SEARCH

SEARCH BY CITATION

Keywords:

  • fluorescent tagging;
  • confocal laser scanning microscopy;
  • infection thread;
  • Rhizobium legume specificity;
  • root nodule endophytes;
  • Vigna radiata

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Legumes develop symbiotic relationships with Rhizobium by a complex exchange of signals. Despite the high specificity between symbiotic partners, the presence of non-rhizobial bacteria in root nodules has been reported. To investigate how these rhizobacteria enter root nodules, fluorescently tagged Pseudomonas fluorescens and Klebsiella pneumoniae were co-inoculated with host-nodulating Ensifer adhaerens to Vigna radiata seedlings and root hair infection was monitored using confocal microscopy at 5 days post inoculation. Pseudomonas fluorescens and K. pneumoniae invaded the root hair only when co-inoculated with E. adhaerens. Recovery of inoculated tagged strains and confirmation through CLSM and 16S rRNA gene sequencing confirmed that the test rhizobacteria occupied nodules. We hereby report with the help of confocal microscopy that rhizobacteria migrate along the length of host-nodulating rhizobial strain and become localized in root nodules. We further report isolation of eight non-rhizobial bacterial genera, predominantly Bacillus spp. and Paenibacillus spp., from nodules of field-grown V. radiata.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Rhizobia encompass a range of bacterial genera including Rhizobium, Bradyrhizobium, Mesorhizobium, Allorhizobium and Azorhizobium that establish symbiosis with legumes for nitrogen fixation. In most legumes (e.g. Vigna radiata), a highly specific signaling process allows rhizobia to enter the roots through root hair curling, formation and progression of infection thread (IT), leading to nodule development (Gage, 2004).

The signaling events between symbiotic partners are highly explicit, such that bacterial species can only nodulate a limited number of leguminous plants (Wang et al., 2012). The lipopolysaccharides (Nod factor) secreted by rhizobial cells initiate the infection process at the binding site on the roots of leguminous plants. The Nod factors carry substitutions unique to each strain that determine the specificity of the rhizobial interaction (Broughton et al., 2000; Perret et al., 2000). Besides the specificity mediated by the Nod factors, the microoxic environment of nodules that ensures the activity of oxygen-sensitive nitrogenase may restrict the growth of most aerobic rhizobacteria (Dixon & Kahn, 2004). Despite of high specificity of the legume–Rhizobium interaction and the selective nodule environment, non-rhizobial nodule rhizobacteria have been reported. Root nodules which traditionally were considered the exclusive niche of rhizobia are being revisited to examine colonization by several free-loaders unrelated to symbiotic nitrogen fixation. Evidence that the healthy nodule interior can contain endophytes not necessarily related to symbiotic or diazotrophic context has been documented, i.e. Bacillus (soybean, Bai et al., 2002), Agrobacterium (tropical grass, de Lajudie et al., 1999), Klebsiella (groundnut, clover, bean, Ozawa et al., 2003) and Pseudomonas (acacia, soybean, Kuklinsky-Sobral et al., 2004; Hoque et al., 2011). Sturz et al. (1997) simultaneously recovered 4.3 × 109 CFU rhizobia and 3 × 105 CFU non-rhizobial endophytes of 12 bacterial genera per gram fresh weight of red clover nodule tissue. Evidence that healthy nodule interiors of wild legumes can contain bacteria not related to rhizobia has also been reported (Zakhia et al., 2006; Deng et al., 2011; Li et al., 2012).

The key question of how these rhizobacteria breach host specificity and enter root nodules remains unanswered. The main questions addressed in this study are whether this ability to breach the specificity barrier is dependent on the population of the rhizobacteria in the rhizosphere or whether rhizobacteria are selected that are capable of withstanding the microoxic conditions prevailing in the nodule.

