Medicago truncatula-Sinorhizobium meliloti-Fusarium oxysporum Tripartite Interaction Alters Nodulation and Nitrogen Fixation

Plants encounter a myriad of microorganisms at the root-soil interface that can invade with detrimental or beneficial outcomes. Promoting a symbiotic relationship may interfere with the restriction responses to pathogens. In this study, we established a tripartite Medicago truncatula (Mt)-Sinorhizobium meliloti (Sm)-Fusarium oxysporum (Fo) interaction to study the effect of the interplay between symbiosis and defense on the symbiosis establishment and efficiency in a resistant (A17) and susceptible (TN1.11) genotype to Fo infection. Our results showed that Sm induced the expression of allene oxide cyclase 1 and 2 (MtAoc1) and (MtAoc2) at 6 and 24 h post infection (hpi) in A17 and at 24 hpi in TN1.11. This up-regulation of MtAoc1 and MtAoc2 was also observed when co-inoculating TN1.11 genotype with Fo and Sm. Moreover, in this later treatment the expression level of the gene of the common signaling pathway doesn’t make infection 3 (DIM3) was reduced compared to Sm-inoculated plants. This reduction of DIM3 transcripts was concomitant to a decrease in nodule number (NN), nodule fresh weight, and nitrogen-fixing rate (NFR), and to an accumulation of spermine in nodules. This defectiveness in symbiosis parameters, in the co-inoculation treatment, was also observed in A17 genotype, along with an accumulation of phytohormones in leaves and polyamines in nodules. Our results suggest that in the tripartite interaction, the high expression of MtAoc and the decrease of DMI3 transcripts along with the accumulation of phytohormones and polyamines in leaves overlapped with symbiosis establishment and functioning and thus, reduced NN and NFR.


