By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Legume–rhizobium interactions have been widely studied and characterized, and the disaccharide trehalose has been commonly detected during this symbiotic interaction. It has been proposed that trehalose content in nodules during this symbiotic interaction might be regulated by trehalase. In the present study, we assessed the role of trehalose accumulation by down-regulating trehalase in the nodules of common bean plants.
We performed gene expression analysis for trehalase (PvTRE1) during nodule development. PvTRE1 was knocked down by RNA interference (RNAi) in transgenic nodules of the common bean.
PvTRE1 expression in nodulated roots is mainly restricted to nodules. Down-regulation of PvTRE1 led to increased trehalose content (78%) and bacteroid number (almost one order of magnitude). In addition, nodule biomass, nitrogenase activity, and GOGAT transcript accumulation were significantly enhanced too.
The trehalose accumulation, triggered by PvTRE1 down-regulation, led to a positive impact on the legume–rhizobium symbiotic interaction. This could contribute to the agronomical enhancement of symbiotic nitrogen fixation.
Biological nitrogen fixation is carried out only by prokaryotes in an enzymatic complex called nitrogenase where atmospheric nitrogen (N2) is reduced to ammonia. The importance of this mutualistic interaction arises from the fact that legumes are an important source of protein in the diet of many countries worldwide, and symbiotic fixed nitrogen makes a major contribution to this, accounting for roughly 200 million tons of nitrogen annually (Graham & Vance, 2003; Kouchi et al., 2010). Leguminous plants are the second largest group of food and feed crop species grown globally, and represent the third largest group of angiosperm species. A key developmental process in this remarkable plant–bacteria interaction is the formation on plant roots of novel organs called nodules, wherein rhizobia differentiate into their intracellular nitrogen-fixing form, known as bacteroids. These structures develop after exchange and recognition of signaling molecules between the rhizobial bacteria and the leguminous plant (Gage, 2004; Ferguson et al., 2010).
The ontogeny of nodules has been characterized genetically, biochemically, and at the physiological and molecular levels throughout all stages, leading to successful and efficient symbiosis (Streeter, 1987; Fougère et al., 1991; Salminen & Streeter, 1992; López et al., 2008; Kouchi et al., 2010). Recently, the metabolites produced at the beginning of the rhizobium–legume symbiotic interaction were analyzed by high-throughput methods (metabolome analysis) and a significant induction in the biosynthesis of 166 metabolites was found, corresponding to flavonoids, amino acids, fatty acids, carboxylic acids, and carbohydrates, with trehalose among the metabolites induced to the greatest extent (Brechenmacher et al., 2010). In Phaseolus vulgaris (common bean) interacting with Rhizobium, enhanced germination, quality, and grain yield have been correlated with trehalose content, and a higher tolerance to abiotic stress has been reported (Farías-Rodriguez et al., 1998; Altamirano-Hernández et al., 2007).
Trehalose (α-D-glucopyranosyl-1,1-α-D-glucopyranoside) is a nonreducing disaccharide. Trehalose biosynthetic and degradation pathways are widespread throughout bacteria, archaebacteria, fungi, metazoans (except for vertebrates), and plants (Avonce et al., 2006). Trehalose (and its precursor trehalose-6-phosphate) is a signaling molecule involved in growth, development, and differentiation in plant cells (Paul et al., 2008). There are five reported biosynthetic pathways for trehalose, but only one for degradation. Trehalose degradation is carried out by trehalase, yielding two glucose moieties (Avonce et al., 2006). Although trehalose is commonly detected during the legume–rhizobia symbiosis, from rhizobia recognition by root hairs through nodule development and nitrogen fixation, trehalose metabolism in leguminous plants is still poorly understood (Müller et al., 2001b; Brechenmacher et al., 2010; Vauclare et al., 2010). Moreover, the importance of trehalose metabolism in plant–microbe interactions, such as legume–rhizobium, plant–mycorrhiza, and plant–pathogen interactions, has recently been highlighted (Müller et al., 2001b; Brodmann et al., 2002; Foster et al., 2003; Ocón et al., 2007; Nehls, 2008; Wilson et al., 2010).
In senescent nodules, trehalose becomes the most abundant nonstructural carbohydrate, as evidenced by the fact that the sucrose content was found to undergo a marked (84%) decrease, while up to one-half of the trehalose remained during nodule senescence (Müller et al., 2001b). Nodule trehalose biosynthesis and catabolism seem to be two separate, albeit interconnected, processes, as trehalose in the symbiosomes is not degraded by the trehalase present in the cytosol of the infected cells of the nodules, and the concentrations of trehalose inside the symbiosomes are kept constant throughout nitrogen fixation (Vauclare et al., 2010).
