Sodium Accumulation in Infected Cells and Ion Transporters Mistargeting in Nodules of Medicago truncatula: Two Ugly Items That Hinder Coping with Salt Stress Effects

The maintenance of intracellular nitrogen-fixing bacteria causes changes in proteins’ location and in gene expression that may be detrimental to the host cell fitness. We hypothesized that the nodule’s high vulnerability toward salt stress might be due to alterations in mechanisms involved in the exclusion of Na+ from the host cytoplasm. Confocal and electron microscopy immunolocalization analyses of Na+/K+ exchangers in the root nodule showed the plasma membrane (MtNHX7) and endosome/tonoplast (MtNHX6) signal in non-infected cells; however, in mature infected cells the proteins were depleted from their target membranes and expelled to vacuoles. This mistargeting suggests partial loss of the exchanger’s functionality in these cells. In the mature part of the nodule 7 of the 20 genes encoding ion transporters, channels, and Na+/K+ exchangers were either not expressed or substantially downregulated. In nodules from plants subjected to salt treatments, low temperature-scanning electron microscopy and X-ray microanalysis revealed the accumulation of 5–6 times more Na+ per infected cell versus non-infected one. Hence, the infected cells’ inability to withstand the salt may be the integral result of preexisting defects in the localization of proteins involved in Na+ exclusion and the reduced expression of key genes of ion homeostasis, resulting in premature senescence and termination of symbiosis.


Introduction
Fabacean root nodules are formed as a result of symbiotic relationships with different genera of soil bacteria, generally known as rhizobia. The root nodule contains some special cells (infected cells) serving as a temporary niche for thousands of rhizobia enclosed in a plant-derived membrane, an organelle-like structure, the symbiosome [1,2]. The root nodule becomes a source of nitrogen in the form of ammonia that is produced by differentiated rhizobia (bacteroids). In the process of performing the functions of the host, the plantinfected cell undergoes significant changes, which include the reduction in lifespan, and changes in the expression of different genes, the cytoskeleton pattern, membrane traffic, and distribution of several proteins [3,4]. In connection with these changes, the infected cell may have preexisting features that should be taken into account and require investigation to improve the efficiency of symbiosis and its resistance to harmful environmental conditions.
It is well known that root nodules are more sensitive than the host plant itself to stress caused by factors such as salinity. Salt stress negatively affects nodulation and damages the root nodule structure and ultrastructure, resulting in early symbiosome senescence and a decrease in nitrogen fixation [5][6][7][8]. To improve the nodule tolerance to salt stress, studies involving natural selection and bioengineering of the host plant [7,[9][10][11] and the microsymbionts [12,13] have been performed. However, the reasons behind the nodule sensitivity to salt stress are not fully understood.
One of the main effects of salt stress in plants is an induced change in cell homeostasis, caused by Na + /K + unbalance. Hence, Na + exclusion constitutes one of the main mechanisms of salt stress responses in plant cells. The efflux of Na + from the plant cell depends on exchangers/antiporters K + (Na + )/H + (NHX), which are included in the monovalent cation/proton antiporters (CPA) family [14][15][16][17]. The NHX potassium/sodium exchangers play significant roles in diverse cellular processes-pH regulation, vesicle trafficking and fusion, growth and development, and salt tolerance-and they are considered the main "players" in salt stress response [17][18][19]. According to the extensive studies in Arabidopsis thaliana, NHX are capable of transporting either K + or Na + into the vacuole or endosome in exchange for H + efflux to the cytosol [NHX1-6], or facilitating a Na + efflux out of the cell in exchange for a H + influx into the cell (NHX7/SOS1) [17,[20][21][22]. The A. thaliana genome contains eight NHX genes. AtNHX1, AtNHX2, AtNHX3, AtNHX4, and AtNHX6 are expressed in roots, shoots, and seedlings, and AtNHX7 (SOS1) is mainly expressed in roots and induced by salt stress [23][24][25].
The maintenance of cellular ion homeostasis in plants also involves K + transporters of the group of KT-HAK-KUP, which include the HAK (high-affinity K + ) and the KUP/KT (K + uptake permease/K + transporter) families that contribute to K + homeostasis in low and high external concentrations, respectively [26,27]. Plants also have a large family of predicted cation/H + exchangers (CHX) with functions not yet fully defined, which may also be involved in pH and cation homeostasis [15,16].
Less is known about the gene expression, location, and functions of the exchangers/antiporters in Medicago truncatula and the information concerning their functionality in root nodules is scarce. The M. truncatula genome contains six NHX genes; MtNHX1, MtNHX6, and MtNHX7 (Medtr2g038400, Chr2) that display high expression in roots and leaves; MtNHX2 and MtNHX3 with the highest expression in flowers and relatively low expression levels in other tissues, and MtNHX4 with the highest expression in pods [28]. Upregulation of NHX7/SOS1 has been associated with salt tolerance in M. truncatula and M. falcata and other Fabaceae [28][29][30]. A decrease in K + levels in symbiosomes and vacuoles during the infected cell lifespan, and the mistargeting and partial depletion from the plasma membrane of mature infected cells of two plant Shaker K + channels-inwardrectifier K( + ) channel AKT1 and potassium outward channel SKOR/GORK in mature infected cells-might also compromise the K + availability in both the host cell and the symbiosome [31].
With the aim to clarify the mechanisms involved in Na + exclusion and K + transport in root nodule cells of M. truncatula subjected to salt stress, we performed: (1) in silico expression analysis of selected ion transporters/channels/exchangers genes of both the host plant and the microsymbiont in nodules; (2) the immunolocalization of Na + /K + exchangers MtNHX6 and MtNHX7 in root nodules; (3) qPCR analysis of selected genes involved in ion translocation; and, (4) low temperature-scanning electron microscopy (LTSEM) and X-ray microanalysis of M. truncatula nodules from plants exposed to salt stress and control plants. Our results suggest that the proteins putatively involved in the exclusion of Na + from the cell cytoplasm in mature infected cells are being lost from their target membranes and expelled to young vacuoles. A downregulation (and even suppression of transcription) of genes related to Na + efflux and K + inward transport occurs in the internal region of the root nodule, making the mature infected cells less protected against salt stress. The combination of these events results in excess accumulation of Na + in infected cells and symbiosomes, thus hampering the ability of infected root nodule cells and the root nodule itself to withstand salt stress.

