Protoplasma DOI 10.1007/s00709-015-0780-y
ORIGINAL ARTICLE
Induction of host defences by Rhizobium during ineffective nodulation of pea (Pisum sativum L.) carrying symbiotically defective mutations sym40 (PsEFD), sym33 (PsIPD3/PsCYCLOPS) and sym42 Kira A. Ivanova & Anna V. Tsyganova & Nicholas J. Brewin & Igor A. Tikhonovich & Viktor E. Tsyganov
Received: 21 November 2014 / Accepted: 12 February 2015 # Springer-Verlag Wien 2015
Abstract Rhizobia are able to establish a beneficial interaction with legumes by forming a new organ, called the symbiotic root nodule, which is a unique ecological niche for rhizobial nitrogen fixation. Rhizobial infection has many similarities with pathogenic infection and induction of defence responses accompanies both interactions, but defence responses are induced to a lesser extent during rhizobial infection. However, strong defence responses may result from incompatible interactions between legumes and rhizobia due to a mutation in either macro- or microsymbiont. The aim of this research was to analyse different plant defence reactions in response to Rhizobium infection for several pea (Pisum sativum) mutants that result in ineffective symbiosis. Pea mutants were examined by histochemical and immunocytochemical analyses, light, fluorescence and transmission electron microscopy and quantitative real-time PCR gene expression analysis. It was
Handling Editor: Adrienne R. Hardham
observed that mutations in pea symbiotic genes sym33 (PsIPD3/PsCYCLOPS encoding a transcriptional factor) and sym40 (PsEFD encoding a putative negative regulator of the cytokinin response) led to suberin depositions in ineffective nodules, and in the sym42 there were callose depositions in infection thread (IT) and host cell walls. The increase in deposition of unesterified pectin in IT walls was observed for mutants in the sym33 and sym42; for mutant in the sym42, unesterified pectin was also found around degrading bacteroids. In mutants in the genes sym33 and sym40, an increase in the expression level of a gene encoding peroxidase was observed. In the genes sym40 and sym42, an increase in the expression levels of genes encoding a marker of hypersensitive reaction and PR10 protein was demonstrated. Thus, a range of plant defence responses like suberisation, callose and unesterified pectin deposition as well as activation of defence genes can be triggered by different pea single mutations that cause perception of an otherwise beneficial strain of Rhizobium as a pathogen.
Kira A. Ivanova and Anna V. Tsyganova contributed equally to this work. K. A. Ivanova : A. V. Tsyganova : I. A. Tikhonovich : V. E. Tsyganov (*) All-Russia Research Institute for Agricultural Microbiology, Podbelsky chaussee 3, Saint-Petersburg, Pushkin 8 196608, Russia e-mail: [email protected] N. J. Brewin John Innes Centre, Norwich Research Park, Norwich NR4 7UH, UK I. A. Tikhonovich Saint-Petersburg State University, Universitetskaya embankment 7-9, Saint-Petersburg 199034, Russia
Keywords Pisum sativum . Symbiotic nodule . Defence response . Suberin . Callose . Pectin
Abbreviations IT(s) Infection thread(s) NF(s) Nod factor(s) PR Pathogenesis-related ROS Reactive oxygen species OG(s) Oligogalacturonide(s)
K.A. Ivanova et al.
Introduction Higher plants are continually exposed to a huge variety of different microbes, including potential pathogens. To prevent invasion by these hostile microbes, plants have evolved inducible defence systems. These involve specific defence responses such as the following: cell wall reinforcement by accumulation of polysaccharides, phenols, lignin, suberin; synthesis of phytoalexins, volatiles, hydroxyproline-rich proteins; the induction of synthesis of Bpathogenesis-related^ (PR) proteins, particularly chitinase, glucanase, protease inhibitors and other stress proteins, such as peroxidases; the generation of reactive oxygen species (ROS) and ultimately hypersensitive cell death to confine the microbial invasion to a limited tissue region (Dixon and Paiva 1995; Hammond-Kosack and Jones 1996; Mauch-Mani and Metraux 1998; Ryan 2000; Nakagawa et al. 2011). In contrast to pathogenic interactions, leguminous plants are able to interact beneficially with soil bacteria that are collectively termed rhizobia, in response to specific lipochitooligosaccharides, called Nod factors (NFs). This results in the formation of symbiotic nitrogen-fixing nodules (Brewin 2004). However, the symbiotic interactions between partners at the early stages of nodule development have many similarities with pathogenesis (Vasse et al. 1993). Purified NFs induced the expression of a number of defence-related genes of Lotus japonicus, such as PR gene homologues, peroxidases, chitinases, ERF and WRKY transcription factors, and genes involved in the biosynthesis of pterocarpans (isoflavone derivatives), which are legume-specific phytoalexins (Nakagawa et al. 2011). NFs also transiently induce the generation of ROS in root hair tips (Cardenas et al. 2008). In pea, a chitinase isoenzyme was identified; the activity of which was enhanced after inoculation with rhizobia and application of purified NFs, as well as various stress conditions (Ovtsyna et al. 2005). Nevertheless, it should be noted that the levels of defence-related gene activation by NFs were quite low compared with those induced by chitin oligosaccharides. Bacterial surface polysaccharides are involved in prevention of plant defence responses during development of an effective legume-Rhizobium symbiosis (Fraysse et al. 2003). For example, lipopolysaccharide-defective mutants were only partially successful in colonisation of nodule tissue and were released into only a minority of the host cells. Accumulation of intercellular matrix material and possibly callose, cell wall modification and sporadic cell death in invasive tissue were previously observed (Perotto et al. 1994). Not only bacterial surface polysaccharides, but also membrane structures of the host plant serve to protect the bacteria from direct contact with the cytoplasm and thereby to prevent active defence reaction in nodules (Kannenberg et al. 1994). In order to incorporate rhizobia efficiently, many legumes utilise a structural pathway for invasion, i.e. an infection thread