Three approaches have been described to study the role of trehalose in the symbiotic process: addition of validamycin A directly to the nodules to inhibit trehalase enzymatic activity; overexpression of the endogenous trehalose phosphate synthase (TPS) from Rhizobium to enhance trehalose accumulation in Rhizobium before and during symbiotic interaction with leguminous plants; and mutating genes for trehalose assimilation in Sinorhizobium to induce trehalose accumulation in Sinorhizobium. Induction of trehalose accumulation in nodules of leguminous plants through validamycin A addition leads to improvement of tolerance to drought and salinity stress. Overexpression of endogenous rhizobial TPS or mutating trehalose assimilation genes enhances trehalose accumulation in bacteria as well as in bacteroids, leading to a significant increase in nodule biomass and nitrogenase activity and conferring abiotic stress tolerance to the plant (Jensen et al., 2005; Suárez et al., 2008; López et al., 2009).
Trehalose accumulation in the cytosol of infected nodule cells strongly suggests that induction of trehalose synthesis and degradation during nodule development are tightly regulated. Furthermore, trehalose metabolism directly impacts photosynthate partitioning and utilization in sink organs such as nodules, which constantly require carbohydrates to fuel growth, N2 fixation and assimilation (Aeschbacher et al., 1999; Müller et al., 2001a; Xie et al., 2003; Van Dijken et al., 2004). Trehalase transcripts are present in expressed sequenced tag (EST) databases of several leguminous plants during symbiosis (Pontius et al., 2003; Ramírez et al., 2005). In the common bean, only three sequences corresponding to trehalase (CV535700, CV537568, and CV535738) were found, from a total of 112 229 sequences deposited to date in the Common Bean Gene Index (DFCI/Common Bean Gene Index, v4.0, http://compbio.dfci.harvard.edu/tgi/). Interestingly, these ESTs are part of the same contig and were found only in the nodule transcriptome.
Herein, transgenic roots (composite plants) induced by Agrobacterium rhizogenes K599 (Estrada-Navarrete et al., 2007; Blanco et al., 2009; Sánchez-López et al., 2011) were used to assess the role of trehalose in common bean symbiosis by knocking down PvTRE1 transcript abundances by RNAi. Furthermore, since genetic transformation in composite plants is restricted to roots and nodules, we evaluated the systemic effect of nodule trehalose accumulation on the untransformed upper part of the plant. This approach helped us to characterize and directly determine the role of trehalose metabolism in nodules of P. vulgaris, and revealed the tight relationship among nodules and the upper parts (leaves) of the plant during symbiotic interaction.
Materials and Methods
Plant materials and growth conditions
Common bean (Phaseolus vulgaris L.) cv Negro Jamapa 81 was used in this study. Seeds were surface-sterilized and germinated under sterile conditions for 2 d and then planted in pots with vermiculite (Estrada-Navarrete et al., 2007). Plants were grown in a glasshouse with a controlled environment (26–28°C, 16 : 8 h light : dark) and were watered with B&D nutrient solution (2 mM CaCl2·2H2O, 1 mM KH2PO4, 1 mM K2HPO4, 20 μM FeC6H5O7, 500 μM MgSO4·7H2O, 500 μM K2SO4, 2 μM MnSO4·H2O, 4 μM H3BO3, 1 μM ZnSO4·7H2O, 4 μM CuSO4·5H2O, 0.2 μM CoSO4·7H2O, 2 μM Na2MoO4·2H2O, 8 mM KNO3; Estrada-Navarrete et al., 2007). The generation of composite common bean plants was done according to the protocol developed by Estrada-Navarrete et al. (2007). Hairy roots (3–6 cm long) emerging from the A. rhizogenes K599 infection site were observed during the second week post-infection. After confirming the presence of reporter genes (by epifluorescence microscopy), the normal (untransformed) root system was excised and composite common bean plants were replanted in pots with vermiculite. Immediately after transferring A. rhizogenes-transformed roots of composite common bean plants into pots with fresh vermiculite, each transformed root was inoculated by adding 1 ml of the Rhizobium etli strain CFN42 culture directly. These were grown for 21 d postinoculation (dpi) under controlled environmental conditions and watered with B&D nitrogen-free nutrient solution (the same reagent chemical composition without 8 mM KNO3). After this period, transgenic roots and nodules were collected in liquid nitrogen and stored at −80°C until they were used for quantitative reverse-transcription polymerase chain reaction (qRT-PCR) analysis and carbohydrate profile analysis by high-performance liquid chromatography (HPLC). Alternatively, transgenic nodules were collected and immediately fixed in LR-White resin for detailed characterization by optical microscopy and transmission electron microscopy (TEM).
Bacterial strains and growth conditions
The Rhizobium etli strain CFN42 was grown in sterilized PY liquid culture (0.5% bactopeptone (w/v), 0.3% yeast extract (w/v), 7 mM CaCl2·2H2O) with 20 μg ml−1 nalidixic acid at 30°C to a cell density of 5–8 × 108 ml−1. A. rhizogenes K599 (K599) was grown in sterilized LB solid culture (1% bactopeptone (w/v), 0.5% yeast extract (w/v), 1% NaCl (w/v), 1.5% agar (w/v)) at 30°C for 24 h. A. rhizogenes K599 with pBGWFS7_PvTRE1PR was grown in sterilized LB solid culture with 200 μg ml−1 spectinomycin at 30°C for 24 h. A. rhizogenes K599 with pTdT-DC-RNAi (empty vector), A. rhizogenes K599 with pTdT-GUS-RNAi, and A. rhizogenes K599 with pTdT-PvTRE1-RNAi were grown in sterilized LB solid culture with 200 μg ml−1 spectinomycin at 30°C for 24 h. All of these A. rhizogenes K599 strains were used according to Estrada-Navarrete et al. (2007) to generate composite common bean plants.
