Soil organic matter attenuates the efficacy of flavonoid-based plant-microbe communication

Abstract

Plant-microbe interactions are mediated by signaling compounds that control vital plant functions, such as nodulation, defense, and allelopathy. While interruption of signaling is typically attributed to biological processes, potential abiotic controls remain less studied. Here, we show that higher organic carbon (OC) contents in soils repress flavonoid signals by up to 70%. Furthermore, the magnitude of repression is differentially dependent on the chemical structure of the signaling molecule, the availability of metal ions, and the source of the plant-derived OC. Up to 63% of the signaling repression occurs between dissolved OC and flavonoids rather than through flavonoid sorption to particulate OC. In plant experiments, OC interrupts the signaling between a legume and a nitrogen-fixing microbial symbiont, resulting in a 75% decrease in nodule formation. Our results suggest that soil OC decreases the lifetime of flavonoids underlying plant-microbe interactions.

Fig. 1. Naringenin bioavailability changes with the total C content of soils and amendment type.

(A) HPLC standard curve for naringenin. The dashed line indicates a linear fit. (B) Biosensor transfer function for naringenin. The CFP emission (λ_ex = 433 nm and λ_em = 475 nm) represents background-subtracted signal normalized to cell growth. The dashed line indicates a fit to the Hill function. For both standard curves, the data represent the average from n = 3, normalized to the average maximum signal in the absence of soil. (C and D) Inceptisol soils collected from three different sites (square, triangle, and circle) and three different land uses (agricultural, crossed circle; meadow, open circle; forest, filled circle) were incubated with naringenin for 24 hours, and the amount remaining in the supernatant was analyzed by (C) HPLC or (D) biosensor. With HPLC analysis (75 μM naringenin added), a fit of the data to y = −0.06172x + 1.022 yields an R² value of 0.88. With the biosensor (0.6 mM naringenin added), a fit of the data to y = −0.06393x + 0.9427 yields an R² value of 0.79. In both cases, higher OC contents correlate with a decrease in naringenin. (E and F) Naringenin was incubated with compost (+POC_comp), maple wood (+POC_plant), and maple wood that had been pyrolyzed at 550°C (+PyOM₅₅₀) or 750°C (+PyOM₇₅₀) for 24 hours. The proportion of naringenin remaining in the supernatant was quantified by (E) HPLC or (F) biosensor. POC_comp and POC_plant decreased naringenin bioavailability [one-way analysis of variance (ANOVA) with Dunnett’s multiple comparisons test, *P < 0.005 and **P < 0.001]. Error bars represent ±1σ calculated using n = 3.

Fig. 2. Naringenin bioavailability is decreased by chemicals in DOC.

Each amendment (25 mg ml⁻¹) was incubated in phosphate buffer (10 mM, pH 7) for 24 hours, and DOC_x (x = source) was obtained by centrifugation. Naringenin was incubated in DOC_x for 24 hours and then quantified using (A) HPLC (40 μM naringenin added to DOC_x) or (B) biosensor (0.6 mM naringenin added to DOC_x). DOC_plant and DOC_comp decreased naringenin compared with the control (one-way ANOVA with Dunnett’s multiple comparisons test, *P < 0.005 and **P < 0.001). (C) Evaluating the role of DOC enzymes using heat treatment. Untreated (+DOC_plant) and heat-treated (+DOC_plant +heat) maple wood DOC were mixed with naringenin (0.6 mM), and bioavailability was measured using the biosensor. (D) Examining the role of DOC enzymes on naringenin (75 μM) using sodium azide (NaN₃), a laccase inhibitor. HPLC did not detect any difference in naringenin availability between the control (+DOC_plant) and azide treatment (+DOC_plant +NaN₃). (E) Naringenin (0.6 mM) availability before (+DOC_plant) and after 1 mM EDTA addition to maple wood DOC (+DOC_plant +EDTA). EDTA increased naringenin compared with DOC_plant (unpaired t test, **P < 0.001), which implicates a metal-mediated reaction as the mechanism for naringenin reduction. Error bars represent ±1σ calculated using n = 3.

Fig. 3. Plant organic matter has varying effects on the availability of different flavonoids.

FdeR activates CFP expression in the presence of (A) naringenin or (B) luteolin. (C) QdoR activates methyl halide transferase (MHT) expression in the presence of quercetin. Biosensor dose-response curves for (D) naringenin, (E) luteolin, and (F) quercetin fit to the Hill function. Standard curves represent background-subtracted signals normalized to the average maximum signal. (G to I) Varying concentrations of POC_plant from maple (ac) and mesquite (pp) were incubated in phosphate buffer (10 mM, pH 7) for 24 hours, including 25 mg ml⁻¹ (25_ac), 50 mg ml⁻¹ (50_ac), and 100 mg ml⁻¹ (100_ac) of maple wood or 25 mg ml⁻¹ (25_pp) of mesquite wood. Samples were centrifuged to obtain DOC, which was mixed with (G) 0.6 mM naringenin, (H) 0.6 mM luteolin, or (I) 120 μM quercetin. After 24 hours, the biosensor was used to quantify bioavailable flavonoid. *P < 0.05 and **P < 0.001 indicate statistical significance for one-way ANOVA test with Dunnett’s multiple comparisons test between each treatment and the positive control. Positive (buffer and flavonoid only) and negative (buffer only) controls are shown in black. Error bars represent ±1σ calculated using n = 3.

Fig. 4. Chromatograms of DOC show a new peak after naringenin addition.

(A) Chromatograms of maple DOC before (+DOC_ac; gray line) and after (+DOC_ac + nar; red line) naringenin addition show that a peak (asterisk) appears with a retention time of 10.7 min that has (B) an m/z of (M + 1) = 573. Naringenin has a retention time of 12.1 min under identical conditions. (C) Chromatograms of mesquite DOC before (+DOC_pp; gray solid line) and after (+DOC_pp + nar; red solid line) naringenin addition reveal that a peak (asterisk) appears at a retention time of 10.3 min with (D) an m/z of (M + 1) = 561. (E and F) Structure of the chemical formed in DOC_pp following the addition of naringenin, which was determined using NMR spectroscopy. The bonding topologies of isolated products were analyzed (see text S1) and determined to be consistent with the structures of mesquitol-C(5)-C(6)-naringenin and mesquitol-C(5)-C(8)-naringenin heterodimers.

Fig. 5. Maple wood decreases the number of nodules in M. sativa plants.

The number of nodules in M. sativa plants grown on agar slants under different conditions is shown. Inoculations lacking E. meliloti (−Em) or containing E. meliloti (+Em) and samples lacking (−POC_plant) or containing maple wood within the agar slants (+POC_plant) are shown. Slants either contained no nitrogen (−N) or 5 mM NH₄NO₃ (+N). The addition of POC_plant significantly decreased the number of nodules in M. sativa (n = 10; two-tailed t test, P < 0.001).

Fig. 6. Environmental properties affecting flavonoid bioavailability in soils.

Following the release of flavonoids into the rhizosphere through passive or active exudation (green), these chemicals can interact with the soil microbiome and different soil components. Abiotic attenuation mechanisms (orange) include sorption by minerals and POC, interactions with DOC as described here, and leaching. Leaching enhances the signal dilution through the soil at high hydration conditions because the signal can readily diffuse through the soil column. Biological processes that specifically attenuate flavonoid communication (white) include consumption and biotransformation through the release of exoenzymes.