Considering the remarkable specificity of rhizobial strains for its symbiotic partner, we hypothesized that host-specific rhizobia initiate the signaling process to form infection thread (IT), which is invaded by rhizobacteria to breach host specificity. The support of advanced fluorescence microscopic techniques to examine nodule symbiosis has gained new importance in studying plant–microbe interactions (Chen et al., 2005; Elliott et al., 2007) and we have coupled this visual approach with the molecular tagging of rhizobacteria.

In the present work, V. radiata seedlings were co-inoculated with a fluorescently tagged host-nodulating rhizobial strain (Ensifer adhaerens) and test rhizobacteria (Pseudomonas fluorescens IAM 12022 and Klebsiella pneumoniae ssp. ozaenae) and colonization of root hair IT was monitored using confocal laser scanning microscopy (CLSM). The tagged rhizobacterial strains were recovered from root nodules of inoculated host plants in pure culture and identified to confirm their presence. The isolation and identification of native root nodule endophytes to support the presence of non-rhizobial rhizobacteria in root nodules of the field grown V. radiata is described.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Isolation of V. radiata-nodulating rhizobial strain

Root nodules from field-grown V. radiata were sampled for isolation of native nodulating rhizobial strain. The nodules were gently washed using commercial detergent under running tap water to remove adhering soil particles. The nodules were then surface-sterilized with 0.1% mercury chloride (15 min) and 70% alcohol (20 min), followed by several washes with sterile distilled water. Nodules were crushed in sterile saline and the suspension plated on Congo red yeast extract mannitol agar (CRYMA; Himedia, India). The white, gelatinous, shiny and dome-shaped colony was subjected to biochemical tests, and the ability to induce root hair curling (Fahraeus, 1957) and nodulation was examined by 16S rRNA gene sequencing with universal primers and amplification of the nodulation gene (nodC F:5′ACGTCGTTGACGACGGTTC 3′; nodC R: 5′ CGAGACAGCCACTCCCTATTG 3′).

Rhizobacterial strains

The rhizobacterial strains selected for the study, P. fluorescens IAM 12022 and K. pneumoniae ssp. ozaenae, were kind gifts from Prof. G. Nareshkumar, Dept. of Biochemistry, M.S. University of Baroda, Gujarat, India, and Dr. Karsten Struve, Statens Serum Institute, Copenhagen, Denmark. The gifted strains were confirmed by 16S rRNA gene sequencing using universal primers prior to transformation. The presence of nifH in rhizobacterial strains was determined by amplification of genomic DNA using nifH specific primers (nifH F: 5′ TAC GGA AAA GGC GGA ATC GGC AA 3′; nifH R: 5′ AGC ATG TCC TCG AGC TCA TCC A 3′).

Fluorescent tagging and inoculation of rhizobacterial strains

The plasmids pDG77 (ptrp-DsRed), which constitutively expressed DsRed (Gage, 2004), and pUCPM18 (placZ-gfp; Buch et al., 2009), were transformed in E. adhaerens and P. fluorescens IAM 12022, respectively, using the calcium chloride method (Sambrook & Russell, 2001). The transformants were subcultured for several generations and confirmed for plasmid stability. Klebsiella pneumoniae ssp. ozaenae carrying cfp integrated in the chromosome (Schroll et al., 2010) was used as received. Five-day-old V. radiata seedlings were placed on microscope slides and covered with 0.4% water agar (Fahraeus, 1957). The seedlings were inoculated with 2% solution (108 × CFU mL−1) of E. adhaerens (positive control), P. fluorescens IAM 12022 (negative control), K. pneumoniae ssp. ozaenae (negative control), co-inoculated with a mixture of E. adhaerens, P. fluorescens IAM 12022 and K. pneumoniae ssp. ozaenae (test). The slides were placed in tubes with sterile micronutrient solution (CaHPO4 1 g L−1; K2HPO4 0.2 g L−1; MgSO4.7H2O 0.2 g L−1; NaCl 0.2 g L−1; FeCl3 0.1 g L−1) and incubated at 28 ± 2 °C in light (light intensity of 200 μE m−2 s−1, 16-h day/8-h night cycle, at a constant temperature of 28 °C and a relative humidity of 50%).

