Interspecies-cooperations of abutilon theophrasti with root colonizing microorganisms disarm BOA-OH allelochemicals

ABSTRACT

A facultative, microbial micro-community colonizing roots of Abutilon theophrasti Medik. supports the plant in detoxifying hydroxylated benzoxazolinones. The root micro-community is composed of several fungi and bacteria with Actinomucor elegans as a dominant species. The yeast Papiliotrema baii and the bacterium Pantoea ananatis are actively involved in the detoxification of hydroxylated benzoxazolinones by generating H₂O₂. At the root surface, laccases, peroxidases and polyphenol oxidases cooperate for initiating polymerization reactions, whereby enzyme combinations seem to differ depending on the hydroxylation position of BOA-OHs. A glucosyltransferase, able to glucosylate the natural benzoxazolinone detoxification intermediates BOA-5- and BOA-6-OH, is thought to reduce oxidative overshoots by damping BOA-OH induced H₂O₂ generation. Due to this detoxification network, growth of Abutilon theophrasti seedlings is not suppressed by BOA-OHs. Polymer coats have no negative influence. Alternatively, quickly degradable 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one can be produced by the micro-community member Pantoea ananatis at the root surfaces. The results indicate that Abutilon theophrasti has evolved an efficient strategy by recruiting soil microorganisms with special abilities for different detoxification reactions which are variable and may be triggered by the allelochemical´s structure and by environmental conditions.

Introduction

Benzoxazinoids, secondary products of several Poaceae, support the control of certain weeds in sustainable organic farming systems. Mulches from young rye, for instance, contain the aglycon 2,4-dihydroxy-2H-1,4-benzoxazin-3(4H)-one (DIBOA) and the more stable derivative benzoxazolin-2(3H)-one (BOA).^1,2 These compounds leading to microbial transformation products, the phenoxazinones, are important for the phytotoxic properties of young rye residues which suppress weeds and crops like Lactuca sativa, Amaranthus retroflexus or Portulaca oleracea.^2-7 A key event in the mode of action is the induction of general stress responses in plant cells, such as the generation of reactive oxygen species (ROS). Oxidative stress induces membrane damage, lipid peroxidation, protein/nucleic acid oxidation and enzyme inhibitions.^7-9 BOA exposure enhances lignification and cell wall rigidity, due to an increased H₂O₂ production and alters proportions of carbohydrate polymers.^10,11 A strong increase of H₂O₂ was found in BOA treated Phaseolus aureus,¹² although the activity of antioxidant enzymes such as superoxide dismutase (SOD), peroxidases (POD) and catalase were significantly enhanced. In Lactuca sativa roots, SOD and POD activity is decreased, while lipid peroxidation is increased after exposure to 1 mM BOA.⁷ Some plants are able to detoxify the allelochemical.^13,14 When glucosylated, BOA cannot function as a precursor for the production of phenoxazinones, compounds which are highly toxic to many organisms including plants. After exposure to BOA, grasses and a few dicots accumulate mainly the BOA derived glucoside carbamate and related compounds whereas most dicotyledonous plants produce BOA-6-O-glucoside and, less widespread, BOA-5-O-glucoside as the major or minor detoxification products via the hydroxylated intermediates BOA-6/5-OH.^4,11 For many plants, the intermediate BOA-6-OH exhibits a higher toxicity than BOA and must be therefore channelled for immediate glucosylation latest after entering root protoplasts.^13,15 However, the resulting glucosides still possess a latent functionality, as β-glucosidases can cleave the sugar moiety and the aglycons are released. BOA-6-OH is therefore a detoxification intermediate of Abutilon when exposed to BOA, a realistic situation when Abutilon grows in maize fields or in other fields after application of mulches from rye plants.

In recent studies, velvetleaf (Abutilon theophrasti), one of the most noxious weeds worldwide, was found to be insensitive to BOA. The seedlings show an unusual detoxification behavior with low accumulations of BOA and of the detoxification product BOA-6-O-glucoside. Seedlings can develop a facultative association with the zygomycete Actinomucor elegans, a frequent fungus in soils rich in organic matter.^3,16,17 The fungus, which colonizes the primary root superficially, secretes tryptophan or tryptophan rich precursor peptides that contribute to plant strengthening. The amino acid is not only a precursor for auxin but inhibits the generation of phytotoxic phenoxazinone from the microbial BOA degradation product 2-aminophenol.^16,5,6 The avoidance of phenoxazinone accumulation is notable since this compound was found to act as histone deacetylase inhibitor.¹⁸

Although Abutilon theophrasti is not sensitive to BOA, more reactive derivatives may overcome the defense barriers. We speculated that direct application of hydroxylated BOA molecules or mixtures of them could be harmful to Abutilon. This consideration led to a study with hydroxylated BOA isomers (Fig. 1). Moreover, we expected insights in OH- position specific alterations of the bioactivity, known from many molecules such as hydroxybenzoic acids, phenylpropanoids or tyrosines.^19-21 However, during the studies it became apparent, that microorganisms associated with the roots contribute to the transformation and elimination of the different BOA-OH isomers. Here, actions of defined microorganisms as well as chemical and biochemical aspects of the interconnected activities with the plant resulting in detoxification and transformation of the compounds are presented.

Figure 1.

Structure of the BOA-OH isomers.

Results

The actinomucor elegans micro-community

Actinomucor elegans, re-isolated from Abutilon roots, was severely affected by BOA-4-OH, showing collapsed hyphae after the treatment when droplets of the BOA-4-OH solution were directly placed on the mycelium. No similar effects were observed with the other isomers. In areas with collapsed hyphae, microorganisms emerged which were previously hidden in the mycelium nexus (Fig. 2). Among the most viable microorganisms, Pantoea ananatis and Stenotrophomonas maltophilia were identified. The unknown unicellular fungus found in the collected blue precipitate after BOA-4-OH incubation was identified as a strain of Papiliotrema (Cryptococcus) baii, a recently described basidiomycetous (Agaricomycotina, Tremellomyceyes, Tremellales, Rhynchogastremaceae) yeast species which is not yet known as a root-colonizing organism or as a member of a microbial association.^22,23

Figure 2.