(IT) (Brewin 2004). In indeterminate nodules, the IT is a tubular structure first initiated in the root hair, where it elongates by cell wall deposition, and it ramifies into the root cortex towards the nodule primordium that developed from the pericycle and inner cortex (Brewin 2004). Investigation of the interactions between plants and pathogens showed that glycoproteins and proline-rich proteins that are similar to matrix components of IT are involved, suggesting that the release of hydrogen peroxide in the extracellular matrix may hamper the penetration of pathogens in host tissue by cross-linking matrix glycoproteins (Bradley et al. 1992). Thus, the localization of similar glycoproteins in the matrix ITs suggests that these components are synthesised as a result of a defensive reaction and their function is to prevent or control invasion of rhizobia in plant tissue. Many aspects of the development of nodules can be considered as an interaction between induced and repressed activities in symbiosomes (a nitrogen-fixing intracellular organelle), which control their progressive development towards eventual senescence. The symbiosome is in a state of dynamic equilibrium, and if the process of nitrogen fixation does not occur, due to bacterial or plant mutations or adverse physiological conditions, bacteroids rapidly senesce, whereupon the host cell dies (Novák et al. 1995; Borisov et al. 1997; Morzhina et al. 2000). The development of an effective symbiotic nodule requires that rhizobia overcome a wide variety of plant defence reactions. During the development of nitrogen-fixing nodules, the synthesis of defence components is relatively low compared to what occurs in a pathogenic state. As a result, the invading microorganisms are not completely inactivated but rather the host defence reactions limit their reproduction and confine their localization to specific compartments provided by the host plant. Studies of ineffective symbiosis are of great interest because they help to identify the barriers being put forward by the plant, which the bacteria must overcome during the normal infection process. These studies provide important clues for unravelling the host regulatory mechanisms of symbiotic nitrogen fixation. In order to study the mechanisms involved in defence responses at the late stages of legume-Rhizobium symbiosis, when symbiotic nodules are formed, a series of single and double pea ineffective mutants was analysed with the obvious morphological defence manifestations, like thickening IT wall, premature degradation of symbiosomes. The series included single mutants in pea symbiotic genes sym33 (PsIPD3/ PsCYCLOPS encoding a transcriptional factor), sym40 (PsEFD encoding a putative negative regulator of the cytokinin response) and sym42 as well as double mutants RBT3 (sym33, sym40) and RBT4 (sym33, sym42). In these mutants, cell wall modifications (suberin, callose and pectin depositions) and defence-related gene expression profiles (7RA84 (peroxidase), HSR203J (a marker of hypersensitive reaction)
Mis-perception of Rhizobium as a pathogen in pea mutants
and ABR17 (PR10 protein)) were analysed in response to invasion by Rhizobium leguminosarum bv. viciae.
Materials and methods Plant material Pea (Pisum sativum L.) ineffective (Fix−) mutants blocked at different stages of nodule development, and corresponding wild-types were used in this study (Table 1).
For immunocytochemical analysis, nodules were harvested at 14 days after inoculation (DAI). For each variant, ten nodules from different plants were analysed. For histochemical analysis, three independent experiments were performed. Nodules were harvested at 21 and 42 DAI. For each experiment, ten nodules per variant from different plants were analysed. Nodules were harvested at 14, 28 and 42 DAI for RNA extraction after washing the root systems in ice-cold water. For each variant, nodules were collected from several (6–10) plants and samples for RNA extraction were frozen in liquid nitrogen.
Bacterial strain
Immunogold labelling
I n a l l ex p e r i m e n t s , pl a n t s w e r e in o c u l a t e d w i t h R. leguminosarum bv. viciae strain 3841 (Wang et al. 1982). Bacteria were cultured on solid nutrient medium TY with streptomycin (600 μg/mL) at 28 °C for 3 days.
The immunogold labelling was described previously (Bradley et al. 1988; Tsyganova et al. 2009b). Briefly, whole nodules were fixed in 2.5 % (v/v) glutaraldehyde in 0.05 M sodium cacodylate, pH 7.2. After overnight incubation at room temperature, nodules were dehydrated in an ethanol series at 35 °C and infiltrated and embedded in London Resin White (Polysciences Europe, Eppelheim, Germany), using benzoin methyl ether as a catalyst for UV polymerisation at -20 °C. For transmission electron microscopy, gold sections, 90–100 nm thick were incubated overnight at 4 °C with primary antibody JIM5 (Knox et al. 1990) (diluted 1/100 in 0.1 % BSA-C). After washing, sections were incubated with secondary antibody, rabbit anti-rat conjugated to 10 nm diameter colloidal gold (Amersham International, Little Chalfont, U.K.), diluted 1/50 in 0.1 % BSA-C in PBS (2.48 g/L NaH2PO4, 21.36 g/L Na2HPO4, 87.66 g/L NaCl, pH 7.4) for 4 h at room temperature. The grids containing sections were counterstained in 2 % (w/v) aqueous uranyl acetate, followed by lead citrate.
Growth conditions and sampling of plant material Seeds were surface-sterilised by treatment for 15 min with 98 % sulphuric acid at room temperature for the lines Finale and RisFixV (sym42), and by a 30-min treatment for lines SGE, SGEFix − -1 (sym40), SGEFix − -2 (sym33), RBT3 (sym33, sym40) and RBT4 (sym33, sym42) and washed with distilled water ten times. The method of surface-sterilisation was chosen on the basis of seed coat properties. Seeds were planted in plastic pots containing 200 mL of vermiculite and 100 mL nutrient solution without nitrogen (Fähraeus 1957). Pots were inoculated with an aqueous suspension of bacterial cells (107–108 cells per seed) at sowing and fertilised every 2 weeks with 100 mL of nutrient solution per pot. Plants were raised in a growth chamber MLR-352H (Sanyo Electric Co., Ltd., Moriguchi, Japan) under controlled conditions: day/ night, 16/8 h; temperature, 21 °C; relative humidity 75 %; photosynthetic photon flux density of ~280 μmol photons m−2 s−1.
Transmission electron microscopy Nodule tissues were viewed and photographed in a JEM– 1200 EM transmission electron microscope (JEOL Ltd., Tokyo, Japan) at 80 kV.
Plant material used in the study
Table 1 Lines SGE −
Phenotype
References
Wild-type
Kosterin and Rozov (1993); Tsyganov et al. (1998)
SGEFix -1 (sym40)
Hypertrophied infection droplet and infection thread formation Tsyganov et al. (1994, 1998); Voroshilova et al. (2009)
SGEFix−-2 (sym33)
Abnormal IT growth inside nodule, no bacterial release
Tsyganov et al. (1994, 1998); Voroshilova et al. (2009)
RBT3 (sym33, sym40) Abnormal IT growth inside nodule, no bacterial release
Tsyganov et al. (2011)
RBT4 (sym33, sym42) Abnormal IT growth inside nodule, no bacterial release
Tsyganov et al. (2014)
Finale
Wild-type
Engvild (1987); Novák et al. (1995); Morzhina et al. (2000)
RisFixV (sym42)
Early senescence of nodule, thickness of ITs
Engvild (1987); Novák et al. (1995); Morzhina et al. (2000); Tsyganov et al. (2001)
K.A. Ivanova et al.
Quantitative analysis of immunogold labelling For statistical analysis, at least 5 different samples of root nodules and at least 20 sectioned walls of infection threads were examined. Morphometrical data were obtained as described by Fernandez-García et al. (2009). Briefly, at least three areas of wall section for each infection thread were evaluated and the number of gold particles per unit area was calculated. The areas and the number of gold particles were measured using software ColorViewII AnalySIS ® (Olympus Soft Imaging Solutions GmbH, Münster, Germany). The data were presented as the number of gold particle/μm2. They were analysed by one-way ANOVA using the software SigmaStat for Windows vesrion 3.5 (Systat Software, Inc, San Jose, California, USA). Means were separated by the Tukey multiple range test (P≤0.001).
To detect callose deposits, sections were washed twice for 10 min with 0.1 M potassium phosphate buffer (pH 8.0). They were stained with 0.1 % Aniline Blue in a potassium phosphate buffer (pH 8.0) for 60 min, washed three times for 10 min in potassium phosphate buffer (Currier and Strugger 1995) and embedded in ProLong Gold® antifade reagent. Light and fluorescent microscopy Analysis was performed on a microscope Axio Imager.Z1 (Carl Zeiss GmbH, Jena, Germany) using bright field, differential interference contrast (DIC) and fluorescence microscopy (filter DAPI). Photos were taken using a digital video camera ColorViewII and software ColorViewII AnalySIS ® (Olympus Soft Imaging Solutions GmbH, Münster, Germany). RNA extraction from plant root nodules
Histochemical staining Nodules were placed into glass vials containing a fixative made of 3 % paraformaldehyde and 0.25 % glutaraldehyde, to which was added 0.3 % Triton X-100 and 0.3 % Tween 20 in diluted 3-fold stabilising buffer MTSB (50 mM PIPES, 5 mM EGTA, 5 mM MgSO4.7H2O, pH 6.9). They were vacuum-infiltrated in a desiccator three times for 5 min with 10-min intervals and then placed at 4 °C overnight. After a rinse in MTSB (three times × 20 min), nodules were dehydrated at room temperature in a series of increasing ethanol concentrations till 96 % (v/v). Samples were then stained with Toluidine Blue (0.1 % in 96 % ethanol) overnight at 4 °C, and rinsed with 96 % ethanol (2×30 min). Nodules were placed in a series of Stidman’s wax+ethanol mixture (from 10 to 100 % of Steedman’s wax (v/v)) at 37 °C (Vitha et al. 1997). Finally, they were embedded in pure wax (37 °C) using a hot plate Präzitherm PZ 28–1 (Harry Gestigkeit GmbH, Düsseldorf, Germany). Sectioning (10–15 μm) was performed at room temperature using a microtome HM360 (Microm International GmbH, Walldorf, Germany). Sections were placed on a slide and were de-waxed in 96 % ethanol twice for 40 min, in 70 % for 10 min, and in 40 % for 10 min. To detect suberin deposits, sections were washed with buffer 1× TBS (50 mM Tris-HCI, 150 mM NaCl, pH 7.4) two times for 10 min. They were stained with Toluidine Blue (0.5 % in 1× TBS buffer) for 45 min, washed in buffer TBS (three times for 10 min), post-stained with 0.1 % Neutral Red in 0.1 M K2PO4 (pH 6.5) for 1 min, washed three times with buffer TBS for 10 min, and then embedded in ProLong Gold® antifade reagent (Life Technologies, Grand Island, NY, USA). The Neutral Red technique appears to be specific for the hydrophobic/lipid domain of suberin (Lulai and Morgan 1992).