Root hair infection and nodule occupancy

Rhizobacterial infection in root hairs of seedlings was observed 5 days post inoculation (DPI) under a CLSM equipped with a long-distance ×10 or ×20 water immersion objective (Leica, Germany). Laser bands of 557, 395 and 458 nm were used for excitation. Specific emission wavelengths of 585, 509 and 480 nm captured signals as red (Rfp), green (Gfp) and cyan (Cfp), respectively.

To monitor nodule occupancy by fluorescently tagged rhizobacteria, surface-sterilized V. radiata seeds imbibed with test strains as described were sown in plastic pots filled with sterile soil. The plants were grown under a controlled environment (light intensity of 200 μE m−2 s−1, 16-h day/8-h night cycle, at a constant temperature of 28 °C and a relative humidity of 50%) and watered regularly with sterile water (Lorteau et al., 2001). At 50 DPI, the sap of crushed, surface-sterilized nodules was plated on selective media and the recovery of tagged rhizobacteria was confirmed by CLSM and 16S rRNA gene sequencing.

Isolation and identification of native nodule endophytes of V. radiata

Thirty undamaged, healthy root nodules of similar size sampled from lateral roots were surface-sterilized. The nodules were crushed in saline and aliquots of suspensions were plated on CRYMA, PCA (plate count agar, Himedia), NA (Nutrient agar, Himedia) and incubated at 28 ± 2 °C for 24–48 h. The validity of surface sterilization was confirmed by plating aliquots of water from the final rinse, pressing and rolling nodules over CRYMA, PCA and NA plates followed by incubation at 28 ± 2 °C. As negative control, a non-sterile nodule was also rolled over a similar set of media plates.

The genomic DNA of nodule endophytes was isolated using a Qiagen bacterial genomic DNA isolation kit and amplified using the universal bacterial primers: forward primer 27F Bacteria (5′ AGA GTT TGA TC (A/C) TGG CTC AG 3′) and reverse primer R1492 (5′ TAC GG(C/T) TAC CTT GTT ACG ACT T 3′). The PCR reaction was performed in 50 μL of the total reaction volume, with 3 U of Taq DNA polymerase, under the following conditions: initial denaturation at 94 °C for 3 min followed by 30 cycles at 94 °C for 30 s, annealing at 55 °C for 30 s and 72 °C for 1 min and 20 s, and final extension at 72 °C for 10 min (Heuer et al., 1997). The PCR products were purified using the QIAquick PCR purification kit (Qiagen) and sequenced using automatic ABI 310 DNA sequencer (Big Dye Terminator cycle sequencing, ready reaction kit, Perkin-Elmer). Analysis of sequences was carried out with the basic sequence alignment blast program run against the database from National Center for Biotechnology Information Blast (www.ncbi.nlm.nih.gov/BLAST). The 16S rRNA gene sequences determined were manually aligned with the published sequences of validated species available from the EMBL/GenBank/DDBJ databases. The phylogenetic relationship was inferred using the neighbor-joining method (Saitou & Nei, 1987). The topography of the constructed tree was evaluated by bootstrap analysis with 1000 replicates (Felsenstein, 1985). The evolutionary analyses, including construction of a phylogenetic tree, were conducted in mega5 using 26 nucleotide sequences (Tamura et al., 2011).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Isolation of V. radiata-nodulating rhizobial strain

The fast growing bacterial isolate from nodule was gram-negative, formed white gelatinous, shiny, dome-shaped colonies on CRYEMA due to exopolysaccharide production and produced acid on bromothymol blue YEMA as reported (Vincent, 1970). Upon reinoculation, the isolated culture induced root hair curling in mung bean seedlings within 5–6 h of inoculation, under a light microscope (Nikon, Japan), which confirmed primary recognition between the symbiotic partners and nodulated V. radiata at 50 DPI (data not shown). blast analysis of the 16S rRNA gene sequence of the rhizobial strain showed 99% similarity with Ensifer adhaerens (GenBank Accession number JX508708) and hence this culture was used as V. radiata-nodulating rhizobia in this study.