In Silico Gene Expression Analysis of Ion Transporters, Channels, and Exchangers Genes in Root Nodule Developmental Zones
The expression of genes encoding ion transporters, channels, and exchangers of sodium and potassium of the host plant and intracellular rhizobia putatively involved in salt stress defense mechanisms in the nodule was estimated by in silico gene expression analysis in the nodule developmental zones using the Symbimics portal [32,33]. This portal contains gene expression data obtained by laser dissection of the cells of the defined developmental zones of the wild-type mature M. truncatula nodule. Developmental zones of root nodules: meristem, infection zone (ZII), where rhizobia are released into the plant cell and start their differentiation process, the interzone II/III (IZ), and the nitrogen fixation zone (ZIII), which is the mature part of the nodule consisting of small non-infected cells and large infected cells that contain the differentiated nitrogen-fixing bacteroids. The basal part of the nodule contains the zone of senescence where the symbiosis is terminated (ZIV).
The expression of M. truncatula and S. meliloti genes in the nodule developmental zones are shown in Tables 1 and 2, respectively, accession numbers listed in Table S1. A substantial number of the analyzed genes showed changes in their expression level in the developmental zones of the nodule, suggesting important adaptations/alterations in ions transport in both symbionts during the development of symbiosis. The spatial expression analysis of the plant genes showed a tendency to downregulation rather than upregulation of the host plant genes in the nitrogen-fixing part of the nodule, as compared with the meristem. All selected genes were expressed in the nodule meristem, except for MtNHX2 and MtNHX3, which were not expressed in either roots or nodules (Table 1). Conversely, in the mature part of the nodule (ZIII), where bacteroids fix nitrogen, 7 of the 20 analyzed genes were either not expressed or substantially downregulated. These genes included the homologs of A. thaliana genes HAK5, KAT1, and NHX7, which are important in salt stress response. In ZIII, six genes were upregulated (MtHAK6, MtHAK7, MtSKOR/GORK, MtCHX18, MtNHX1, MtNHX6). The upregulation of MtNHX1 and MtHAK6 was also reported by Limpens et al. [34] Table S2).The expression of MtKT3 and MtNHX8 was stable in all nodule zones, and the rest of the genes showed a downregulation with respect to the meristem (Table 1).
These results indicate that the K + /Na + transport systems involved in Na + sequestration in the mature part of the nodule might be compromised. The expression of rhizobial genes involved in the inward transport of K + and the sodium/proton antiporter Na + /H + NhaA showed upregulation in ZIII as compared with the gene expression levels in the infection zone (FIId, FIIp) ( Table 2). However, such an increase probably reflects the drastic increase in the rhizobia number per host cell in mature infected cells that were up to 25 times higher in comparison with the host cell at the start of infection [35].

Localization of MtNHX7 and MtNHX6 in Root Nodule Cells by Confocal and Electron Microscopy
Two M. truncatula NHX exchangers, MtNHX7 and MtNHX6, were used for immunolocalization analysis. MtNHX7 is a homolog of the AtNHX7 (SOS1) gene. AtSOS1 is located on the plasma membrane and involved in Na + exclusion from the plant cell via the plasma membrane to the apoplast. The upregulation of NHX7/SOS1 has been associated with salt tolerance in M. truncatula and M. falcata [28,29], and other fabacean species [30]. The exchanger MtNHX6 is a homolog of AtNHX6 [20]. AtNHX6, a Golgi/endosome/PVC resident, participates in K + or Na + sequestration into endosomal lumen [17,25]. This endosomal exchanger has been selected for the immunolocalization study due to the known changes in the symbiosome membrane identity, and the appearance of some PVC endosome markers on the membrane during the symbiosis development [35]. Wild-type and transgenic nodules carrying the ProMtNHX7:MtMTNHX7:GFP construct were used for the localization study of MtNHX6 and MtNHX7, using anti-MtNHX6 and anti-GFP antibodies, respectively. A secondary Alexa 488-conjugated antibody was used in both cases.
The MtNHX7 protein signal was present in the plasma membrane (PM) of cortex cells, vascular bundles, meristematic cells, and young nodule cells in the infection zone ( Figure 1A-D). In mature infected cells, the pattern of the protein localization was altered, and the signal was lost from the PM and was detected in the cytoplasm, forming a dot-like pattern that reflects a process of internalization of the protein from the PM ( Figure 1C).
The localization analysis of MtNHX6 in nodules showed the presence of the protein in the cytoplasm (Figure 2A), especially strong on the tonoplast membranes ( Figure 2B), reflecting the fusion of endosomes with the tonoplast membrane in the meristem and in newly infected cells ( Figure 2B). At high magnification, the immunosignal was discernible in small bodies with the size of endosomes/vesicular bodies budding from the trans-Golgi network (50-100 nm) ( Figure 2C). At the tissue level, the signal was observed in meristematic cells, distal and proximal infection zones, and in vascular bundles ( Figure 2D,E). In the mature infected cells, however, the pattern of protein distribution was changed, the signal was observed over big dots (500-1000 nm), and young lytical vacuoles, which may suggest the process of mistargeting and the depletion of the protein from its target membranes. The presence of the immunosignal, labeled by gold particles, in vacuole-like bodies (500-1000 nm) was confirmed by electron microscopy, the signal was present in the lumen of young vacuoles ( Figure 3A,B).

Ion Compartmentalization Analysis
Ion compartmentalization analysis was performed in the mature nitrogen-fixing pa of root nodules. High magnification was used to distinguish individual symbiosomes an vacuoles in nodules from control plants and from plants subjected to salt stress (Figu 4A). The distribution of sodium (Na), potassium (K), magnesium (Mg), phosphorus (P sulfur (S), chlorine (Cl), calcium (Ca), iron (Fe), copper (Cu), and molybdenum (Mo) w analyzed. The content and distribution of sodium (Na) and potassium (K) are presente in Figure 4B,C; the data of other elements are presented in Table S3. In nodules subjecte to salt treatment, an increase in Na + content 'on spot' was statistically significant in th cytoplasm of both infected and non-infected cells, and in bacteroids ( Figure 4B) as com pared with nodules of control plants. K + content diminished 2-3 times in the vacuoles non-infected cells ( Figure 4C).
Concerning other ions, salt stress significantly decreased the contents of Ca 2+ in th vacuoles of non-infected cells, and of S in the vacuoles of infected cells. The chlorine co tent increased in the vacuoles and in the cytoplasm of non-infected and infected cells, an the content of Fe increased in the cytoplasm of infected cells (Table S5). According to th results, in the mature part of the nodules of salt-treated plants, Na + accumulated in th cytoplasm and in the bacteroids, while there was a decrease of K + content in the vacuole These results suggest that in salt-stressed nodules, infected and non-infected cells pr sented an insufficient ability to expel Na + from the cytoplasm and to prevent Na + accum lation in the symbiosomes.

Ion Compartmentalization Analysis
Ion compartmentalization analysis was performed in the mature nitrogen-fixing part of root nodules. High magnification was used to distinguish individual symbiosomes and vacuoles in nodules from control plants and from plants subjected to salt stress ( Figure 4A). The distribution of sodium (Na), potassium (K), magnesium (Mg), phosphorus (P), sulfur (S), chlorine (Cl), calcium (Ca), iron (Fe), copper (Cu), and molybdenum (Mo) was analyzed. The content and distribution of sodium (Na) and potassium (K) are presented in Figure 4B,C; the data of other elements are presented in Table S3. In nodules subjected to salt treatment, an increase in Na + content 'on spot' was statistically significant in the cytoplasm of both infected and non-infected cells, and in bacteroids ( Figure 4B) as compared with nodules of control plants. K + content diminished 2-3 times in the vacuoles of non-infected cells ( Figure 4C).
Concerning other ions, salt stress significantly decreased the contents of Ca 2+ in the vacuoles of non-infected cells, and of S in the vacuoles of infected cells. The chlorine content increased in the vacuoles and in the cytoplasm of non-infected and infected cells, and the content of Fe increased in the cytoplasm of infected cells (Table S5). According to the results, in the mature part of the nodules of salt-treated plants, Na + accumulated in the cytoplasm and in the bacteroids, while there was a decrease of K + content in the vacuoles. These results suggest that in salt-stressed nodules, infected and non-infected cells presented an insufficient ability to expel Na + from the cytoplasm and to prevent Na + accumulation in the symbiosomes.

qPCR Expression Analysis of Nodules under Salt Stress
With the aim to detect the effect of salt stress on the expression of plant genes in volved in the exclusion of Na + and translocation of K + , we analyzed nodules from sal treated and control plants. The effect of the salt treatment on the expression of genes re lated to the exclusion of Na + was analyzed: MtNHX1 and MtNHX6, involved in seques tration of Na + into vacuoles and endosomes, MtNHX7, the homolog of SOS1 of A. thalian involved in Na + efflux out of the cell via the plasma membrane to the apoplast, and re cently annotated MtNHX7-like and MtNHX8 genes, as putative homologs were evaluate ( Figure 5A). Salt stress led to significant upregulation of MtNHX1, MtNHX7, an MtNHX7-like genes. Genes involved in the translocation of K + were also investigated: Th expression of the high-affinity potassium transporter MtHAK7, the H + /K + exchange MtCHX18, the potassium inward channel MtAKT1, and the potassium outward channe SKOR/GORK were not affected by salt; however, the expression of MtHAK6 was signif cantly downregulated ( Figure 5B). The expression analysis of the nifH gene encoding on of the subunits of nitrogenase [6] showed a tendency to downregulation that can be ex pected as a well-known effect of salt treatment on nitrogen-fixing activity ( Figure 5C). I summary, salt stress induced certain genes from the NHX clade, involved in Na + seques tration, as could be expected, but had no effect on the expression of K + translocators gene

qPCR Expression Analysis of Nodules under Salt Stress
With the aim to detect the effect of salt stress on the expression of plant genes involved in the exclusion of Na + and translocation of K + , we analyzed nodules from salt-treated and control plants. The effect of the salt treatment on the expression of genes related to the exclusion of Na + was analyzed: MtNHX1 and MtNHX6, involved in sequestration of Na + into vacuoles and endosomes, MtNHX7, the homolog of SOS1 of A. thaliana involved in Na + efflux out of the cell via the plasma membrane to the apoplast, and recently annotated MtNHX7-like and MtNHX8 genes, as putative homologs were evaluated ( Figure 5A). Salt stress led to significant upregulation of MtNHX1, MtNHX7, and MtNHX7-like genes. Genes involved in the translocation of K + were also investigated: The expression of the highaffinity potassium transporter MtHAK7, the H + /K + exchanger MtCHX18, the potassium inward channel MtAKT1, and the potassium outward channel SKOR/GORK were not affected by salt; however, the expression of MtHAK6 was significantly downregulated ( Figure 5B). The expression analysis of the nifH gene encoding one of the subunits of nitrogenase [6] showed a tendency to downregulation that can be expected as a wellknown effect of salt treatment on nitrogen-fixing activity ( Figure 5C). In summary, salt stress induced certain genes from the NHX clade, involved in Na + sequestration, as could be expected, but had no effect on the expression of K + translocators genes, other than the downregulation of MtHAK6.

Discussion
Root nodule infected cells, which are housing thousands of symbiosomes, are physi-

Discussion
Root nodule infected cells, which are housing thousands of symbiosomes, are physiologically different from non-infected cells. They have a short lifespan not exceeding four weeks and other special biological features, such as the modification of the endomembrane system, the transformation of the cytoskeleton pattern, and changes of vectors and destination membranes in membrane trafficking [2,4,[35][36][37][38]. One of the most important features of infected cells is the change in vesicular traffic from the plasma membrane, endosomes, and tonoplast toward the symbiosome membrane, the interface surrounding the bacteria in the host cytoplasm [2][3][4]. Such cell-specific modification of some proteins targeting infected cells favors the propagation and maintenance of the intracellular microsymbiont colony [2,3,35,36,39]. However, the presence of a huge bacterial colony inside the host cells, as it seems, has some negative effects that now start to be revealed. Among the examples of these negative effects, we can indicate the partial loss of functionality of the infected cell vacuoles, manifested by the loss of acidic pH, and a great part of the volume [35] and the drastic changes in the distribution of channel proteins involved in K + transport (AKT1, SKOR/GORK) [31] in mature infected cells. The specific environment of the infected cell, caused by the presence of an intracellular bacterial colony, may have other consequences, making the root nodule more vulnerable to salt stress than other plant organs. Here, we report the location of the MtNHX6 and MtNHX7, which supports the above hypothesis.
In plants, salinity causes osmotic stress at the tissue and the cellular level and decreases turgor pressure, which results in a competition between Na + and K + for the major binding sites in key metabolic processes, affecting enzymatic reactions, protein synthesis, ribosome functions, and cell metabolism [40,41]. Salinity has also a negative effect on photosynthesis [42] which affects root nodules, they consume more than 25% of the total photosynthates [43].
The NHX group exchangers, which are involved in the efflux of Na + , are important regulators of endomembrane K + /Na + homeostasis due to the process of Na + efflux in response to salt stress [29,[44][45][46][47]. Apoplastic Na + depolarizes the plasma membrane, making K + uptake by potassium channel AKT1 thermodynamically impossible, and activating the gated outward rectifying K + SKOR/GORK efflux channels [46] which leads to a diminishing of K + content in the cells. The loss of K + is also one of the symptoms of stress caused by pathogens elicitors, reactive oxygen species (ROS), salinity, and drought [48,49].
Similar to other plant tissues, the root nodule responds to the changing environment by adapting to the dynamic endomembrane system and the transport of necessary substances, including ions. The root nodule tissue similar to other plant cells is equipped with the mechanisms to adapt or diminish the negative effects of salt stress. However, the high sensitivity of the root nodule to salinity is an undeniable fact manifested as rapid senescence of the symbiosomes, drop in nitrogen fixation, and termination of symbiosis [5]. In the present paper, we describe a new aspect of symbiosis development: the changes in the localization of NHX exchanger proteins in the root nodule and the response of nodule cells to the salt treatment in terms of ion distribution and gene expression of ion exchangers, channels, and transporters.
Some of the intrinsic features of the root nodule, according to in silico gene expression analysis (Table 1), is the downregulation or termination of expression of several genes involved in the maintenance of optimal cytosolic K + /Na + ratio in the mature part of the nodule in comparison with the apical part of the nodule, including the meristem and the infection zone ( Table 1). The differential expression of cation transporters in infected and non-infected cells of root nodules was reported before [34] (Table S2). Genes encoding some metal ion transporters (Cu, Zn), show high expression in non-infected cells, but Mn 2+ and Fe 2+ transporters of the NRAMP family and Ca 2+/ H + antiporter VCX1 were highly expressed in infected cells of the zone of nitrogen fixation (34 , Table S3). According to gene expression, data obtained by laser-dissection analysis of infected and non-infected cells in nodule developmental zones, MtNHX1 and MtHAK6, were upregulated in infected cells in the zone of nitrogen fixation. The expression data from the Gene Atlas portal for MtNHX6 and MtNHX7-like genes (Mtr.13430.1.S1_at [log2_rma]) show high expression in young nodule cells of zone II, but the moderate expression in non-infected cells and non-infected cells of nitrogen fixation zone [50].
We speculate that the hypoxic conditions in the mature part of the nodule might be a cause for the spatial differences in gene expression. In root nodules, the central part and especially infected cells have a low O 2 pressure that facilitates nitrogenase activity [36], but in the root nodule meristem and the cortex, the level of oxygen is not so limited. As it was shown for nodules of alfalfa and sweet clover, both higher O 2 permeability and higher respiratory capacity are associated with the nodule tip, which contains meristem and zone of initial infection [51]. It has been reported that the expression of some ion channels is downregulated in hypoxic conditions, for example, ion channel GORK in A. thaliana roots was downregulated up to 3 times within 1 h of hypoxic treatment [52]. The negative effects of hypoxia on the expression of ion channels are being also intensively studied in animal cells [53].
In A. thaliana, Na + influx into the cell is counterbalanced by the active removal of this ion by the mechanisms of Na + extrusion via plasma membrane by AtNHX7/SOS1 [15], or the sequestration of Na + in the vacuole by AtNHX1 and AtNHX2 and in endosomes by AtNHX5 and AtNHX6 [15,20,22,23]. In root nodules, the functional status of these genes may be impaired. According to our data, the spatial pattern and the correct location of exchangers MtNHX6 and MtNHX7 were maintained in nodule meristem, in very young, infected cells, and in most of the non-infected cells. However, in mature nitrogen-fixing cells, exchanger proteins were mistargeted, removed from their destination membranes, and expelled to young vacuolar compartments. The immunolocalization analysis of MtNHX7 and MtNHX6 proteins evidenced comparable dynamics in protein relocation in mature infected cells despite the differences in gene expression in the nodule developmental zones, and their diverse destination membranes: plasma membrane for MtNHX7 and endosome/tonoplast for MtNHX6. Similar changes in localization patterns have also been reported for the MtAKT1 protein involved in K + transport [31] in both mature and senescent infected nodule cells. The reasons for the changes that we observed in the pattern of NHX proteins localization cannot be unequivocally explained. The mistargeting of the proteins might indicate initial signs of degradation of the infected cells or it may be associated with hypoxia in these cells. However, the change in the pattern of ion exchanger localization may be one of the reasons for the low tolerance of infected cells toward the accumulation of Na + . It could be expected that the loss of the proteins from their target membranes may be detrimental to the functional activity of MtNHX7 and MtNHX6, specifically in infected cells, which may result in diminished ability to expel or sequester Na + ions and likely affect other combined ion transport pathways. The efficiently functioning MtNHX7 and MtNHX6 proteins seem to be spatially restricted to non-infected cells of the vascular bundles, meristem, and the external cortex where the proteins display a proper localization pattern (Figure 1). The expression of genes found to be upregulated according to qPCR analyses after NaCl application is probably also spatially restricted to these tissues. On the contrary, the hypoxic central nodule tissue, especially infected cells, seems to be less protected against Na + accumulation due to the loss of proteins from the target membranes.
According to the data presented in this paper, mature infected cells accumulate more Na + than non-infected interstitial cells, as confirmed by the ion distribution analysis. The salt treatment led to an increase in Na + content 'on the spot' approximately four times in the cytoplasm of infected and non-infected cells, and three times in bacteroids (Figure 3) as compared to the cells of control nodules. We should keep in mind that mature infected host cells, due to the drastic increase in the number of symbiosomes, have a volume that is more than five times that of non-infected cells [35]. That means that, according to IDA analysis, an infected cell accumulates at least 5-6 times more Na + per cell than a neighboring noninfected one, which supports the assumption of the insufficient ability of mature infected cells to remove excess Na + and prevent its accumulation in the bacteroids.
In general, the effect of salinity stress in the central part of the root nodule would be similar to combined waterlogging and salinity stress modeled in A. thaliana roots [54,55]. Combined hypoxic conditions and salinity in A. thaliana usually lead to more damage to plant growth than salinity stress alone [54]. According to Wang et al. [54], the response to combined hypoxia and salt stress in A. thaliana is regulated by RBOHD (respiratory burst oxidase homolog protein D). In salt-treated plants, the onset of hypoxia leads to increased uptake of Na + and Cl − in RBOHD mutant plants, linking the regulation of salt stress response to reactive oxygen species (ROS), produced in the hypoxic environment, which causes the disturbances in K + homeostasis [54][55][56]. To confirm the model in salt-stressed nodules, further analyses of RBOHD are needed.
Although our study was focused on a fraction of the ion exchangers involved in Na + and K + transport in nodules our results reflect the changes that may be attributable to other proteins in the cells housing bacteria. The change in protein location in infected cells may possibly also affect the functionality of other proteins involved in the transport of necessary compounds belonging to other pathways. Such a situation should be considered a preexisting norm for these cells, taken under consideration in bioengineering works.
In conclusion, in M. truncatula nodules, the cells of the mature infected zone, containing the nitrogen-fixing bacteria, have a limited capability to withstand salt stress. The reasons for this are the inefficient mechanisms of Na + exclusion as a consequence of the mistargeting of the proteins involved in Na + exclusion in mature nitrogen-fixing cells and the partial downregulation of relevant genes in the mature part of the nodule, leading to the accumulation of Na + especially in the infected cell and in symbiosomes and to the premature termination of symbiosis.
M. truncatula seeds were sterilized and germinated. Seedlings were sown in vermiculite supplemented with Fahraeus medium and inoculated with S. meliloti 2011 3 d after sowing. Plants were grown under long-day conditions (16 h:8 h, day:night) at 20 • C.
Eight-week-old plants were subjected to salt stress by watering with a nutrient solution containing 100 mM NaCl for two weeks. For the analysis, we used nodules from the second generation, which was elicited on lateral roots around 4 weeks later than the first generation that is formed on the main root. These nodules did not display a visible senescent zone.