Actinomucor elegans culture plugs treated with A: BOA-4-OH, B: BOA-6-OH, C: BOA-7-OH and D: BOA-5-OH (black arrows). BOA-4-OH led to collapsed hyphae and hitherto hidden microorganisms emerge in jelly like colonies (red arrow heads).

Further PCR analyses using different DNA loci verified the yeast unequivocally as a species associated with Actinomucor elegans in all cultures of the micro-community used in this study (Fig. S2). Thus, on Abutilon roots, the yeast P. baii, the 2 Gammaproteobacteria Pantoea ananatis and Stenotrophomonas maltophilia, zygomycete fungus Actinomucor elegans, and not yet identified microorganisms converge to build up a rhizosphere micro-community. In further experiments we focus on the behavior of the yeast, Pantoea ananatis, of the entire micro-community, and on combinations with Abutilon.

Hydroxylated BOA isomer products in root extracts

The BOA-OH isomers were used for incubations of Abutilon seedlings (Fig. 3). Analyses of the methanolic root extracts indicated the formation of BOA-6-O-glucoside.^3,16 The amount of BOA-6-O-glucoside accumulating in the seedlings was similar to the one found after BOA incubation. BOA-5-OH is glucosylated yielding BOA-5-O-glucoside, also known from BOA incubated plants.¹⁴ Incubation of seedlings grown from purchased seeds (Herbiseed) with BOA-4-OH and BOA-7-OH led to new compounds which were not yet further identified as the isomers were no substrates for the NaCl extractable glucosyltransferase activity present in plant roots (see below). Seedlings from seeds harvested from the Meleti field produced more BOA-6-O-glucoside but less of BOA-5-O-glc and of the BOA-7-OH product. With BOA-4-OH no product was identified. Incubations of Herbiseed seedlings with combinations of the isomers reduced the accumulation of the BOA-6-O-glucoside dramatically, when BOA-7-OH and BOA-4-OH were present. BOA-5-O-glucoside accumulation was most affected by BOA-7-OH. Thus, the 5- and 6-OH isomers compete for glycosylation and the 4/7-OH isomers reduce glycosylation of the 5/6-OH isomers. Taking into account results described later, the reduction of glucosylations suggests potential alternative detoxification strategies. In contrast to our assumption, combinations of the isomers did not inhibit the growth of the Abutilon seedlings.

Figure 3.

Accumulation of BOA-6- and BOA-5-O-glucoside in roots and shoots of Abutilon seedlings. BOA-4-OH: -4-OH; BOA-5-OH: -5-OH; BOA-6-OH: -6-OH; BOA-7-OH: -7-OH (A: Herbiseed, B: Meleti-field). Compounds occurring with BOA-4-OH and BOA-7-are not identified. C: Combinations of the BOA-OH isomers led to a reduction of BOA-6-O-glucoside and BOA-5-O-glucoside accumulation. D: NaCl^-extracted glucosyltransferase activity with BOA-OH isomers as substrates. The fungus Actinomucor elegans favors the glucosylation of BOA-5-OH. BOA-4-OH and BOA-7-OH were not accepted as substrates.

Protein extracts prepared from roots of seedlings (from Herbiseed seeds) were assayed for UGT activity with the isomers as sugar acceptors. Corroborating the former finding that BOA-6-OH is glucosylated by a salt extractable glucosyltransferase,¹⁶ we observed that BOA-5-OH was the preferred substrate, when seeds were inoculated with the Actinomucor elegans micro-community. BOA-5-O-glucoside is, however, a minor detoxification product of Abutilon, when the plant is incubated with BOA. Thus, hydroxylation in position 5 is not preferred for efficient BOA detoxification in Abutilon. Neither BOA-4-OH nor BOA-7-OH was accepted for glucosylation. The results are in line with the former assumption that BOA detoxification via O-glucosylation has a subordinate role in the resistance of Abutilon against BOA and that BOA-6-O-glycosylation is reduced in the inoculated seedlings. Instead, BOA-6-O-glucoside can be detected occasionally in substantial amounts in the incubation medium (Fig. 7). Taken together, this study provides direct experimental evidence that detoxification processes may occur predominantly outside of plant protoplasts.

Figure 7.

HPLC-chromatograms of root surface compounds and BOA-6-OH incubation medium of Pantoea ananatis. Root surface detoxification compounds can vary. Their occurrence may depend on the incubation time, the diversity of the root colonizing microbiome and the viability of the seedlings. A: DSMO root wash after 4 h incubation with BOA-6-OH non-inoculated Abutilon seedlings. *: BOA-6-O-glucoside; Δ: BOA-6-OH dimer; •: 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one. B: For comparison, BOA-6-OH and derived compounds (traces of the BOA-6-OH dimer and 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one) in a Pantoea ananatis incubation medium. C: Medium of inoculated Abutilon seedlings after 4 h of incubation with BOA-6-OH. BOA-6-O-glucoside is present only in traces; Instead the dimer and small amounts of the nitro-compound could be found.