Identification of trehalase gene in the common bean and phylogenetic analysis
We identified ESTs corresponding to the trehalase gene, which is directly involved in trehalose degradation, in the whole Common Bean Gene Index v.4.0 (DFCI, http://compbio.dfci.harvard.edu/tgi/). In addition, we identified the promoter and complete coding sequence of trehalase by analyzing the whole common bean draft genome (Phytozome v8.0, http://www.phytozome.net; Mazorka, http://mazorka.langebio.cinvestav.mx/blast/). The coding sequence of trehalase (PvTRE1) was analyzed with ‘Pfam’ at the Sanger Institute website, and was compared with other trehalase sequences deposited in GenBank using the ‘BLASTX’ tool at the National Center for Biotechnology Information (NCBI) website. Multiple sequence alignment and phylogenetic analysis, with the neighbor-joining method and 1000 bootstrap replicates, of the reported trehalase protein sequences were performed using MEGA 5.0 (Tamura et al., 2011).
Protein alignment and modeling of PvTRE1
Alignment of the proteins was performed with ClustalX (2.0.10). The alignments were adjusted based on the modeling data for PvTRE1. Modeling of PvTRE1 was performed using Swiss model (http://swissmodel.expasy.org) based on the known three-dimensional structure of trehalase from Escherichia coli (Gibson et al., 2007; PDB code 2WYN). Visualization and calculations of root-mean-square deviation (RMSD) values for the modeling of PvTRE1 were performed with PyMol (DeLano, 2002).
Constructs and plant transformation
To generate a promoter analysis construct, primers were designed based on the PvTRE1 promoter sequences obtained from the draft genome. The promoter region of PvTRE1 (1900 bp) was amplified using the forward primer PvTRE1PromFor: CACCGGGTACATGTGTAATTACAATAAACTTCAA and the reverse primer PvTRE1PromRev: GAAAATGTGGCAATGAATTGATGAAAGAGAAG. The amplified fragment was cloned in the pENTR/SD/D-TOPO vector (Invitrogen) and sequenced. The resulting pENTR-PvTRE1PR plasmid was recombined into the pBGWFS7 binary vector (Karimi et al., 2007), producing the pBGWFS7_PvTRE1PR construct. The 3′-coding sequence of PvTRE1 (PvTRE1, 220 bp) was amplified with the forward primer PvTRERiFor: CACCGAAGGCCTTCTAAAATCTGGGTTGC and the reverse primer PvTRERiRev: CAAACGTTTCCCTAGACCAAAAAGAATAATGTTG. The amplified fragment was cloned in the pENTR/SD/D-TOPO vector (Invitrogen) and sequenced. The resulting pENTR-PvTRE1 plasmid was recombined into the pTdT-DC-RNAi binary vector (Valdés-López et al., 2008). The correct orientation was confirmed by PCR using the WRKY-5-Rev primer (GCAGAGGAGGAGAAGCTTCTAG) or WRKY-3-Fwd primer (CTTCTCCAACCACAGGAATTCATC) and PvTRERiFor primer for the pTdT-PvTRE1-RNAi plasmid. The resulting pBGWFS7_PvTRE1PR (for expression analysis) and pTdT-PvTRE1-RNAi (for PvTRE1 silencing by RNAi) plasmids were introduced by electroporation into A. rhizogenes K599 and were then used for plant transformation. Composite common bean plants were generated as described by Estrada-Navarrete et al. (2007). Putative transgenic hairy roots were confirmed by checking for the presence of green/red fluorescence, resulting from the expression of the green fluorescence protein (GFP)/tdTomato reporter gene, by epifluorescence microscopy. The composite common bean plants were grown in the glasshouse for 21 d as described earlier.
PvTRE1 promoter fusion to GUS to determine PvTRE1 expression in nodules
After 21 dpi, the transgenic roots with nodules of composite common bean plants harboring pBGWFS7_PvTRE1PR (pPvTRE1::GUS) were collected for beta-glucuronidase (GUS) assays to perform the PvTRE1 promoter analysis. The roots with nodules were incubated in X-Gluc buffer (5× GUS buffer, 100 ml: 250 mg X-Gluc, 0.1 g Triton X-100, 82.3 mg K-ferricyanide (K3Fe(CN)6, M = 329.9 g mol−1), 105.6 mg K-ferrocyanide (K4Fe(CN)6·3H2O, M = 422.4 g mol−1) in 50 mM phosphate buffer) at 37°C for 10 h before clearing in acetone : methanol (1 : 3).
