Introduction

Plants and microorganisms have been revealed to deliver compounds that act as signals to invoke various responses in the recipient. The quality and dynamics of such interactions are actively driven by both parties. However, the identification of signaling compounds is only the tip of the iceberg. Indeed, the signaling machinery, which includes the production and export of chemical signals, their fate in the soil environment, the recipients' perception, and the modes of the answer, are topics requiring thorough research as the outcomes could provide useful tools to benefit humanity (Faure et al. 2009; Herridge et al. 2008; Sasse et al. 2018). Current research on plant–microbe signaling pathways raises intriguing questions about subterranean and aboveground life, such as how to distinguish between beneficial and harmful partners or the role of miRNA in establishing symbiotic relationships (Bozsoki et al. 2020; Middleton et al. 2020; Zhou et al. 2020). This review attempts to provide an overview of the below-ground signaling machinery, with a special focus on the particular aspects of signal sensing and transduction in plants and microbes. In the final chapter, we also discuss current approaches to the exploitation of plant-microbiome signaling in agronomy and environmental restoration.

Exploring the rhizosphere

Due to its highly heterogeneous and dynamic character, soil is a habitat for a vast range of microorganisms. These, in turn, alter the soil environment and participate in the creation of biological niches at multiple scale levels (Erktan et al. 2020). Plant-vegetated soils are abundantly occupied and modified by plant roots and their exudates representing a continuous influx of both organic and inorganic substances (Thies and Grossman 2006). In the rhizosphere, i.e., a thin layer of soil surrounding plant roots, multipartite inter- and intra-kingdom interactions occur between the plant and the present macrobiota, such as other plant roots and soil fauna, and microbiota, the so-called rhizomicrobiome, via the exchange of specific chemical compounds (Hiltner 1904; Lynch et al. 2001). Other definitions describe the rhizosphere as “the area around a plant root that is inhabited by a unique population of microorganisms influenced by the chemicals released from plant roots” (McNear 2013) or “that part of the soil ecosystem where plant roots, soil, and the soil biota interact with each other” (Lynch et al. 2001). Plant-derived organic compounds enrich the rhizosphere, making it a favorable habitat for microorganisms and, at the same time, shaping the present microbial communities (Thies and Grossman 2006). Consequently, the rhizosphere hosts a higher microbial biomass than bulk soil, although its microbial diversity is lower (Dennis et al. 2010).

Rhizodeposition

The constant influx of newly formed compounds from the above-ground plant parts into roots is compensated by either a passive or active release of a broad range of substances from the root to the rhizosphere, in a process referred to as rhizodeposition (Jones et al. 2004). Rhizodeposits are introduced into the soil matrix through various mechanisms, such as root exudation, root cap cell and border cell loss and lysis, gas leakage, and deposition of root necromass (Bais et al. 2006; Jones et al. 2009). The composition of rhizodeposits is determined by both abiotic (e.g., nutrient status, water regime, climate conditions) and biotic factors (plant genotype, physiological state, phenological stage, or distribution of roots in the soil) (Jones et al. 2004). Rhizodeposited molecules can be divided into two groups, i.e., low-molecular-weight compounds represented by a wide variety of organic acids, sugars, amino acids, phenols, and other secondary metabolites (SPMs, secondary plant metabolites) such as alkaloids, flavonoids, and terpenes along with inorganic substances such as various ions, water, and protons. The high-molecular-weight compounds are chemically less diverse and include mucilage and proteins (Jones et al. 2009).

As rhizodeposition accounts for a loss of up to 40% of fixed carbon (Badri and Vivanco 2009), it creates a decoy for heterotrophic microorganisms present in the soil. Consequently, rhizospheric microorganisms act as a sink for rhizodeposits and form spatial gradients that fuel the flux of solutes from roots into the soil. Nevertheless, plants exhibit a certain level of active control over root exudation by regulating internal processes such as vacuolar storage or the expression of efflux carriers (Canarini et al. 2019). At the cellular level, molecules are exuded from the roots through passive or facilitated diffusion, as well as active transport. (Oburger and Jones 2018). Diffusion, driven by a concentration and electrochemical gradient, only permits the transport of small nonpolar compounds across the plasmalemma or tonoplast and is of great importance in the transport of solutes, particularly in young root tissues (Sasse et al. 2018). On the other hand, substrate-specific membrane transporters facilitate the diffusion of larger or polar molecules such as SPMs or sugars (Yang and Hinner 2015). Active efflux of other root exudate components is linked to either primary or secondary active transport (Badri et al. 2008; Canarini et al. 2019; Hassan and Mathesius 2012; Sasse et al. 2018; Sugiyama et al. 2007). Additionally, certain SPMs can be exuded by vesicles (Weston et al. 2011). Moreover, plants can actively recapture exudates and thus fortuitously perceive a wide array of molecules from other plant roots and microorganisms (Jones et al. 2009).

2.2 Mechanisms of signal sensing and transduction

The underground communication network comprises multipartite inter- and intra-kingdom chemical signaling. Signaling pathways include the biosynthesis of signaling molecules, their transport from the originator to the recipient organism, followed by signal perception, and, finally, induction of physiological response (Venturi and Keel 2016). Given the vast phylogenetic and functional diversity of plants and microorganisms, signaling must exhibit a certain level of specificity to ensure it reaches the desired target organism. This allows plants or microbial cells to differentiate between beneficial, neutral, and pathogenic partners and trigger specific responses. Apart from the chemical structure of signaling molecules per se, such specificity is provided by various recognition principles that will be further elaborated (Bozsoki et al. 2020; Laub and Goulian 2007; Zhou et al. 2020).