Samples were ground into a fine powder in ceramic mortars in liquid nitrogen and transferred into 1.5-mL Eppendorf tubes. Total RNA was isolated from nodules (0.1–0.15 g) using PURE-ZOL reagent (Bio-Rad, Hercules, California, USA) according to the manufacturer’s protocol. Samples were collected from different plants in order to minimise the individual plant variation in gene expression. The integrity of the total RNA was checked by electrophoresis on a 1.5 % standard agarose gel, and its quantity as well as purity were determined using a microchip electrophoresis system for the study of nucleic acids MultiNA (Shimadzu Corporation, Kyoto, Japan). All RNA samples were then adjusted to the same concentration, measured and adjusted again in order to homogenise RNA input in the subsequent reverse transcription reaction. Reverse transcription-PCR (RT-PCR) cDNA was synthesised from 1.5 μg DNA-free RNA using 200 U RevertAid ™ M-MuLV Reverse Transcriptase, 4 μL 5× RT buffer, 1 μg oligo(dT)18 primer, 40 U RiboLock ™ RNase Inhibitor and 0.5 μL 25 mM dNTPs (MBI Fermentas, Lithuania) and DEPC water. Reverse transcription reactions were carried out in a volume of 20 μL and the resulting cDNAs were diluted five times for use in real-time PCR. Temperature conditions were 40 °C for 60 min and 70 °C for 10 min. All cDNA samples were stored at −20 °C until used for PCR. The reaction was carried out in an automated thermocycler C1000™ Thermal Cycler (Bio-Rad). Real-time PCR In the first stage of the study, four reference genes were chosen based on their previous use as internal controls in plant gene expression studies. Our results showed that glyceraldehyde 3-
Mis-perception of Rhizobium as a pathogen in pea mutants
phosphate dehydrogenase (GAPC1) was the most stably expressed gene. Primers were designed with the following criteria: Tm of 56 ±2 °C and PCR amplicon lengths of 60–400 bp, yielding primer sequences with lengths of 18–25 nucleotides and GC contents of 45–65 % (Table 2). Polymerase chain reactions were performed in a 96-well plate with a CFX96™ real-time PCR System combined with a C1000™ thermal cycler using SYBR Green (Bio-Rad) to monitor dsDNA synthesis. Reactions contained 5 μL 2x iQTM SYBR® Green Supermix, 3.4 μL H2O, 0.8 μL of cDNA and 0.4 μL of each gene-specific primer (10 μM) in a final volume of 10 μL. The following standard thermal profile was used: polymerase activation (95 °C for 3 min), amplification and quantification cycles repeated 40 times (95 °C for 30 s, 56 °C for 30 s, 72 °C for 40 s). The specificity of the amplicons was checked by electrophoresis in 2 % (w/v) agarose gel and a melting curve analysis performed by the PCR machine after 40 amplification cycles (55°–95 °C with one measurement of fluorescence every 0.5 °C). All investigated qPCR products showed only single peaks and there were no primer-dimer peaks or other artefacts. Negative control with a quantification cycle (Cq) less than 38 was not observed. Two biological repetitions were used for the measurement, and two technical replicates were analysed for each biological repetition. Data analysis Data were analysed using the CFX Manager™ software version 2.0. Primer efficiency values with an R2 value less than 0.997 were ignored. The expression levels of the gene of interest (GOI) relative to the GAPC1 were calculated for each cDNA sample using the PCR efficiency and the equation: relative ratio ¼ E
ΔC control sample –ΔC test sample T T
The values of samples were used in a Student’s t test and Dunnett’s criterion to calculate probabilities of distinct Table 2
induction or repression, and the average ratio of these values was used to determine the fold change in transcript level in mutant lines compared with parental lines at the different stages of nodule development. Student’s t test was used for pairwise comparisons and Dunnett’s test for comparison of many experimental groups with one control.
Results Immunocytological localisation of unesterified pectin in IT wall in the symbiotic nodules of ineffective pea mutants The composition of the IT wall in 14-day-old pea wild-type and mutant nodules was studied using antibodies JIM5 specific for unesterified pectin. In nodules of wild-type SGE antibody JIM5 labelled uniformly the IT wall (Fig. 1a, b). Similar distribution was characteristic for wild-type Finale and mutant SGEFix−-1 (sym40) (data not shown). In nodules of the mutant line SGEFix−-2 (sym33) (Fig. 1c, d) and double mutant RBT3 (sym33, sym40) (Fig. 1e) the thickened ITs were extensively labelled by JIM5 in comparison with SGE (Table 3). In mutant RisFixV (sym42) JIM5 labelled the IT wall (Table 3) but label was unevenly distributed with many obvious clusters (Fig. 1g, h). In the double mutant RBT4 (sym33, sym42), the distribution of JIM5 labelling was similar with SGEFix−-2 (sym33) and double mutant RBT3 (sym33, sym40) (Fig. 1f; Table 3). The additional abnormality in JIM5 localisation was observed in RisFixV (sym42) mutant, when degrading bacteroids in zone IV were surrounded by deposition of unesterified pectins (Fig. 1i). Histochemical localisation of suberin and callose depositions in the symbiotic nodules of ineffective pea mutants Histological organisation in 21-day-old nodules of the parental pea line SGE was typical for wild-type nodules, including a meristem, a zone of infection and a nitrogen-fixation zone (Fig. 2a) (Tsyganov et al. 1998; Voroshilova et al. 2009).
Description of primer sequences for qPCR
Accession no.