In Silico Gene Expression Analysis of S. meliloti and M. truncatula Ion Transporters, Channels, and Exchangers Putatively Involved in Salt Stress Response
The expression level of putative M. truncatula and S. meliloti exchangers and K + transporters and channels was estimated in the different nodule developmental zones using the Symbimics portal database (https://iant.toulouse.inra.fr/symbimics/, (accessed on 2 September 2022)) [33,34] obtained by laser-capture microdissection of the root nodule zones and infected and non-infected cells.
Arabidopsis thaliana genes were retrieved from the available genomic and cDNA sequence databases (https://www.uniprot.org, (accessed on 2 September 2022)) and used as query sequences. Selected protein sequences were used for the search of M. truncatula homologs in public bioinformatic resources (https://phytozome.jgi.doe.gov/pz/portal (accessed on 2 September 2022), https://www.ncbi.nlm.nih.gov/ (accessed on 2 September 2022)). Accession numbers or gene identifiers used in this study are given in Table S1. Recently, two new exchanger homolog genes found in the M. truncatula Annotation Release 102 (MtrunA17r5.0, February 2021) were referred to in the NCBI database, that is, "NHX7like" and "NHX8". In this study, we name them MtNHX7-like and MtNHX8, respectively, and we maintain MtNHX7 for the gene reported earlier [28]. To avoid misunderstanding, we follow the established nomenclature proposed earlier [28], but their equivalences in several databases are shown in Table S1 for consultation.