Bacterial perception of signal

In bacteria, chemical signal perception is primarily mediated by two-component or by later discovered one-component systems (Fig. 1). Both systems were identified in bacteria, archaea, and exceptionally in eukaryotes, and comprise the transmission of an extracellular signal through an interaction cascade to the output domain of a cytosolic response regulator (RR) that mediates the physiological response (Faure et al. 2009; Mascher et al. 2006). The majority of output domains are DNA-binding domains such as those activating σ54-dependent transcription (Ulrich et al. 2005). Other output domains possess enzymatic activity, e.g., phosphorylation or the alteration of c-di-GMP levels, or they bind directly to RNA or proteins (Zschiedrich et al. 2016). For example, in Escherichia coli, EnvZ (a histidine kinase, HK) responds to changes in osmolarity by phosphorylating a transcriptional regulator OmpR. In turn, OmpR regulates the expression of outer membrane porin proteins OmpF and OmpC, that allow for the diffusion of small polar compounds across the membrane (Cai and Inouye 2002). In Burkholderia multivorans, OmpR affects also the transcription of genes involved in stress response, biofilm formation, or antibiotic resistance (Silva et al. 2018).

Fig. 1
figure 1

Schematic representation of two-component and phosphorelay systems. The signal perception leads to the autophosphorylation of a histidine residue (His) of a histidine kinase (HK). The phosphoryl group (P) is then transferred to the aspartate residue (Asp) of the cognate response regulator (RR) and activates its output domain. The phosphorelay system involves a hybrid HK that is capable of phosphoryl transfer to its receiver-like domain; histidine phosphotransferase (HPT) further conveys the phosphoryl group from the receiver-like domain to the Asp residue of RR. Some HKs have both kinase and phosphatase activity and can dephosphorylate the RR, hence terminating the signal transduction (Laub and Goulian 2007). Ways of responding to signal perception such as the regulation of gene expression, protein/RNA binding, or alteration of enzymatic activity are also shown along with major protein families of RR involved

Full size image

The components of different signal perception and transduction systems generally show significant sequence and structural conservation. Hence, since one- and two-component systems are involved in responses to a wide array of environmental stimuli and signals, crosstalk is likely to occur in recipient bacterial cells. To prevent crosstalk among the numerous transduction pathways present in a single cell simultaneously, finely tuned signal specificity of system components is required. Such specificity may be mediated by the HK phosphatase activity or higher affinity towards cognate RR (Laub and Goulian 2007). Finally, the non-conserved residues of HKs are supposedly crucial for the interaction with cognate RR (Casino et al. 2009). However, interference between some two-component systems may be beneficial for the host bacteria. Such multipartite systems can be exemplified, e.g., by the quorum sensing (QS) bioluminescence control in Vibrio harveyi. In this system, two extracellular HKs, LuxN and LuxQ, individually recognize two distinct QS autoinducers. In the absence of autoinducers, the HKs retain their kinase activity resulting in the phosphorylation of a single cognate RR, LuxO (Henke and Bassler 2004; Laub and Goulian 2007). Phosphorylated LuxO then promotes the expression of small regulatory RNAs which, in turn, hinder the expression of LuxR, a QS gene regulator. On the other hand, in the presence of autoinducers, the phosphatase activity of HKs is activated, and LuxO is dephosphorylated and cannot interfere with LuxR expression. Consequently, the LuxR-controlled luxCDABE gene cluster is expressed resulting in bioluminescence (Henke and Bassler 2004). The concerted activation of either LuxO or LuxR by both HKs is supposedly crucial: in the absence of just a single ligand, the kinase activity of the respective HK is sufficient to negate the phosphatase activity of the remaining HKs, and, as a result, the Vibrio bioluminescence is impaired (Mok et al. 2003).

Plant perception of signals

Various innate and exogenous molecules in the vicinity of the plant cell surface are sensed by so-called pattern recognition receptors (PRRs). The signal is generated by the interaction of signal molecules with the PRR’s extracellular ligand-binding domains; the molecules recognized by the ligand-binding domains can be of microbial origin, i.e., microbe- and pathogen-associated molecular patterns (MAMPs and PAMPs, respectively), or linked to cell lysis, i.e., damage-associated molecular patterns (DAMPs). The interaction and subsequent signal transduction lead to the induction of the plant immune system, so-called PAMP-triggered immunity (PTI). In addition, as some pathogens have overcome such defense, plants utilize intracellular receptors (so-called R-proteins) that recognize virulence factors secreted into the cell and lead to effector-triggered immunity (ETI) (Fig. 2) (Boller and Felix 2009; Zipfel 2008). The triggering of PTI and ETI results in the induced systemic resistance in the entire plant upon the local attack of either beneficial microbes or a pathogen, then referred to as systemic acquired resistance, (Boller and Felix 2009; Pieterse et al. 2014).