Reactions at the root surface

All BOA-OH isomers undergo polymerization reactions that differ depending on the position of the OH group. A generation of bubbles was observed with all isomers during the early phase of incubation (up to 4 h) at defined roots and was not restricted only to root hair zones. Bubble development was most dramatic and long lasting with BOA-7-OH but with BOA-6-OH only when additional P. baii was added (Fig. 4). With 2 mM BOA-7-OH, incubation media changed their color to a yellowish-brown, and root turned to brownish-yellow. The precipitate at the root surfaces was not soluble in DMSO. Application of 0.5 mM BOA-7-OH produced no polymer coats, indicating that the compound was degraded and polymers occurred only with high BOA-7-OH concentrations. Incubation with 0.5 mM BOA-4-OH led first to dark brownish spots or ring-like areas on the surface of the main roots and at break through sites of lateral roots. After application of higher concentrations of BOA-4-OH, all roots turned black within 4 h, but even after 24 h the viability of the seedlings was not negatively affected. Only BOA-4-OH led to sapphire-blue coloring of incubation media (Fig. 4). The black BOA-4-OH polymer coat on root surfaces could be completely removed with DMSO, which may indicate a melanin-like structure. Assays with laccase from Trametes versicolor and BOA-4-OH led to a suspension containing blue particles and oligomers similar to the ones found in the DMSO root wash. Therefore, we assume that laccases at root surfaces can participate in polymer generation.

Figure 4.

Upper panel - A: Blue medium and black roots after incubation with BOA-4-OH. Hypocotyls stay bright. Insert: Additional C. baii inoculation decolors the medium after BOA-4-OH incubation. B: Assays with 0.5 mU Trametes versicolor laccase and 250 µl 0.5 mM BOA-4-OH led to particles of a blue polymer (arrows). C: Polymer coats on roots after incubation with BOA-5-OH (left) and BOA-6-OH (right). Lower panel: A: Incubation with 0.5 mM BOA-7-OH led to severe bubble formation at some root surfaces within 30 min, but roots show no browning or blackening. B: Incubation with 2 mM BOA-7-OH turned roots reddish-brown without stop of bubble formation. B,C: Arrows point to roots with and without bubble formation. C: Abutilon seedlings inoculated with P. baii show a strong bubble development in presence of BOA-6-OH at defined sites of some roots (green arrows). Roots without bubble formation develop a brownish coat (purple arrow). The pinkish color of the medium results from the starting accumulation of the BOA-6-OH dimer.

Polymers generated from BOA-5-OH and BOA-6-OH colored roots in reddish or dark brown tones. Incubation media with BOA-6-OH turned brownish within 24 h. Unlike the BOA-6-OH polymer, the BOA-5-OH derived polymer could be removed from the roots with DMSO. The strong chemical responses at the root surfaces distinguished the extracellular space and the root surface as sites most important for the elimination of BOA-OH isomers by polymerization reactions.

Compounds in incubation media

Supporting the above described results, analyses of the incubation media revealed that oligo-/polymerization of the isomers starts almost immediately after the application of a compound (Fig. S3).

BOA-6-OH was first converted to a more hydrophobic compound with shifted UV maxima (BOA-6-OH: 230, 290 nm; new compound: 221, 302 nm), (Fig. S3), which could not be glucosylated in UGT assays. The compound subsequently polymerized to an insoluble product. For this reason, further attempts to purify the compound failed and it was not possible to determine the exact structure by NMR analyses. The concentrated medium was therefore subjected to UPLC-PDA -mass spectrometry to study the nature of the compound. According to the UPLC-PDA-MS measurements (Table S1), the new compound is a dimer of BOA-6-OH and represents the first intermediate of an oxidative polymerization process.

Media with BOA-5-OH did not contain higher amounts of transformation products. Several small peaks indicate the generation of a possible dimer and hypothetic oligomers. The major amount of the polymerized compound precipitated at the root surface, but the transformation was slower in comparison with the other isomers. With BOA-4-OH, only traces could be detected after 24 h incubation. Instead, numerous compounds, possibly oligomers, were found in media. BOA-7-OH media contained small amounts of a possible dimer and oligomers. The high polymerization activity at the root surface led to the consideration that A. elegans and the associated microorganisms take part in BOA-OH elimination. Moreover, polymerization reactions highlight the involvement of detoxification enzymes (e.g. peroxidases, laccases, polyphenol oxidases), H₂O₂/O₂ production and radical formation in the elimination of bioactive BOA-OH molecules. Thus, the application of the BOA-OH isomers seems to elicit first a process at the root surface which resembles an oxidative burst.

Enzyme activities in crude protein extracts from abutilon roots

Inoculation with the micro-community substantially decreases polyphenol oxidase activity in soluble protein extracts from roots of seedlings incubated with the BOA-OH isomers (Fig. 5). Peroxidase activity in soluble protein extracts from the roots was not significantly affected after compound application, except for BOA-5-OH, which increased the activity in inoculated seedlings. With syringaldazine as substrate, laccase activities in the root protein extracts were rather similar in protein preparations from inoculated and non-inoculated roots.

Figure 5.

Comparison of root enzyme activities of inoculated and non-inoculated plants. BOA-4-OH: -4-OH; BOA-5-OH: -5-OH; BOA-6-OH: -6-OH; BOA-7-OH: -7-OH; -: without inoculation, + with inoculation (complete micro-community). A: Cytosolic polyphenoloxidase: Except for BOA, all combinations (+/−) differ significantly. B: Root surface polyphenol oxidase: Except for the control, BOA-5-OH and -6-OH, differences between +/− samples are not significant. Inoculation led to lower activity in + samples in the control and BOA-6-OH, but higher activity with BOA-5-OH and in tendency with BOA-7-OH. C: Cytosolic laccase activity in soluble root protein extracts did not differ significantly except for control samples. D: Inoculation with the fungus resulted in a significant increase of laccase activity with BOA-4-OH. With the other compounds, differences between +/− samples were not significant. Compared to the controls, all compounds except for BOA-6-OH increased soot surface laccase activity with ABTS as substrate significantly in +/− samples. E: With syringaldehyde as substrate only BOA-7-OH increased root surface laccase activity significantly. F: Cuvette incubation of an inoculated seedling for laccase activity with ABTS as substrate. Streaks of the green colored ABTS product are emitted from the root surface. The assay may capture also some peroxidase activity because of endogenous H₂O₂ development. G: Cytosolic peroxidase: BOA-5-OH +/− differ significantly; BOA-7-OH samples (+/− have a significantly lower activity than the controls. (−) BOA-6-OH + has a significantly lower activity than the corresponding control. H: Root surface peroxidase activity. BOA-4-OH +samples show a significant lower activity than the + controls; the opposite effect is found with BOA-6-OH. All other samples do not differ significantly. All other differences are not significant. Bars present means ± SD; * = p < 0.05, ** = p < 0.001; *** = p < 0.0001.

