Motility, Adhesion and c-di-GMP Influence the Endophytic Colonization of Rice by Azoarcus sp. CIB

Proficient crop production is needed to ensure the feeding of a growing global population. The association of bacteria with plants plays an important role in the health state of the plants contributing to the increase of agricultural production. Endophytic bacteria are ubiquitous in most plant species providing, in most cases, plant promotion properties. However, the knowledge on the genetic determinants involved in the colonization of plants by endophytic bacteria is still poorly understood. In this work we have used a genetic approach based on the construction of fliM, pilX and eps knockout mutants to show that the motility mediated by a functional flagellum and the pili type IV, and the adhesion modulated by exopolysaccarides are required for the efficient colonization of rice roots by the endophyte Azoarcus sp. CIB. Moreover, we have demonstrated that expression of an exogenous diguanylate cyclase or phophodiesterase, which causes either an increase or decrease of the intracellular levels of the second messenger cyclic di-GMP (c-di-GMP), respectively, leads to a reduction of the ability of Azoarcus sp. CIB to colonize rice plants. Here we present results demonstrating the unprecedented role of the universal second messenger cyclic-di-GMP in plant colonization by an endophytic bacterium, Azoarcus sp. CIB. These studies pave the way to further strategies to modulate the interaction of endophytes with their target plant hosts.


Introduction
Plant associated bacteria play a fundamental role in the healthy state of the plant by contributing to its nutrition, development and defense [1][2][3]. A good number of the plant-associated bacteria colonize the plant tissues as endophytes being the roots the major target for endophytic colonization [2]. The endophytes are microorganisms that spend most of their life cycle inside plant tissues without causing symptoms of plant damage [4], but some endophytes are able to live outside of the plant without losing the capacity to colonize seedlings [5]. In addition, some endophytic bacteria exhibit beneficial effects on the host plant, such as plant growth promotion, the induction of increased resistance to pathogens, as well as the supply of fixed nitrogen to the host plant [6]. The effective root colonization by endophytic bacteria involves the coordinated expression of a number of gene factors, many of which are still poorly understood [7][8][9][10]. To analyze the genes involved in plant colonization several genomes from endophytic bacteria were compared [2,[11][12][13][14]. Moreover, to establish the determinants that differentiate endophytes from non-endophytic bacteria able to interact with plants, genomes from endophytic bacteria were compared with those from bacterial plant pathogens, bacteria that colonize the rhizosphere and bacteria from the rhizobia group [12]. Some properties such as motility, chemotaxis and adherence were considered crucial for initiating endophytic colonization. Bacterial motility plays an important role in plant colonization. Since endophytes are attracted to the root by plant exudates [2], most of them have flagella [14]. Chemotaxis towards the plant varies greatly between the different species of endophytes according to the rank of their preferred of industrial value [61,62]. All these features make strain CIB significantly different to the typical Azoarcus endophytic strains, and more closely related to members of the new genus Aromatoleum [63].
The aim of this work is to explore the role of the flagellum, pili type IV and exopolysaccarides on the colonization of rice roots by Azoarcus sp. CIB. Since the second messenger c-di-GMP is connected with motility and adhesion, in this work we investigated if the levels of c-di-GMP also control the ability of the bacterium to colonize plant, an aspect scarcely explored in endophytes.

Strains, Seeds and Plasmids Used
The bacterial strains and plasmids used in this work are listed in Table 1. Azoarcus sp. strain CIB was deposited in the Spanish Type Culture Collection (CECT #5669). Seeds from Oryza sativa L. GLEVA were kindly provided by Castells Seeds Co. (Tarragona, Spain). Km r , ori pBBR1, harbors the gfp gene under the control of the PlexA promoter [66] pIZ1016 Gm r , ori pBBR1MCS-5 derivative vector for cloning and expression harboring the Ptac promoter and the lacI gene [67] pIZ2133 Gm R , pIZ1016 derivative containing the gene PA2133 from P. aeruginosa PAO1 under the control of the Ptac promoter [68] pIZ4959 Gm R , pIZ1016 derivative containing the gene PP4959 from P. putida KT2440 under the control of the Ptac promoter [68] pK18mob Km R , oriColE1, Mob + , lacZα, suicide vector for the construction of insertion mutants by homologous recombination [69] pK18mobpilX

Bacterial Growth Conditions
Azoarcus strains were grown on MC medium (MA basal medium plus trace elements and vitamins). MA basal medium is detailed in Table 2 [59]. As carbon source, 0.2% (w/v) pyruvate was added. When needed, bacterial strains were also grown on a variant of the VM-ethanol rich medium [70] with the composition detailed in Table 2. E. coli strains were grown in lysogeny broth (LB) medium [71] at 37 • C. When required, kanamycin (Km) or gentamicin (Gm) was added at 50 µg mL −1 or 10 µg mL −1 , respectively. The growth of the cultures was monitored by measuring the absorbance at 600 nm (A 600 ) using a Shimadzu UV-260 spectrophotometer or by counting viable cells. Cell morphology was analyzed with a Nikon OPTIPHOT-2 phase contrast microscope.

Molecular Biology Techniques
Standard molecular biology techniques were performed as previously described [72]. DNA fragments were purified with Gene-Turbo (BIO101 Systems). Plasmids and PCR products were purified with a High Pure Plasmid and PCR Product Purifications kits (Roche), respectively. Oligonucleotides were supplied by Sigma Co and they are detailed in Supplementary Materials Table S2. All cloned inserts and DNA fragments were confirmed by DNA sequencing with fluorescently labeled dideoxynucleotide terminators [73] and AmpliTaq FS DNA polymerase (Applied Biosystems) in an ABI Prism 377 automated DNA sequencer (Applied Biosystems, Foster City, CA, USA). Transformations of E. coli were carried out by using the RbCl method or by electroporation (Gene Pulser, Bio-Rad, Hercules, CA, USA) [72]. Transformation of Azoarcus sp. CIB was done by biparental conjugation using the strain E. coli S17-1λpir as donor, following a protocol previously established [64] with slight modifications: donor cells were grown to A 600 of 5 and the receptor CIB strain was grown on MC medium supplemented with pyruvate 0.2%, and concentrated to reach an A 600 of 35. The transconjugants were selected on MC medium supplemented with 10 mM glutarate plus the corresponding antibiotic. For insertional disruption of pilX, fliM and epsF genes in the genome of Azoarcus sp. CIB we used single homologous recombination, that usually generates polar effects downstream of the the mutated gene, with PCR-amplified DNA fragments obtained with the primer pairs 5 pilX/3 pilX; 5 fliM/3 fliM and 5 epsF/3 epsF, respectively (Table S2). The obtained fragments were double-digested with the appropriate restriction enzymes, generating the pK18mobpilX, pK18mobfliM and pK18mobepsF recombinant plasmids (Table 1). These plasmids were transferred from E. coli S17-1λpir (donor strain) to Azoarcus sp. CIB (recipient strain) by biparental filter mating [72], and exconjugants strains Azoarcus sp. CIBdpilX, Azoarcus sp. CIBdfliM and Azoarcus sp. CIBdepsF were isolated on kanamycin-containing MC agar plates harboring 10 mM glutarate as the sole carbon source for counterselection of donor cells. The mutant strains were verified by PCR amplification using a standard oligonucleotide (F24 or R24) flanking the multiple cloning site of the plasmid and a second specific oligonucleotide external to the fragment used to perform the homologous recombination (Table S2). RT-PCRs were used to ensure that the mutated genes were not expressed in their corresponding host mutant strains ( Figure S1).

RNA Extraction and Quantitative Reverse Transcription-PCR (qRT-PCR) Experiments
RNA was purified from bacterial cells grown up to the middle of the exponential phase and resuspended in a solution containing TE buffer (Tris-HCl 10 mM pH 8.0, EDTA 1 mM) and lysozyme 50 mg mL −1 (Sigma). Total RNA was obtained using High Pure RNA Isolation Kit (Roche). The DNA was removed with DNAse and Removal Treatment Kit (Ambion). The concentration and purity of RNA was spectrophotometrically determined at A 260 and calculating the A 260 /A 280 ratio, respectively. The cDNA was obtained by using the Transcriptor First Strand cDNA Synthesis kit (Roche). Each RT reaction (20 µL) contained 1 µg RNA, 10 U reverse transcriptase, RNAse inhibitor 20 U, dNTPS 1 mM and 60 µM random hexamer primers. The standard procedure of cDNA production includes a 10 min incubation at 25 • C followed by 1 cycle of 30 min at 55 • C and another incubation of 5 min at 85 • C using the Mastercycler Gradient equipment (Eppendorf). Then, 1 µL of the obtained cDNA was used as template for the PCR. The analysis was performed in three technical replicates from three biological samples. Reactions (20 µL) contained 1 µL of cDNA, 0.25 µM of each of the three target-specific primers, and 10 µL of SYBR Green I Master Mix (Roche). Oligonucleotides 5 RTpilY1 and 3 RTpilY1 (Table S2) were used to amplify transcripts from the pilY1 gene. Oligonucleotides 5 RTfliC and 3 RTfliC (Table S2) were used to amplify transcripts from the fliC gene. Oligonucleotides 5 RTepsF and 3 RTepsF (Table S2) were used to amplify transcripts from the epsF gene. Oligonucleotides 5 POLIIIHK and 3 POLIIIHK (Table S2) were used to amplify transcripts from the dnaE gene (αDNApol III subunit gene) used as an internal control to normalize the sample data as reported before [74]. PCR amplifications were carried out with one denaturation cycle (95 • C for 5 min), followed by 30 cycles of amplification (95 • C for 10 s, 60 • C for 10 s, and 72 • C for 10 s). After amplification, 7 melting curves were generated to confirm amplification of a single product. For relative quantification of the fluorescence values, a calibration curve was constructed by seven-fold serial dilutions of Azoarcus sp. CIB genomic DNA sample ranging from 1 ng to 0.5 × 10 −7 ng. This curve was then used as a reference standard for extrapolating the relative abundance of the cDNA target within

Inoculation of Rice Seedlings with Bacteria
Dehulled rice seeds (O. sativa L. GLEVA) were surface sterilized by shaking them at 25 • C for 30 min in 30 mL 1% (v/v) sodium hypochlorite. After rinsing them three times for 10 min in sterile water, the seeds were incubated in VM-ethanol for 48 h. The germ-free seedlings were selected for inoculation. The germination of the seeds continued on humidified filter papers for 24 h prior to inoculation with bacterial cells. Bacteria were obtained as described previously [56]. After growth up to mid-exponential phase, bacterial cells were collected by centrifugation, washed with sterile 0.9% NaCl (w/v) solution, resuspended in 1 mL of sterile distilled water, and the cell suspension was inoculated onto the surface of each seedling in aseptic conditions. After inoculation, the seedlings were grown at 25 • C under natural daylight conditions (10 h of light and 14 h of darkness) for 5-10 days.

Recovery and Quantification of Endophytes
Rice seedlings were collected 5 days after inoculation with bacterial strains. The roots were weighed, proceeding to its surface disinfection by immersion for 3 min in a 1% (v/v) solution of sodium hypochlorite and three washes with sterile water. The roots were then homogenized by adding 1 mL of saline (0.85% NaCl) solution to a Potter-Elvehjem [76]. Once the extract was obtained, serial dilutions were made and plated in VM medium plates (1.5% agar) supplemented with kanamycin in order to determine the number of Colony Forming Units (CFUs). Each experiment was performed by triplicate (n = 3) and the total number of roots collected for each replica was 20.

Motility Tests
Swimming-type motility assays were performed on semi-solid VM-agar 0.3% medium plates using cultures grown in VM-ethanol liquid medium up to an A 600 of 0.6. For swarming motility assays, plates of MC medium with 0.2% pyruvate (w/v) and semi-solid 0.4% agar were used employing cultures grown in liquid MC medium with 0.2% pyruvate (w/v) up to an A 600 of 0.6. The plates were incubated for 7 days at 30 • C and the diameter of the growth halos of the different Azoarcus sp. CIB strains was measured.

Colony Morphology
The morphological study of the colonies of Azoarcus sp. CIB strains was performed on VM medium plates supplemented with 0.5 g L −1 of Congo Red dye (Sigma-Aldrich, San Louis, MO, USA). The bacteria were grown in VM-ethanol liquid medium up to an A 600 of 0.6. Subsequently, serial dilutions were made in saline solution (0.85% NaCl) and plated on VM-Congo Red plates. The plates were incubated for 7 days at 30 • C and the morphology of isolated colonies was analyzed with a Leica MZ16FA stereomicroscope.

Sequence Data Analyses
Nucleotide sequence analyses were done at the National Center for Biotechnology Information (NCBI) server (https://www.ncbi.nlm.nih.gov/, accessed on 7 March 2021).

Statistical Analysis
The data were analyzed with the GraphPad software package (San Diego, CA, USA) using paired or unpaired t-test. Statistical differences were presented as p < 0.01 (**) or p < 0.05 (*).

A Functional Flagellum and Pili Type IV Are Required for Rice Colonization of Azoarcus sp. CIB
As mentioned in the Introduction, bacterial motility plays an important role in plant colonization. The flagellum has traditionally been regarded as the main responsible for bacterial motility. Three gene clusters predicted to control the synthesis and regulation of the flagellum have been identified at three different positions of the genome of Azoarcus sp. CIB ( [58], Table S1). We then checked whether a mutation that abolish flagellum functionality may have an effect on the endophytic colonization of rice by strain CIB.
To try to generate a motility minus phenotype in Azoarcus sp. CIB, we constructed a mutant strain in the fli operon. The Azoarcus sp. CIBdfliM strain (Table 1) harbors a disruptional insertion of the fliM (AzCIB_0962) gene, encoding the FliM protein of the C ring which has been shown to be directly involved in flagellum movement and the flagellar export apparatus [77][78][79][80]. To check whether the fliM mutant strain showed an altered motility phenotype, we performed bacterial swimming assays. As shown in Figure 1A, the movement of the mutant cells was significantly diminished (around 50%) with respect to that of the parental strain, hence suggesting an important loss of functionality of the flagellum in Azoarcus sp. CIBdfliM strain.

Statistical Analysis
The data were analyzed with the GraphPad software package (San Diego, CA, USA) using paired or unpaired t-test. Statistical differences were presented as p < 0.01 (**) or p < 0.05 (*).

A Functional Flagellum and Pili Type IV Are Required for Rice Colonization of Azoarcus sp. CIB
As mentioned in the Introduction, bacterial motility plays an important role in plant colonization. The flagellum has traditionally been regarded as the main responsible for bacterial motility. Three gene clusters predicted to control the synthesis and regulation of the flagellum have been identified at three different positions of the genome of Azoarcus sp. CIB ( [58], Table S1). We then checked whether a mutation that abolish flagellum functionality may have an effect on the endophytic colonization of rice by strain CIB.
To try to generate a motility minus phenotype in Azoarcus sp. CIB, we constructed a mutant strain in the fli operon. The Azoarcus sp. CIBdfliM strain (Table 1) harbors a disruptional insertion of the fliM (AzCIB_0962) gene, encoding the FliM protein of the C ring which has been shown to be directly involved in flagellum movement and the flagellar export apparatus [77][78][79][80]. To check whether the fliM mutant strain showed an altered motility phenotype, we performed bacterial swimming assays. As shown in Figure 1A, the movement of the mutant cells was significantly diminished (around 50%) with respect to that of the parental strain, hence suggesting an important loss of functionality of the flagellum in Azoarcus sp. CIBdfliM strain.  We then checked the ability of the mutant strain to colonize rice seedlings. Interestingly, the number of endophytes recovered after inoculation of rice seedlings with Azoarcus sp. CIBdfliM strain was around an order of magnitude lower (3.4 × 10 3 per g of roots) than those recovered from seedlings inoculated with the wild type CIB strain (4.5 × 10 4 per g of roots) ( Figure 1B). These results indicate an important role of the flagellum at some stage of the endophytic colonization of rice by Azoarcus sp. CIB. These results are in agreement with previous reports showing that deletion of the fliC3 gene, encoding the main flagellin (FliC) of the flagellum, in the endophyte Azoarcus sp. BH72 strain leads to a significant decrease of around 80% in the plant colonizing capacity on competition assays between the wild type and the mutant strain [16], hence supporting that the flagellum is involved in the Azoarcus-plant interaction.
Different studies conducted on Azoarcus spp. strains have concluded that pili type IV are needed for plant colonization since they are responsible of adhesion and movement to the plant roots [7,16,18,81]. The genes predicted to be responsible for the synthesis of pili type IV in Azoarcus sp. CIB are organized in several gene clusters distributed along the genome ( [58], Table S1). In order to analyze if the pil genes from Azoarcus sp. CIB are involved in the colonization of rice root, we constructed an Azoarcus sp. CIBdpilX mutant strain containing a disruption of the pilX gene (AzCIB_3117) ( Table 1). The motility (swarming pattern) of the pilX mutant strain was reduced by 40% compared to that observed with the wild-type strain (Figure 2A), which is in agreement with the already observed lower motility in pil mutant bacterial strains [82,83] and confirms that the pil genes indeed encode proteins that participate in the motility apparatus in Azoarcus sp. CIB.
We then checked the ability of the mutant strain to colonize rice seedlings. Interestingly, the number of endophytes recovered after inoculation of rice seedlings with Azoarcus sp. CIBdfliM strain was around an order of magnitude lower (3.4 × 10 3 per g of roots) than those recovered from seedlings inoculated with the wild type CIB strain (4.5 × 10 4 per g of roots) ( Figure 1B). These results indicate an important role of the flagellum at some stage of the endophytic colonization of rice by Azoarcus sp. CIB. These results are in agreement with previous reports showing that deletion of the fliC3 gene, encoding the main flagellin (FliC) of the flagellum, in the endophyte Azoarcus sp. BH72 strain leads to a significant decrease of around 80% in the plant colonizing capacity on competition assays between the wild type and the mutant strain [16], hence supporting that the flagellum is involved in the Azoarcus-plant interaction.
Different studies conducted on Azoarcus spp. strains have concluded that pili type IV are needed for plant colonization since they are responsible of adhesion and movement to the plant roots [7,16,18,81]. The genes predicted to be responsible for the synthesis of pili type IV in Azoarcus sp. CIB are organized in several gene clusters distributed along the genome ( [58], Table S1). In order to analyze if the pil genes from Azoarcus sp. CIB are involved in the colonization of rice root, we constructed an Azoarcus sp. CIBdpilX mutant strain containing a disruption of the pilX gene (AzCIB_3117) ( Table 1). The motility (swarming pattern) of the pilX mutant strain was reduced by 40% compared to that observed with the wild-type strain (Figure 2A), which is in agreement with the already observed lower motility in pil mutant bacterial strains [82,83] and confirms that the pil genes indeed encode proteins that participate in the motility apparatus in Azoarcus sp. CIB. Interestingly, the plant colonization ability of the Azoarcus sp. CIBdpilX strain was more than one order of magnitude lower than that of the parental CIB strain ( Figure 2B). Therefore, these results allowed us to conclude that pili type IV are involved in the interaction of the CIB strain with rice. Our results in strain CIB are in agreement with those Interestingly, the plant colonization ability of the Azoarcus sp. CIBdpilX strain was more than one order of magnitude lower than that of the parental CIB strain ( Figure 2B). Therefore, these results allowed us to conclude that pili type IV are involved in the interaction of the CIB strain with rice. Our results in strain CIB are in agreement with those reported in Azoarcus sp. BH72 where a deletion of pilX did not inhibit the synthesis of the pili but affected its functionality [81][82][83][84][85] and decreased by 80% the plant colonizing capacity of the mutant strain [7,18].
In summary, all these results taken together indicate that both, flagellum and pili type IV are needed for the interaction of Azoarcus sp. CIB with rice.