Fig. 2
figure 2

Adapted from Pieterse et al. (2014) and Charpentier (2018)

Scheme of the plant signal perception via pattern recognition receptors (PRRs). These are represented by either receptor-like kinases (RLKs) or receptor-like proteins (RLPs) (Liebrand et al. 2013). The recognition of microbe-, pathogen, or damage-associated molecular patterns (MAMPs, PAMPs, DAMPs, respectively) by PRRs, or intracellular effectors of bacterial or fungal origin by R-proteins, leads to PAMP-triggered immunity (PTI) or effector-triggered immunity (ETI). PTI and ETI trigger calcium uptake, production of reactive oxygen species (ROS), salicylic acid (SA), jasmonic acid (JA), or ethylene, or activates the mitogen-activated protein kinase (MAPK) cascades leading to the regulation of gene expression. The signal perception machinery is exemplified by the common symbiotic signaling pathway involved in the formation of root nodule symbiosis. The signaling MAMP molecule is perceived by the Nod factor receptor (NFR1) and the signal is proposed to be further transduced via G-proteins or the activation of the Does not make infection 2 (DMI2) co-receptor promoting the synthesis of mevalonate (MVA) by 3-hydroxyl-3-methylglutaryl-CoA-reductase (HMGR1). Then, mevalonate or a G-protein subunit are proposed to trigger the formation of an as-yet unidentified second messenger that interacts with CNGC15-DMI1 (Ca2+/K+ channel). As a result, Ca2+ is pumped into the cell from the nuclear envelope-endoplasmic reticulum storage (ER), and its influx is balanced by K+ efflux. The Ca2+ oscillations are maintained by nuclear membrane-localized calcium efflux ATPase (MCA8). The Ca2+ signal is decoded by calcium and calmodulin-dependent protein kinase (CCaMK), which phosphorylates CYCLOPS transcriptional factors. Finally, Gene expression driven by CYCLOPS leads to the onset of symbiosis formation. Dashed arrows represent the proposed steps of the pathway. Regulator of G-protein signaling (RGS); Cyclic nucleotide-gated calcium channels (CNGC15); Does not make infection 1 (DMI1); Nuclear pore complex (NPC); nitrogen-fixing root nodule symbiosis (RNS); Plant cell wall degrading enzymes (PCWDEs); lipopolysaccharide (LPS).

Full size image

The ongoing question is how plants distinguish between microbial commensals and pathogens. In the study of Zhou et al. (2020), pathogen-related damage to differentiated root cells triggered the expression of PRRs in neighboring cells. Therefore, for non-pathogenic bacteria, preventing damage to root cells may be a cornerstone of successful root colonization (Zhou et al. 2020). Furthermore, certain plant mutualists, such as ectomycorrhizal fungi, seem to produce lower levels of plant cell-wall degrading enzymes (PCWDEs), thereby reducing the production of DAMPs and thus promoting colonization and residence within the plant hosts’ roots (Kohler et al. 2015). The production of low levels of PCWDEs by beneficial microbes in contrast to non-beneficial ones or even pathogens is in agreement with the above-mentioned damage avoidance proposed by Zhou et al. (2020). Additionally, several studies have addressed that plant micro-RNAs (miRNA) may be the mediators of such friend-or-foe recognition via RNA interference (Liao et al. 2023; Miao et al. 2022; Pradhan et al. 2023). For instance, in multiple mycorrhizal plant species, arbuscular mycorrhiza is negatively controlled by the expression of miRNAs from the miR171 family targeting the LOM1 (lost meristems 1) gene, a positive regulator of arbuscule formation. A member of this family miR171b was found to contain a conserved mismatch in the LOM1-recognition site hindering the cleavage of LOM1 mRNA and even protecting it against degradation mediated by other members of the miR171 family (Couzigou et al. 2017).

The focus on ligand-binding sites of PRRs may also help to elucidate how plants differentiate between beneficial and pathogenic interactors. For instance, extracellular ligand-binding domains LysM are believed to recognize fungal chitin hexamers in chitin elicitor receptor kinases (CERKs), or lipochitooligosaccharides (LCOs) in Nod factor receptors (NFRs), respectively. The study conducted on Lotus japonicus (Regel) Larsen revealed that the recognition of ligands that elicit the immune response or nodulation is determined solely by LysM structure. A comparison of crystal structures indicated that residues within the LysM domain of CERK are highly conserved, as the chitin hexamer structure remains invariant. In contrast, NFRs show higher sequence degeneration as the Nod factors differ according to the bacterial species (Bozsoki et al. 2020).

Intrakingdom signaling

The vicinity of microorganisms and plant roots has resulted in a vast array of ecological relationships, thus, when considering the ecology of soil-dwelling organisms, these should not be approached as isolated entities or populations but rather whole communities should be taken into account. In other words, an intensive cross-talk occurs between but also within both plant and microbial communities and populations. Recent technological and methodological advances have clarified some signaling pathways and highlighted the need for further research, particularly in the area of often neglected non-mutualistic partnerships (Sasse et al. 2018; Venturi and Keel 2016). Nevertheless, several findings in the field of plant and microbial crosstalk have already been practically exploited and represent environmentally friendly approaches to the utilization of natural resources (Dayan and Duke 2014; Zhao et al. 2019).

Roots dialogue

The broad spectrum of root exudates includes substances that affect physiological processes and potentially the fitness of other plants as broadly reviewed, e.g., in Blande and Glinwood (2016). For instance, hydrojuglone, the oxidized form of juglone produced by black walnut (Juglans nigra L.), is known for its phytotoxic activities toward weeds and was one of the first to be considered to suppress the growth of weeds in agriculture as a natural alternative to artificial herbicides (Topal et al. 2007). Similarly, some flavonoids such as tricin possess allelopathic activity and might also be exploited as herbicides (Kong et al. 2004; Zhao et al. 2019). Apart from allelochemicals with a direct negative impact (Bais et al. 2004; Putnam 1988), the exudation of certain compounds may also positively affect the root distribution of surrounding plants (Semchenko et al. 2014), or trigger the immune response of the target plant and the production of allelochemicals (Li et al. 2020). Moreover, root-to-root signaling may also affect the aboveground plant parts such as flower development or flowering time. For instance, Moricandia moricandioides Boiss. was demonstrated to increase both the number of flowers and petal mass when grown in the presence of plants of the same species thus impacting the success of plants’ reproduction as pollinators are believed to select large flower patches over individuals (Parachnowitsch and Kessler 2010; Torices et al. 2018).