Root surface enzyme activities

Polyphenol oxidase activity on root surface was less influenced by inoculation of the seeds with the micro-community than the cytosolic activity (Fig. 5). Only the controls and roots treated with BOA-6-OH differ significantly.

ABTS-dependent laccase activity with BOA, BOA-4, -5-, and -7-OH on surfaces of roots of pre-incubated seedlings revealed a strong increase, with the highest activity in inoculated seedlings (+). As an exception, pre-incubation with BOA-6-OH had no effect. The root surface ABTS-dependent laccase activities are opposed to the cytosolic polyphenol oxidase activities of roots inoculated with the micro-community. With syringaldazine as substrate, no or no significant difference between inoculated and non-inoculated roots was found in controls, BOA-4-OH, BOA-5- and -6-OH pre-incubated roots. Only after BOA-7-OH treatments, inoculated roots showed a strongly enhanced activity (200%). BOA pre-incubation led to a higher activity at (+) and (−) root surfaces.

The substrate-dependent activities may indicate the participation of several laccases at the root surface. Assays with the commercial laccase from T. versicolor verified that BOA-7-OH can be degraded by the fungal laccase (not shown) while BOA-4-OH is converted to a bluish polymer (Fig. 4).

The root surface peroxidase activity profile was very similar to the polyphenol oxidase activities. Peroxidase activity at the root surface was considerably higher than the activity measured with root protein extracts. The enzyme activity determinations supported the results of polymer formation at the root surface and showed again that the largest amount of the BOA-OH isomers was eliminated by polymerization. BOA-4-OH is no substrate for commercial horseradish peroxidase (HRP), but the BOA-6-OH dimer could be produced in vitro with HRP (data not shown) underlining the suitability of BOA-6-OH as a substrate for peroxidases.

The polymerization shut down the glucosylation of BOA-6-OH by the plant glucosyltransferase. On the other hand, exuded BOA-6-O-glucoside may be a precursor of subsequent microbial transformations after deglucosylation.

H₂O₂ production at the root surface and by microorganisms

BOA-OHs elicited H₂O₂ production and catalase activity at the root surface, indicated by bubble formation (Fig. 4). Since high H₂O₂ production indicates not only a stress response but also supplies the substrate for peroxidases and catalase activities, the H₂O₂ contents of the incubation media were determined. The catalase inhibitor 3-amino-1,2,4-triazole was used to mirror the H₂O₂ depletion due to catalase activity and O₂ generation. The measurements revealed a strong increase in H₂O₂ production in presence of the BOA-OH isomers and a less pronounced one in the presence of BOA (Fig. 6). The increase was extenuated for BOA, BOA-4-OH and in tendency for BOA-7-OH by inoculation with the micro-community. Inoculation with BOA-5-OH and with BOA-6-OH showed no effect, and H₂O₂ release was even higher. Supplementation with the catalase inhibitor mirrored that in the presence of BOA-4-OH and BOA-7-OH a considerable amount of H₂O₂ is converted by catalase, yielding O₂, a substrate for laccase. The inhibitor had no effect in presence of BOA-5- and BOA-6-OH, indicating that H₂O₂ might be directly used for polymerization by peroxidases.

Figure 6.

H₂O₂ contents and indication for catalase activity. A: Application of the BOA-OH isomers and BOA increased H₂O₂ development dramatically, with highest values in presence of BOA-5- and BOA-6-OH. The increase is highly significant except for C+/BOA+ and C-/BOA-4-OH-. With BOA-6-OH, H₂O₂ is higher in cuvettes with non-inoculated seedlings, BOA+ led to higher amounts than BOA-. For calculation of significance, inoculated and non-inoculated sets were compared. B: Measurements of H₂O₂ in presence of the catalase inhibitor (I-samples) indicate higher catalase activity, particularly with BOA-4-OH and BOA-7-OH. For calculation of significance, inoculated sets with and without catalase inhibitor were compared. C: Additional application of Papiliotrema baii and Pantoea ananatis to the incubation medium with Abutilon seedlings increases H₂O₂ development dramatically, with an extreme increase in presence of BOA-5- and BOA-6-OH. For calculation of significance inoculated sets with additional P.baii or P. ananatis were compared with the controls (inoculated sets). D: P. baii produced per se H₂O₂ and possesses catalase activity. There was no significant difference between controls and sets with one of the BOA-OH isomers or BOA. E: P. ananatis developed more H₂O₂ in presence of BOA-5-OH and BOA-6-OH. The catalase inhibitor indicates higher catalase activity in the control, BOA and BOA-4/6/7-OH. Bars present means ± SD; * = p < 0.05, ** = p < 0.001; *** = p < 0.0001. All other differences are not significant.

Addition of Pantoea ananatis or Papiliotrema baii to the plant incubation media roughly doubled the H₂O₂ release when BOA and the BOA-4/5/7-OH isomers were present, but in case of BOA-6-OH, the increase was 3 to 4-fold. Thus, the micro-community could contribute dramatically to an increase of the H₂O₂ concentration at the root surface and within the rhizosphere. Incubation with pure isolates of the 2 microorganisms indicated that Papiliotrema baii produced generally high amounts of H₂O₂, which is not influenced considerably by BOA or the BOA-OH isomers (Fig. 6). Some H₂O₂ is eliminated by catalase, as concluded from the effect of the inhibitor. Pantoea ananatis reacted with a higher H₂O₂ production when exposed to the benzoxazolinones. The increase of H₂O₂ amounts in presence of the catalase inhibitor pointed to a higher catalase activity after BOA-6-OH exposure. When Abutilon (inoculated) incubation sets were supplemented with additional Papiliotrema baii or Pantoea ananatis, the H₂O₂ amount increased dramatically in presence of BOA-6-OH (Fig. 6). This response led to the assumption that BOA-6-OH might be the most harmful isomer for organisms.

Incubation of the Actinomucor elegans micro-community with BOA-6-OH and 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one

Incubation of the Actinomucor elegans micro-community with BOA-6-OH resulted in a yellow-orange compound of transient presence. This compound is identical with 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one, produced and then degraded by pure Pantoea ananatis cultures when incubated with BOA-6-OH (Schulz et al.2017).²⁴ Tests on agar plates showed that Actinomucor elegans was apparently not damaged by the nitro compound, but hyphae did first not overgrow drops of it and finally the fungus started to sporolate in the neighborhood of the drops (Fig. S4). When a pure culture of Papiliotrema baii was incubated with BOA-6-OH, large amounts of the dimer were generated, but no 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one. Pure cultures of Papiliotrema baii did not transform any of the other BOA-OH isomers.

Low amounts of the nitro-compound could be detected in Abutilon incubation media and on the root surface (Fig. 7). Incubation of Abutilon seedlings with the nitro-compound did not result in any extractable BOA related detoxification product. Previous studies showed that isolated proteins from the Abutilon root surface are able to degrade the compound.²⁴ Papiliotrema baii is also able to degrade the compound at low concentrations, but the degradation is very slow and the optimal conditions for the degradation were not yet further investigated. The yeast does not survive when the compound is applied in the mM range. The efficiency of the isolated Actinomucor elegans micro-community to degrade the nitro-compound is assumed to depend on the composition of bacteria present in the mycelium plugs used for incubations, to oxygen availability and further still unknown factors, as the time necessary for degradation varied highly. When colonizing the root, the degradation activity seems to be optimally synchronized with plant activities.

Discussion

The importance of interactions between microorganisms and plants is increasingly recognized.²⁵ However, interactions of the plant-microbiome communities are not yet well understood. Protection of plants against deleterious xenobiotics, including allelochemicals, by the rhizosphere community was rarely taken into account, although plant-microbe associations are used in green bioremediation technologies.^26,27 The generation of polymer coats on root surfaces has been described. L-3,4-dihydroxyphenylalanine (L-DOPA), the main allelochemical exuded by roots of velvetbean, is converted to protecting melanin which precipitates on soybean roots.²⁸ Racemic catechin causes browning of the hypocotyl and roots of Arabidopsis thaliana.²⁹ The present study provides additional insights into mechanisms underlying plant and root-microbiome interactions directed to the elimination of potentially harmful allelochemicals. Supported by microorganisms, the root surface is regarded as the first barrier preventing considerable uptake into the protoplasts.

Plants seem to establish their specific microbial environment, starting with the attraction of microorganisms by root exudates during germination thereby completing the seed-born microbiome. The Abutilon microbial association was suggested to be facultative and nonspecific such that the composition of microbial communities would depend on environmental parameters including the nature of exuded secondary metabolites by the root, organic manure application, use of mineral fertilizer and many other soil properties.^16,17 The mechanisms actuated for allelochemical elimination are thought to be common defense responses, including the early oxidative burst at the root surface. Since incubations of the complete isolated micro-community with BOA-6-OH did not result in another product profile than found with isolated Pantoea ananatis and Papiliotrema baii, the 2 microorganisms seem to have a leading role in triggering reactions for polymerization and degradation of this isomer. First of all, oxidative processes are initiated by H₂O₂ release, leading to polymerization and precipitation of polymers derived from BOA-6-OH and the other BOA-OH isomers on the root surface or to a complete degradation of the polymers.

The production of H₂O₂ by the root and by associated microorganisms, here especially by Papiliotrema baii, is a prerequisite for peroxidase reactions. H₂O₂ decomposition by catalases, resulting in O₂ and H₂O, is necessary to include polymerizations and degradations by laccases and polyphenol oxidases. Since both types of enzymes act in an O₂ dependent manner, they compete for O₂. Abutilon roots exposed to BOA-OHs (BOA-6-OH, BOA-7-OH) show strong bubble formation, a marker for catalase activity. However, bubbles occurred only at defined roots and root parts. It is known that microbial colonization of the root depends on developmental stages. Thus, H₂O₂ and O₂ producing microorganisms are presumably not uniformly distributed at the same zones of different roots. It is unknown whether the composition of the microbiome can differ from root to root, but it is known that plant growth-promoting rhizobacteria colonize the root surface heterogeneously²⁷ On the other hand, differences in the colonization density from plant to plant have been described and might be the reason for relatively large standard deviations found with enzyme activities at root surfaces of individual plants. Polymer coats at the root surface have no growth suppressing effect; in contrast, especially melanin like coats, may protect root cells from ROS caused damage of cellular structures as also suggested by Soares et al.²⁸

The Actinomucor elegans microbial community reduces the polyphenol oxidase activity in the cytosol and buffers an increase of peroxidase and laccase activities within the plant protoplasts. In turn, polymerization reactions are accelerated at the root surface. Although it is not yet clear, which organism secretes which enzyme, the elimination of the BOA-OHs is unequivocally a cooperative action of plant surface enzymes and those secreted by the microorganisms. Among the open questions the management and regulation of the root detoxification process is of high importance. Whether dirigent proteins are necessary to manage, for instance laccase activities, is presently speculative. Dirigent proteins assist in stipulating radical reactions.³⁰ The chemical nature of the compounds and the life time of radicals occurring during the detoxification processes are essential. Depending on the position of the OH group of the synthetic isomers, the composition of participating enzymes apparently varies: BOA-4-OH, BOA-7-OH laccases, BOA-5-OH polyphenol oxidases, laccases and peroxidases, BOA-6-OH peroxidases. Oxi-reductases are known to exist as extra cellular enzymes.³¹ Low amounts of BOA-7-OH might be completely degraded by laccases, as found for other phenols.^32,33 Peroxidases may be responsible for polymer coat formation at lower H₂O₂ concentrations but these enzymes are also known to destroy polymers at high H₂O₂ concentrations.

The BOA-OH isomers differ in their substrate acceptance by the above mentioned enzymes. Because transformation of BOA-5-OH is slowest, this isomer seems to be a less suitable substrate for all involved enzymes. However, it is not obvious to make an estimation of the reactivity of the OH group at the aromatic ring because the combination of functional groups in the heterocycle is a hybrid of a lactone with a lactam. Thus, at least it has to be concluded that the N-atom has no basicity due to a mesomeric interaction with the carbonyl group. Similarly, the oxygen in the ring will not be a good donor atom by the same electron withdrawing effect of the carbonyl group. Hence, the acidity of the OH group should be in the range of phenols at around pKa 10 with no big difference in relation to its position. An OH at C-6 should however, increase electron density at C-5 and C-7 due to its own donor effect, which may be causative for a higher bioactivity. The most interesting question is the role of the molecular shape in the interaction with the different enzymes.

Plant detoxification by BOA-6-OH glucosylation could be an alternative, when the microbial/ root surface enzyme activities are overstrained or their setting up and coordination failed. Glucosylation may also take place to slow down oxidative processes. In addition glucosylation is a valuable tool to deprive laccases and other enzymes of substrates. This could be important for the plant, since laccases are known virulence factors.³⁴ In addition, glucosylation may avoid the accumulation of high amounts of 6-hydroxy-5-nitrobenzo[d]oxazo-2(3H)-one, when its rapid degradation is hampered.

The nitration of BOA-6-OH in position 5 by Pantoea ananatis could depend on the nitrate concentration in the rhizosphere. The apoplast can contain high amounts of nitrite, produced by nitrate reductase.³⁵ As a competing reaction, nitration of BOA-6-OH may reduce temporarily the subsequent formation of highly reactive NO in the apoplast, a signal transduction molecule with many functions including programmed cell death. The radical nitrogen dioxide (NO₂^.) acts as nitrating agent for proteins and lipids.³⁶ For instance, nitration of ionically-bound cell-wall peroxidase happens as a stress response.³⁷ Nitro-linolenic acid was identified as a signaling molecule important in plant defense mechanism against stress.³⁸ In a recent study we found that Pantoea ananatis not only produces but also degrades 6-hydroxy-5-nitrobenzo[d]oxazo-2(3H)-one. Nitro-aromatic compounds have been described as completely degradable by certain bacteria.³⁹ Therefore, under natural conditions the compound should be only temporarily and in low concentration present. Nitro aromatic compounds and their degradation intermediates are highly toxic to many organisms and most likely, Actinomucor elegans and the yeast are affected by the compound at higher concentrations.

Regarding the BOA-6-OH detoxification, occurrence of the BOA-6-OH dimer as a polymer precursor, BOA-6-O-glucoside and short-lived 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one production depends on multiple conditions and switches between the different ways are possible. Although the different reactions can compete for substrates, they finally complement each other. All of the identified products can be found either in the medium or at the root surface, with 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one in low amounts or only in traces (Fig. 7).

The results indicate the evolution of a highly advanced strategy of Abutilon theophrasti by recruiting soil microorganisms with special capabilities suitable for directed detoxification processes. The colonization of plant root with those microbial compositions increases the survival of the plant when exposed to BOA derived allelochemicals. In particular, Abutilon´s microbial community prevents the accumulation of highly toxic phenoxazinones. Plant-microbe interspecies-cooperation for detoxification processes opens, therefore, an additional facet of allelopathic interactions.

Material and methods

Plant material and analyses of detoxification products

Abutilon theophrasti Medik. seeds were purchased from Herbiseed (Twyford, UK). Others were collected from the “Meleti” field of the research station, Piacenza, Italy. Seeds were surface sterilized if not otherwise noted, germinated, cultivated for 10 d and inoculated with Actinomucor elegans as described in Haghi Kia et al. (2014).¹⁶ All experiments performed with plants and microorganisms described here and further in the text were done at least 3 times with 3 technical replications.

Controls were not inoculated. BOA-OH isomers were from recently synthesized stocks (D. Sicker). A total of 15–30 seedlings were incubated with 1 ml/seedling of 0.5 mM and with 2 mM BOA or the respective BOA-OH isomer only or in combinations mentioned in the text. The incubation times were either 24 h or 2 h, if not otherwise noted. Methanolic extracts from roots and shoots were prepared and analyzed by HPLC–DAD (Shimadzu).³ The media of the incubations with the BOA-OH isomers were analyzed likewise after centrifugation to pellet polymer particles derived from the BOA-OH isomers. The pellets were treated with DMSO and centrifuged (10.000 g, 10 min) to separate insoluble material. Roots of the BOA-OH isomer incubated seedlings were washed with DMSO (1 ml/10 roots). The solutions and incubation media supernatants were used for HPLC analyses.