RNA Extraction and qPCR Gene Expression Analysis
Total RNA was isolated from control and treated nodules using the RNeasy Plant-Minikit (Qiagen, Hilden, Germany). The reverse transcription reaction was performed using the High-Capacity cDNA Reverse-Transcription Kit (Applied Biosystems, Foster City, CA, USA). Gene-specific primers were designed using the PRIMER 3-PLUS software [59] and the Primer Express v3.0 software (Applied Biosystems, Foster City, CA, USA). The primers used are listed in Table S2. The gene identifiers or locus names of the studied genes are shown in Table S1. Gene annotation is according to the NCBI Medicago truncatula Annotation Release 102 (February 2021).

Cloning
The MtNHX7 open reading frame and its 2.5-kb regulatory sequence were amplified via PCR from nodule cDNA and genomic DNA, respectively, using Phusion high-fidelity polymerase (Finnzymes). The coding sequence of MtNHX7 was directionally cloned into a modified pENTR vector (pENTR2) containing a multiple cloning site. Entry clones for MtNHX7 promoters were generated by TOPO cloning (Invitrogen, Waltham, MA, USA). The Gateway cloning system (Invitrogen) was used to create a genetic construct with the GFP fusion [59]. The pENTR clone of MtNHX7 was recombined into the pKGW-GGRR destination vector using LR Clonase (Invitrogen): The ProMtNHX7:MtNHX7:GFP translational fusion driven by the 2.5 kb native 5 regulatory sequences with its own promoter was generated by multiple cloning of gene and promoter sequences into the pKGW-MGW vector. Primers are listed in Table S1. Gene identifiers of MtNHX6 (Medtr2g028230) and MtNHX7 (Medtr2g038400), respectively. Primers are listed in the Table S6.

Confocal Microscopy
A Leica MZFLIII fluorescence microscope with filter cubes for GFP (excitation, 470/40; Dichroic: 495; emission, 525/50) and DsRed (excitation, 545/30; emission, 620/60) was used for the selection of transgenic tissue. Minimal 10 nodules per trial were fixed in 1% formaldehyde, cut along the longitudinal axis, and blocked with 2% BSA. The GFP signal in the sections carrying the ProMtNHX7:MtNHX7:GFP construct was enhanced by using an antibody (Ab) against GFP developed in rabbit (Molecular Probes) at 1:50 dilution. For immunolocalization of MtNHX6, a custom-made antibody was developed in rabbit by GenScript (Piscataway, NJ, USA) from the peptide sequence SENEISPADVHKAP (peptide-KLH conjugate) and used in a 1:100 dilution. The antibody recognized a protein of 58 kDa ( Figure S1). An anti-rabbit Alexa 488 (excitation max. 490, emission max. 525 nm; Molecular probes), diluted 1:200, was used as a secondary antibody. A mix of 0.5% skim-milk powder with 2% BSA was used as a blocking solution. Sections were counterstained with 0.01% propidium iodide (PI). Confocal imaging was done on hand-sectioned nodules with a Meta LSM 510 microscope (Leica) in a single image mode.

Ion Distribution Analysis (IDA)
IDA was performed by low-temperature scanning electron microscopy (LTSEM) and energy-dispersive X-ray microanalysis (EDXRMA) as described before [31]. Scanning electron cryomicroscopy (cryo-SEM) was performed with a Versa 3D DualBeam (FEI, US), equipped with an energy dispersive X-ray (EDX) spectroscopy system (EDAX, Mahwah, NJ, USA) and a Quorum Tech PP3010T cryogenic system (Quorum Tech, Laughton, East Sussex, UK). Prior to the study, the nodules were glued to the metal holders, then specimens were frozen in supercooled liquid nitrogen and transferred into the Quorum Tech PP3010T preparation chamber under vacuum, where nodules were fractured. Afterward, the specimens were transferred to the cryo-stage cooled by a continuous nitrogen flow at −140 • C. IDA analysis was performed in a spot acquisition mode using an accelerating voltage of 20 kV, 4 nA current, and a resolution of 133 eV. Quantitative element analysis was obtained using standard ZAF (atomic number, absorption, and fluorescence) correction procedures with TEAM 4.4 (EDAX, Mahwah, NJ, USA) and with Link Isis, version 3.2 (Oxford, UK) For each analysis, 10 nodules were used, 10-15 cells were analyzed per nodule, and 7-12 symbiosomes and vacuoles were examined.

