Boltzmann distribution (Fig. 3A). All other structures had values equal to or less than 1%. In the majority structure, there were double bonds between the N atoms of the nitro groups and the neighboring N atoms. A 3D representation revealed a non-planar molecule with a dihedral angle HO-N-N-CO of 11.7 ° (Figs. 3B,C).

Figure 3:

N-formyl-D-aspartic acid, in all protonation states (ASP*), showed equal (p < 0.05) affinities to biuret hydrolase, although their values were lower (p < 0.05) than those of biuret and its tautomers (BIU*) (Fig. 4). Regarding the calculated values for 1,5-dinitrobiuret and its tautomers (DNB*), most of them were lower (p < 0.05) than those calculated for BIU*.

Figure 4: