An endophyte is a microorganism that lives, at least for a portion of its life cycle, inside plant tissues without producing any symptoms of disease. Among endophytic microorganism are bacteria, fungi, actinomycetes, and viruses (Bao and Roossinck, 2013; Stępniewska and Kuźniar, 2013), and they are found in almost every plant (Partida-Martínez and Heil, 2011). Endophytes can be classified as systemic (or true endophytes) or non-systemic (or transient endophytes), depending on their life cycle and the type of relationship they establish with the plants. Indeed, there is a huge variability of symbiotic lifestyles, from mutualism to parasitism, depending on genotypic and/or environmental factors. True endophytes co-evolved with their hosts, creating mutualistic relationships, and are often vertically transmitted, while transient endophytes could shift from a pathogenic to a mutualistic behaviour, depending on external conditions (Wani et al., 2015). In any case, several factors may influence the host response to endophytic interactions: mainly host genotype, nutrient availability, environment, field management practices, and microorganism strain (Hesse et al., 2003; Malinowski and Belesky, 2006; Qawasmeh et al., 2012).

Endophytes can be found in the root tissues, where they are more abundant, but also in the aerial parts of the plant (leaf, flower) and in the seeds (Hallmann, 2001).

Endophytic bacteria could be considered as a subgroup of the rhizospheric bacteria, that acquired the ability of colonizing their host plants (Reinhold-Hurek and Hurek, 1998). In fact, rhizosphere is a highly competitive environment (Raaijmakers et al., 2002), while the internal tissue of the host may represent a protected ecological niche with minor perturbations from the external conditions of the soil or, in general, of the environment (Kasmir et al., 2011), and this could have created an evolutionary drive from the first to the second. Endophytic and rhizospheric bacteria implement very similar strategies to promote plant growth, but usually endophytic microorganisms have a higher beneficial potential.

Among bacterial endophytes, Proteobacteria are the most widely represented, including α-, β-, and γ-Proteobacteria; the latter taxon being the most diverse and widespread (Miliute et al., 2015; Santoyo et al., 2016). Other classes frequently isolated from plant tissues are Actinobacteria, Bacteroidetes, and Firmicutes (Reinhold-Hurek and Hurek, 2011). Rarer, but still present, are Acidobacteria, Planctomycetes and Verrucomicrobia (Santoyo et al., 2016). The most common bacterial genera are Bacillus (Firmicutes), and Pseudomonas (Proteobacteria). Rhizobia spp. are also included among endophytic bacteria, as they colonize internal root tissues of Fabaceae, developing the typical nodules for nitrogen fixation.

Among endophytic fungi, on the other hand, we can find the families Clavicipitaceae (associated with grasses), Cladosporiaceae, Glomerellaceae, Sebacinaceae, Pleosporaceae, and Hypocreaceae, among which the most representative genus is Trichoderma (Hardoim et al., 2015).

Key molecular and metabolic pathways at the base of host-microbe recognition and strain selection by different plant genotypes are starting to be elucidated. It is clear how different genotypes grown in different soil/environmental combinations are enriched with different endophytic strains (Granér et al., 2003; Rashid et al., 2012; Edwards et al., 2015). The role of plant genotype on strain selection and microbial population composition has been widely studied and even if it is not the main force driving microbial diversity, it is able to modulate it (Weinert et al., 2011; Bulgarelli et al., 2015; Walters et al., 2018). Interestingly, it has been shown that plant domestication, and lately the development of high-yielding genotypes, caused a reduction in the plant capacity of associating with useful microorganism (Porter and Sachs, 2020; Valente et al., 2020), since human-centered breeding neglected the traits related to microbiota association.

Considering endopythes role in promoting plant growth, especially in nutrient-deprived conditions, and in increasing plant defence against pathogen attack, either directly or indirectly, they are now considered a tool for crop management. These could sustain agricultural practices with fewer chemical inputs. However, research is still needed to further the knowledge both on the plant-side, trying to identify the genetic factors responsible for a more efficient microbial colonization, and on the microbic-side, to isolate the most promising micro-organisms and the most effective synthetic communities. Once these aspects are clarified, it will be possible to engineer plants and microbes to make their interaction more effective. It will also be possible to explore the use of root exudates or organic compounds that might serve as pre-biotics. Finally, we should be able to overcome and the bottleneck, as of the applicability of this research to open fields conditions, that remains challenging.

Another relevant application of this category of microorganisms is phytoremediation. It has been shown that some bacterial strains are tolerant to high concentrations of heavy metals, as Cd, Cu and Zn, and other pollutants. These strains favour their accumulation into the plants, promoting their growth (Ullah et al., 2015; Ma et al., 2016), even in a stressful environment.

Mycorrhizae are likely to have played a crucial role in the evolution of terrestrial plants (Bonfante and Genre, 2008; Partida-Martínez and Heil, 2011). Today, mycorrhizal fungi can be divided into ectomycorrhizae, when the hyphae colonize the root intercellular spaces, and endomycorrhizae, when they penetrate inside the plant cells. Endomycorrhizae are further divided into orchid (OM), ericoid (ERM) and arbuscular mycorrhizae (AMs) (Bonfante and Genre, 2010; Chen et al., 2021).

Ectomycorrhizae are mostly associated with woody perennial trees such as Pinaceae, Fagaceae, Dipterocarpaceae and Caesalpinoidaceae, contributing to the wellness of most forest ecosystems (van der Heijden et al., 2015). EM fungi are phylogenetically diverse and belong to Basidiomycetes, Ascomycetes and Zygomycetes, representing the orders Pezizales, Agaricales, Helotiales, Boletales, and Cantharellales. EM hyphae grow partially inside the root intercellular space and partially outside, creating a mantle covering the tip of colonized lateral roots, called the Hartig net (Shi et al., 2023).

EM fungi live in symbiosis with their hosts, but are also facultative saprotrophs, decomposing complex organic matter present in the soil and making nitrogen and phosphate available for the plants (Martin and Nehls, 2009). In turn, the EM fungi receive photosynthates from the plant. Genomic data and functional studies show the presence of specialized families of phosphate, ammonium, and nitrate transporters (Jargeat et al., 2003; Casieri et al., 2013; Becquer et al., 2014; Stuart and Plett, 2019), supporting evidence of their fundamental role in different genomes of EM fungi. The entire metabolic chain transporting N/P from the soil to the plants through EM hyphae has been fully elucidated in recent years, and many advances have been made thanks to high-quality genomic sequences available (Nehls and Plassard, 2018).

Among endomycorrhizae, Arbuscular Mycorrhiza (AM) are the most represented, as these symbioses are formed by the 70-90% of terrestrial plants species, while the fungi all belong to the monophyletic phylum Glomeromycota (Schüβler et al., 2001). AM fungi are obligate biotrophs, considered organisms with no or rare sexual reproduction, and present aseptate hyphae developing inside the plant cells, where they form the characteristic tree-shaped hyphal structure. In AM symbiosis, the fungi support the plants by supplying mainly P-based nutrients and water, while the plant supplies the fungi with carbon nutrition (Parniske, 2008). It is estimated that almost 20% of the photosynthetic products of terrestrial plants are allocated to AM (Bago, 2000). The N contribution is less pronounced in AM compared to EM, even though some publications have shown that the portion of N transported to the plant cells from AM is not negligible and depends on soil pH, moisture, and nutrient concentration (Govindarajulu et al., 2005; Jin et al., 2005; Tanaka and Yano, 2005).

Typically, a soil ecosystem characterized by mycorrhizal symbiosis features a wide variety of plant-fungi relationships, thus offering a broad functional diversity (van der Heijden et al., 1998; Burleigh et al., 2002). This diversity arises from the presence of different plant species and fungi, each with the potential to select the most cooperative partner (Werner and Kiers, 2015). Plant roots tend to be enriched with fungal species or isolates that ensure optimal growth benefits (Bever et al., 2009; Kiers et al., 2011; Verbruggen et al., 2012), while AM fungi typically select plants that can allocate the highest amount of C nutrients (Lekberg et al., 2010; Kiers et al., 2011). However, the genetic and ecological mechanisms underlying this partner selection remain unclear, and understanding them could reveal crucial insights for future agricultural applications.

The success of the mutualistic relationship depends on many factors, such as the combination of plant-fungi genotypes, their molecular and metabolic regulation, and soil characteristics (pH, structure, moisture), nutrient availability, and colonization rate. Different plant genotypes, under controlled conditions, may respond differently to AM in terms of plant growth and yield, as shown in crops such as maize (Zea mays L.) (Ramírez-Flores et al., 2019) and sorghum (Sorghum bicolor (L.) Moench) (Watts-Williams et al., 2019), or in terms of stress resistance, as demonstrated in rice (Oryza sativa L.) (Chareesri et al., 2020) and bread wheat (Triticum aestivum L.) (Lehnert et al., 2018). The results of these studies represent milestones for future breeding programs, supporting more sustainable agriculture.

Besides AM, endomycorrhizae are also represented by ericoid (ERM) and orchid (OM) mycorrhizae. As for ERM, the fungi colonize plants of the Ericaceae family, such as Calluna, Vaccinium and Erica, typically found on nutrient-poor and acidic soils (Soudzilovskaia et al., 2019).Thus, they represent an essential way to mobilize organic material in infertile soils. In OM, the fungi belong to Basidiomycetes, mainly to Rhizoctonia species, with which most orchids are associated (Favre-Godal et al., 2020). Orchids strongly depend on the nutrients coming from the fungi, especially for the initial stages of seed germination and growth.

It should be noted that there are also plants that establish different types of mycorrhiza, either spatially, temporally, or simultaneously, within the same root system. For example, this is the case with plants from Populus, Fraxinus and Eucalyptus genera (Ambriz et al., 2010; Teste et al., 2020), as well as other plants known as dual-mycorrhizal plant species. It must be noted that there are also plants establishing different types of mycorrhiza, in a spatially or temporally distinguished manner, or simultaneously, within the same root system. For example, this is the case of plants from Populus, Fraxinus and Eucalyptus genera (Ambriz et al., 2010; Teste et al., 2020) and other plants that are called dual-mycorrhizal plant species.

It is worth mentioning that often a single fungus may connect the root systems of several plants, creating what are known as common mycorrhizal networks (Figueiredo et al., 2021). This facilitates the exchange of signalling compounds and nutrients, increases pathogen resistance, and promotes plant growth.

Some endophytes are seed-borne and are present in germinated plants, thus representing a bridge across host plant generations (Coombs and Franco, 2003). Also, plants with vegetative propagation can transmit their endophytic microbiota to the next generation (Kamran et al., 2022). In others, a chemotactic response of endophytes to host plant root exudates has been observed. These exudates are rich in biomolecules (including nutrients and water), and thus attract or are recognized by friendly endophytic microbes (Compant et al., 2010; Brader et al., 2014). Flavonoids are one such group of metabolites secreted by several plants and categorized as chemo-attractants, playing an important role in endophytic interaction with the root hair. Flavonoids are used in bioformulations to affect successful infection of legume roots by rhizobia (Arora and Mishra, 2016). They are also reported to play a role with non-rhizobial endophytes, and it has been proven that in the presence of these metabolites, the colonization of roots in rice and wheat by the endophytic Serratia sp. EDA2 and Azorhizobium caulinodans ORS571 is far more effective (Webster et al., 1998; Balachandar et al., 2006). Lipo-ChitoOligosaccharides (LCO), also called Nod factors, are well-characterized signal molecules activating the Common Symbiotic Pathway (CSP) in rhizobia-legume and arbuscular mycorrhizal associations (Gough and Cullimore, 2011). Recently, StrigoLactone (SL) secreted by roots of Arabidopsis thaliana was found to act as a signal molecule for colonization of endophytic Mucor sp (Rozpądek et al., 2018). SL treatment may also activate the synthesis and release of short-chain chitin oligomers, whose perception by the plant can stimulate the symbiotic signalling pathway during early stages of host colonization (López-Ráez et al., 2017). Additionally, ArabinoGalactan Proteins (AGPs), which are highly glycosylated members of the Hydroxyproline-Rich GlycoProteins (HRGPs) superfamily of plant cell wall proteins, play a crucial role in establishing the interaction of plant with microbes (including endophyte) at several stages (Nguema-Ona et al., 2013). Several other root exudates, including sugars, amino acids, organic acids, phenolic compounds, and other secondary metabolites, are now known to be secreted by plant roots, which selectively invite the mutualistic microbes, particularly the endophytes (Chagas et al., 2017). A bacterial endophyte can also utilize the hyphae of a fungal pathogen to gain access from the soil to plant roots, thereby protecting the host from infection (Palmieri et al., 2020). showed that the endophytic rhizobacterium Rahnella aquatilis utilizes hyphae of the fungal pathogen Fusarium oxysporum to access and colonize plant roots. Metabolomic and multi-omics approaches, as those described below, would most likely increase the knowledge about metabolites released by plant seeds and roots involved in attracting favourable endophytic microorganisms. This information is necessary to address the realization of bioformulations or genetic engineering approaches to increase the production of chemo-attractants and thus the colonization of plant tissues by bacterial endophytes under normal and stressful plant growth conditions (Arif et al., 2020).

The strategies that plants use to distinguish beneficial microbes, such as endophytes, from pathogens, are still a matter of research and not completely understood. Plants possess various PRRs (Pattern Recognition Receptors) that recognize M/PAMP (Microbe/Pathogen-Associated Molecular Patterns) ligands and initiate immune reactions (Hacquard et al., 2017). The most characterized MAMPs include flagellin, elongation factor Tu, peptidoglycan, lipopolysaccharides, bacterial cold shock proteins, bacterial superoxide dismutase, Beta-Glycan, β-glucans from oomycetes, and chitin (Newman et al., 2013). These MAMPs are recognized on the surface of plant cells by PRRs, which include receptor-like kinases and receptor-like proteins (Tang et al., 2017). Both pathogens and symbionts can be recognized by PRRs, because the M/PAMPs are not specific to pathogens. To avoid recognition by the host plant and the subsequent immune response, pathogens and symbionts have evolved complex extracellular invasion strategies. Due to the similarity of pathogen and symbiont genomes (Reinhardt et al., 2021), common extracellular strategies exist between them. They can be divided into three categories: avoiding accumulation of MAMP precursors, reducing hydrolytic MAMP release, and preventing MAMP perception (Buscaill and van der Hoorn, 2021). These strategies can involve different microbial effectors. Symbionts have developed various strategies to allow their potential hosts to better distinguish them from pathogens during the recognition phase. For example, LCO Nod Factors are perceived by legumes, activating the symbiotic pathway (Radutoiu et al., 2003; Bozsoki et al., 2020). In rice, short-length chitooligosaccharide (CO4) triggers symbiotic signal transduction with the symbiotic complex receptor MYR1–CERK1.This suppresses the formation of the CEBiP-CERK1 heteromer that would mount the immune response, while long-chain chitooligosaccharide (CO8) induces immune signalling through CEBiP-CERK1 (Chiu and Paszkowski, 2021; Zhang et al., 2021). It has also been observed that fungal endophytes produce chitin deacetylases, which deacetylate chitosan oligomers that are thus not perceived by plant receptors (Cord-Landwehr et al., 2016). There is also evidence where endophytic bacteria are known to produce their own MAMPs, which are either not recognized by PRRs of plants or trigger in plants a comparatively weak and transient defence reaction compared to pathogenic interactions (Vandenkoornhuyse et al., 2015). Along this line, it was shown that in grapevine (Vitis vinifera L.), thanks to an alteration in sequence, the perception of flagellin from an endophytic Burkholderia phytofirmans by LRR-RLK (Leucine-Rich Repeat-Receptor-Like Kinase) FLAGELLIN-SENSITIVE 2 (FLS 2) PRR was different from the perception of those of bacterial pathogens, such as Pseudomonas aeruginosa or Xanthomonas campestris (Trdá et al., 2015). Still, there are knowledge gaps about the genetic mechanisms that differentiate recognition strategies deployed by beneficial with respect to pathogenic microbes, that need to be filled through comparative genomic studies between the two microbial categories, complemented with functional analyses. Availability of complete information on gene functions involved in the endophytic recognition would allow targeted modifications of favourable strains through gene editing and/or over-expressing approaches that would improve the plants capability in recognizing microbial symbionts and protecting them from the immune response.

Potential entry points for endophytes are cracks formed at the emergence of lateral roots, zones of root elongation, root hair cells, and wounds. Other sources include stomata, particularly of young stems and leaves, lenticels, and germinating radicles. For successful colonization, some bacteria must find their way to these apertures. Klebsiella pneumoniae 342 (Kp342) can colonize the lateral root junctions in wheat and alfalfa (Medicago sativa L.) (Dong et al., 2003). Similarly, Herbaspirillum seropedicae and Gluconacetobacter diazotrophicus dominate colonization at lateral root junctions (James et al., 1997; Luna et al., 2012). Some endophytes enter through infection colonization, where cellulolytic and pectinolytic enzymes produced by endophytes come into play (Miliute et al., 2015), such as pectate lyase, which has been implicated in the colonization of Klebsiella strains (Kovtunovych et al., 1999). Symbionts can colonize hosts while overcoming the response to Damage-Associated Molecular Patterns (DAMPs) and MAMPs, while a response against pathogens is still possible in the presence of non-pathogenic microbes (Zhou et al., 2020). Several studies have proven that there is a downregulation of plant defence pathways during the colonization of plants by mutualistic partners, such as rhizobia or AMF (Fouad et al., 2014; Benhiba et al., 2015). In the case of an oxidative burst or generation of Reactive Oxygen Species (ROS) as plant defence system, endophytes protect themselves by producing enzymes such as superoxide dismutases, catalases, peroxidases, alkyl hydroperoxide reductases, and glutathione-S-transferases (Zeidler et al., 2004). The root endophytic fungus Serendipita indica secretes a histidine-rich protein to improve its access to micronutrients and to influence oxidative stress and reactive oxygen homeostasis to facilitate the colonization of the host plant (Nostadt et al., 2020). Also, symbionts could induce Jasmonic Acid (JA) and suppress Salicylic Acid (SA) formation to Induced Systemic Resistance (ISR), whereas pathogens typically enhance the SA biosynthesis to mediate Systemic Acquired Resistance (SAR) in plants (Martínez-Medina et al., 2017). Moreover, during mutualistic interactions, late induction of SA/JA/ET signalling pathways prevents the microbe from ‘overstepping’ and ‘overpowering’ the plant (Plett and Martin, 2018). It is reported that most miRNAs induced in the host during the establishment of endophytes also target hormone-response pathways (Formey et al., 2014). During AMF infection, the miRNA E4D3Z3Y01BW0TQ is upregulated and disrupts Gibberellic Acid (GA) signalling pathway, known for repressive action against mutualistic associations (Formey et al., 2014; Martín-Rodríguez et al., 2015; Wu et al., 2016). The plant may also induce the expression of different groups of genes during colonization by diverse sets of microbes. For example, during the establishment of symbiosis, the majority of pathways targeted by miRNAs for plant defence system are turned off, thus preventing the obstacle to the proliferation of endophytes (Plett and Martin, 2018). For AMF, two receptor-like kinases called Arbuscular Receptor-like Kinase 1 (ARK1) and ARK2 are required for the sustenance of the symbiotic interaction in several plant species (Montero et al., 2021). Moreover, AMF are separated from the plant cytoplasm by a specialized host-derived membrane, which represents the main interface facilitating the bidirectional exchange of nutrients and information and protects the microbial symbionts from the immune response (Huang et al., 2021). The biosynthesis of this peri-arbuscular membrane is controlled by a gene called GLUCOSAMINE INOSITOL PHOSPHORYLCERAMIDE TRANSFERASE1 (GINT1) (Moore et al., 2021). Protein Secretion Systems (SSs) in bacteria also modulate the plant immune system. Among all known SSs, Type III Secretion System (T3SS) and Type IV Secretion System (T4SS) are essential for delivering Effector Proteins (EFs) by the pathogenic bacteria into the plant, but these are either absent or present in low abundance in mutualistic endophytic bacteria (Green and Mecsas, 2016; Liu et al., 2017). Notable exceptions can be seen in some rhizobial strains where T3SS is important for nodulation of some legumes (Ausmees et al., 2004; Okazaki et al., 2016, 2013). The T3SS is also a determinant for rice endophyte colonization by non-photosynthetic Bradyrhizobium spp (Piromyou et al., 2015). Furthermore, the Type 2 Secretion System (T2SS) was demonstrated to be required for suppressing MAMP-triggered immunity in efficient root colonizer bacteria and, notably, enhanced the colonization capacity of other tested commensal bacteria in Arabidopsis (Teixeira et al., 2021). On the other hand, in mutualistic proteobacterial endophytes, Type VI Secretion Systems (T6SSs) are present, and are also commonly found in commensal and pathogenic plant-associated bacteria. However, they are associated with important functions, which are apart from virulence, usually such as competition against other bacteria (Reinhold-Hurek and Hurek, 2011; Bernal et al., 2018). From this picture, it emerges that colonization involves a plethora of traits from both, plants and microorganisms and available data most likely shed light only on a small fraction of the involved processes. Considering the plant side, in addition to the information provided above, recent investigations highlighted that plant genes can shape the microbiota composition (Zhang et al., 2019; Deng et al., 2021; Escudero-Martinez et al., 2022; Oyserman et al., 2022; Escudero-Martinez and Bulgarelli, 2023) and that wild germplasm is supposed to support higher microbiome diversity than domesticated counterparts (e.g (Pérez-Jaramillo et al., 2018; Nerva et al., 2022). Taken together, these results indicate that there is room for genetics interventions addressed to increase both plant and beneficial microbial aptitude in establishing favourable interactions and that further functional and multi-omics investigations can increase the available targets for improving endophytic colonization by plant growth promoting microorganisms. Once plants and microbial effective targets are identified, these could be modified/introgressed/engineered into their respective genomes (Arif et al., 2020; Nerva et al., 2022; Escudero-Martinez and Bulgarelli, 2023).

The development of NGS technologies has allowed to perform whole genome sequences of numerous fungi and bacteria. Overcoming the limit of traditional culture-dependent identification approaches, it has enabled the identification of as much microbial diversity as possible. Meta-genomic approaches nowadays almost routinely make use of DNA extraction from the whole soil/tissue microbial population, allowing the analysis of its gene/taxa content using next generation sequencing (Allan, 2014). The sequencing can involve the whole genome, which is then tentatively assembled and annotated, or only the 16S rRNAs. These data can be used to study the microbial diversity and to evaluate the absolute abundance of different bacterial strains, taking into account the different copy number of 16S rRNA genes in distinct bacterial genomes (Case et al., 2007; Wang et al., 2020).

It is worth mentioning that the availability of several AM fungi genomes has allowed for the study of the evolution of these organisms, which are considered as living fossils and ancient asexuals (Parniske, 2008). Their genome size is highly variable, from the 39.6 Mb of Paraglomus occultum (Malar C et al., 2022) to 784 Mb of Gigaspora margarita (Venice et al., 2020), with large genomes hosting a higher number of genes and a high proportion of transposable and active elements (Venice et al., 2020; Dallaire et al., 2021); differences that could explain their intra-specific variability.

In parallel, the study of the epigenome variability is emerging as a tool to understand a hidden layer of variability (Chaturvedi et al., 2021). Another interesting example of recent scientific advances given by the most recent sequencing technologies concerns the use of long-read sequencing and chromatin conformation capture techniques that made it possible to understand the genomic organization of multi-nucleate coenocytic hyphae of AM, demonstrating their heterokaryotic nature and supporting rare sexual reproduction events (Yildirir et al., 2022; Sperschneider et al., 2023).

This approach, coupled to advanced bioinformatic pipelines, for example using algorithms of artificial intelligence, could be considered as the most useful omic science for understanding the network of interactions. It has largely benefited from NGS technologies, whose recent advances have significantly increased the sensitivity to catch the rarest transcripts. Moreover, long-read sequencing technologies in the Iso-Seq approach, among others, allow to cover the entire transcript length thereby distinguishing rare isoform resulting from alternative splicing events (Li et al., 2017; Zhang et al., 2019).

Transcriptomics has been successfully applied to uncover the plant molecular strategies used to recruit the most favourable microbial organisms in response to diverse abiotic and biotic stresses, and to understand the microbial molecular networks used to successfully establish the symbiotic relationships (Sheibani-Tezerji et al., 2015). Furthermore, transcriptomic studies applied to bacterial cells may help decipher which strains and cells, among the endophytic or rhizospheric population, are transcriptionally active (Sharma et al., 2004; Knauth et al., 2005; Sheibani-Tezerji et al., 2015), surmounting DNA-based technologies that cannot distinguish non-viable cells. The completion of whole genome sequencing of new microbial species and strains will be crucial allowing the identification of the microbial response to different soil characteristics and plant genotypes.

RNA-seq has also been applied to the population of small RNAs, to identify and characterize plant miRNAs involved in host-microbiota communications. These small non-coding RNAs are important key regulators of different plant biological pathways, from organ development to stress response. It has been shown that microorganisms might stimulate the expression of plant miRNAs, modulating drought tolerance response, nutrient uptake, or facilitating symbiosis establishment (Mohsenifard et al., 2017; Pentimone et al., 2018; Kord et al., 2019; Tao et al., 2023). Besides plant miRNAs, other small RNA-like molecules are coded by fungi and bacteria, that could be involved in an intriguing system of cross-kingdom RNAi-mediated regulation, for example during mycorrhizal colonization (Silvestri et al., 2019), or nodule formation (Ren et al., 2019). The intriguing hypothesis of small RNAs as mobile cell-to-cell signalling molecules (Huang et al., 2019) has been explored in detail thanks to the possibility to purify Extracellular Vesicles (EV). In fact, EVs have been shown to transport small RNAs between plant host and microorganisms in both pathogenic and mutualistic interactions (Cai et al., 2019), thus their further analysis will deepen our understanding of below-ground inter- and intra-kingdom communications. Single cell transcriptomics, coupled with enhanced microscopy techniques will greatly improve our understanding of endophyte bacteria and AM fungi lifestyle inside the plants (Yin et al., 2023) and of root cells regulatory network.

In parallel with next-generation and third-generation (or single molecule) (Schadt et al., 2010) sequencing technologies, which have been successfully applied to the study of all the DNA/RNA populations present in a tissue, the analyses aimed at characterizing the entire set of proteins, with their post-translational modifications, and metabolites have evolved. This evolution is to comprehensively study all the molecules in a microbe/plant biological system, thereby increasing their sensitivity and throughput. Proteomics has been applied to plant tissues to understand how the presence of an endophyte, for example, may modulate the synthesis of different plant proteins (Lery et al., 2011), revealing their role in cellular recognition. The analysis of the protein-protein interactions, also called an interactome, is essential to unveil molecular mechanisms at the base of symbiotic relationships.

Metabolomics and proteomics have been used to analyse root exudates, containing both primary and secondary metabolites, to understand how biotic and abiotic factors might modulate their composition, and as a result, attract and associate with different microorganisms. However, analysing either the metabolites or the proteins present in a colonized plant tissue, or both, is challenging. This is because it is difficult to distinguish between molecules produced by either the plants or the fungi/bacteria. Recently, to resolve this issue, several techniques have been developed to narrow the analyses to the single-cell level, such as Mass Spectrometry Imaging (Boughton and Thinagaran, 2018), Laser ablation electrospray ionization (Bhattacharjee et al., 2020), live single-cell mass spectrometry (Masuda et al., 2018), and the spatial metabolomics pipeline (Geier et al., 2020).

In addition to soluble metabolites, plants can diffuse Volatile Organic Compounds (VOCs).Metabolomics is essential to uncover the role of these signalling molecules and their modulation in response to environmental stimuli and genotype interactions. However, their role in soil matrices could be less abundant and relevant than in aerial open-air environments.

Metabolomic analyses have shown that plants can influence their microbiota by secreting various metabolites. In turn, the microbiome can influence the metabolome of the host plant (Haichar et al., 2008).

It is now clear that to acquire global information on the interconnections existing among plants and the microbiota, single omics technologies should be integrated into a multi-omics approach (Chen et al., 2021). To this end, the development of bioinformatic tools and networking models that can integrate and visualize information is essential. This will provide a comprehensive view of the regulatory network connecting all the molecular levels from the genome to the metabolic pathways. By employing a multi-omics approach, it will be possible to deepen our knowledge on the complex interactions between plants and their growth-promoting microbial counterparts. This will be fundamental to understand how to engineer microbial communities and plants for a more sustainable agriculture. In this scenario it is fundamental to develop high-throughput phenotyping platforms, to measure and analyze qualitative and quantitative traits on a large scale, developing suitable phenomics approaches, that could be non -invasive and able to work in the field as well, in order to fill the gap with other omics techniques (Großkinsky et al., 2015).

Temperature changes are not necessarily harmful to plants; they can play a crucial physiological role in regulating internal clocks and controlling processes like the opening and closing of flower corollas. Some species require exposure to low temperatures to initiate important developmental processes, such as vernalization for flowering or germination (Ruelland and Zachowski, 2010). However, temperatures that are too low can induce a range of physiological responses that can be detrimental to their survival. Chilling, intended as a few degrees above 0°C air temperature, can cause reductions in enzymatic activity, rigidification of membranes, destabilization of protein complexes, and stabilization of RNA secondary structure, while also promoting the accumulation of ROS. Additionally, chilling can impair photosynthesis and increase membrane permeability, resulting in the leakage of cellular contents. Freezing (below 0°C) stress, on the other hand, can cause more severe damage, as ice formation within cells leads to mechanical disruption and cell/tissue/plant death (Ruelland and Zachowski, 2010).