In contrast to the interactions conveyed by molecules released into soil, a major insight into the means of root-to-root communication might be provided by considering common mycorrhizal networks (CMNs), i.e., the interconnection of roots of multiple plant individuals through fungal hyphae. Such interaction through CMNs is thought to be of great importance, as suggested, e.g., by the fact that roughly 80% of higher plants are associated with arbuscular mycorrhizal fungi (AMF) (Soudzilovskaia et al. 2020). Indeed, multiple studies have proven that CMNs transport plant signaling molecules (Barto et al. 2012; Van Der Heijden and Horton 2009). For instance, in the study of Song et al. (2014), hyphae of the fungi Funneliformis mosseae transduced jasmonic acid between the roots of tomato plants (Lycopersicon esculentum L.). The recipient tomato plant, in reaction to the herbivore attack on the donor plant, upregulated the expression of jasmonic acid-signaling pathway genes thus repelling the herbivores (Song et al. 2014).

Microbial crosstalk

Microbes, similarly to plants, have developed mechanisms of communication that enhance their capacity to explore and exploit the soil environment. The cell density-dependent manner of microbial talk, quorum sensing (QS, Fig. 1) has been one of the most investigated and discussed mechanisms of microbial signaling in the rhizosphere (Henke and Bassler 2004; Mascher et al. 2006; Venturi and Keel 2016). In Gram-negative bacteria, QS is primarily mediated by acyl-homoserine lactones (AHLs). AHLs in the environment are sensed by the cell’s cytosolic cognate LuxR-like receptors that, after AHLs reach a threshold concentration, interact with DNA and regulate the expression of the genes responsible for physiological responses (Henke and Bassler 2004; Mascher et al. 2006).

AHLs are not the only known QS signaling molecules in Gram-negative bacteria. For example, Burkholderia cenocepacia, besides AHLs, employs diffusible signaling factors (DFSs, cis-2-unsaturated fatty acids) (Ryan et al. 2009). The perception of DSFs occurs via two-component systems and may result, for instance, in the formation of extracellular polymeric substances (Ryan et al. 2006; Ryan et al. 2015). Moreover, the DSF QS system regulates the production of AHLs (Deng et al. 2013). The spectrum of QS signals in Gram-negative bacteria has been recently expanded by photopyrones in Photorhabdus luminiscens and dialkylresorcinols in Photorhabdus asymbiotica (Brameyer et al. 2015). Additionally, some bacterial QS signaling molecules may originate from plants. For instance, Rhodopseudomonas palustris uses exogenous p-coumaric acid, a common SPM, to synthesize p-coumaroyl-homoserine lactones that are recognized by LuxR-like receptors (Schaefer et al. 2008).

In Gram-positive bacteria, the QS signaling compounds are peptides, the perception of which is also mediated by two-component systems. For instance, the ComP/ComA competence system in Bacillus subtilis employs isoprenylated peptides as signals such as the DNA-encoded decapeptide ComX. (Mascher et al. 2006; Solomon et al. 1995). The cell density-dependent talk in Lactococcus lactis and Bacillus subtilis is mediated by the production of lantibiotics such as nisin or subtilisin (both antimicrobial peptides), respectively (Kleerebezem 2004). Lastly, staphylococcal virulence is also regulated via the production and perception of autoinducing peptides (Ji et al. 1995).

In addition to QS signaling, volatile organic compounds (VOCs) are a large group of low-molecular-weight metabolites that enable microbial crosstalk (Audrain et al. 2015). For instance, rhizospheric Pseudomonas fluorescens and Serratia plymuthica release dimethyl disulfide, which was proposed to suppress Agrobacterium tumefaciens crown gall formation in tomato plants (Dandurishvili et al. 2011). Furthermore, Streptomyces albidoflavus produces albaflavenone, an odorous terpene with antibiotic properties against Bacillus subtilis (Gurtler et al. 1994). The genus Streptomyces is also known to produce geosmin (Scholler et al. 2002), a compound believed to attract soil arthropods, and, in turn, help to spread Streptomyces spores attached to the surface of their cuticle (Becher et al. 2020). Lastly, VOCs may alter antibiotic resistance in bacteria. For instance, glyoxylic acid and 2,3-butanediol emitted by Bacillus subtilis were suggested to affect the resistance of E. coli cells to ampicillin (Kim et al. 2013).

Interkingdom signaling

Interkingdom interactions between plants and microbes vary from parasitism to mutualism and are induced and maintained by secondary metabolites (Bais et al. 2004; Venturi and Keel 2016). Some of these alliances are highly specific cooperations, such as nitrogen-fixing root nodule symbiosis (RNS) or arbuscular mycorrhizal symbiosis (AMS), both noteworthy in terms of soil ecology. However, some alliances are not as tight, such as those between plants and plant-growth-promoting bacteria and fungi (PGPB/F) (Bailly et al. 2014; Charpentier et al. 2016). Although their importance in biotechnology and agriculture is widely acknowledged, the comprehension of signaling pathways behind plant–microbe associations might enhance their further exploitation.

Molecules mediating a specific plant–microbe cross-talk can be exemplified by flavonoids, a large group of SPMs. In soil, as significant constituents of plant root exudates, flavonoids can alter nutrient acquisition as some are capable of chelating metal ions such as iron; in plants, they also act as pigments and scents that attract pollinators, protect against UV radiation, and reduce oxidative stress (Cesco et al. 2010; Samanta et al. 2011). The role of flavonoids as signals in soil has been mostly emphasized in the regulation of mutualistic associations such as the above-mentioned RNS and mycorrhizae including AMS. Although RNS and AMS symbioses differ in terms of the range of host plants, microorganisms involved, and outcomes for the interacting partners, they share several aspects. Such “backbone” of RNS and AMS is also known as the common symbiotic signaling pathway (CSSP) and generally involves microbial signaling molecules derived from chitin, such as Nod or Myc factors, that are perceived by plant cognate RLKs further mediating signal transduction which leads to repeated oscillation of intracellular calcium concentration (Bozsoki et al. 2020; Charpentier 2018). Details on the CSSP signal perception and transduction cascade are given in Fig. 2.

In AMS (Fig. 3), the onset of mutualism resulting in nutrient exchange in arbuscules is triggered by phosphor deficiency in the soil. Plant-released flavonoids, terpenoids, strigolactones, or cutin monomers attract fungi hyphae and induce their branching, thus increasing the likelihood of physical contact with the plant root. In the fungus, the response consists of the production of Myc factors, which induce the CSSP in the plant counterpart. As a result, the structure of root cortex cell walls is altered to favor fungal hyphae penetration, followed by the formation of arbuscules (Charpentier 2018; Hassan and Mathesius 2012; Pimprikar et al. 2016). On the other hand, the chemical signals can be perceived by multiple recipients with contrasting interests. For instance, strigolactones also trigger the germination of parasitic plants that can harm the host plant (Bouwmeester et al. 2007).

Fig. 3
figure 3

Adapted from Nishida and Suzaki (2018)

Representation of interkingdom signaling between plant and microbiota. Plant root exudates pose a selective pressure over rhizospheric microorganisms and proposedly select the beneficial ones that lead to RNS, AMS, and phytostimulation via promoted nutrient acquisition, regulation of hormone levels, or induced pollutant degradation in soil. Mutually beneficial interactions are exemplified by the nitrogen-fixing RNS. Perception of root-released flavonoids by rhizobia leads to the secretion of Nod factors that induce morphological changes in the root hair that enable infection by rhizobial cells and the formation of an infection thread (IT). Nodule maturation is completed with nitrogen fixation in the nodule. Similarly, the onset of AMS is depicted as well as the use of exogenous DNA/RNA in crop control management. Root nodule symbiosis (RNS); arbuscular mycorrhizal symbiosis (AMS); plant-growth-promoting bacteria/fungi (PGPB/F; polychlorinated biphenyls (PCBs), 1-aminocyclopropane-1-carboxylic acid (ACC), indole-3-acetic acid (IAA).

Full size image

The nature and the establishment of RNS have been reviewed in detail elsewhere (Debellé, 2019; Liu and Murray 2016). Briefly, during the onset of RNS, micromolar levels of flavonoids released by plant roots are perceived by rhizobacterial NodD proteins that trigger the transcription of nod genes (Liu and Murray 2016; Morel et al. 2015). This results in the formation of Nod factors whose structural diversity determines the host specificity. Activation of the CSSP by Nod factors leads to the root curling around the rhizobial cell, formation of an infection thread, and lastly, upon infection, the development of a nodule (Fig. 3) (Singh et al. 2014). Rhizobia present in nodules fix molecular dinitrogen by reducing it to ammonia, which is then available to the host plant, primarily legumes of the Fabaceae family (Cooper 2004; Nishida and Suzaki 2018).

However, the specificity of the plant's response to the interacting partner is still not fully understood due to the structural similarity of the rhizobial and fungal signals. It is unclear whether the answer lies in the structural diversity of RKLs, reduced degradation of the cell wall by the infecting mutualist, or early nutrient flow monitoring. Thus, further investigation is required (Bozsoki et al. 2020; Plett and Martin 2018; Zhou et al. 2020).

Intra- and interkingdom signaling: an opportunity for mankind?

Plant–microbe mutualistic interactions including RNS, AMS, and PGPB/F have a potential for exploitation in agriculture, as the health and growth of plants involved in these interactions are boosted (Morel et al. 2015). For instance, phosphorus, a key plant growth nutrient, is mostly present in soil in inorganic or organic forms unavailable to plants. To cope with phosphorus deficiency, plants profit from phosphate-solubilizing bacteria in their vicinity that release low molecular weight organic acids such as citric or malic acid to chelate phosphate ions and convert them into bioavailable forms. In the study of Sharma et al. (2020), the co-inoculation of tomato plants by phosphate solubilizing bacteria, particularly Pseudomonas spp., led to increased plant biomass (Sharma et al. 2020). Similarly, plant iron deficiency is balanced by iron-chelating siderophores produced by plant-associated microorganisms such as in the case of endophytic fungi of the aloe-leafed cymbidium (Cymbidium aloifolium L.); moreover, the extracted siderophores exhibited antimicrobial activities against Ralstonia solanacearum, a common plant pathogen (Chowdappa et al. 2020). A comprehensive review of PGP microorganisms as potential biofertilizers can be found in Mącik et al. (2020). The input of beneficial microorganisms into their association with the host plant does not necessarily lie solely in nutrient acquisition, but also in phytostimulation via the release of hormones such as cytokines, gibberellins, auxins, and indole derivatives, or elicitors of plant immune response (Fig. 3) (Rosier et al. 2018). For instance, in Arabidopsis thaliana L., a VOC indole released by rhizospheric PGPB such as Burkholderia spp. was found to increase plant biomass (Bailly et al. 2014). Alternatively, another VOC 2,3-butanediol from Bacillus subtilis was proposed to function as a biopesticide, as it promotes the immune response of plants to plant pathogens such as Erwinia sp. (Ryu et al. 2004). In addition, bacteria can outcompete fungi via nutrient competition or the production of antifungal signals such as 2,4-diacetyl phloroglucinol or phenazines (Lugtenberg et al. 2002).