Role of the Genes Involved in Exopolysaccharide Modification in the Interaction between Azoarcus sp. CIB and Rice
As mentioned in the Introduction, exopolysaccharides have an important role in plantbacteria interactions [23,25,26]. In the genome of Azoarcus sp. CIB we have identified three gene clusters putatively involved in the synthesis and modification of exopolysaccharides ( [56], Table S1). Previous results have shown that the gene cluster AzCIB_0813-0833 was strongly induced at the transcriptional level in the presence of rice extracts [57]. To analyze the potential role of this gene cluster in the colonization of rice roots by the CIB strain, it was inactivated by insertional disruption at the AzCIB_0818 (epsF) gene. The surface of the colonies of the Azoarcus sp. CIBdepsF mutant strain grown on a solid culture medium supplemented with Congo Red dye [86] showed a rugose phenotype in contrast to the smooth phenotype shown by colonies of the wild-type strain ( Figure 3A). These changes in colony morphology strongly suggest the existence of significant alterations in the exopolysaccharide of the epsF mutant strain [87], hence indicating that the eps genes are indeed involved in exopolysaccharide synthesis/modification. The next step was to study the ability of the strain Azoarcus sp. CIBdepsF to colonize rice roots as an endophyte. The plant colonization results showed that the number of Azoarcus sp. CIBdepsF endophytes recovered from rice roots (1.65 × 10 4 bacteria per g of roots) was 36% lower than that obtained from roots inoculated with the parental strain Azoarcus sp. CIB (4.57 × 10 4 bacteria per g of roots) ( Figure 3B). Therefore, the results obtained reveal that the inactivation of the eps cluster reduces the ability of Azoarcus sp. CIB to interact with the rice plant. These results provide strong experimental support on the role of the exopolysaccharide(s) promoting the endophytic lifestyle in members of the Azoarcus/Aromatoleum group.
In summary, all these results taken together indicate that both, flagellum and pili type IV are needed for the interaction of Azoarcus sp. CIB with rice.

Role of the Genes Involved in Exopolysaccharide Modification in the Interaction between Azoarcus sp. CIB and Rice
As mentioned in the Introduction, exopolysaccharides have an important role in plant-bacteria interactions [23,25,26]. In the genome of Azoarcus sp. CIB we have identified three gene clusters putatively involved in the synthesis and modification of exopolysaccharides ( [56], Table S1). Previous results have shown that the gene cluster AzCIB_0813-0833 was strongly induced at the transcriptional level in the presence of rice extracts [57]. To analyze the potential role of this gene cluster in the colonization of rice roots by the CIB strain, it was inactivated by insertional disruption at the AzCIB_0818 (epsF) gene. The surface of the colonies of the Azoarcus sp. CIBdepsF mutant strain grown on a solid culture medium supplemented with Congo Red dye [86] showed a rugose phenotype in contrast to the smooth phenotype shown by colonies of the wild-type strain ( Figure 3A). These changes in colony morphology strongly suggest the existence of significant alterations in the exopolysaccharide of the epsF mutant strain [87], hence indicating that the eps genes are indeed involved in exopolysaccharide synthesis/modification. The next step was to study the ability of the strain Azoarcus sp. CIBdepsF to colonize rice roots as an endophyte. The plant colonization results showed that the number of Azoarcus sp. CIBdepsF endophytes recovered from rice roots (1.65 × 10 4 bacteria per g of roots) was 36% lower than that obtained from roots inoculated with the parental strain Azoarcus sp. CIB (4.57 × 10 4 bacteria per g of roots) ( Figure 3B). Therefore, the results obtained reveal that the inactivation of the eps cluster reduces the ability of Azoarcus sp. CIB to interact with the rice plant. These results provide strong experimental support on the role of the exopolysaccharide(s) promoting the endophytic lifestyle in members of the Azoarcus/Aromatoleum group.