Description
Primer sequence (5′–3′)
PCR product size (bp)
Reference
L07500.1
Glyceraldehyde-3-phosphate dehydrogenase (GapC1) Peroxidase 7RA84
222
AAGAACGACGAACTCACCG TTGGCACCACCCTTCAAATG 89 TGTTTGAATCAGATGCTGCATTG 140 CATTTGATTGAAGATGTTGTGCAA
188
Martin et al. (1993)
75
Pérez-de-Luque et al. (2006); Die et al. (2009)
a
Hypersensitive reaction 203 J
70
Die et al. (2007, 2009)
Z15128.1
ABA-responsive protein
TGTTTGAATCAGATGCTGCATTG CA TTTGATTGAAGATGTTGTGCAA 21 TGGGTGTCTTTGTTTTTGATGATGA 436 TATGGCCTTGATAAGTCCAGTTCCT
440
Iturriaga et al. (1994)
AF396465 TC107779
a
– Identifier in the TIGR M. truncatula Gene Index
390
K.A. Ivanova et al. Fig. 1 Immunogold labelling of ultrathin sections through infection zone of pea wild-type and mutant nodules. Infection thread wall (a– h) and pectin depositions around bacteroid (i) labelled by JIM5 antibody (arrow), which is specific for unesterified pectin. IT – infection thread, ITW – infection thread wall, ba – bacterium, Ba – bacteroid. a Wild-type SGE. b High magnification of the boxed area in (a). c Mutant SGEFix−-2 (sym33). d High magnification of the boxed area in (c). e Double mutant RBT3 (sym33, sym40). f Double mutant RBT4 (sym33, sym42). g, i Mutant RisFixV (sym42). h High magnification of the boxed area in (g). Bar= 500 nm (a, c, e, f, g, i), bar= 250 nm (b, d, h)
Additionally, in 42-day-old nodules a senescence zone was observed below the nitrogen-fixation zone (data not shown). Suberin depositions in the 21- and 42-day-old nodules were only identified in the nodule endodermis, which separates the inner from the outer layers of the nodule cortex in postmeristematic regions (Fig. 2b). Nodules of cv. Finale were similar to those of SGE with respect to suberin deposition in the nodule endodermis (data not shown). In 21- and 42-day-old white nodules of the mutant line SGEFix−-1 (sym40) (Fig. 2c, e) suberised cells were not observed in the central tissue. However, there were suberin depositions in cell walls of both the nodule endodermis and the vascular endodermis (Fig. 2f), and the nodule endodermis enclosed the tip of the nodule, leading to the isolation of the
entire nodule from the rest of the plant (Fig. 2d). An agedependent increase of suberin accumulation in the cell walls was not revealed. In 21-day-old white nodules of the mutant line SGEFix−-2 (sym33) (Fig. 3a) suberin deposits were observed in the walls of host cells that were traversed by ITs (Fig. 3b, d). Suberin also accumulated in the thickened walls of the ITs (Fig. 3c, d). The increase of suberisation was observed in 42-day-old nodules (data not shown). In 21-day-old nodules of the double mutant line RBT3 (sym33, sym40), suberin distribution was similar to that of the mutant line SGEFix−-2 (sym33) and it was also increased in 42-day-old nodules (data not shown). In 21-day-old nodules of the mutant line RisFixV (sym42) suberin deposits in
Mis-perception of Rhizobium as a pathogen in pea mutants Table 3 Distribution of gold particles in infection thread walls in pea wild-type and mutant lines Genotype
Mean value
SE
SGE SGEFix−-1 (sym40) SGEFix−-2 (sym33) RBT3 (sym33, sym40)
139a 112a 506b 569b
20.6 18.8 25.0 39.4
depositions were revealed in the central tissues of 21- and 42day-old nodules of the mutant line RisFixV (sym42) (Fig. 4c, d). Callose accumulated in the walls of mature ITs and in the cell walls (Fig. 4 e, f). Changes associated with age of the nodule were not found. In the other mutant lines, callose was not found at all (Fig. 4g, h) or was registered only in the form of small deposits in phloem elements in 21- and 42-day-old nodules (Fig. 4 i).
RBT4 (sym33, sym42) Finale RisFixV (sym42)
575b 157 a 231c
35.7 22.4 23.7
Real-time PCR for 7RA84 (peroxidase), HSR203J (a marker of hypersensitive reaction) and ABR17 (PR10 protein)
Results are presented as the number of gold particles/μm2 Mean value±SE (n=20-25) are shown. Means without a common letter are significantly different by the Tukey multiple range test (P≤0.001)
the thickened walls of ITs were not observed. Suberin was found only in the nodule endodermis, as in the nodules of the wild-type Finale (data not shown). Callose distribution was also analysed in order to investigate the composition of thickened cell walls surrounding mature ITs in greenish nodules of the mutant line RisFixV (sym42). In contrast to wild-type Finale (Fig. 4 a, b), callose Fig. 2 Histochemical localization of suberin deposition in the 21-day-old nodules of the wild-type SGE (a, b) and mutant SGEFix−-1 (sym40) (c–f). Sections stained with Neutral Red. I – nodule meristem, II – zone of infection, III – nitrogen-fixation zone, indicated where they are identified, arrows indicates nodule endodermis, arrowheads – vascular endodermis, * – vascular bundles, Bar=200 μm. a, c, e DIC (differential interference contrast), b, d, f fluorescence microscopy. a–d Sagittal sections, e, f transverse sections made in the middle of nodule, in the region corresponding to infection zone in wild-type
The expression of three genes known to be involved in host defence responses was monitored by real-time PCR: 7RA84 (peroxidase), HSR203J (a marker of hypersensitive reaction) and ABR17 (PR10 protein). For the gene 7RA84, the maximum level of expression was observed at 14 DAI in wild-type and in all mutant nodules examined (Fig. 5a). However, in the mutant lines SGEFix−-1 (sym40) and SGEFix−-2 (sym33), the levels of expression were significantly higher (at least 2.3fold) than in wild type SGE for each day of analysis (Fig. 5a). In nodules of wild-type Finale and the mutant line RisFixV (sym42), the levels of 7RA84 expression were similar
K.A. Ivanova et al. Fig. 3 Histochemical localisation of suberin depositions in a 21-day-old nodule of the mutant SGEFix−-2 (sym33). Sections stained with Neutral Red. I – nodule meristem, arrowheads mark the infection threads, arrows indicate suberised walls of infection threads. Note suberised cell walls. Bar=200 μm (a, b), bar=20 μm (c, d). a, c DIC (differential interference contrast), b, d fluorescence microscopy. a, b Sagittal sections
at 14 and 28 DAI but significantly different (5-fold higher) at 42 DAI. The level of expression of the gene HSR203J in nodules of the mutant line SGEFix−-1 (sym40) was significantly higher (at least 2.5-fold) than in nodules of wild-type SGE at all stages analysed (Fig. 5b). In the mutant line SGEFix−-2 (sym33), the level of expression was significantly increased at 28 and 42 DAI in comparison with wild-type SGE (at least 2.5-fold). In the mutant line RisFixV (sym42), the timedependent increase in the level of expression HSR203J in nodules was observed and the levels of expression were significantly higher (at least 8.1-fold) than in wild-type Finale at 28 and 42 DAI (Fig. 5b). The level of expression of the gene ABR17 was significantly higher in the mutant lines SGEFix−-1 (sym40) than in nodules of wild-type SGE at 14 and 28 DAI. The highest level of expression was at 28 DAI in this mutant line; this was also true in the line SGEFix−-2 (sym33) in contrast to the wild-type SGE (14.2- and 5-fold respectively) in which some decrease in the level of expression was observed at this stage (Fig. 5c). Interestingly, the level of expression of this gene in mutant line SGEFix−-2 (sym33) was significantly lower than with wildtype SGE (3.6-fold) at 42 DAI. In the mutant line RisFixV (sym42), the time-dependent increase in the level of expression of the gene ABR17 was observed and the levels of expression were significantly higher than in wild-type Finale at 28 and 42 DAI (6- and 3.8-fold respectively) (Fig. 5c).