Nevertheless, the soil environment is highly dynamic and dependent on local climate conditions. To sustain the long-term positive effect of PGP microorganisms on host plants, the application of endophytic PGP biofertilizers should be considered. A significant increase in plant biomass production together with the doubled number of florals was observed on Physalis ixocarpa Brot, inoculated with non-native host endophytes (Méndez-Bravo et al. 2023). Similarly, Ghimire et al. (2023), cross-inoculated oat (Avena sativa L.) with endophytes isolated from Brassica carinata Braun resulting in the stimulation of oat-biomass growth (Ghimire et al. 2023). Though these studies depict endophytic PGP biofertilizers as efficient across plant species, which is a remarkable advantage to technological production and in situ application, the identity and diversity of plant-derived molecules responsible for the shaping of the rhizospheric and supposedly endophytic microbiome have been neglected. Recently, p-coumaric acid was found to be involved in the recruitment and maintenance of microbiome homeostasis in the rice phyllosphere (Su et al. 2024). Hence, further research is needed to identify such key molecules shaping microbiome structure in other agriculturally important crops as well as to elucidate the mechanisms of plant-PGP endophyte interactions in planta. Additionally, the role of vertically transmitted endophytes should be considered in the assembly of plant microbiomes as they interact with the plant’s acquired PGP microbiota. These aspects should be therefore reflected in the development of PGP-based biofertilizers (Pirttilä et al. 2021).

On the other hand, rhizosphere bacteria also encounter fluctuating and often toxic concentrations of a variety of SPMs as a part of root exudates, so, several mechanisms for alleviating their toxicity are involved. In addition to the expression of specific efflux pumps (Burse et al. 2004) and alterations of membrane permeability (Parniske et al. 1991), some bacteria can catabolize and utilize SPMs as an additional energy and carbon source. Importantly, certain bacterial species have been found to possess the ability to co-metabolically degrade organic pollutants when exposed to SPMs (Pham et al. 2015; Zubrova et al. 2021). For instance, in crude oil-contaminated sites in Alaska, plant metabolites were crucial for site restoration (Leewis et al. 2013). Monoterpenes such as l-carvone or p-cymene enhanced the co-metabolization of PCBs (polychlorinated biphenyls) by Arthrobacter sp. B1B in soil contaminated by Aroclor 1242, a dielectric medium from transformers (Gilbert and Crowley 1997). Similarly, industrial solvents trichloroethenes were co-metabolized by Pseudomonas sp. JR1 and Rhodococcus sp. BD1 in the presence of cumene (Dabrock et al. 1992). Other studies have examined the role of flavonoids and terpenes in the induction of biphenyl 2,3-dioxygenase (bphA) in representatives of ubiquitous rhizospheric bacterial genera Pseudomonas and Rhodococcus (Toussaint et al. 2012; Zubrova et al. 2021). The combination of bacterial degraders with convenient SPM-exuding plants represents an environmental-friendly biotechnological tool for the degradation of chemicals that pose risks to human and ecosystem health, as well as food production, and thus the restoration of polluted soils (Fraraccio et al. 2017; Mackova et al. 2006; Singer et al. 2003). Last but not least, PGP microorganisms can be utilized in phytoremediation strategies as they stimulate plant health and growth. Vice versa, plant metabolites might help to enrich pollutant-degrading PGP bacteria such as in the case of Gordonia sp. S2RP-17 in the vicinity of Zea mays L. on a diesel-contaminated site (Hong et al. 2011).

The above-mentioned examples highlight the potential of beneficial microorganism application to improve nutrition, health, and subsequently yield of commercial crops. Moreover, the use of the concept of intra- or interspecies companion plants deserves thorough research in the field of sustainable agronomy. Such plants might enhance neighboring plant fitness, produce precursors of pesticides or elicitors of plant immune response, and not only warn but also protect their neighbors from biotic stress. For instance, (-)-loliolide from barnyard grass (Echinochloa crus-galli L.) upregulated the biosynthesis of the allelochemicals momilactone B (a diterpene) and tricin (a flavonoid) in neighboring rice plants (Oryza sativa L.) roots (Li et al. 2020). The addition of a synthetic tricin-derived herbicide into paddy fields was also investigated as an alternative to currently used pesticides. The derivative inhibited the growth of several paddy weeds without affecting rice, moreover, it had no toxic effects on zebrafish and earthworms and was quickly degraded in soil (Zhao et al. 2019). Recently, intercropping with green manure such as Vicia villosa Roth was demonstrated to alter soil properties via root exudation, leading to a decrease in Fusarium oxysporum abundance in soil, a fungus preying on monocultures of Cavendish banana (Musa acuminate AAA Cavendish cv. Brazil) (Yang et al. 2022). Thus, the extensive application of beneficial companion plants could raise crop production, limit the use of synthetic pesticides, and also enhance soil quality (Dayan and Duke 2014; Ren et al. 2014). Last but not least, the resulting biomass of companion plants might be further commercialized or reused as fertilizer or forage.