TEM and optical microscopy
Nodules (nine nodules per experiment) were embedded in LR-White resin after being fixed with 2% ρ-formaldehyde and 0.4% glutaraldehyde in phosphate-buffered saline (PBS), and subjected to a short dehydration ethanol series. Samples for optical microscopy were stained with toluidine blue. Sections of 7 μm were prepared with an ultramicrotome (Leica Ultracut R, Vienna, Austria). The optical microscopy analyses were performed with a light microscope (Motic BA300, Xiamen, China) and photographed with a digital camera (Motic M1000, Xiamen, China). Samples for TEM were stained with uranyl acetate. Thin sections of 60 nm were prepared with an ultramicrotome (Leica Ultracut R). The scanning electron microscopy analyses were performed with a Zeiss EM900 transmission electron microscope dual vision coupled cam system (Gatan, Inc., Pleasanton, CA, USA).
Rhizobium reisolation from nodules and determination of colony-forming units (CFUs)
Nodules (nine nodules per experiment) were isolated from roots and surface-sterilized by immersion for 10 min in sodium hypochlorite (10% v/v). Each nodule was then homogenized in five volumes of 100 mM MgCl2 using a plastic pipette. Serial dilutions (100–10−8) were plated (100 μl) onto PY solid medium with 20 μg ml−1 nalidixic acid at 30°C and colonies were counted after 24 h.
Analysis of the carbohydrate profile by HPLC
Nodules and leaves from composite common bean plants at 21 dpi were collected, weighed, frozen in liquid nitrogen, and ground for extraction of total soluble carbohydrates (i.e. sugars) with 1 ml of ethanol 80% (v/v) for 10 min under agitation at 80 °C. The supernatant was dried and dissolved in 1 ml of HPLC-grade water and immediately filtered with a 0.22 μm membrane to remove impurities. These extracts were analyzed by HPLC in a Waters-600E system controller (Waters, Milford, MA, USA) equipped with a Waters 410 refractive index detector (Waters) and a carbohydrate analysis column (Kromasil NH2–5 μm, Supelco, PA, USA). The temperature of the column was kept at 35°C, and the mobile phase used was acetonitrile : water (80 : 20) at a flow rate of 1.2 ml min−1. Glucose, fructose, sucrose, maltose, and trehalose (Sigma-Aldrich) solutions were used as standards. Standard curves used to carry out the quantifications had an index correlation ranging from 0.9865 to 0.9928. Results were expressed as μg mg−1 FW.
Biological nitrogen fixation
Nitrogen fixation was assayed using the acetylene reduction method (Vessey, 1994). Transgenic nodulated roots (21 dpi) were placed (nine individual roots per experiment) in a 160 ml vial closed with a serum cap. Immediately, air (4 ml) was withdrawn from the closed vial and replaced by acetylene gas. Ethylene production was assayed in a gas chromatograph and expressed as nmol ethylene min–1 per nodule FW.
cDNA synthesis and qRT-PCR analysis
TRIzol reagent (Invitrogen) was used to isolate RNA from nodules at 21 dpi. For qRT-PCR analysis, RNA was treated with DNaseI (Invitrogen) to remove genomic DNA. The absence of DNA was confirmed by performing PCR (40 cycles, similar to the real-time PCR program) on the DNaseI-treated RNA using Taq-DNA polymerase. A LightCycler® 480 Real-time PCR system (Roche, Penzberg, Germany) was used for real-time PCR quantifications. qRT-PCR was performed according to the standard First Strand cDNA Synthesis kit with the Maxima® SYBR Green qPCR Master Mix (2×) protocol (Fermentas Life Sciences, Waltham, MA, USA). A ‘no DNA’ template control was used in each analysis. A list of primers used is given in Table S1. The results presented are from six independent (n =6) transgenic roots (or another organ) from different composite common bean plants; numbers in brackets indicate the composite common bean plant used in this study, and statistical significance was determined with an unpaired two-tailed Student's t-test. Each biological replicate was tested by triplicate and data were normalized to the elongation factor 1-α (PvEF1α) reference gene (Livak & Schmittgen, 2001).
Identification of the trehalase gene in common bean
Three ESTs (CV535700, CV537568, and CV535738) corresponding to a single trehalase gene were identified in GenBank. With these, we were able to identify the complete coding sequence (scaffold00002) in the common bean draft genome (Phytozome v8.0, http://www.phytozome.net; Mazorka, http://mazorka.langebio.cinvestav.mx/blast/). The trehalase gene was analyzed and its identity confirmed using the ‘Pfam’ and ‘BLASTX’ tools at The Sanger Institute and NCBI websites, respectively. The corresponding gene for trehalase indeed encodes for a protein that was found to belong to the trehalase protein family (with an E-value of 4.8 × 10−162), and is homologous to the trehalase protein from Glycine max, with 81% identity and 86% similarity, and an E-value of 0.0; therefore the trehalase gene was designated as PvTRE1. PvTRE1 has 10 exons and nine introns (Figs 1a, S1a). We carried out three-dimensional modeling of PvTRE1 (Fig. 1a), and the RMSD value for alignment with the three-dimensional structure of trehalase from E. coli (EcTRE) was 0.080 Å (DeLano, 2002; Gibson et al., 2007). Furthermore, we performed a phylogenetic analysis of PvTRE1 with trehalases from plants, fungi, animals, and bacteria (Tamura et al., 2011). The PvTRE1 amino acid sequence grouped with trehalases from plants, specifically with trehalases from leguminous plants (Medicago truncatula and G. max; Fig. 1b).
PvTRE1 is strongly expressed in nodules of common bean plants
In addition to identifying the coding region for PvTRE1, the promoter region was also cloned. The regulatory sequence region of PvTRE1 was analyzed with PlantPan (http://plantpan.mbc.nctu.edu.tw/seq_analysis.php) and found to contain eight nodule-specific regulatory motifs related to those found in the promoter region of common bean leghemoglobin A (PvLeghA). Interestingly, the promoter region from PvTRE1 has three more nodule regulatory motifs than does the promoter region of PvLeghA, and also more than the promoter regions of M. truncatula trehalase (three) and GmTRE1 from G. max (six; Supporting Information, Fig. S1b). In order to determine the expression pattern of PvTRE1, the promoter region (1.9 kb of upstream sequence) of PvTRE1 (pPvTRE1) was fused to the reporter gene GUS, and transformed into A. rhizogenes K599. At 21 dpi, GUS assays were performed on transgenic roots and nodules. GUS staining showed that expression of PvTRE1 in nodulated roots is mainly restricted to nodules (Fig. 2b, and Fig. S1c). This result is in line with the bioinformatic analysis of the PvTRE1 promoter region. In addition, we carried out a PvTRE1 transcript accumulation analysis to determine the expression level of this gene in several organs (nodules, roots, leaves, pods, and seeds). PvTRE1 is expressed at different levels in all the organs analyzed, except roots that have five times lower transcript abundance than nodules (Fig. S2a).
Determination of PvTRE1 transcript abundance during nodule development
Since PvTRE1 is expressed mainly in nodules than roots, we decided to assess PvTRE1 transcript abundance during nodule ontogeny. We performed nodulation analysis at 14, 21, 25, and 28 dpi in wildtype common bean plants interacting with R. etli CFN42. Transcript abundance of PvTRE1 in nodules remained unchanged from 14 to 21 dpi. At 25 dpi PvTRE1 transcript abundance underwent a decrease of 16.19 ± 0.5%, relative to 14–21 dpi. However, at 28 dpi a dramatic increase in PvTRE1 transcripts of up to 264.3 ± 3.68% (compared with 14–21 dpi; Fig. 2c) was observed. By contrast, PvTRE1 transcript abundance remained low and constant in the roots at the same time points. In comparison, sucrose synthase (PvSUS1) transcript abundance (Silvente et al., 2003) underwent an increase of 40 ± 1.52% at 21 dpi relative to 14 dpi. At 25 dpi a further increase of 123 ± 9.8% (compared with 14 dpi) was observed. However, the PvSUS1 transcript abundance decreased down to 57.9 ± 1.78% at 28 dpi (relative to 14 dpi), contrary to the pattern observed for PvTRE1 (mirror-image symmetry).
Carbohydrate profile during common bean nodule development
Soluble carbohydrates, including fructose, glucose, sucrose, and trehalose, were quantified at each step of the nodulation process. Glucose was present at 21 and 25 dpi, with its highest concentration at 25 dpi. Fructose was found only at 21 dpi. Sucrose concentration showed peaks at 21 and 28 dpi, and troughs at 14 and 25 dpi. Trehalose peaked at 25 dpi; however, the lowest trehalose concentrations were detected at 14 and 28 dpi, with similar concentrations at those two points of the nodulation process (Fig. 2d). The trehalose concentration peak at 25 dpi was coincident with the decrease in PvTRE1 transcript abundance. Interestingly, the highest PvTRE1 transcript accumulation occurred at 28 dpi, perfectly matching the lowest trehalose concentration during nodule development (Fig. 2c,d). Overall, the concentrations of the soluble carbohydrates trehalose and sucrose during the nodulation process negatively correlated with the PvTRE1 and PvSUS1 transcript abundances, respectively.
Transcript accumulation of genes involved in growth, carbon metabolism and autophagy in loss-of-function PvTRE1 transgenic nodules
To follow up on the results from the pPvTRE1::GUS expression, PvTRE1 transcript abundance, and soluble carbohydrate profile experiments, PvTRE1 gene silencing by RNA interference (RNAi) was performed in order to explore the role of this gene in nodule development and physiology. Composite common bean plants expressing the PvTRE1-RNAi construct (pTdT-PvTRE1-RNAi) were generated and hairy roots were inoculated with R. etli CFN42 (Estrada-Navarrete et al., 2007). PvTRE1 transcript abundances were determined at 21 dpi by qRT-PCR in nodules expressing the RNAi construct, as well as in those from control lines (K599 and pTdT-DC-RNAi). PvTRE1 transcript abundances were reduced (Fig. 3a) up to 81.9 ± 0.37% in PvTRE1-RNAi transgenic nodules, as compared with transgenic nodules harboring a control RNAi construct (pTdT-GUS-RNAi, unable to silence any endogenous gene), and nodules from K599 hairy roots. In these latter two controls, PvTRE1 transcript abundance was quite similar (Fig. S2c). We also assessed transcript accumulation of genes involved in growth and carbon meta-bolism (PvSUS1, PvHXK1 (hexokinase), PvTOR (target of rapamycin), PvSnRK1 (Sugar non-fermentable related kinase1)) and autophagy (PvATG3 (autophagy related gene 3), PvBeclin (Beclin or autophagy related gene 6)) in the PvTRE1-RNAi transgenic nodules. Down-regulation of PvTRE1 inversely correlated with increased transcript abundance of PvSUS1, PvHXK1, PvTOR, PvSnRK1, and PvATG3 (Fig. 3b), whereas PvBeclin remained without significant change.