Western Blot Analysis
Proteins were extracted from root tips in 25 mM Tris-HCl buffer containing 1 mM EDTA, 1 mM DTT, and protease inhibitors cocktail (Roche, Basel, Switzerland). Sixty µg of protein were loaded per well. Proteins were separated by 12% sodium dodecyl sulfatepolyacrylamide gel electrophoresis and were blotted onto a nitrocellulose membrane (Bio-Rad). The membrane was incubated in 3% BSA, as a blocking agent, followed by a primary anti-NHX6-specific antibody, 1:50 dilution, followed by the secondary antibody, an antirabbit immunoglobulin G peroxidase produced in goat (Sigma, Saint Louis, MI, USA), 1:5000 dilution. The immunosignal was visualized by incubation with the Immuno-Star Western chemiluminescent kit (Bio Rad, Hercules, CA, USA). The antibody recognized a protein of 58 kDa ( Figure S1). Blots were photographed with a Molecular Imager Chemi Doc XRS+, using Image Lab software in Chemi mode. The pre-stained protein ladder was photographed in normal light.
Phylogenetic analysis was inferred using the neighbor-joining method [63,64]. The evolutionary distances were obtained by the maximum composite likelihood method [65].
Node robustness was assessed by the bootstrap method (n = 1000 pseudo-replicates). Evolutionary analyses were conducted in MEGA X [61]. Phylogenetic Analysis of M. truncatula Homologs of the NHX Group.
Due to the important role of NHX exchangers in plant ion homeostasis, the various NHX genes recently annotated in the M. truncatula genome, the high expression of certain members of this group in the mature part of the nodule, and the upregulation of certain NHX genes found in salt-stressed nodules, we performed a phylogenetic analysis of these genes (especially focused on NHX6-7-8 genes) to establish homologies with A. thaliana AtNHXs and other functionally characterized homologs from other plant species (Oryza sativa, Phoenyx dactylifera, Triticum aestivum, Beta vulgaris, Populus trichocarpa). Moreover, phylogenetic relationships with NHX6-7-8 homologous genes from other legume species were also characterized.
The phylogenetic analysis showed that NHX homologs clustered into three mean clades ( Figure 6). Nodes in most clades displayed strong support, with bootstrap values of 72 or higher. The first main clade (where only A. thaliana and M. truncatula genes were included) contains NHX1-4 genes from A. thaliana and their corresponding homologous from M. truncatula. MtNHX1 and MtNHX2 genes were closely associated with AtNHX1 and AtNHX2, all of them included in the same clade. MtNHX3 and MtNHX4 are also closely associated with AtNHX3 and AtNHX4. This clade topology agrees with that obtained by [28].
In the second main clade, MtNHX6 clustered together with previously functionally characterized homologs from the included species A. thaliana (AtNHX5 and AtNHX6) [20,28], Phoenix dactylifera [66], Oryza sativa [67], Beta vulgaris [68], and Populus trichocarpa [69], suggesting similar functions for MtNHX6 as endosome-associated antiporters involved in ion homeostasis in vacuoles. MtNHX6 and all other legume genes clustered together in a clade, according to a classical legume phylogeny ( Figure 6). It is interesting to note that M. truncatula and the other Hologalegina legumes have one NHX6-like homolog as is also the case for Genistoids (Lupinus angustifolius). However, Mimosoid and Millettioid legumes have at least two NHX6 homologs. The cultivated peanut, the Dalbergioid, Arachis hypogaea, displays two homologs, probably due to its origin from the hybridization of A. duranensis and A. ipaensis, both of them containing only one NHX6 homolog (not shown). The third main clade included NHX7 and NHX8 homologs of M. truncatula, A. thaliana, and other legume and non-legume species ( Figure 6). We included the MtNHX7-like and MtNHX8 new genes reported in the NCBI Medicago truncatula Annotation Release 102 (Assembly Mtru-nA17r5.0-ANR). Both genes were closely associated with MtNHX7, AtNHX7, and AtNHX8 homologs. These genes clustered together with functionally characterized homologs from the included species Oryza sativa [70], Triticum aestivum [71], Beta vulgaris [68], Populus trichocarpa [69], and the legume Glycine max [72], suggesting the involvement of MtNHX7 in salt tolerance. M. truncatula genes and all other legume genes clustered together in a clade, according to a classical legume phylogeny. In this case, Hologalegina species display one, two, or three (for M. truncatula) homologs. The Genistoid L. angustifolius has two homologs. The Mimosoid Prosopis alba has four homologs, and most Millettioids display one homolog (except Cajanus cajan, containing two homologs) ( Figure 6). The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. For each gene, the gene name is indicated. Plant orders (for non-legume species) and legume clades are also indicated.

Statistical Methods
IDA analysis was performed twice in two independent experiments. Ten nodules from different plants per experiment were used. Minimal ten infected and ten non-infected cells per nodule were analyzed. Gene expression analysis was performed twice in two independent experiments. The Student's t-test was used to determine the significance of the differences between the means of data sets.