c-di-GMP Levels Influence the Endophytic Lifestyle of Azoarcus sp. CIB
As mentioned in the Introduction, c-di-GMP is a second messenger involved in the regulation of many bacterial functions. However, its role in beneficial plant-bacteria interactions and, more specifically, in their endophytic relationships, is poorly studied and far to be understood. Since the production and/or functionality of pili type IV, flagellum and exopolysaccharide (all of which were shown above to be involved in the Azoarcus sp. CIB-rice interactions) have been related with the intracellular levels of c-di-GMP in bacteria [88][89][90], it was tempting to speculate that c-di-GMP could be also regulating the endophytic lifestyle of strain CIB. To study further the proposed new function of c-di-GMP in bacterial endophytes, we checked recombinant Azoarcus sp. CIB strains containing different intracellular levels of c-di-GMP. To this end, we used Azoarcus sp. CIB (pIZ2133), expressing the PA2133 c-di-GMP phosphodiesterase from P. aeruginosa PAO1 [26], and Azoarcus sp. CIB (pIZ4959), expressing the PP4959 diguanilate cyclase from P. putida KT2440 [91], two strains described for their reduced or increased levels of c-di-GMP, respectively, with respect to those of the parental strain Azoarcus sp. CIB (pIZ1016) ( [68], Table 1). To confirm that the two recombinant strains showed the phenotypes anticipated according to the variations in c-di-GMP levels [28], we analyzed first their swarming and swimming motility. As expected, whereas Azoarcus sp. CIB (pIZ2133) (reduced c-di-GMP levels) showed an increased swarming and swimming motility, the strain Azoarcus sp. CIB (pIZ4959) (increased c-di-GMP levels) showed a reduced swarming and swimming motility with respect to that of the control strain ( Figure 4). The growth rates of the CIB control strain and its derivative strains CIB (pIZ2133) and CIB (pIZ4959) were similar (data not shown), so the observed different motility cannot be attributed to differences in their cell growth. The morphology of the colonies was later analyzed and whereas Azoarcus sp. CIB (pIZ2133) generated smoother and more plain colonies than the control strain, Azoarcus sp. CIB (pIZ4959) showed rough colonies (Figure 4), which is consistent with a higher exopolysaccharide synthesis [28].