Rhizobacterial strains

The strains of P. fluorescens IAM 12022 (GenBank Accession number JX657840) and K. pneumoniae ssp. ozaenae (GenBank Accession number JX657841) were confirmed by 16S rRNA gene sequencing. PCR amplification confirmed the presence of nifH genes in E. adhaerens and K. pneumoniae ssp. ozaenae; these genes were absent in P. fluorescens. nodC amplification and nodulation of V. radiata confirmed the presence of functional nod genes in E. adhaerens (data not shown).

Root hair infection and nodule occupancy by fluorescently tagged rhizobacteria

Progression of the infection within root hairs of seedlings was observed under CLSM at 5 DPI. Red ITs in the positive control (Fig. 1a), no ITs in the negative controls (Fig. 1b and c) and mixed ITs (red, green and cyan) were visible along the length of co-inoculated root hairs (Fig. 1d–f). The images in Fig. 1d–f demonstrate the presence of all the inoculated rhizobacteria within the same root hair when excited with different wavelengths specific for fluorescent proteins employed for tagging test organisms.

image

Figure 1. Inoculation of Ensifer adhaerens-Rfp, Pseudomonas fluorescens IAM 12022-Gfp and Klebsiella pneumoniae ssp. ozaenae-Cfp. Growing ITs expressing Cfp, Rfp and Gfp confirm the presence of inoculated strains within the root hairs of Vigna radiata at 5 DPI. Confocal micrographs: (a) E. adhaerens (positive control), presence of red IT; (b) P. fluorescens; and (c) K. pneumoniae (negative controls), no ITs. Cells adhere only to surface and base of the root hair. (a), the arrow shows the location of infection thread inside root hair. (d) Ensifer adhaerens, (e) P. fluorescens and (f) K. pneumoniae (co-inoculation) recently initiated mixed ITs (red, green and cyan, respectively) within the same root hair. In (d–f) arrows show the location of infection thread inside root hair. Note: Seedlings were gently washed with sterile Mili-Q water to remove unadhered cells to minimize background fluorescence (noise).

Download figure to PowerPoint

In this study we demonstrate in vivo rhizobacterial (non-rhizobial) infection within the root hairs of V. radiata where the presence of native E. adhaerens allowed Pseudomonas and Klebsiella spp. to enter root hairs. The ability of native E. adhaerens to infect root hair remained unaltered irrespective of the presence of other rhizobacteria (Fig. 1a and d). Pseudomonas fluorescens IAM 12022 and K. pneumoniae ssp. ozaenae adhered to the surface and base of root hairs but failed to enter IT in the absence of E. adhaerens (Fig. 1b and c), which confirmed their inability to form an infection thread and invade root hairs independently. However, when co-inoculated with E. adhaerens, P. fluorescens IAM 12022 (Fig. 1e) and K. pneumoniae ssp. ozaenae (Fig. 1f) successfully colonized root hairs. The infection process was initiated by host-specific E. adhaerens, forming IT which was invaded by test rhizobacteria. To monitor whether rhizobacteria that infected root hair remained restricted therein or survived to become nodule occupants, nodules from plants inoculated with similar rhizobacterial treatments were examined for occupancy at 50 DPI. Of 30 nodules collected from experimental plants, 18 harbored inoculated tagged rhizobacteria that were recovered on selective media. The characteristic fluorescence under CLSM (Supporting Information, Data S1) and 16S rRNA gene sequences confirmed that E. adhaerens, P. fluorescens IAM 12022 and K. pneumoniae ssp. ozaenae survived within the root nodules of V. radiata at 50 DPI.