Enzyme assays

Cytosolic enzymes and salt extractable glucosyltransferase activity

For glucosyltransferase assays, proteins of the roots from BOA incubated seedlings were extracted in presence of 1 M NaCl as described in Haghi Kia et al.¹⁶ The protein extracts contained 1 mg/ml protein in average after desalting with PD-10 columns using the assay buffer. Assays contained 25 µl protein eluate, 1 mM UDPG, 400µM of the respective BOA-OH isomer as sugar acceptor and MES-KOH buffer pH 5.8 with 100 µM NaCl to a final volume of 100 µl. The incubation at 30°C was stopped after 30 min by boiling for 10 min. The boiled assays were centrifuged at 10,000 g and analyzed by HPLC, using standard curves of the appropriate sugar acceptor molecule for calculation of the enzyme activity.

Polyphenoloxidase (PPO) was assayed in presence of the substrate 4-methylcatechol as described by Broothaerts et al.⁴⁰ with protein extracts from 0.5 g Abutilon roots. The extracts were prepared by homogenization of the roots with ice cold 100 mM phosphate buffer pH 6.8. After centrifugation of the homogenate, the supernatant was used for PPO assays. The same extracts were used for peroxidase assays performed according to Kukavica et al.⁴¹ with pyrogallol as substrate and laccase was measured as published by Eichlerová et al.⁴² with syringaldazine as substrate. Calculation of enzyme activities was as published by the cited authors. Substrates were purchased from Sigma.

Enzyme activities on root surfaces

Laccase activity on root surfaces was determined using the experimental design of Courty et al. ⁴³ with several modifications. Ten seedlings were preincubated with the isomers (0.5 mM, 15 ml) for 30 min, then washed with water and placed into cuvettes (1 seedling/microcuvette) containing 1 ml tap water, 0.2 ml 100 mM acetate buffer pH 5 supplemented with 0.2 ml 2 mM ABTS. The seedlings were incubated for 30 min, the incubation medium transferred into a fresh cuvette and the color intensity immediately measured at 415 nm with a spectrophotometer. Controls contained no seedling and seedlings without preincubation. Corrected absorptions (minus absorption of control without seedling) were used to estimate laccase activity referring to standard curves established with laccase from Trametes versicolor (Sigma). An absorption increase of 0.0216 / min was defined as 1 mU activity. In addition, root surface laccase activity was measured in cuvettes containing 1 ml tap water, 200 µl acetate buffer and 5 µl of a saturated syringaldazine solution in methanol. The absorption was determined after 20 min and the absorption of the control was subtracted.

Peroxidase and polyphenol oxidase activities on root the surface were measured using the same experimental design. For peroxidase determination, the pre-incubated seedlings were placed directly into cuvettes containing 1 ml H₂O, 162 µl acetate buffer pH 5.0, 162 µl 10 mM pyrogallol and 80 µl H₂O₂ (1:60 dilution of 50% H₂O₂, Sigma). Catalase activity was inhibited by adding 3-amino-1,2,4-triazole (Ueda et al. 2003).⁴⁴ The assay mixture for polyphenol oxidase contained 1 ml H₂O, 162 µl 100 mM phosphate buffer pH 6.8 and 162 µM 10 mM 4-methylcatechol. After 20 min of incubation at 25°C, the seedlings were removed and the absorption immediately measured. Peroxidase activity was estimated with a standard curve established with horseradish peroxidase (Sigma). PPO activity was expressed as change in absorbance per min and 4 seedlings. All assays were performed in triplicates with 3 biologic replicates. H₂O₂ was determined according to Liu et al.⁴⁵ with titanium oxysulfate solution (Sigma). Acceptance of BOA-OH isomers as substrates for laccases and peroxidases were estimated by assays with the purchased enzymes.

UPLC-electrospray-mass spectrometry

A BOA-6-OH derived transformation product, which accumulates temporarily in the incubation medium, was not stable enough for purification in amounts allowing NMR spectroscopy. The compound was extracted with ethyl acetate, the organic phase concentrated by evaporation and the residual solution immediately used for UPLC-Electrospray-Ionization-Mass Spectrometry. Ultra-Performance Liquid chromatography (UPLC)-electrospray-ionization-mass spectrometry (MS) was performed using the ACQUITY UPLC system equipped with a Xevo TQ-S triple quadrupole mass spectrometer (Waters, Eschborn, Germany). The UPLC system was equipped with an autosampler, binary pump, reversed-phase PFP column (100 × 2.1 mm, 2.5 µm core-shell Kinetex, Phenomenex, Aschaffenburg, Germany), column oven (40°C) and photodiodearray (PDA) detector (190–330 nm). Solvent A was pure water, solvent B pure methanol (Biosolve, Valkenswaard, The Netherlands). The elution was performed at a flow rate of 0.2 ml/min and with the linear gradient from 5% B to 100% B in 12 min. The injection volume was 10 µl for each sample.

The mass spectra were recorded with an ESI source in positive mode. Full scan MS were recorded in the mass range 50–800 m/z, MS/MS spectra in the mass range 50–350 m/z with argon as collision gas. The main ion source parameters are capillary voltage 1.4 kV, source temperature 150°C, desolvation temperature 600°C, desolvation gas flow 650 l/h, nebulizer gas flow 5 bar, cone gas flow 150 l/h, collision gas flow 0.12 ml/min, collision energy 5 (MS) and 20 (MS/MS), respectively. Fragmentation (table S1) and MS data interpretation are given in the supplementary data.