Introduction
In nature, plants employ sophisticated systems to distinguish pathogens from mutualistic microorganisms.Recognition of microbes triggers corresponding plant genetic programs resulting in the restriction of detrimental, or the promotion of mutualistic interactions respectively (Feng et al. 2019).Interacting with both microbes simultaneously is a precarious balance, which is controlled by two opposite, yet overlapping, systems: the symbiotic and the immunity systems.Pathogenic attackers are perceived through their pathogen-associated molecular patterns (PAMPs) by cell wall plant receptors, which trigger a downstream signaling cascades leading to immune response and damping of the pathogen (Dangl and Jones 2001).Similarly, symbionts, like nitrogen-fixing rhizobia and arbuscular mycorrhizae, utilize their microbial features, i.e.Nod factors (NFs) and Myc factors, to communicate with the plant receptors, bypass the immunity system, and extensively infect their plant hosts (Mus et al. 2016).
Legumes (Leguminosae or Fabaceae) are important crops for both human and animal diet.They provide fiber, protein, carbohydrate, B vitamins, iron, copper, magnesium, manganese, zinc, and phosphorous (Polak et al. 2015).In several farming systems, they are often used as an inter-crop (e.g., combined with cereals) or in crop rotation.The benefit of such practice is a decrease in the excessive use of fertilizers.
This benefit is due to the capability of most legume species to establish nitrogen-fixing symbiosis with rhizobia to fix atmospheric nitrogen (N 2 ) through the nitrogenase activity that reduces N 2 to NH 3 , located in root nodule bacteria (Gourion et al. 2015;Masson-boivin and Sachs 2018;Sachs et al. 2018).This association is a promising approach to increase the production of legumes and allow increases of the productivity in other crops.However, legumes represent suitable hosts for a plethora of pathogens, which are simultaneously present with rhizobia in the rhizosphere.Thus, legume plants, specifically the model plant Medicago truncatula, provide a model organism to address the plantsymbiont-pathogen tripartite interaction topic and decipher the overlapping between symbiotic and pathogenic signaling pathways.
Several questions in this tripartite interaction have been investigated in the recent years.However, the main focus of these studies is generally limited to the positive effect of symbiosis on enhancing defense responses and resistance in infected plants.Hitherto, studies investigating the interplay of symbiont-induced and pathogen-induced responses and the limiting effect of the pathogen on the symbiont, when present simultaneously, are still scarce.Comparative transcriptome analyses of M. truncatula-symbiont and M. truncatula-pathogen interactions revealed several parallels in the two responses namely calcium influx (Tian et al. 2020), ROS production (Rey et al. 2019), and activation of a MAPK cascade (Chen et al. 2017).Besides, phytohormones are key players in defense and involved as negative regulators of symbiosis establishment from the very first stages of interactions (Akamatsu et al. 2016;Ma and Ma 2016).However, these responses are regulated differently during symbiosis and are transient and local.
How do plants coordinate downstream responses when simultaneously challenged by symbionts and pathogens, and is there a molecular switch that can either promote the Rhizobium-legume symbiosis or activate plant immunity ?Does the plant routinely lower its defenses to allow symbiotic partners in ?Since microbial signatures are perceived at the cell periphery, is there any overlap between immunity and symbiotic receptors ?To approach these questions, scientists utilized mutant plants in key defense and/or symbiotic genes to decipher the involvement of one signaling pathway in the other.Studies on M. truncatula, or Lotus japonicus, mutants in one or more genes of the common symbiotic signaling pathway (dmi1, dmi2, and dmi3), in NF perception (nfp and lyk3), and in transcription factors (nsp1, nsp2, and nin) corroborate the symbiotic pathway being involved in pathogen interaction beyond the restricted group of symbiotic interactions (Rey et al. 2013(Rey et al. , 2015;;Zhang et al. 2015Zhang et al. , 2019;;Gavrin and Schornack 2019;Skiada et al. 2020).Other reports challenged this hypothesis and showed no effect of these genes on pathogenic infection and immunity signaling (Huisman et al. 2015;Kelly et al. 2018).Moreover, mutants in phytohormones pathways, namely ethylene and cytokinin (sickle, efd, and cre1) were shown to be affected in their susceptibility/resistance response to pathogens (Varma Penmetsa et al. 2008;Ben et al. 2013;Rey et al. 2015;Laffont et al. 2015).Regarding jasmonic acid, it was reported that low levels of JA in M. truncatula plants facilitates Aphid infestation and colonization (Lu et al. 2020;Pandharikar et al. 2020).In Symbiosis, Sun et al. (2006) showed that JA suppresses nodule formation in M. truncatula plants, including their responsiveness to Nod factor, by interfering with Ca spiking and the expression of MtRIP1 and MtENOD11.Other efforts in this field studied these questions through a co-inoculation procedure, either by a culture filtrate of the pathogen and/or the symbiont, or by the two microbes (Feng et al. 2019;Benezech et al. 2020).Lopez-Gomez et al. (2012) and Chen et al. (2017) showed an antagonistic effect of the defense pathway on the nodule formation in the initial steps of rhizobium-legume interaction by flg22 treatment or infection of Pseudomonas syringae pv.tomato DC3000, respectively.Tallying with previous data, a study of Medicago-Sinorhizobium-Ralstonia tripartite interaction showed an inhibition of Medicago nodulation (Benezech et al. 2021).However, Feng et al. (2019) discussed that a combination of chitooligosaccharides and lipochitooligosaccharides recognition would promote a symbiotic outcome.Regarding filamentous pathogens, i.e. fungi and oomycete, no co-inoculation studies investigating the regulation of symbiotic responses in Legume-Rhizobium interaction, when simultaneously present with filamentous pathogen, were reported.The few reported studies utilized mutant plants in key symbiotic genes to decipher how root colonization by root-infecting oomycetes like Phytophthora palmivora and Aphanomyces euteiches, would proceed (Reyet al. 2013(Reyet al. , 2015)).Both studies suggested a role for symbiosis genes, namely RAM2, NIP/LATD, NFP, LYK3, ERN, EFD, and LIN, in disease development.
Here, we propose to deepen our understanding of plantsymbiont-filamentous pathogens interaction.To that end, we conducted in vitro and in vivo experiment with the tripartite interaction M. truncatula-F.oxysporum-S.meliloti.We investigated the expression patterns of MtAoc1 and MtAoc2 genes, involved in the production of allene oxide cyclase, a precursor of jasmonic acid (JA), and of the LRR kinase receptor doesn't make infection2 (DIM2) and the calcium-calmodulin-dependent protein kinase doesn't make infection3 (DIM3) from the common symbiotic signaling pathway.In parallel, phytohormones and polyamines were quantified in leaves and nodules.The hypothesis behind this research is that the overlapping between defense-induced and symbiosis-induced responses will alter symbiotic interaction and establishment and thus nitrogen-fixing capacity.

Bacterial Strain
Sinorhizobium meliloti TII7 (Sm) (Zribi et al. 2004) was used to inoculate M. truncatula plants.The bacterial strain was grown in liquid yeast extract mannitol medium (YEM) (Vincent 1970) for 48 h at 28 °C, and the final concentration of the inoculum was adjusted to 10 8 CFU/ml.