Down-regulation of PvTRE1 transcript abundance leads to alteration in bacteroid morphology
Nodule structure and bacteroid morphology were analyzed by optical microscopy and TEM, respectively. At the optical microscopy level, we did not observe obvious morphological differences in sections of PvTRE1-RNAi nodules stained with toluidine blue compared with control nodules (K599, pTdT-DC-RNAi; Fig. 4a,c). To assess the effects of PvTRE1 silencing on bacteroid morphology, we analyzed ultrathin nodule sections using TEM. There was an electrodense zone surrounding bacteroids in PvTRE1-RNAi transgenic nodule sections, which was labeled as ‘electrodense boundary’ (edb) (Fig. 4i); this structure was absent in bacteroids of control nodule sections (Fig. 4g,h). In addition, the number of bacteroids per visual camp (577.5 μm2) in PvTRE1-silenced nodules (Fig. 4f) was increased (36 ± 2.11%) with respect to those in control nodules (K599, pTdT-DC-RNAi; Figs 4d–e, S3).
Down-regulation of PvTRE1 has a direct influence on nodule trehalose concentrations, bacterial viability, nitrogen assimilation, and nodule biomass
The trehalose content of PvTRE1-RNAi transgenic nodules was increased by 78.0 ± 5.72% compared with that in control nodules (K599, pTdT-DC-RNAi; Fig. 5a). By contrast, sucrose was decreased up to 19.90 ± 4.98% in PvTRE1-RNAi nodules compared with control nodules (K599, pTdT-DC-RNAi; Fig. 5a). There were no significant differences in the fructose and glucose contents between PvTRE1-RNAi and control nodules (K599, pTdT-DC-RNAi; Fig. 5a). The soluble carbohydrate profile was assessed by HPLC at 21 dpi. In order to ensure that trehalose concentrations were accurately quantified, we included maltose as an internal control to assess the retention times of this soluble carbohydrate and trehalose, as both disaccharides are formed from two glucose molecules linked by a glycosidic bond: α,α-1,4 for maltose and α,α-1,1 for trehalose. The retention time for maltose was 15.8 min, and for trehalose it was 18.8 min (Fig. 5a, upper inset).
To determine the effect on bacterial survival of down-regulation of PvTRE1, Rhizobium bacteria were reisolated from PvTRE1-RNAi and control nodules. CFUs recovered from PvTRE1-RNAi transgenic nodules (Fig. 5b) were almost one order of magnitude higher than those from control nodules (K599, pTdT-DC-RNAi). In addition, we determined the nitrogenase activity in PvTRE1-RNAi transgenic nodules and control nodules (K599, pTdT-DC-RNAi). The PvTRE1-RNAi transgenic nodules displayed 71.01 ± 7.19% more nitrogenase activity than control nodules (Fig. 5c). Just as acetylene reduction is indicative of nitrogenase activity in bacteroids, the increase in transcript accumulation of genes involved in nitrogen assimilation (GOGAT (NADH-Glutamate synthase II )) provides information about the amount of fixed nitrogen mobilized into plant cells (Silvente et al., 2002; Cordoba et al., 2003; Blanco et al., 2008). GOGAT transcript abundance was increased in PvTRE1-RNAi transgenic nodules by 82 ± 0.02% compared with that of control nodules (Fig. 5d). By contrast, GS (glutamine synthetase) transcript accumulation did not show any significant change between PvTRE1-RNAi transgenic nodules and control nodules (Fig. S4b). We also quantified nodule biomass, as it is indicative of the amount of carbon allocated to nodule cells. The net nodule DW, or biomass, of PvTRE1-RNAi transgenic nodules increased threefold (Fig. 5e) compared with control nodules, without a significant increase in nodule number (Fig. S4a).
PvTRE1 down-regulation in nodules does not have negative side-effects compared with leaves of composite common bean plants
As PvTRE1-RNAi transgenic nodules exhibited changes in the content of soluble carbohydrates (sucrose and trehalose), nodule biomass, bacterial survival, nitrogen fixation, and nitrogen assimilation, we assessed the trehalose content in leaves of the composite common bean plants (Fig. 6a). Interestingly, in leaves of composite plants with PvTRE1-RNAi transgenic nodules, trehalose was increased by 27.32 ± 3.40% compared with leaves of control composite plants (K599, pTdT-DC-RNAi). However, sucrose, fructose, and glucose contents in leaves of plants with PvTRE1-RNAi transgenic nodules did not show any significant changes compared with control plants (K599, pTdT-DC-RNAi; Fig. 6a). In addition, PvTRE1 transcript accumulation in leaves did not show any significant change.