Native root nodule endophytes of V. radiata

In all, 26 distinct colonies were isolated from the surface-sterilized root nodules of field-grown V. radiata on CRYMA, PCA and NA. The isolation procedure used a thorough surface sterilization of nodules and thus, no colony was observed on plates rolled with surface-sterilized nodules or plates spread with water from a final rinse for a sterility check even after prolonged incubation. As expected, plates rolled with non-sterile nodules developed bacterial colonies of epiphytes, which confirmed that colonies present on agar plates rolled with surface-sterilized nodules were endophytes. One gram fresh nodule tissue yielded 4 × 109 CFU of rhizobial and 3 × 105 CFU of non-rhizobial nodule endophytes. The phylogenetic distribution of nodule endophytes based on sequences of 16S rRNA genes is presented in Fig. 2. The phylogenetic tree highlighted the predominance of gram-positive Paenibacillus spp. and Bacillus spp. together with gram-negative species of Klebsiella, Ensifer, Agrobacterium, Blastobacter, Dyadobacter and Chitinophaga, which represented the major bacterial nodule population.

image

Figure 2. Phylogenetic tree. Neighbor-joining tree based on 16S rRNA gene sequences, showing relationships among mung bean root nodule endophytes. Scale bar: 0.1 nucleotides substitution per site. Numbers on branches indicate confidence limits estimated from bootstrap analysis of 1000 replicates. Accession numbers are mentioned within the phylogenetic tree corresponding to respective endophytes.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

In this article we demonstrate that invasion of rhizobial IT by strains of Pseudomonas and Klebsiella led to nodule colonization. Ensifer adhaerens isolated from field-grown V. radiata root nodules was inoculated as host-nodulating rhizobia. Pseudomonas fluorescens was selected because of its abundance in rhizosphere, and K. pneumoniae, being a facultative aerobe, was more suitable for survival in the microoxic environment of nodules. Ensifer adhaerens infected root hair through IT formation irrespective of the presence of other rhizobacteria (Fig. 1a and d). Pseudomonas fluorescens IAM 12022 and K. pneumoniae ssp. ozaenae attached to the base of root hair but failed to enter root hair in the absence of E. adhaerens (Fig. 1b and c), which confirmed their inability to form IT and invade root hairs independently. However, when E. adhaerens was present in the vicinity of root hairs, P. fluorescens and Kpneumoniae invaded root hairs (Fig. 1e and f). The presence of tagged rhizobacteria within the same IT was confirmed by exciting root hairs from each inoculated treatment with laser bands specific to Gfp, Rfp and Cfp. Signals from all the fluorescent proteins were captured only in the co-inoculated root hairs, confirming the presence of E. adhaerens, P. fluorescens and K. pneumoniae within the same root hair (Fig. 1d–f). The plasmids pDG77 and pUCPM18 used for tagging E. adhaerens and P. fluorescens were stably inherited. The colE1 origin of replication of pUCPM18 does not serve effectively in E. adhaerens (Simon et al., 1983) and absence of oriT would prevent its mobilization to E. adhaerens, confirming that fluorescence in confocal images originated from designated tags of rhizobacteria.

The inability of P. fluorescens and K. pneumoniae to enter root hairs may be attributed to their inability to secrete cellulase and pectinase. The role of cellulase (CelC2), the key cell wall-degrading enzyme, in facilitating the primary infection process is reported to be essential in Rhizobium (Robledo et al., 2008) and degradation of pectin layers by pectate lyase in favoring the entry of Klebsiella strains in plant tissues is documented (Kovtunovych et al., 1999). Some endophytes may enter independently from other bacteria through cracks formed at the emergence of lateral roots or at the zone of elongation and differentiation of the root (Rosenblueth & Martinez-Romero, 2006) but their localized movement within plant cells requires controlled release of pectinase and/or cellulase (Bekri et al., 1999). Therefore, even if Pseudomonas and Klebsiella entered roots through cracks, their movement from one cell to the other would be restricted, as the strains used in the experiment were cellulase- and pectinase-negative phenotypes (data not shown). Most importantly, known strains of Pseudomonas and Klebsiella lack Nod factors which allow rhizobia to attach to the tip of growing root hair and signal the plant to form nodules. The entry of host-specific rhizobia alone is facilitated through release of Nod factor in response to plant flavonoids (Perret et al., 2000).