Signaling in disguise

Both plants and microorganisms have developed deceptive mechanisms how to overcome the inter- or intrakingdom signaling for their profit. These mechanisms include not only compound mimicry and degradation of chemical signals, but also the modulation of the host responsiveness via RNA interference (Middleton et al. 2020; Nievas et al. 2017; Washington et al. 2016). However, such deceptive strategies represent not only obstacles but also opportunities for the target organisms, resulting in a chemical “arms race” between host plants and both their pathogens and mutualists.

Quorum quenching (QQ) impairs QS through mechanisms such as compound mimicry or enzymatic cleavage of QS signals. Bacteria and plants can both biosynthesize mimicking molecules, which can confuse the target microorganism and either promote or inhibit its residence in the root zone. Nievas et al. (2017) concluded that peanut (Arachis hypogaea L.) responded differently to bacterial QS signals by producing either mimics of QS signal molecules (e.g., long acyl chains AHLs, lac-AHLs), or QS inhibitors (e.g., short acyl chains AHLs, sac-AHLs). Therefore, A. hypogaea was proposed to shape its rhizomicrobiome by selecting beneficial bacteria such as Bradyrhizobium spp. using lac-AHLs signals over pathogenic bacteria that are targets of the sac-AHLs language (Nievas et al. 2017). Similarly, pathogenic microorganisms have also developed mechanisms how to interfere with the host plant immune system or perception machinery by secreting various effector molecules. For instance, pathogens can repress the plant defense response such as in the case of the type 3 secretion system-effector HopAF1 of Pseudomonas syringae: HopAF1 blocks ethylene production in the host plant thus hindering plant immune defense and promoting infection (Washington et al. 2016). On top of that, mutualists can also effectively employ a similar deception strategy to promote the colonization of their host plant. For instance, the AM fungi Rhizophagus irregularis secretes the SP7 effector, which interacts with the pathogenesis-related transcriptional factor ERF19 in the nucleus of plant cells, thereby repressing the ethylene signaling pathway. This results in the enhanced formation of mycelium during the initiation of AMS, as the expression of genes involved in plant immune defense is blocked (Kloppholz et al. 2011). One could therefore question the sustainability of inter- or intrakingdom signaling: at what point, if ever, does such an “arms race” end? The understanding of the origin and evolution of both mutualistic relationships and pathogenesis, molecular mechanisms of how the partners recognize each other, and the identification of potentially active molecules and their targets are of the highest interest, not only from the perspective of advances in areas of evolutionary biology and ecology, but the answers might also provide a theoretical basis for strategies of agricultural crop protection. For instance, a combined effect of plant growth promotion and QQ activity of Pseudomonas segetis P6 was demonstrated, resulting in the biocontrol of multiple plant pathogens (Rodriguez et al. 2020). Similarly, AidB, an AHL lactonase from the soil bacterium Bosea sp. F3-2, which disrupts QS in Pseudomonas aeruginosa and Pectobacterium carotovorum subsp. carotovorum was suggested as a potential biocontrol agent (Zhang et al. 2019).

DNA and RNA language and tools

Recently, a novel mechanism of signaling between plants and microorganisms was proposed, represented by small RNAs (sRNAs) carried by extracellular nanovesicles (Middleton et al. 2020). For instance, Arabidopsis thaliana sRNA transferred by exosome-like extracellular vesicles was shown to silence fungal pathogen Botrytis cinerea virulence genes (Cai et al. 2018). Vice versa, Botrytis cinerea sRNA, Bc-siR37, was found to interfere with the Arabidopsis thaliana immune system and silence genes involved in plant defense (Wang et al. 2017). Middleton et al. (2020) thus proposed such a mode of communication to mediate interactions between plants and non-pathogens. Furthermore, given the ubiquity of sRNAs in both unicellular and multicellular organisms, the authors proposed their potential role in mediating communication across kingdoms (Middleton et al. 2020; Middleton et al. 2022; Zhao et al. 2007).

Interestingly, the focus on sRNA may further broaden the knowledge of seemingly notoriously known interactions. For instance, the study of Ren et al. (2019) of RNS between soybean (Glycine max L.) and Bradyrhizobium japonicum revealed a positive regulation of nodulation by tRNA-derived small RNA fragments (tRFs): B. japonicum tRFs targeted soybean genes involved in root hair development, a key aspect of infection thread and subsequent nodule formation. Moreover, tRFs from Rhizobium etli were predicted to target a different set of genes in common bean (Phaseolus vulgaris L.) suggesting the involvement of tRNAs in the regulation of multiple physiological responses of host plants (Ren et al. 2019). Furthermore, Formey et al. (2014), showed that 31% of studied miRNA genes from Medicago truncatula were upregulated in the presence of either a mutualist or pathogen (Formey et al. 2014), and together with Ren et al. (2019), their studies indicate that the answer to the question of symbiont differentiation may also lie in deciphering the sRNA language.