There are not many reports available in literature about the protective effects of endophytes against the low temperature stresses, perhaps because such conditions also limit the growth and multiplication of microorganisms. As an example of protective effects towards cold stress, the endophytic fungus Epichloe gansuensis increases the biosynthesis of alkaloids and unsaturated fatty acids during the seed germination of Drunken horsegrass (Achnatherum inebrians), thereby increasing tolerance to cold stress (Chen et al., 2016). It was also reported that the endophytic rhizobacterium Parabulkholderia. phytofirmans PsJN induced the upregulation of some cold stress-related genes in grapevine (Theocharis et al., 2012; Chen et al., 2021).

Heat stress is defined as a temperature rise of 10-15°C above ambient, with the optimal range for plant growth about 15-24°C. Among all the different abiotic stress factors, it has the most detrimental effects on plants, reducing flower fertility and modifying crop growth with a detrimental final effect on yield (Shaffique et al., 2022). Its effects include an increase in membrane fluidity, the formation of ROS, and alterations to photosynthesis and respiration processes (Banerjee and Roychoudhury, 2018; Shaffique et al., 2022). Heat stress triggers a cascade of physiological responses that result in the release of Heat Shock Proteins (HSPs), a class of molecular chaperones that facilitate protein folding and prevent aggregation under conditions of cellular stress. These proteins can assist unfolded proteins in refolding into their proper conformation or direct them towards degradation through ubiquitination processes (Ruelland and Zachowski, 2010; Banerjee and Roychoudhury, 2018). The often simultaneous presence of heat and drought stress exhibits holistic features, as the combined effects are greater than those caused by each stress individually (Lipiec et al., 2013).

The inoculation of the endophytic bacterium Brevibacterium linens RS16 in rice plants reduced the emission of the plant stress hormone ethylene due to its 1-aminocyclopropane-1-carboxylate (ACC) deaminase activity. Plants inoculated with B. Linens RS16 also showed increased levels of small HSPs (Choi et al., 2022).

A recent report demonstrates that the application of the plant growth-promoting root endophyte Paraburkholderia phytofirmans PsJN enhances the development of the Root System Architecture (RSA) in Arabidopsis thaliana, under both normal and high-temperature conditions. This allows the plant to access a larger soil area, thereby better managing abiotic stresses such as heat and drought (Macabuhay et al., 2022).

The simultaneous presence of heat and salinity stress can greatly impact crops. The inoculation of the endophytic fungus Trichoderma virens SB10, along with Glycine Betaine (GB) treatments, conferred significant tolerance in soybean (Glycine max L.) plants against these two stresses. In presence of GB, T. virens SB10 enhanced the production of hormones like gibberellins, IAA, and SA. The co-treatment with the fungus and GB also led to a reduction in proline accumulation and Na⁺ uptake and an increase in macronutrient (N, Ca, K) uptake. Effects on GmHKT1 and GmSOS1 gene expression, two major genes involved in salt tolerance (Singh and Roychoudhury, 2021), were also recorded, leading to the maintenance of a high K⁺/Na⁺ ratio. Treated plants exhibited higher growth rates and an increase in antioxidant activities due to the upregulation of Ascorbate PeroXidases (APX), SuperOxide Dismutases (SOD), PerOXidases (POD) and reduced Glutathione (GSH) enzymes (Bilal et al., 2023).

These two stresses represent the main abiotic stress factors that limit crop production globally (Trenberth et al., 2014; Rodriguez and Durán, 2020).

Drought is defined as a period when the available water is insufficient for an organism or environment to function at its best (Kamran et al., 2022). Drought stress represents one of the most critical threats to plant productivity, and thus to global food production, affecting all the stages of plant growth. It leads to a reduction in turgor pressure, affecting cell division, enlargement, and differentiation (Farooq et al., 2009, 2009).

Drought thus affects plants in many ways: typical symptoms include stunted plants, scorching, rolling, and yellowing of leaves, and permanent wilting (Seleiman et al., 2021). Moderate drought stress can induce modifications in the RSA and in the allocation of resources by the plant. In the case of severe drought stress, the roots shrink, and alterations occur at PhotoSystem II (PS II) (Ma et al., 2020).

Soil salinity is defined as the increased amount of sodium (Na⁺) and especially chloride (Cl^-) ions in soils, resulting primarily from natural events such as weathering of parent rocks, seawater, or atmospheric deposition. Other than that, anthropogenic processes, such as poor drainage facilities, irrigation with brackish groundwater, unsuitable water management, and ‘cultural’ errors in irrigated agriculture, can increase soil salinity (Evelin et al., 2019). Salinity causes ionic imbalance and alters metabolic pathways in plant cells, like the synthesis of proteins and the function of some enzymes and ribosomes. Besides, Na⁺ competes with other essential nutrients like phosphate, nitrate and potassium (Angon et al., 2022).

From several reports, it became evident that the plant microbiome can play a role in protecting against high salinity and drought (Yang et al., 2009; Rolli et al., 2015; Berg et al., 2016).

The endophytic fungus Piriformospora indica enhances the expression of genes involved in the drought stress response of maize hosts by increasing auxin, ABA (ABscissic Acid), SA, and cytokinin levels (Zhang et al., 2018). Also, Trichoderma harzianum was shown to improve drought tolerance in rice, by modulating the activity of genes for aquaporin and dehydrin, Dehydration responsive element-Binding Protein (DBP), and SOD (Pandey et al., 2016).

Symbiotic relationships between plants and endophytic fungi such as Piriformospora indica can enhance the adaptation of plants to drought stress by regulating amino acid and soluble sugar metabolism. For instance, P. indica was found to improve the adaptation of barley (Hordeum vulgare L.) to drought stress (Ghaffari et al., 2019). Soybean inoculated with the endophytic fungus Porostereum spadiceum AGH786 under saline conditions showed reduced effects of salinity. The endophyte caused decreasing levels of JA and ABA and increasing levels of GA3, leading to improved plant growth (Hamayun et al., 2017). Similarly, researchers observed a positive effect of the interaction between the endophytic fungus strain Yarrowia lipolytica and maize under saline conditions, which improved plant growth attributes such as leaf relative water content, levels of oxidative enzymes and chlorophyll content through the enhancement of metabolism and hormones (ABA and IAA) secretions (Gul Jan et al., 2019).