The nodule occupancy experiment comprised similar treatments of tagged rhizobacteria to V. radiata plants grown in sterile soil. At 50 DPI, nodules were formed only in plants inoculated or co-inoculated with E. adhaerens. The test rhizobacteria were recovered in pure culture on selective media with appropriate antibiotics. Colony morphology, characteristic fluorescence of Rfp/Gfp/Cfp under CLSM (Supporting Information, Fig. S1) and 16S rRNA gene sequencing confirmed that test rhizobacteria proceeded past the infection thread and colonized root nodules. These observations suggested that a similar mechanism might be employed by other rhizobacteria for invading root hairs, to colonize nodules in hosts where rhizobial infection proceeds through IT formation.

Since root hair infection and nodule occupancy experiments were performed under in vitro conditions, we further explored nodule endophyte diversity of field-grown V. radiata by isolating the maximum number of non-rhizobial species. The isolation yielded 26 distinct isolates, which were distributed among eight major bacterial genera (Fig. 2) with a predominance of gram-positive Bacillus and Paenibacillus. The results were in disagreement with those of Dubey & Gupta (2012), who reported isolation of 12 gram-negative and four gram-positive endophytes from Vigna mungo, a species close to V. radiata. Rajendran et al. (2008) reported that the majority of isolated root nodule endophytes of pigeon pea were gram-positive, whereas Sturz et al. (1997) found an equal distribution of gram-positive and gram-negative endophytes in nodules of red clover. A preponderence of gram-positive bacteria among the isolated endophytes raises the possibility that bacterial endospores bound to the surface of nodules might have survived the surface-sterilization procedure, but this does not seem to be the case as a thorough validity check of nodule surface sterilization was carried out and the results were conclusive that sterilization was adequately performed and isolated cultures were nodule endophytes. Similarly, isolation of Paenibacillus and Bacillus strains from other legume nodules have been reported previously (Muresu et al., 2008). These findings emphasize the significant role of the host plant in defining the composition of rhizospheric communities through root exudates, in addition to the ability of rhizobacteria to survive under stress caused by biotic and abiotic factors.

The predominance of Pseudomonas in the nodule suggest rhizosphere population-based selection, as it is profusely present in rhizosphere and establishes easily in a wide variety of niches (Spiers et al., 2000). However, the presence of Paenibacillus and Bacillus as major nodule colonizers might indicate a role of rhizospheric preponderance as one of the selection criteria, together with the ability to form resistant spores able to thrive in harsh conditions (Li et al., 2012). The presence of K. pneumoniae in the IT and field-grown nodules raises the possibility that, once within nodule, this strain might fix nitrogen, as the environment in root nodules is microaerobic (Trzebiatowski et al., 2001) and expression of nif occurs under nitrogen-limiting anaerobic or microaerobic conditions (Buck, 1991). The intricate selection and survival strategy of natural rhizobacterial occupants and their role in nodules remains to be explored.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The authors gratefully acknowledge the financial grant received from the Department of Biotechnology (DBT) (BT/PR4153/AGR/21/351/2011), Ministry of Science and Technology, Government of India, New Delhi and infrastructure provided by Nirma Education and Research Foundation (NERF), Ahmedabad, Gujarat, India.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
fml12245-sup-0001-Supporting-information.docWord document790K

Data S1. Recovery of tagged rhizobacteria from V. radiata root nodules at 50 DPI.

Fig. S1. Confocal micrograph of (g) Ensifer adhaerens-Rfp (h) Pseudomonas fluorescens-Gfp and (i) Klebsiella pneumoniae-Cfp isolated in pure culture from experimental root nodules of Vigna radiata at 50 DPI. Note: noise (autofluorescence) was removed prior to capturing images.

Table S1. Best-match identification of phylotypes (NCBI Database) of root nodule endophytes, obtained with 16S rRNA gene universal primers.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.