Fusarium oxysporum Infection of M. truncatula
Fusarium oxysporum (Fo) strain (KLR13) from the laboratory collection (Haddoudi et al. 2021), and characterized for its pathogenicity against M. truncatula plants ( Batnini et al. 2021), was used in this study.To prepare the spore suspensions, six fungal agar-discs (6 mm in Ø) from 21 days-old culture were sub-cultured in 100 mL liquid Potato Dextrose Broth medium (PDB) at 27 °C with shaking at 160 rpm for 7 days.The obtained suspension was filtered and adjusted to 10 6 conidia ml −1 .

Plant Materials and Growth Conditions
Two M. truncatula genotypes (A17 and TN1.11) previously characterized as respectively resistant and susceptible to F. oxysporum infection (Batnini et al. 2021) were used.Seeds were scarified and germinated as previously described (Mhadhbi et al. 2005), and similar sized germinated seedlings were transplanted in pots containing sterilized vermiculite:perlite mixture (3:1).Plants were incubated in controlled environment rooms at 25/22 °C day/ night temperatures, 60-70% relative humidity, and 16/8 h photoperiod.Pots were regularly watered with a free-nitrogen nutritive solution Vadez et al. (1996).Nitrogen-fertilized plants were watered with KNO 3 enriched solution (2 mM).48 h old seedlings were inoculated with S. meliloti culture adjusted to 10 8 CFU/ml.Three days later, seedlings were infected with F. oxysporum following Haglund (Haglund 1989) description with some modifications.Roots were trimmed and dipped in the conidial suspension (10 6 conidia ml −1 ) for 10 min.Control plant were immersed in sterile Potato Dextrose Broth (PDB).Then, seedlings were re-planted in the same pots.Four treatments for each genotype were established as follow: C: KNO 3 -fertilized plants, Sm: Plants inoculated with S. meliloti, Fo: KNO 3 -fertilized plants infected with F. oxysporum, Sm + Fo: Plants inoculated with S. meliloti and infected with F. oxysporum.Plants were harvested 6 weeks after inoculation.

Gene Expression Analyses
For the expression analysis of DMI2 (MTR_5g030920), DMI3 (MTR_8g043970), MtAoc1 (allene oxide cyclase 1: MTR_5g053950), and MtAoc2 (allene oxide cyclase 2: MTR_7g417750) genes in roots, M. truncatula plants were grown for 2 weeks in glass tubes in liquid nitrogen-free B&D medium (Broughton and Dilworth 1971).The lower half of the tube was wrapped with a black cover to shield the roots from light.Each root was treated with 1 ml of S. meliloti cell suspension (10 8 CFU/ml) and/or dipped in F. oxysporum spore solution (10 6 conidia ml −1 ) for 10 min.Mock treatment consisted of water.Total RNA was extracted using an RNeasy plant mini kit (Macherey-Nagel, Düren, Germany) followed by treatment with DNase I (Ambion, Austin, TX, USA) to remove genomic DNA.300 ng of total RNA was used to synthesize First-strand cDNA using iScript reverse transcriptase (Bio-Rad, Hercules, CA, USA).Real-Time quantitative Reverse Transcription PCR (qRT-PCR) was performed using Power SYBR Green master mix (Applied Biosystems, Foster City, CA, USA) and 1 µl of a twofold diluted cDNA template.Results were quantified using the Ct method (Livak and Schmittgen 2001).Standardization of transcript levels was carried out based on the expression of the gene M. truncatula elongation factor 1-alpha EF-1α (MTR_6g021800).Four independent biological replicates were processed for each treatment.Primers were designed using the NCBI Primer BLAST tool.
ACC was organically extracted.200 mg of leaf tissue was ground in liquid nitrogen and homogenized with 1.5 ml of 100% methanol, 10 μl 2-4D as standard internal, and 2% BHT (w/v).The homogenate was ultrasonicated during 25 min at 4 °C, stirred for 2 h at 4 °C, and centrifuged at 8000 g for 10 min at 4 °C.The supernatant was collected in falcon tubes with 1 ml of HCL 1N, centrifuged (8000 g for 10 min at 4 °C), and the organic phase was collected with 750 μl of ethyl acetate.After centrifugation (8000×g for 10 min at 4 °C), the supernatant was collected and dried completely under nitrogen stream, suspended in 500 μl of CAN, and filtered through nylon filter (0.22 µm).
To quantify the phytohormones, an Acquity-class ultrahigh performance liquid chromatography system UPLC was used to deliver solvent and introduce samples.Samples were injected into an Acquity UPLC HSS T3 1.8 μm, 2.1 mm × 100 mm column.The column was eluted at a constant flow rate of 0.6 ml min −1 , and a gradient elution with water containing 0.01% formic acid (solvent A) and with acetonitrile containing 0.01% formic acid (solvent B).The gradient profile was applied as follows (t (min); %B): (0; 5%), (6; 100%), (6.10; 5%), (8; 5%).Eluates were detected using a Xevo TQ-S Triple Quadrupole Mass Spectrometer (UPCL/ESI-MS/MS, Waters) in a negative electrospray ionization mode.The ion sputtering voltage was set at − 2 kV and the source temperature at 150 °C.Phytohormones were detected in the multiple-reaction monitoring mode of the tandem mass spectrometer with the following transitions:

Free Polyamines (PAs) Contents
Analysis of free PAs was performed as described in (López-Gómez et al. 2016).0.2 g of ground leaves and nodules were incubated with 0.6 ml of 5% (v/v) cold perchloric acid (PCA) for 24 h at 4 °C.The homogenate was centrifuged (3000×g, 5 min, 4 °C) and 200 µl aliquots of the supernatant were dansylated as described below.The derivatization was carried out by mixing 200 µl aliquots of the extracts with 400 µl of dansyl chloride (prepared in acetone, 10 mg ml −1 ) and 200 µl of saturated sodium carbonate.After brief vortexing, the mixture was incubated in the dark at room temperature overnight.To eliminate excess dansyl reagent, the mixture was incubated with 100 µl of proline (100 mg ml −1 ) for 30 min.The dansylated polyamines were extracted in 500 µl of toluene.The organic phase was collected, dried completely under nitrogen stream, and dissolved again in 100 µl of acetonitrile.The analysis of the free PAs: putrescine, spermidine, homospermidine, and spermine was carried out by HPLC (Agilent Technologies 1260) using a Hewlett-Packard system equipped with a reverse phase column (4.6 mm × 250 mm C18).The column flow rate was 1.5 ml min −1 and the elution gradient was prepared with water and acetonitrile.The column was equilibrated with 70% acetonitrile and 30% water before injecting 10 µl samples.This was followed by a linear gradient ending in 100% acetonitrile after 9 min.The final step was held for 4 min before regenerating the column.Detection was performed with a fluorimeter using excitation and emission wavelengths of 415 nm and 510 nm, respectively, according to (Flores and Galston 1982).A relative calibration procedure was used to determine the PAs in the samples, using 1,7-diaminoheptane (HTD) as an internal standard and quantities of PA standards ranging from 0.3 to 1.5 nmol purchased from Sigma.The results were expressed in nmol g −1 fresh weight.

Statistical Analyses
The experiment had a completely randomized block design.One-way analysis of variance (ANOVA) was performed using the SPSS 18 program, and means were separated according to the LSD Tukey test at P ≤ 0.05.Data shown for dry matter, NN, NFW, NFR, phytohormones accumulation, polyamines contents in leaves and nodules are means of 6 replicates (each replicate with six plants).For genes expression analyses data are means of 4 replicates (each replicate represents 3 plants pooled together).All data in this paper are expressed as mean ± standard error (SE).

Relative Expression of DMI2 and DMI3
An in vitro culture was conducted to determine the relative expression levels of DMI2 and DMI3, two genes of the common signaling pathway essential for the establishment of the symbiosis.Expression patterns were determined at 6 hpi: corresponds to brunching and deformation of root hairs, and curling of root hairs, division of cortical cells, and 24 hpi: corresponds to the formation of infection threads and initiation of primordial nodule.
1 3 DMI2 and DMI3 showed different expression patterns in the roots of the two genotypes (Fig. 1).At 6 hpi, DMI2 was slightly repressed in roots of A17 inoculated with S. meliloti (Sm) and co-inoculated with S. meliloti and F. oxysporum (Sm + Fo) compared to control conditions.At 24 hpi, DMI2 was still downregulated in roots inoculated with Sm.Regarding the susceptible genotype TN1.11, no significant changes were observed at 6 hpi.At 24 hpi, DMI2 was downregulated in roots inoculated with Sm, while roots infected with Fo showed approximately onefold induction, and no changes in the co-inoculation treatment (Fig. 1).
Regarding DMI3, it was strongly upregulated at 6 hpi, in roots of A17 plants inoculated with Sm and infected with Fo (increased by 33-fold).Co-inoculated roots showed an induction of DMI3 by 22 fold (Fig. 1).After 24 h, DMI3 was slightly downregulated in the roots infected with F. oxysporum and in the co-inoculated roots.However, roots of TN1.11, showed a different expression profile with low induction of the expression of DMI3 (about onefold change), 6 hpi, inoculation with S. meliloti while the infection with F. oxysporum poorly reduced DMI3 expression.At 24 hpi, a slight upregulation was observed in roots infected with Fo and inoculated with Sm, yet the co-inoculated roots showed a lower level of expression of DMI3 compared to the roots in symbiosis only.

Relative Expression of MtAoc1 and MtAoc2
Two genes encoding the biosynthesis of jasmonic acid (MtAoc1 and MtAoc2: allene oxide cyclase1, and 2) were analyzed.The genotype A17 showed opposite expression profiles regarding all treatments (Fig. 1) for MtAoc1 and MtAoc2.Overall, MtAoc1 was upregulated at 6 hpi, by 1.8 and 4.3 times in Sm-inoculated and Fo-infected roots, respectively, and at 24 hpi, by 1.4 fold in Fo-infected roots.In Sm + Fo treatment and at 6 hpi, MtAoc1 was strongly induced by approximately 11.47 times.However, MtAoc2 was downregulated 4 and 2 folds, when compared to control, in roots interacting with both microbes, at 6 and 24 hpi, respectively.Concerning the susceptible genotype TN1.11, the expression of MtAoc1 showed a slight repression by 2 and 1.5 folds in Sm-inoculated roots at 6 and 24 hpi, respectively, while MtAoc2 was highly upregulated in all treatments and at both time points.In fact, MtAoc2 was highly induced, by 10 and 25 folds, in Fo-infected and Sm + Foinoculated roots, respectively, at 6 hpi.At 24 hpi, MtAoc2
With respect to ABA content, leaves of both genotypes showed similar levels under control and Sm inoculation conditions with 30 ng g −1 FW (Fig. 2a).Infection with Fo increased the accumulation of ABA in leaves of both genotypes by approximately 74% when compared to control and Sm treatments.The same behaviour was observed in leaves of co-inoculated plants, but with more prominent accumulation in leaves of TN1.11.ABA content increased by 81.2% and 88.9% when compared to control and Sm treatments, respectively.Analysis of salicylic acid (SA) accumulation showed similar levels in all the treatments of A17 plants when compared to control conditions (Fig. 2b).While for TN1.11 genotype, all treatments showed a significant increase over control conditions, by 1.92, 1.47, and 3 times in leaves of plants inoculated with Sm, infected with Fo, and co-inoculated with Sm + Fo, respectively.In addition, coinoculated plants showed higher accumulation of SA when compared to Sm and Fo treatment (0.39 and 0.64 times), respectively.Regarding JA levels, the most prominent accumulation was observed in leaves of A17 plants infected by Fo where it increased by approximately 26 times when compared to all treatments within this genotype (Fig. 2c).However, JA in plants inoculated with Sm and co-inoculated with both microorganisms showed similar levels to control conditions.For TN1.11, only inoculation with Sm increased the accumulation of JA by 2 times, relative to control conditions.
The phytohormone indole-3-acetic acid (IAA) in leaves of A17 plants with Sm + Fo increased by 1.3 times compared to all the treatments.However, in TN1.11,only the plants infected with Fo showed a significant increase by 63% over the other treatments.
The ethylene precursor 1-aminocyclopropane-1-carboxylic acid (ACC) showed generally higher concentrations in TN1.11 than in A17, where an increase of ACC accumulation by approximately 47%, was observed in leaves of the co-inoculated plants when compared to the rest of the treatments.In the TN1.11 genotype ACC increased by 68% and 42% in Fo and Sm + Fo co-inoculation, respectively, when compared to control and Sm inoculation treatments.

Polyamines Accumulation in Leaves
Spermine (SPM) and spermidine (SPD) were quantified in the leaves of both genotypes.Overall, the values were comparable between the two genotypes, ranging from 3.9 to 6.3 nmol g −1 FW for spermine and from 13.8 to 21.6 nmol g-1 FW for spermidine (Fig. 3).The level of SPM showed non-significant increase in A17 leaves when co-inoculated with both microorganisms compared to the control and Sm treatment.Within TN1.11 genotype, the more prominent accumulation of SPM was observed following the co-inoculation with Sm + Fo, SPM content increased by 20% when compared to other treatments.

Polyamines Accumulation in Nodules
The analyses of PAs in the nodules of both genotypes revealed different accumulation patterns in both genotypes (Fig. 4).In the co-inoculation treatment (Sm + Fo), the A17 nodules showed a significant increase in the level of all PAs compared to the Sm inoculation.PUT, SPD, HSPD, and SPM increased by 1.17 times, 1.6 times, 0.76, and 3.3 times, respectively.In TN1.11, the co-inoculation treatment increased SPM accumulation in nodules by 50%.However, it caused a reduction of SPD and HSPD accumulation by 67% and 69%, respectively, and no changes were observed for the accumulation of PUT (Fig. 4).

Nitrogen Fixation
The impact of Fo infection on the nodule number (NN), nodule fresh weight (NFW), and nitrogen fixation rate (NFR) was assessed on both genotypes.Overall, in the inoculation treatment (Sm) NN and NFR were comparable in both genotypes and no significant differences were observed (Table 1).In the co-inoculation treatments, all nitrogen fixation parameters decreased significantly within both genotypes but with more prominent effect in the susceptible genotype TN1.11. in which the infection with Fo reduced NN by 73%, NFW by 55%, and NFR by 71%.Likewise, A17 showed a decrease by 41%, 45%, and 72% of NN, NFW, and NFR respectively.

Discussion
Plants in their natural environment are exposed to a myriad of microorganisms ranging from potential mutualists to pathogens.Hence, they need to monitor these interactions to coordinate appropriate responses, whether it is  the restriction of a pathogen or the promotion of a mutualist.Two ways of signaling, namely symbiosis signaling that promotes microbial associations and immunity signaling that hinder them, are the main decision-making regulators.This study analyze the impact of the tripartite interaction Medicago truncatula-Sinorhizobium meliloti-Fusarium oxysporum on the symbiosis establishment from the signaling pathways to the metabolites involved in the defense and symbiosis within A17 and TN1.11, M. truncatula genotypes previously characterized as resistant and susceptible, respectively, towards F. oxysporum (Batnini et al. 2021).

M. truncatula-S. meliloti-F. oxysporum Interaction Alters DMI2 and DMI3 Expression
DMI2 encodes for a transmembrane receptor like kinase which is of an absolute importance in the transduction of signals downstream of the perception of nod factors to initiate events of infection and nodule organogenesis and is usually induced after infection by Rhizobium (Wais et al. 2000).In our study, the levels of expression of DMI2 in the roots of both genotypes inoculated with S. meliloti (Sm) were either slightly downregulated or unmodified which is consistent with reports that showed no induction of DMI2 in roots interacting with rhizobia at similar time points (Lohar et al. 2006).This can be partly due to the fact that rhizobia induce DMI2 expression at a very early stage of symbiosis (Pan et al. 2018) and then the gene might be constitutively expressed and not modified by rhizobia inoculation (Vernié et al. 2016).Regarding pathogenic interactions, some attempts used M. truncatula mutants in the genes of the common signaling pathways, i.e.DMI1, DMI2, and DMI3, to explore whether these genes are involved in pathogenic interactions (Genre et al. 2009;Genre et al. 2009;Rey et al. 2015).No effect on the colonization with the different pathogens, Colletotrichum trifolii; Phytophthora medicaginis; Phytophthora palmivora, was observed however other reports showed that pathogens could exploit the symbiosis-signaling pathway to promote their own colonization (Ben et al. 2013).
In this scenario, in our study, roots of the susceptible genotype TN1.11 showed a moderate induction of DMI2 at 24 h post-infection with Fo, respectively.Downstream of the activation of DMI2, another key component of the symbiosis signaling pathway is DMI3 which encodes for a calcium and calcium/calmodulin-dependent serine/ threonine-protein kinase.This gene decodes upstream calcium oscillations to induce downstream transcription factors and thus initiating nodule organogenesis (Charpentier and Oldroyd 2013).Generally, DMI3 after inoculation with rhizobia is strongly induced at the early stages of interaction and relatively induced at later stages.Indeed, our results showed that DMI3 was strongly expressed in Sm-A17 inoculated roots but slightly induced in TN1.11 roots.These results are consistent with other studies that have shown an increase in DMI3 transcripts (Catoira 2000;Lohar et al. 2006;Rival et al. 2012).However, in the coinoculation treatment, the level of DMI3 transcripts within both genotypes was reduced compared to the treatment of S. meliloti, which can result in alteration of downstream events and, ultimately, of the formation and the number of nodules (Tkacz et al. 2022).In addition, it has been reported that the dmi3 mutants are either altered in cytoplasmic responses or more sensitive to infection with pathogens (Ben et al. 2013;Rey et al. 2015;Genre and Russo 2016).Aligning with this, our results showed that DMI3 was strongly induced, at 6 dpi, in Fo-infected A17 roots.However, at 24 hpi, DMI3 was downregulated in A17 and slightly induced in TN1.11.Such observations extend the role of DMI3 from symbiosis to pathogenesis and suggest that the involvement of DMI3 in response to F. oxysporum may be time-controlled and genotype-dependent.