Root associated microorganisms

The BOA-4-OH derived, pelleted polymers were streaked on plates with Czapek medium and incubated at 25°C for 3 d. Actinomucor elegans was re-isolated from the plates and cultured on Petri dishes with Czapek medium. Plugs of the newly grown mycelium were cut and placed on new plates. The plugs were cultured until the colony diameter was about 1.5 cm, then 200 µl of the respective BOA-OH isomer (2 mM) was applied on the top of the plugs. The plates were then incubated for additional 24 h. Areas of the BOA-4-OH treated plugs showing a breakdown of Actinomucor elegans hyphae were cut and placed on Petri dishes with Czapek medium. The plates were cultured for 2 d. Bacterial strains were purified on trypticase soy broth agar. Identification of bacteria has been performed with partial 18 S rRNA sequencing. Purification and identification has been performed with the assistance of the identification services of Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany). Pantoea ananatis was identified as one of the bacteria occurring in break down plugs after BOA-4-OH treatment (see below). Stenotrophomonas maltophilia was identified by the same way, but was not included in further studies.

Cultures of the Actinomucor elegans plugs contained also an unknown fungus. Microscopic examination suggested A. elegans is growing together with a basidiomycetous yeast. The yeast culture was purified by streaking on the surface of yeast mold agar supplemented with 200 mg/L chloramphenicol. Plates were incubated at 4 ºC to slow down the development of A. elegans. A single colony was picked, purified and identified using partial sequencing of the 26/28S rRNA gene (D1 and D2 domains) and sequencing of the complete ITS region. This culture is preserved in the open collection of the Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany) as DSM 100638.

Nucleotide sequences of LSU and ITS fragments were compared with those deposited in GenBank and MycoBank, and showed that the yeast represents a species comprising Papiliotrema (Cryptococcus) flavescens species complex.²² Independent alignments for ITS and LSU rRNA were made using online version of MAFFT algorithm with the default parameters.⁴⁶ Phylogenetic relationships, shown in Fig. S1, were inferred from the concatenated ITS-LSU data set by the maximum likelihood (ML) method based on the general time reversible (GTR) model with RaxML (version 7.4.2) using raxmlGUI 1.3 and the GTRGAMMA option with 100 rounds of bootstrap replicates.⁴⁷ The phylogenetic analysis showed that the yeast culture belongs to the recently described species Papiliotrema (Cryptococcus) baii,²² (Fig. S1). The yeast P. baii is a rare cryptic species, which has been isolated from plant material and soil.²²

The co-existence of P. baii with Actinomucor elegans was verified in the cultures used for further experiments by treating Petri dish cultures with 10 mM salicylic acid. After 1 day, only the yeast was left. Cells of the colonies were transferred to new petri dishes with Czapek medium and cultured for 2 weeks. Colonies were picked and used for PCR analyses with the primers published in Yurkov et al.²² PCR products with the expected size were gel extracted using the NucleoSpin® Gel and PCR Clean-up Kit (Macherey Nagel), according to the manufacturer`s instructions. PCR amplification conditions used in this study are presented as supplemental material (Fig. S2). Primers were synthesized by Integrated DNA Technologies (IDT). The purified PCR amplification products were sequenced by BECKMAN COULTER GENOMICS (Essex, UK). Primers used for PCR amplification were also used for sequencing.

Incubation of the Actinomucor elegans micro-community and Papiliotrema baii with hydroxylated BOA

The yeast was pre-cultured in Sabouraud liquid medium for one week under shaking at 25°C. An aliquot 1 ml of the culture was transferred to flasks with 25 ml Czapek medium supplemented with the respective BOA-OH isomer. Controls were without benzoxazolinones. The cultures were checked for new compounds after 6 h and further every 24 h for 2 weeks.

Plugs of the re-isolated Actinomucor elegans grown on Petri dishes for 3 d were transferred into 50 ml flasks with 25 ml Czapek medium and incubated with BOA and BOA-6-OH at 23°C without shaking in the dark. Amounts of 1 up to 4 mg of BOA-6-OH or BOA, dissolved in 500 µl methanol, were applied to the cultures when started. Controls contained 500 µl methanol or no additive. Luminous yellow color development of the media in presence of BOA-6-OH started within 3 days, extended incubation led to orange colored media since the media shifted to alkaline pH.²⁴ Aliquots of the cultures were also analyzed by HPLC for the occurrence of new compounds.

Incubation of Abutilon theophrasti, Papiliotrema baii and the Actinomucor micro-community with 6-Hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one

Ten-day-old Abutilon theophrasti seedlings were incubated with 30 ml/9 seedlings of 0.5 mM 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one for 24 h at 23°C under greenhouse conditions. Roots were harvested and extracted as described above for HPLC analysis. Flasks with 25 ml Czapek medium were inoculated with 1 ml of a Papiliotrema baii culture (OD 1.6) and 0.5 mM 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one at 23°C without shaking in the dark for 14 d. Plugs of the Actinomucor micro-community were incubated in Czapek medium with 4.5µmol 6-hydroxy-5-nitrobenzo[d]oxazol-2(3H)-one over a period of 3 d. Aliquots were taken every 3 hours during the first 12 h of incubation, then every 24 h. They were centrifuged and analyzed by HPLC or frozen at −20°C until analysis.

Statistical data analysis

Error bars are based on SD. For statistical analysis of all data, the unpaired t-test (with Welch´s correction under assumption of normality) was used. Analysis of variance was performed with ANOVA using non-transformed data. In figures, data are presented as mean ± standard deviation. Each data point is based on at least 3 biologic replicates from 3 independent experiments, if not otherwise noted.

Acknowledgment

The authors thank the Università Cattolica del Sacro Cuore, Italy, for cooperation within the framework of the WEA Program for students exchange promotion.