Leaf biomass in composite common bean plants with PvTRE1-RNAi transgenic nodules (Fig. 6d) was comparable to that of control plants (K599, pTdT-DC-RNAi). Additionally, leaf area measurements did not reveal any negative effect of PvTRE1-RNAi nodules on leaf biomass (Fig. 6e).
Herein the complete sequence of the trehalase gene (PvTRE1) found in the common bean genome (Phytozome v8.0 and Mazorka) was obtained and compared with homologous plant (GmTRE1) and E. coli (EcTRE) trehalase genes. Prediction of the three-dimensional structure of the encoded protein based on the known structure of the E. coli homolog led us to designate this gene PvTRE1. The structural analysis revealed a high degree of conservation, and moreover, they are grouped (Fig. 1) with other plant trehalases, suggesting a high degree of sequence conservation and perhaps of structure/function identity among plant trehalases, as was reported for trehalases from insects and fungi, and other phylogenetic groups (Doehlemann et al., 2006; Lunn, 2007; Silva et al., 2009; Chen et al., 2010). The PvTRE1 gene structure consists of 10 exons and nine introns, similar to the gene structure of GmTRE1 and MtTRE1, both of which have 11 exons and 10 introns, and different from AtTRE1, which contains six exons and five introns (Fig. S1a). Furthermore, in nodulated roots, the expression of PvTRE1 was restricted mainly to nodules, which resembled the expression of nodulin genes, for example, leghemoglobin A and sucrose synthase. Nodule-enhanced expression of PvTRE1 (Fig. 2a,b) suggests that PvTRE1 could have an active role in carbon metabolism in nodules, as previously suggested for its homolog (GmTRE1) in soybean (Aeschbacher et al., 1999).
PvTRE1 transcript accumulation in roots was lower than that in nodules, but remained constant throughout nodule ontogeny. By contrast, the PvTRE1 transcript accumulation profile in nodules was inversely proportional to trehalose concentration during nodule development, and it is worth noting that the highest PvTRE1 transcript accumulation coincided with the beginning of nodule senescence, correlating with the up-regulation of AtTRE1 during senescence in Arabidopsis (Fernandez et al., 2010). Interestingly, PvSUS1 transcript accumulation paralleled trehalose content (Fig. S2b) and was opposite to accumulation of the PvTRE1 transcript (Fig. 2c). This supports the idea that PvTRE1 regulates trehalose content, as inhibition of trehalase activity by addition of validamycin A in nodules resulted in trehalose accumulation (López et al., 2009). Additionally, it seems that trehalose content influences PvSUS1 expression directly, which is consistent with a previous report that addition of trehalose into nonnodulated soybean roots strongly induces sucrose synthase activity, and trehalase inhibition by validamycin A leads to a decrease in sucrose content (Müller et al., 1995, 1998).
We decided to assess the impact of trehalose on nodule physiology by down-regulating PvTRE1 expression via RNAi in common bean composite plants. The resulting 82% decrease in PvTRE1 transcript (Fig. 3a) accumulation seems to have a higher impact on trehalose content (78% increase; Fig. 5a) than that attained by inhibiting trehalase activity with validamycin A in soybean nodules; this led to a 73% inhibition of enzyme activity but only a 46% increase in trehalose (Müller et al., 1995). In addition to the significant enhancement of trehalose content in PvTRE1-RNAi transgenic nodules, the expression of genes directly involved in growth and carbon metabolism (PvSUS1, PvHXK1, PvTOR, PvSnRK1) and autophagy (PvATG3) were also increased (Fig. 3b), suggesting that the up-regulation was triggered by PvTRE1 down-regulation, which in turn gave rise to a higher trehalose content. The increase in transcript accumulation for sucrose synthase (203%) and hexokinase (134%) reflects an enhancement in carbon assimilation, most probably driven by an increased rate of sucrose degradation, as evidenced by the specific decrease in sucrose in these nodules without any significant changes in soluble fructose and glucose concentrations (Fig. 5a; Müller et al., 1998; Moore et al., 2003). Furthermore, the increase in PvTOR (43%), PvSnRK1 (81%), and PvATG3 (26%) transcript (Fig. 3b) accumulation suggests a direct positive impact on autophagy and cellular cytoplasm recycling, as these genes are actively involved in regulating cell development and protein turnover. Previously, it has been reported that trehalose can induce autophagy (trehalose-dependent autophagy) by a target of rapamycin (TOR)-independent pathway, suggesting that plant and animal cells possess similar mechanisms to promote cellular recycling instead of triggering programmed cell death (Sarkar et al., 2007; Aguib et al., 2009; Jossier et al., 2009; Wang et al., 2009; Casarejos et al., 2011; Hofius et al., 2011).