c-di-GMP Levels Influence the Endophytic Lifestyle of Azoarcus sp. CIB
As mentioned in the Introduction, c-di-GMP is a second messenger involved in the regulation of many bacterial functions. However, its role in beneficial plant-bacteria interactions and, more specifically, in their endophytic relationships, is poorly studied and far to be understood. Since the production and/or functionality of pili type IV, flagellum and exopolysaccharide (all of which were shown above to be involved in the Azoarcus sp. CIB-rice interactions) have been related with the intracellular levels of c-di-GMP in bacteria [88][89][90], it was tempting to speculate that c-di-GMP could be also regulating the endophytic lifestyle of strain CIB. To study further the proposed new function of c-di-GMP in bacterial endophytes, we checked recombinant Azoarcus sp. CIB strains containing different intracellular levels of c-di-GMP. To this end, we used Azoarcus sp. CIB (pIZ2133), expressing the PA2133 c-di-GMP phosphodiesterase from P. aeruginosa PAO1 [26], and Azoarcus sp. CIB (pIZ4959), expressing the PP4959 diguanilate cyclase from P. putida KT2440 [91], two strains described for their reduced or increased levels of c-di-GMP, respectively, with respect to those of the parental strain Azoarcus sp. CIB (pIZ1016) ( [68], Table 1). To confirm that the two recombinant strains showed the phenotypes anticipated according to the variations in c-di-GMP levels [28], we analyzed first their swarming and swimming motility. As expected, whereas Azoarcus sp. CIB (pIZ2133) (reduced c-di-GMP levels) showed an increased swarming and swimming motility, the strain Azoarcus sp. CIB (pIZ4959) (increased c-di-GMP levels) showed a reduced swarming and swimming motility with respect to that of the control strain ( Figure 4). The growth rates of the CIB control strain and its derivative strains CIB (pIZ2133) and CIB (pIZ4959) were similar (data not shown), so the observed different motility cannot be attributed to differences in their cell growth. The morphology of the colonies was later analyzed and whereas Azoarcus sp. CIB (pIZ2133) generated smoother and more plain colonies than the control strain, Azoarcus sp. CIB (pIZ4959) showed rough colonies (Figure 4), which is consistent with a higher exopolysaccharide synthesis [28].  To analyze whether the altered levels of c-di-GMP in Azoarcus sp. CIB (pIZ2133) and Azoarcus sp. CIB (pIZ4959) strains might influence their plant colonization abilities, rice seedlings colonization tests were performed. As shown in Figure 5, the number of endophytes recovered from seedlings inoculated with Azoarcus sp. CIB (pIZ2133) cells (1.46 × 10 3 bacteria per g of root), was more than one order of magnitude lower than the number of endophytes recovered after inoculation with the CIB control strain (2.5 × 10 4 bacteria per g of root). On the other hand, inoculation with the strain Azoarcus sp. CIB (pIZ4959) led to 7.63 × 10 3 endophytes per g of root, which represents a loss of efficiency in colonization of 70% with respect to the control strain. Therefore, these data suggest that either the artificial decrease or increase of c-di-GMP levels in Azoarcus sp. CIB affect negatively the endophytic colonization of rice by this bacterium. These results are in agreement with previous observations that any modification in the homeostasis of c-di-GMP negatively affects the interactions between the plant and pathogenic or symbiont bacteria [41][42][43]. These results can be explained by taken into account that both motility and adhesion are needed for successful plant colonization [2]. Thus, an increase in c-di-GMP levels leads to greater adhesion but, at the same time, to a decrease in bacterial motility; conversely, a decrease of c-di-GMP levels enhance bacterial motility but are detrimental for adhesion and formation of microcolonies needed for the entrance of the bacteria to the internal tissues of the plant [2,92].
Azoarcus sp. CIB (pIZ4959) strains might influence their plant colonization abilities, rice seedlings colonization tests were performed. As shown in Figure 5, the number of endophytes recovered from seedlings inoculated with Azoarcus sp. CIB (pIZ2133) cells (1.46 × 10 3 bacteria per g of root), was more than one order of magnitude lower than the number of endophytes recovered after inoculation with the CIB control strain (2.5 × 10 4 bacteria per g of root). On the other hand, inoculation with the strain Azoarcus sp. CIB (pIZ4959) led to 7.63 × 10 3 endophytes per g of root, which represents a loss of efficiency in colonization of 70% with respect to the control strain. Therefore, these data suggest that either the artificial decrease or increase of c-di-GMP levels in Azoarcus sp. CIB affect negatively the endophytic colonization of rice by this bacterium. These results are in agreement with previous observations that any modification in the homeostasis of c-di-GMP negatively affects the interactions between the plant and pathogenic or symbiont bacteria [41][42][43]. These results can be explained by taken into account that both motility and adhesion are needed for successful plant colonization [2]. Thus, an increase in c-di-GMP levels leads to greater adhesion but, at the same time, to a decrease in bacterial motility; conversely, a decrease of c-di-GMP levels enhance bacterial motility but are detrimental for adhesion and formation of microcolonies needed for the entrance of the bacteria to the internal tissues of the plant [2,92]. To try to determine if the presumed modification of c-di-GMP levels in Azoarcus sp. CIB affected the transcription of genes involved in bacterial motility and/or exopolysaccharide formation, an expression analysis of the pilY1 gene (encodes a protein for the synthesis and stabilization of pili type IV), fliC gene (encodes a structural protein of the flagellum) and the epsF gene (encodes a protein for exopolysaccharide modification) was performed. The results obtained revealed that the expression of the epsF gene was not To try to determine if the presumed modification of c-di-GMP levels in Azoarcus sp. CIB affected the transcription of genes involved in bacterial motility and/or exopolysaccharide formation, an expression analysis of the pilY1 gene (encodes a protein for the synthesis and stabilization of pili type IV), fliC gene (encodes a structural protein of the flagellum) and the epsF gene (encodes a protein for exopolysaccharide modification) was performed. The results obtained revealed that the expression of the epsF gene was not significantly altered in any of the two strains that have modified levels of c-di-GMP, i.e., Azoarcus sp. CIB (pIZ2133) and Azoarcus sp. CIB (pIZ4959) (Figure 6). These results suggest that c-di-GMP does not control the exopolysaccharide