Discussion Pectin is one of the most abundant components of a cell wall and different modifications of pectin occur during nodule
development as well as during pathogenic infection. In legumes, a pectate lyase was recently identified that is inducible by NFs and apparently involved in the penetration of rhizobia in IT (Xie et al. 2012). The involvement of pectin in defence responses probably involves the production of oligogalacturonides (OGs), which are the result of degradation of homogalacturonan by the action of polygalacturonases and pectate lyases, produced by both plants and microbes. OGs trigger calcium-mediating ROS production and activation of defence genes (Ferrari et al. 2013). In the legume-Rhizobium symbiosis, OGs inhibit flavonoid-induced nod gene expression, which could be important for fine-tuning of the NF concentration during IT growth (Moscatiello et al. 2012). However, pectin may be involved not only in signalling during defence responses, but it may also participate together with other components in cell wall modification. Structural reinforcement of cell walls is one of the most prominent manifestations of defence responses triggered by pathogens (Hammond-Kosack and Jones 1996). Different components like lignin, suberin, callose and pectin have been shown to be involved in wall cross-linking (Matern et al. 1995). The modification of pectin, caused by the activity of pectin methylesterases dramatically alters the physical properties of demethylesterified forms, causing an increase in cell wall rigidity and the concomitant formation of gel-structures to support damaged walls (Pelloux et al. 2007; Wolf and Greiner 2012; Bethke et al. 2014). In this study, the high labelling of the IT wall by JIM5, which labels unesterified pectins, was observed for single and double mutants carrying a mutation in the gene sym33. For mutant RisFixV (sym42) JIM5 labelling was unevenly dispersed in the IT wall with some clusters of intense labelling. The other striking feature of the RisFixV (sym42) was the presence of JIM5 label around degrading
Mis-perception of Rhizobium as a pathogen in pea mutants Fig. 4 Histochemical localisation of callose depositions in the 21-day-old nodules of the wild type Finale (a, b), mutant RisFixV (sym42) (c–f) and mutant SGEFix−-2 (sym33) (g–i). Sections stained with Aniline Blue. I – nodule meristem, II – zone of infection, III – nitrogen-fixation zone, IV – senescence zone indicated where they are identified, ic – infected cell. Arrows indicate callose depositions around infection threads, arrowheads in cell wall, * in vascular bundles. h Note autofluorescence of suberised cell walls and walls of infection threads in mutant SGEFix−-2 (sym33). Bar=200 μm (a–d, g, h), bar=20 μm (e, f), bar=50 μm (i). a, c, e, g DIC (differential interference contrast), b, d, f, h, i fluorescence microscopy. a–d, g, h Sagittal sections
bacteroids, indicating that unesterified pectin encapsulated ineffective bacteroids. To our knowledge, this phenotype has never been observed previously in pea wild-type nodules (Rae et al. 1992) or in mutants; however, the reason for this observed encapsulation of degrading bacteroids is still unknown. Suberin deposits in the nodule endodermis of wild-type SGE and Finale nodules are physiologically important for nodule protection from excessive dehydration, salt stress and the penetration of pathogens (Schreiber et al. 1999). The intense level of suberisation observed in nodules of the mutant line SGEFix−-2 (sym33) is an example of a plant defence response which prevents the release of bacteria into the plant cytoplasm and their subsequent differentiation into bacteroids.
These host defence responses indicate that the mutation in the pea gene sym33 leads to the perception of nodule bacteria not like microsymbionts, but as pathogens. Previously, it was shown that the accumulation of polyphenolic substances in walls of thickened ITs is typical for M. truncatula mutant TE7 in the gene Mtsym1 (Benaben et al. 1995). Recently, it was shown that Sym33 and MtSym1 are orthologues, encoding a protein IPD3, which interacts with calcium-dependent and calmodulin-dependent kinase CCaMK, activating the expression of a nodule-specific remorin, which is necessary for the growth and formation of ITs and symbiosomes (Ovchinnikova et al. 2011). In L. japonicus, the ortholog of Sym33 and MtIPD is LjCYCLOPS, which encodes a transcriptional factor that
K.A. Ivanova et al.