RNA signals might also be an effective tool how to reduce the use of synthetic pesticides in the field of crop production. Exogenous RNA application, so-called spray-induced gene silencing (SIGS), has been recently applied to induce plant resistance to pathogens via RNA interference. This fast, flexible, and easy-to-apply method not only leads to local or systemic resistance in plants but can also target multiple pathogens simultaneously. Nevertheless, due to the low stability of RNA, the feasibility of SIGS depends on external conditions; moreover, the method is limited by the bioavailability of RNAs to the plant. However, both of these obstacles could be solved by using nanoparticles as a vector (Morozov et al. 2019; Taliansky et al. 2021). Therefore, deciphering the RNA language together with artificial RNA and nanoparticle production drives the advance of yet another approach to sustainable crop control. Moreover, we suggest that sRNA-encoding genes could be considered for the genetic engineering of crops resulting in pathogen-resistant varieties similarly to RNA silencing-based crop protection against plant viruses such as in the study of Khalid et al. (2017).

In addition to RNA, the use of exogenous DNA has been drawing attention as a biotechnological agent in crop-pest management either as an inhibitor of their proliferation or elicitor of plant immune response (so-called DNA-vaccines) (Ferrusquía-Jiménez et al. 2021). The former exploits the self-inhibitory effect of fragmented exogenous DNA (eDNA) released from the pathogen. Thus, once identifying the plant pathogen, its DNA can be exogenously applied to the plant to suppress pathogen growth. In DNA vaccines, eDNA is of the host plant origin, i.e., acting as a DAMP priming the host plant immunity response and inducing its resistance to pathogens. The major advantage of DNA-based strategies of crop protection is the accessibility and ease of application of the effector molecule, i.e., DNA, compared to the conventional labor-, time-, and cost-demanding screening of bioactive compounds. Apart from DNA, a wide variety of other DAMPs can be utilized, however, the choice is limited by certain aspects such as the feasibility of the biotechnological production, agent stability in the environment, or ease of application (Ferrusquía-Jiménez et al. 2021; Mazzoleni et al. 2014; Quintana-Rodriguez et al. 2018).

Conclusions and future perspectives

A wide array of compounds of both plant and microbial origin have been reported to act as signals in inter- and intra-kingdom communication, yet more are to be identified (Venturi and Keel 2016). Due to the increasing need to effectively utilize, preserve, and renew natural resources, especially in agriculture, it is crucial to fully comprehend the range of signaling compounds and corresponding plant-mutualist and plant-pathogens signaling pathways and their specificity (Dayan and Duke 2014; Ryan et al. 2009; Singer et al. 2003). However, due to the intricate and multifaceted interactions among soil, plants, and microorganisms, the notion of plants and microorganisms as separate entities has gradually been abandoned in favor of the concept of a plant holobiont. This concept reflects the fact that plants and associated microbiomes have shared their evolutionary history and they create a niche for the other partner; consequently, they interact with the environment in concert as a single entity (Vandenkoornhuyse et al. 2015). The implementation of such a holistic view rather than a reductionistic approach focused solely on one partner is essential for comprehension of the intricate signaling networks between plants, their microbiota, and the environment. Moreover, interactions between rhizospheric and endophytic microbiomes as well as their dynamics throughout the plant’s phenology and response to environmental stimuli such as extreme temperatures, drought, attack of herbivores, or intensive UV light should come into focus, especially in connection with the ongoing climate change (Patel et al. 2022). Additionally, for the sake of sustainable agriculture, the potential of plant-associated microorganisms genetically engineered towards sustained PGP traits should be investigated including their stability in host plants and vertical transmission to next generations. The application of such PGPB/F represents a versatile, environmental- and cost-friendly, and legislatively more feasible alternative to crops genetically modified, e.g., toward resistance to pathogens or herbivores (Ghimire et al. 2023; Méndez-Bravo et al. 2023; Pirttilä et al. 2021). Thus, (meta)genomic and (meta)transcriptomic analyses of genetic determinants of PGP and endophytic microbial traits are crucial for the identification of candidate genes and microorganisms, in combination with state-of-the-art imaging techniques to decipher the fate of these in the host plant and environment. Similarly, transcriptomics of the host plants would shed more light on their response to the infection by both mutualists and pathogens. Moreover, the application of metabolomics, i.e., optimally non-targeted qualitative and quantitative analyses of both plant and microbial low- and high-molecular-weight metabolites, should be considered as they mediate the microbial community assembly and the quality of interactions (Compant et al. 2021; Pirttilä et al. 2021).

On the other hand, the repeatability and, consequently, the informative value of studies dealing with plant endophytes is often hindered by inconsistent methodologies, such as plant tissue surface sterilization, DNA isolation, and PCR amplification. Thus, a unification is strongly recommended to enable comparability across studies (Compant et al. 2021). Similarly, there are still certain methodological challenges in studying both rhizospheric microorganisms caused by the intricate nature of the soil matrix and the microscale on which the rhizospheric plant–microbe interactions take place, such as analyses of the chemical composition of root exudates under the conditions of real soil and tracking of the fate of exuded molecules (Oburger and Jones 2018). These obstacles can be solved by the application of isotope labeling-based techniques, e.g. studies dealing with 13C-, 14C, or 15N-labelled plant biomass and metabolites (Uhlik et al. 2013). Moreover, conducting plant–microbe interaction studies in real soil in situ is necessary as it reveals the contribution of viable but nonculturable cells (VBNC) to these processes as well as reflects the fluctuations of often harsh climatic conditions and other environmental factors (Izgordu et al. 2022). Finally, focusing on the RNA language as a communication tool within a plant holobiont may revolutionize the field of plant–microbe ecology in terms of filling gaps in the understanding of mechanisms of signaling between plants and microorganisms as well as the sustainability of crop management (Middleton et al. 2020; Middleton et al. 2022).