genes at the transcriptional level, pointing to a post-transcriptional control, as already described for exopolysaccharide genes in other bacteria [93], as a more likely regulatory mechanism. Similarly, the expression of the pilY1 and fliC genes was not significantly altered in Azoarcus sp. CIB (pIZ2133) (reduced c-di-GMP levels) when compared with their expression in the control strain ( Figure 6), suggesting that the increased motility observed in the former strain might be controlled by a c-di-GMP dependent post-transcriptional regulation as already shown in P. aeruginosa PAO1 [94]. In contrast, the expression of pilY1 and fliC genes was 3-3.5 times lower in the strain Azoarcus sp. CIB (pIZ4959) than in the control strain ( Figure 6), suggesting that the increase of c-di-GMP levels leads to a transcriptional repression of such genes. The repression of the pil and fli genes in response to elevated cellular levels of c-di-GMP was previously observed in other bacteria such as V. cholerae [95,96], and is in agreement with the reduced motility observed in swarming and swimming assays (Figure 4). It is known that the c-di-GMP mediated transcriptional regulation of motility genes takes place through the FleQ regulator in P. aeruginosa and P. putida [37,97,98], or through alterations in chemotactic machinery affecting receptors with PilZ domains [99,100]. Since Azoarcus sp. CIB does not have a FleQ ortholog and only a protein with a putative PilZ domain has been identified by genome search, it is currently difficult to predict the signaling cascade involved in the transcriptional control mediated by c-di-GMP on the pil and fli genes in the CIB strain.
level, pointing to a post-transcriptional control, as already described for exopolysaccharide genes in other bacteria [93], as a more likely regulatory mechanism. Similarly, the expression of the pilY1 and fliC genes was not significantly altered in Azoarcus sp. CIB (pIZ2133) (reduced c-di-GMP levels) when compared with their expression in the control strain ( Figure 6), suggesting that the increased motility observed in the former strain might be controlled by a c-di-GMP dependent post-transcriptional regulation as already shown in P. aeruginosa PAO1 [94]. In contrast, the expression of pilY1 and fliC genes was 3-3.5 times lower in the strain Azoarcus sp. CIB (pIZ4959) than in the control strain ( Figure 6), suggesting that the increase of c-di-GMP levels leads to a transcriptional repression of such genes. The repression of the pil and fli genes in response to elevated cellular levels of c-di-GMP was previously observed in other bacteria such as V. cholerae [95,96], and is in agreement with the reduced motility observed in swarming and swimming assays (Figure 4). It is known that the c-di-GMP mediated transcriptional regulation of motility genes takes place through the FleQ regulator in P. aeruginosa and P. putida [37,97,98], or through alterations in chemotactic machinery affecting receptors with PilZ domains [99,100]. Since Azoarcus sp. CIB does not have a FleQ ortholog and only a protein with a putative PilZ domain has been identified by genome search, it is currently difficult to predict the signaling cascade involved in the transcriptional control mediated by c-di-GMP on the pil and fli genes in the CIB strain. In summary, all the results presented here suggested for the first time how the modification of intracellular levels of c-di-GMP in an endophytic bacterium affects its ability to colonize the host plant, and they reveal an unnoticed role of c-di-GMP in bacteria controlling the endophytic lifestyle. It is tempting to speculate that, as described in Figure 6. Expression of the pilY1, fliC and epsF genes in Azoarcus sp. CIB strains producing different levels of c-di-GMP. The transcript levels of the pilY1, fliC and epsF genes in the Azoarcus sp. CIB (pIZ2133) (red columns) and Azoarcus sp. CIB (pIZ4959) (yellow columns) strains grown in VMethanol medium for 24 h, were determined by quantitative RT-PCR using dnaE gene as internal control, as detailed in Methods. The expression of the genes in each strain is shown relative to their expression in the control strain Azoarcus sp. CIB (pIZ1016). The mean value and the standard deviation corresponding to the results of three independent experiments are indicated. Values of expression of pilY gene (p = 0.0174) in Azoarcus sp. CIB (pIZ2133) and Azoarcus sp. CIB (pIZ4959), were significantly different (p = 0.0174; one asterisk). Values of expression of fliC gene in Azoarcus sp. CIB (pIZ2133) and Azoarcus sp. CIB (pIZ4959), were very significantly different (p = 0.0095; two asterisks). Values of expression of epsF gene in Azoarcus sp. CIB (pIZ2133) and Azoarcus sp. CIB (pIZ4959) were not significantly different (p = 0.0516). All the statistic data were obtained according to unpaired t-test.
In summary, all the results presented here suggested for the first time how the modification of intracellular levels of c-di-GMP in an endophytic bacterium affects its ability to colonize the host plant, and they reveal an unnoticed role of c-di-GMP in bacteria controlling the endophytic lifestyle. It is tempting to speculate that, as described in the endophyte bacterium Azoarcus sp. CIB, the modification of c-di-GMP levels may modulate pathogenesis of plant pathogenic bacteria. In this sense, there have been described some examples where changes in the levels of c-di-GMP control pathogenesis, e.g., modulating flagellar swimming motility and plant disease progression in Erwinia amylovora [101] or Pseudomonas syringae pv. tomato DC3000 [102]. However, these findings could not be generalized to all pathogenic bacteria. Thus, increasing c-di-GMP levels in P. savastanoi pv. savastanoi might reduce the necrosis observed on tomato or bean but the development of other disease symptoms did not seem significantly affected by high c-di-GMP [45]. More work needs to be done to elucidate the signaling cascade that responds to the levels of c-di-GMP in Azoarcus sp. CIB.

Conclusions
The results of the present study describe the participation of the motility mediated by the flagellum and pili type IV, the adhesion modulated by exopolysaccharides and, remarkably, the unnoticed participation of the intracellular levels of c-di-GMP in the colonization of rice roots by Azoarcus sp. CIB. More experiments need to be developed to understand the complete landscape of molecular determinants involved in the signaling cascade that responds to the levels of c-di-GMP in Azoarcus sp. The results of these experiments might be used to set strategies to modulate the interaction of endophytes with their target plant hosts.

Data Availability Statement:
The data presented in this study are available within this article.