Fig. 5 Gene expression dynamics: 7RA84 (peroxidase), Hsr203J (a marker of hypersensitivity reactions), ABR17 (PR10 protein) in nodules of pea wild types SGE, Finale and corresponding mutants. * – significant differences from the wild-types of corresponding mutants within one time-period (t test, p
ORIGINAL ARTICLE
Induction of host defences by Rhizobium during ineffective nodulation of pea (Pisum sativum L.) carrying symbiotically defective mutations sym40 (PsEFD), sym33 (PsIPD3/PsCYCLOPS) and sym42 Kira A. Ivanova & Anna V. Tsyganova & Nicholas J. Brewin & Igor A. Tikhonovich & Viktor E. Tsyganov
Received: 21 November 2014 / Accepted: 12 February 2015 # Springer-Verlag Wien 2015
Abstract Rhizobia are able to establish a beneficial interaction with legumes by forming a new organ, called the symbiotic root nodule, which is a unique ecological niche for rhizobial nitrogen fixation. Rhizobial infection has many similarities with pathogenic infection and induction of defence responses accompanies both interactions, but defence responses are induced to a lesser extent during rhizobial infection. However, strong defence responses may result from incompatible interactions between legumes and rhizobia due to a mutation in either macro- or microsymbiont. The aim of this research was to analyse different plant defence reactions in response to Rhizobium infection for several pea (Pisum sativum) mutants that result in ineffective symbiosis. Pea mutants were examined by histochemical and immunocytochemical analyses, light, fluorescence and transmission electron microscopy and quantitative real-time PCR gene expression analysis. It was
Handling Editor: Adrienne R. Hardham
observed that mutations in pea symbiotic genes sym33 (PsIPD3/PsCYCLOPS encoding a transcriptional factor) and sym40 (PsEFD encoding a putative negative regulator of the cytokinin response) led to suberin depositions in ineffective nodules, and in the sym42 there were callose depositions in infection thread (IT) and host cell walls. The increase in deposition of unesterified pectin in IT walls was observed for mutants in the sym33 and sym42; for mutant in the sym42, unesterified pectin was also found around degrading bacteroids. In mutants in the genes sym33 and sym40, an increase in the expression level of a gene encoding peroxidase was observed. In the genes sym40 and sym42, an increase in the expression levels of genes encoding a marker of hypersensitive reaction and PR10 protein was demonstrated. Thus, a range of plant defence responses like suberisation, callose and unesterified pectin deposition as well as activation of defence genes can be triggered by different pea single mutations that cause perception of an otherwise beneficial strain of Rhizobium as a pathogen.
Kira A. Ivanova and Anna V. Tsyganova contributed equally to this work. K. A. Ivanova : A. V. Tsyganova : I. A. Tikhonovich : V. E. Tsyganov (*) All-Russia Research Institute for Agricultural Microbiology, Podbelsky chaussee 3, Saint-Petersburg, Pushkin 8 196608, Russia e-mail: [email protected] N. J. Brewin John Innes Centre, Norwich Research Park, Norwich NR4 7UH, UK I. A. Tikhonovich Saint-Petersburg State University, Universitetskaya embankment 7-9, Saint-Petersburg 199034, Russia
Keywords Pisum sativum . Symbiotic nodule . Defence response . Suberin . Callose . Pectin
Abbreviations IT(s) Infection thread(s) NF(s) Nod factor(s) PR Pathogenesis-related ROS Reactive oxygen species OG(s) Oligogalacturonide(s)
K.A. Ivanova et al.
Introduction Higher plants are continually exposed to a huge variety of different microbes, including potential pathogens. To prevent invasion by these hostile microbes, plants have evolved inducible defence systems. These involve specific defence responses such as the following: cell wall reinforcement by accumulation of polysaccharides, phenols, lignin, suberin; synthesis of phytoalexins, volatiles, hydroxyproline-rich proteins; the induction of synthesis of Bpathogenesis-related^ (PR) proteins, particularly chitinase, glucanase, protease inhibitors and other stress proteins, such as peroxidases; the generation of reactive oxygen species (ROS) and ultimately hypersensitive cell death to confine the microbial invasion to a limited tissue region (Dixon and Paiva 1995; Hammond-Kosack and Jones 1996; Mauch-Mani and Metraux 1998; Ryan 2000; Nakagawa et al. 2011). In contrast to pathogenic interactions, leguminous plants are able to interact beneficially with soil bacteria that are collectively termed rhizobia, in response to specific lipochitooligosaccharides, called Nod factors (NFs). This results in the formation of symbiotic nitrogen-fixing nodules (Brewin 2004). However, the symbiotic interactions between partners at the early stages of nodule development have many similarities with pathogenesis (Vasse et al. 1993). Purified NFs induced the expression of a number of defence-related genes of Lotus japonicus, such as PR gene homologues, peroxidases, chitinases, ERF and WRKY transcription factors, and genes involved in the biosynthesis of pterocarpans (isoflavone derivatives), which are legume-specific phytoalexins (Nakagawa et al. 2011). NFs also transiently induce the generation of ROS in root hair tips (Cardenas et al. 2008). In pea, a chitinase isoenzyme was identified; the activity of which was enhanced after inoculation with rhizobia and application of purified NFs, as well as various stress conditions (Ovtsyna et al. 2005). Nevertheless, it should be noted that the levels of defence-related gene activation by NFs were quite low compared with those induced by chitin oligosaccharides. Bacterial surface polysaccharides are involved in prevention of plant defence responses during development of an effective legume-Rhizobium symbiosis (Fraysse et al. 2003). For example, lipopolysaccharide-defective mutants were only partially successful in colonisation of nodule tissue and were released into only a minority of the host cells. Accumulation of intercellular matrix material and possibly callose, cell wall modification and sporadic cell death in invasive tissue were previously observed (Perotto et al. 1994). Not only bacterial surface polysaccharides, but also membrane structures of the host plant serve to protect the bacteria from direct contact with the cytoplasm and thereby to prevent active defence reaction in nodules (Kannenberg et al. 1994). In order to incorporate rhizobia efficiently, many legumes utilise a structural pathway for invasion, i.e. an infection thread