F. oxysporum Induced High Expression Levels of Allene Oxide Cyclase and Increased the Accumulation of Phytohormones in TN1.11-S. meliloti Symbiosis
Jasmonic acid (JA) together with other phytohormones (ethylene, salicylic acid, abscisic acid), is a key player in both symbiosis and defense in the early and advanced stages of interaction (Liu et al. 2018;Lin et al. 2020).Analyses of MtAoc1 and MtAoc2, coding for the enzyme of the JA biosynthetic pathway, allene oxide cyclase, revealed an upregulation of MtAoc1 in Fo-infected A17 roots and of MtAoc2 in the roots of TN1.11.These results align with several reports that have shown a strong induction of JA-related genes after pathogen attacks and confirmed the pivotal role of JA in defence responses (Yang et al. 2019;Yuan et al. 2019).Moreover, when infected with pathogens, plants accumulate SA, JA, and ethylene to modulate systemic plant defense response (Checker et al. 2018).Indeed, the quantification of phytohormones in Fo-infected leaves revealed a significant accumulation of ABA, SA, and ACC within TN1.11 and ABA and JA within A17.Considering rhizobial inoculation, phytohormones, JA, SA, IAA, ABA and ethylene, are involved in all steps of interaction from perception of nod factors to nitrogen fixation (Buhian and Bensmihen 2018;Liu et al. 2018) and are generally reported as negative regulators of symbiosis (Lin et al. 2020).Thus, concentrations of these hormones need to be tightly regulated and sustained high concentrations can lead to a complete abolishment of symbiosis or too few nodules (Ferguson et al. 2019).Moreover, Tsyganova and Tsyganov (2018) reported that JA negatively regulates the growth of the infection threads and the formation of nodules primordia.Indeed, in the coinoculation treatment, the transcripts of MtAoc1 and MtAoc2 were upregulated in the roots of TN1.11, with greater levels compared to Sm inoculation and similarly to Fo-infected roots at both time points.Consistently, the accumulation of MtAoc1 and 2 transcripts may ultimately limit the number of nodules and decrease the rate of nitrogen fixation.In fact, in concomitance with the severe reduction of nodule number observed within the susceptible genotype TN1.11, the quantification of phytohormones in the leaves revealed a strong accumulation of ABA, SA, and ACC.On the other hand, and within the resistant genotype A17, the regulation of MtAoc genes expression in Sm + Fo-inoculated roots was similar to Sm inoculation.In this scenario, MtAoc1 and MtAoc2 induction 6 h post inoculation may disrupts calcium oscillations which will eventually result in a strictly regulated nodules number (Tsyganova and Tsyganov 2018)