Endophytic microbes can alleviate the salt-generated oxidative stress in plants by activating genes for ion transporters, ROS scavenging, and by activating the production and signalling of phytohormones such as auxin, JA, and Ethylene (ET) (Brotman et al., 2013; Ghaffari et al., 2016; Qin et al., 2018; Eida et al., 2019).

Finally, seed bio-priming is a novel beneficial technique aiming to employ bio-stimulating agents like growth-promoting microorganisms to improve the physiological functioning of seeds and their stress resilience (Chakraborti et al., 2022). Two salt-tolerant endophytic fungi, Paecilomyces lilacinus KUCC-244 and Trichoderma hamatum Th-16 were used for bio-priming wheat and mung bean (Vigna radiata L.) seeds. Results showed that both endophytes, in particular T. hamatum Th-16, increased the growth and chlorophyll content of wheat and mung bean plants under extreme salinity conditions. The primed plants also exhibited increased activities of antioxidant enzymes and enhanced photosynthetic attributes (Irshad et al., 2023).

Phytopathogenic bacteria are predominantly represented by the genera Agrobacterium, Bacillus, Burkholderia, Clavibacter, Erwinia, Pantoea, Pseudomonas, Ralstonia, Streptomyces, Xanthomonas, and Xylella (Muthu Narayanan et al., 2022). Bacterial diseases can be systemic or localized, with the most common symptoms in plants being galls and overgrowth, wilting, rot, scabs, necrosis, chlorosis, and blights (Nazarov et al., 2020).

Fungal metabolites produced by Trichoderma harzianum have shown strong antibacterial activity against Ralstonia solanacearum, a phytopathogenic bacterium that causes bacterial wilt disease in tomato (Solanum lycopersicum L.) plants (Yan and Khan, 2021).

Endophytic actinobacteria were isolated from Chilean native potatoes (Solanum tuberosum subsp. tuberosum L.) and they were demonstrated to act against Pectobacterium carotovorum subsp. Carotovorum and P. atrosepticum, bacterial pathogens that cause tissue maceration symptoms in potato tubers. One of the isolates, Streptomyces sp. TP199, was found to inhibit the growth of Pectobacterium sp., reducing tissue maceration symptoms. Moreover, strain TP199 showed metal-dependent Acyl Homoserine Lactones (AHL) quorum quenching activity, which can inhibit communication between bacterial cells (Padilla-Gálvez et al., 2021).

The most common plant pathogenic fungi are Alternaria spp., Aspergillus spp., Colletotrichum spp., Fusarium spp., Phytophthora spp., Pythium, and Pyricularia spp., while anthracnose, dieback, gall, powdery mildew, blight, rust, rot, wilt, and smut are examples of diseases caused by these fungal phytopathogens (Muthu Narayanan et al., 2022).

Some Bacillus strains significantly increased antioxidant enzymes like SOD, PerOXidase, PolyPhenol Oxidase and Phenylalanine Ammonia Lyase in leaves and roots of the rice plant, contrasting the fungus Pyricularia oryzae. They also secreted different biocontrol molecules such as proteases, glucanases, and siderophores in the rice rhizosphere (Rais et al., 2017). In wheat, Fusarium graminearum is the cause of Fusarium head blight as well as Fusarium foot and root rot. Colombo and colleagues (Colombo et al., 2019) studied the biocontrol activity of Streptomyces spp. on F. graminearum in spring wheat and one strain, DEF09, effectively inhibited FHB under controlled and field conditions by blocking the spread of the pathogen at the infection site.

Endophytes with biocontrol potential against Rhizoctonia solani, a fungal pathogen causing sheath blight disease in maize, were isolated from Stevia rebaudiana plants. Three bacterial isolates, identified as Ochrobactrum ciceri SR1EB1, Achromobacter spanius SR1EB11 and Bacillus licheniformis SR2EB5, showed growth inhibition effects against R. solani (Vyas and Singh, 2023).

Furthermore, the gram-positive bacterium Micromonospora, isolated from nitrogen-fixing nodules of leguminous plants, showed biocontrol effects on fungal pathogens and stimulation of plant immunity in tomato. Root inoculation with Micromonospora strains showed reduced infection from the fungal pathogen Botrytis cinerea, and investigations on defence mechanisms highlighted an increased induction in JA-related defence pathways (Martínez-Hidalgo et al., 2015).

Plant viruses are globally diffused plant pathogens, obligatory parasites due to their need for replication within plant cells. Plant viruses are primarily RNA viruses, and the ones considered most important are Tobacco Mosaic Virus (TMV), belonging to the family Virgaviridae, Tomato Spotted Wilt Virus (TSWV) (Tospoviridae), Tomato Yellow Leaf Curl Virus (TYLCV) (Geminiviridae), Cucumber Mosaic Virus (CMV) (Bromoviridae), Potato Virus Y (PVY) (Potyviridae), Cauliflower Mosaic Virus (CaMV) (Caulimoviridae), African Cassava Mosaic Virus (ACMV) (Geminiviridae), Plum Pox Virus (PPV) (Potyviridae), Brome Mosaic Virus (BMV) (Bromoviridae), Potato Virus X (PVX) (Alphaflexiviridae) (Scholthof et al., 2011). Symptoms of viral diseases in plants include growth suppression, deformation, discoloration, necrosis, and impaired reproduction (Nazarov et al., 2020). Investigating the contribution of microbiota towards viral infection (Zhou et al., 2015), identified two new butyrolactones (aspernolides C and D) along with two previously known butyrolactones (aspernolides A and B) from a culture of the endophytic fungus Aspergillus versicolor. When tested against viruses, both aspernolides C and D exhibited moderate anti-TMV activity. Similarly, Kiarie et al. (2020) tested the ability of fungal endophytes to contrast Maize Lethal Necrosis (MLN), a serious disease affecting maize crops in eastern Africa. This disease is caused by the co-infection of maize plants with Maize Chlorotic Mottle Virus (MCMV) (Tombusviridae) and Sugarcane Mosaic Virus (SCMV) (Potyviridae). Maize plants inoculated with T. harzianum and Metarhizium anisopliae showed reduced severity and titer of SCMV, respectively, indicating their potential to induce resistance against SCMV. However, no significant effect was observed on the MCMV.