The accumulation of trehalose, mediated by PvTRE1 down-regulation, triggered sucrose degradation and increased bacterial survival, nodule biomass, nitrogen fixation, and GOGAT transcript accumulation (Fig. 5). The enhanced bacterial proliferation inside nodules could be fueled by the trehalose accumulation and its translocation from the infected cells to Rhizobium. Rhizobia possess two operons and two genes directly involved in trehalose transportation and assimilation (thuEFGK, aglEFGAK, thuA, thuB) mainly induced by trehalose (Jensen et al., 2002, 2005; Ampomah et al., 2008; Schmeisser et al., 2009), suggesting that the trehalose-processing metabolism could be active in bacteroids. Furthermore, the increased trehalose in nodules not only triggered an increase in nodule biomass, but also had a positive impact on trehalose accumulation in leaves. This suggests a systemic effect of trehalose accumulation in nodules, as the plant is only partially genetically modified: only the roots and nodules are transgenic. The trehalose content in leaves was increased by c. 27%, without any detriment to leaf area or leaf DW (net biomass; Fig. 6a,d,e). The increased trehalose in leaves could be a result of the mobilization of this disaccharide from nodules to leaves, as previously reported and suggested to occur in seeds from common bean plants inoculated with Rhizobium (Altamirano-Hernández et al., 2007). Therefore, this could suggest that there is direct bidirectional transport of disaccharides between nodules and leaves in common bean plants during symbiosis. In addition, the accumulation of trehalose caused by down-regulating PvTRE1 in nodules led to an increased proliferation of bacteroids (by almost one order of magnitude), the enhancement of nitrogen fixation (70%) and a higher nodule biomass (threefold). It is worth noting that increased nodule biomass and bacteroid proliferation did not have a negative effect on leaf area or leaf biomass. In supernodulating legumes, a negative impact has been observed on the aerial part of the plant (length and biomass reduced). Interestingly, this increase in nodule number does not lead to higher nitrogen fixation rates (Penmetsa et al., 2003; Schnabel et al., 2010). Thus, PvTRE1 down-regulation in nodules leads to an increase in carbon consumption, reflected in enhanced nodule biomass, and in turn an increased trehalose content in nodules as well as in leaves. Trehalose is known to play a central role in the regulation of sucrose content in plant cells (Lunn, 2008; Paul, 2008; Schluepmann & Paul, 2009). The results obtained in this work suggest that trehalose is able to induce PvSUS1 expression for sucrose degradation, and to regulate the sucrose content in plant cells (Müller et al., 1998; Schluepmann et al., 2003, 2004; Gómez et al., 2006; Ramon & Rolland, 2007). Altogether, the data suggest that by modifying the trehalose content in nodules, it is possible to alter metabolic carbohydrate flux, changing the sink-and-source state of these organs (Schluepmann & Paul, 2009).
Interestingly, changes in the bacteroid number and morphology of the bacteroids inside these nodules, including an edb structure surrounding the bacteroids (Fig. 4i), were observed. These effects could be derived from carbohydrates, as previously reported for another Rhizobium species (Fernandez-Aunión et al., 2010). The modification of the trehalose content in Rhizobium by the overexpression or mutation of a trehalose biosynthesis (OtsA) gene directly affected the distribution of infected cells inside the nodules, and triggered the expression of several genes involved in nitrogen assimilation, carbon assimilation, oxygen transport, and hydrogen peroxide or reactive oxygen species detoxification in common bean during the symbiotic interaction, which gave rise to improvements in grain yield, nitrogen fixation, and stress tolerance (Suárez et al., 2008). These findings strongly suggest that modification of the trehalose content in the nodules triggers physiological alterations that enhance carbon and nitrogen metabolism, as well as bacteroid fitness (greater survival) and nitrogen fixation, which in turn positively impact the symbiotic interaction. Finally, we propose that breeding for the genetic modification of trehalose degradation in nodules could be a valuable mechanism for improving agricultural symbiotic nitrogen fixation.
PvTRE1 down-regulation in common bean nodules was achieved using RNAi technology, giving rise to a 78% increase in trehalose accumulation. The resulting enhancement in nodule biomass and bacterial proliferation strongly supports a central and active role for trehalose in orchestrating carbon metabolism in plant cells. In spite of the nodule biomass increase, there was no detrimental effect on leaf area or leaf biomass. Since PvTRE1 down-regulation has a remarkably positive effect on the symbiosis, this approach could represent a new promising and viable strategy for genetic screening of trehalase mutants that could enhance symbiotic nitrogen fixation.
We thank Olivia Santana, Noreide Nava, Fernando Gonzalez, Maria Luisa Tabche-Barrera, José Alberto Hernández-Eligio, Guadalupe Zavala, Xochitl Alvarado-Affantranger, and Ana Isabel Bieler for technical advice and the use of equipment at the Instituto de Biotecnología and Facultad de Ciencias, Universidad Nacional Autónoma de México, and Paul Gaytán, Jorge Yañéz and Eugenio López for primer synthesis and DNA sequencing at the Instituto de Biotecnología, Universidad Nacional Autónoma de México. This work was supported by grant nos 83324 and 177744 from CONACyT, and PAPIIT no. IN222012. A.B. was supported by a PhD scholarship (169219) from CONACyT.