Accumulation of Polyamines in the Nodules Under F. oxysporum Infection
Other metabolites that are important actors in symbiosis and defense responses are polyamines (PAs) (Walters 2003;Hidalgo-castellanos et al. 2019) since they interact with phytohormones and function synergistically in the stress responses (Ugena et al. 2018).It has been reported that the pathogen may modulates PAs metabolism to promote symptoms development and the colonization of the host plants (Stes et al. 2011).In this scenario, PAs accumulation is generally related to H 2 O 2 production, which is a key player in inducing programmed cell death (Marina et al. 2008;Lazzarato et al. 2009).However, when interacting with necrotrophic pathogens, cell death promotes plant colonization and disease development.In fact, Fo-infected leaves of the susceptible genotype TN1.11 accumulated SPD, and we have reported previously that Fo causes wilt disease symptoms development in leaves of TN1.11 in concomitance to a decrease in antioxydant activities (Batnini et al. 2020(Batnini et al. , 2021)).
On the other hand, PAs accumulation in leaves and exogenous application are reported to have negative effect on symbiosis interaction in terms of root nodule number and biomass (Vassileva and Ignatov 1999;Terakado et al. 2006).
Our study showed an accumulation of SPD in the leaves of Sm + Fo-inoculated TN1.11 plants in concomitance with the observed accumulation of phytohormones.This accumulation may reduce nodulation as reported by Terakado et al. (2006).Moreover, SPD accumulation is correlated with a negative effect on photosynthesis by inducing stomatic closure and ABA accumulation which may ultimately reduce plant resources to establish an efficient symbiotic interaction (Shu et al. 2016;Liu et al. 2019).Besides controlling nodule number and biomass, PAs are involved in regulating nodule functioning under stress (Jiménez-Bremont et al. 2014;Hidalgo-Castellanos et al. 2022).In fact, lowering of the accumulation of PAs titers in the nodules are correlated with the NFR and may be one of the reasons for the appearance of nodule senescence syndrome.In this scenario, nodules of the susceptible genotype TN1.11, when infected with Fo, exhibited reduced levels of SPD and HSPD and no accumulation of PUT which might be a sign of the senescence of the infected nodules.On the other hand, several reports argued that PAs act as scavengers of excessive produced reactive oxygen species, mainly H 2 O 2 (Papadakis 2005;Arthikala et al. 2017;López-Gómez et al. 2017;Hidalgo-castellanos et al. 2019).In addition, PAs are implicated in nitrogen fixation and in maintaining a favorable nodular environment for the bacteroid.Indeed, Becerra-Rivera et al. (2020) reported that an ornithine decarboxylase2 mutant strain of S. meliloti showed a reduction in the efficiency of nitrogen fixation, and Flores-Tinoco et al. (2020) highlighted the participation of PUT in the co-catabolism of arginine and succinate in a novel metabolic pathway that maintain the bacteroid life in the nodule.Tailoring with these studies, we found that the nodules of the resistant genotype, A17, in the coinoculation treatment exhibited increased levels of PUT, SPD, HSPD, and SPM, however the NFR was reduced at similar levels within both genotypes.Thus, the accumulation of PAs in A17 nodules may indicate its involvement in mitigating the excessive production of reactive oxygen species, thereby providing a more viable nodule environment for the bacteroid.
Overall, the mixed inoculation with Fo and Sm led to a decrease in the expression levels of DMI2 and DMI3 and to a long standing increment of MtAoc1 and MtAoc2 transcripts at the very early stages of interaction.Moreover, this tripartite interaction increased the level of the negative regulators of symbiosis, i.e. phytohormones and polyamines, in leaves and nodules, which suggest that the host plant steered its responses towards defence over symbiosis establishment with a drastic reduction of nodule number and nitrogen fixing rate.This study presents M. truncatula-S.meliloti-F.oxysporum as a potential model to understand plant-symbiont-filamentous pathogen interaction and gives new insights on "How do plants engage with beneficial microorganisms while at the same time restricting pathogens?".

Fig. 1
Fig. 1 The expression patterns and hierarchical clustering of a DMI2 and DMI3; b MtAoc1 and MtAoc2 in A17 and TN1.11 roots at 6 hpi and 24 hpi.The data were subjected to TBtools Software processings (Chen et al. 2020).Transcript levels were determined by quantitative RT-PCR.The rows represent change rates of expression at both time points along with their average expression difference (log2 fold change): both genotypes under control conditions (A17-C/TN1.11-

Table 1
Effect of the co-inoculation with F. oxysporum (Sm + Fo) on nodule number (NN), nodule fresh weight (NFW), and nitrogen fixation rate (NFR) of M. truncatula plants inoculated with S. meliloti (Sm) Data are mean ± SE In each column, means denoted with different letters differ significantly at p ≤ 0.05 based on HSD Tukey test . However, ABA and ACC were significantly accumulated in the leaves of Sm+Fo-infected plants which agree with several studies in which exogenous application of phytohormones reduces the number and limits the development of nodules(Lin et al. (Lin et al. 2010;chen et al. 2017ganov 2018)ov 2018).Furthermore, phytohormones, i.e.JA, SA, ABA, auxin, and ethylene, are involved in a systemic shoot-derived negative regulation of nodulation called autoregulation of nodulation (AON) (Oka-Kira and Kawaguchi 2006;Lin et al. 2010;Ryu et al. 2012).Our data suggest that, in the co-inoculation treatment, Fo infection caused high accumulation of stress related hormones mainly in leaves of the susceptible genotype TN1.11, which may inhibit nodulation events(Lin et al. 2010;chen et al. 2017).