Invasive insects cause at least $70 billion in crop losses every year (Bradshaw et al., 2016). Panaccione et al. (2014) summarized a series of studies on the role of bioactive alkaloids produced by endophytic fungi in protecting plants against herbivores. Four major types of bioactive alkaloids (ergot alkaloids, indole-diterpenes, lolines and peramine) are produced by fungi from the Clavicipitaceae family. Symbioses between plants and these endophytes have significant effects on both insects and mammalian herbivores, largely due to the production of these bioactive alkaloids. Ergot alkaloids contribute to herbivore resistance, also affecting nematodes. They also act through feeding deterrence, delayed development, and increased mortality of insects.

Loline alkaloids exhibit insecticidal and feeding-deterrent activity. Lolines are often present in grasses with fungal endophytes of the genera Epichloë and Neotyphodium (Wilkinson et al., 2000). Genetic analyses to determine whether the production of lolines in plants is active against aphids highlighted that the endophyte Epichloë festucae showed heritable variation in the expression of loline alkaloids. Analyses on Lol+ (alkaloid expression) and Lol- (no expression) linked alkaloid expression to activity against aphids, and the levels of alkaloids in the plants were correlated with the level of anti-aphid activity (Wilkinson et al., 2000).

Peramine is the most widely distributed of the four classes of epichloae-derived secondary metabolites. It is another alkaloid that acts as a strong feeding deterrent for different insects. Peramine is water-soluble and is dispersed throughout the plant (Rowan, 1993; Panaccione et al., 2014).

Additional information about the role of Endophytic EntomoPathogenic Fungi (EEPFs) was recently made available by (Samal et al., 2023).

On the other hand, different endophytes could be used not only to prevent herbivore damage in plants but, in some cases, to favour this phenomenon for domestic herbivores, reducing undesirable molecules present in plants for livestock nutrition (Bluett et al., 2005a, 2005b).

Nematodes are small, non-segmented invertebrates. They are the most abundant animals on Earth and are fundamental to the soil-food web (Gamalero and Glick, 2020). The phylum Nematoda comprises more than 30,000 species (Bernard et al., 2017), classified into five groups: bacterivores, fungivores, herbivores, omnivores, and predators (van den Hoogen et al., 2020).

Nematodes include the so-called Plant-Parasitic Nematodes (PPN), among which some of the most important are root-knot nematodes (Meloidogyne spp.), cyst nematodes (Heterodera spp. and Globodera spp.), and root-lesion nematodes (Pratylenchus spp.) (Kumar and Dara, 2021). Over 4,100 species of PPNs have been identified, causing an estimated $80–$118 billion dollars per year of damage to crops (Bernard et al., 2017). PPNs can damage the host plant through a needle-like oral structure called stylet, used to release specific enzymes inside plant tissues (Pulavarty et al., 2021, p. 202), and the group of root-knot nematodes develop root knots by forming a complex of multinucleate hypertrophied giant cells, which cause visible knots or galls at the root level (Martínez-Medina et al., 2017). More detailed information about nematodes, their characteristics, and modes of action can be found in Jones et al. (2011).

The main way to fight PPNs is to use chemical nematocides, but these are expensive and harmful to the environment, and EU regulations are constantly reducing the nematocides available for agriculture (Poveda et al., 2020). Therefore, the biocontrol of nematode infection is becoming a promising possibility.

Endophytes isolated from banana (Musa acuminata AAA Cavendish) plant roots infected with Meloidogyne spp. were tested against Meloidogyne javanica, and one strain, named SA and identified as Streptomyces spp., showed an inhibition rate of more than 50% in vitro and a biocontrol efficiency of more than 70% in sterile soil against the nematode (Su et al., 2017).

Fungi could also represent a valuable source of biocontrol agents against plant-infecting nematodes. The fungus Trichoderma harzianum strain BI was used against M. javanica (Sahebani and Hadavi, 2008), reducing the infection rates of the nematode through penetration inside the nematode egg mass matrix, leading to reduced hatch levels. T. harzianum BI also increased the activity of resistance-related enzymes in plants, such as POX, PPO, and PAL. Further investigation showed that chitinase activity in T. harzianum BI culture filtrates increased in the presence of colloidal chitin and nematode eggs, implying its potential for degradation of chitin present in nematode eggs.

The root endophyte strain T. harzianum T-78 was used to study the protective effects on tomato plants against the root-knot species M. incognita (Martínez-Medina et al., 2017) using a split-root system, in which the two halves of the plant roots were allowed to grow in two different pots, one for the treatment and the second as a control. T-78 inoculation decreased the amount of root galls, and a significant reduction in the number of nematode egg clusters was observed in systemic tests. Moreover, the expression profile of the SA-responsive marker genes Pathogenesis-Related protein 1a (PR1a) and Pathogenesis-Related protein P6 (PR-P6) was higher in T-78-pretreated plants infected with the nematode and SA concentrations were higher compared with the non-pretreated ones. Finally, the expression analysis of the JA-responsive genes Proteinase Inhibitors II (PI II) and MultiCystatin (MC) after M. incognita infection showed that JA signalling was down-regulated in plants not inoculated with T-78, while in tomato plants pre-inoculated with T-78, the inhibition in genes PI II and MC showed a significant reduction.

More specific reviews are available to deepen the knowledge about the use of endophytes as biocontrol agents against nematodes (e.g., Gamalero and Glick, 2020; Kumar and Dara, 2021).