(IT) (Brewin 2004). In indeterminate nodules, the IT is a tubular structure first initiated in the root hair, where it elongates by cell wall deposition, and it ramifies into the root cortex towards the nodule primordium that developed from the pericycle and inner cortex (Brewin 2004). Investigation of the interactions between plants and pathogens showed that glycoproteins and proline-rich proteins that are similar to matrix components of IT are involved, suggesting that the release of hydrogen peroxide in the extracellular matrix may hamper the penetration of pathogens in host tissue by cross-linking matrix glycoproteins (Bradley et al. 1992). Thus, the localization of similar glycoproteins in the matrix ITs suggests that these components are synthesised as a result of a defensive reaction and their function is to prevent or control invasion of rhizobia in plant tissue. Many aspects of the development of nodules can be considered as an interaction between induced and repressed activities in symbiosomes (a nitrogen-fixing intracellular organelle), which control their progressive development towards eventual senescence. The symbiosome is in a state of dynamic equilibrium, and if the process of nitrogen fixation does not occur, due to bacterial or plant mutations or adverse physiological conditions, bacteroids rapidly senesce, whereupon the host cell dies (Novák et al. 1995; Borisov et al. 1997; Morzhina et al. 2000). The development of an effective symbiotic nodule requires that rhizobia overcome a wide variety of plant defence reactions. During the development of nitrogen-fixing nodules, the synthesis of defence components is relatively low compared to what occurs in a pathogenic state. As a result, the invading microorganisms are not completely inactivated but rather the host defence reactions limit their reproduction and confine their localization to specific compartments provided by the host plant. Studies of ineffective symbiosis are of great interest because they help to identify the barriers being put forward by the plant, which the bacteria must overcome during the normal infection process. These studies provide important clues for unravelling the host regulatory mechanisms of symbiotic nitrogen fixation. In order to study the mechanisms involved in defence responses at the late stages of legume-Rhizobium symbiosis, when symbiotic nodules are formed, a series of single and double pea ineffective mutants was analysed with the obvious morphological defence manifestations, like thickening IT wall, premature degradation of symbiosomes. The series included single mutants in pea symbiotic genes sym33 (PsIPD3/ PsCYCLOPS encoding a transcriptional factor), sym40 (PsEFD encoding a putative negative regulator of the cytokinin response) and sym42 as well as double mutants RBT3 (sym33, sym40) and RBT4 (sym33, sym42). In these mutants, cell wall modifications (suberin, callose and pectin depositions) and defence-related gene expression profiles (7RA84 (peroxidase), HSR203J (a marker of hypersensitive reaction)
Mis-perception of Rhizobium as a pathogen in pea mutants
and ABR17 (PR10 protein)) were analysed in response to invasion by Rhizobium leguminosarum bv. viciae.
Materials and methods Plant material Pea (Pisum sativum L.) ineffective (Fix−) mutants blocked at different stages of nodule development, and corresponding wild-types were used in this study (Table 1).
For immunocytochemical analysis, nodules were harvested at 14 days after inoculation (DAI). For each variant, ten nodules from different plants were analysed. For histochemical analysis, three independent experiments were performed. Nodules were harvested at 21 and 42 DAI. For each experiment, ten nodules per variant from different plants were analysed. Nodules were harvested at 14, 28 and 42 DAI for RNA extraction after washing the root systems in ice-cold water. For each variant, nodules were collected from several (6–10) plants and samples for RNA extraction were frozen in liquid nitrogen.
Bacterial strain
Immunogold labelling
I n a l l ex p e r i m e n t s , pl a n t s w e r e in o c u l a t e d w i t h R. leguminosarum bv. viciae strain 3841 (Wang et al. 1982). Bacteria were cultured on solid nutrient medium TY with streptomycin (600 μg/mL) at 28 °C for 3 days.
The immunogold labelling was described previously (Bradley et al. 1988; Tsyganova et al. 2009b). Briefly, whole nodules were fixed in 2.5 % (v/v) glutaraldehyde in 0.05 M sodium cacodylate, pH 7.2. After overnight incubation at room temperature, nodules were dehydrated in an ethanol series at 35 °C and infiltrated and embedded in London Resin White (Polysciences Europe, Eppelheim, Germany), using benzoin methyl ether as a catalyst for UV polymerisation at -20 °C. For transmission electron microscopy, gold sections, 90–100 nm thick were incubated overnight at 4 °C with primary antibody JIM5 (Knox et al. 1990) (diluted 1/100 in 0.1 % BSA-C). After washing, sections were incubated with secondary antibody, rabbit anti-rat conjugated to 10 nm diameter colloidal gold (Amersham International, Little Chalfont, U.K.), diluted 1/50 in 0.1 % BSA-C in PBS (2.48 g/L NaH2PO4, 21.36 g/L Na2HPO4, 87.66 g/L NaCl, pH 7.4) for 4 h at room temperature. The grids containing sections were counterstained in 2 % (w/v) aqueous uranyl acetate, followed by lead citrate.
Growth conditions and sampling of plant material Seeds were surface-sterilised by treatment for 15 min with 98 % sulphuric acid at room temperature for the lines Finale and RisFixV (sym42), and by a 30-min treatment for lines SGE, SGEFix − -1 (sym40), SGEFix − -2 (sym33), RBT3 (sym33, sym40) and RBT4 (sym33, sym42) and washed with distilled water ten times. The method of surface-sterilisation was chosen on the basis of seed coat properties. Seeds were planted in plastic pots containing 200 mL of vermiculite and 100 mL nutrient solution without nitrogen (Fähraeus 1957). Pots were inoculated with an aqueous suspension of bacterial cells (107–108 cells per seed) at sowing and fertilised every 2 weeks with 100 mL of nutrient solution per pot. Plants were raised in a growth chamber MLR-352H (Sanyo Electric Co., Ltd., Moriguchi, Japan) under controlled conditions: day/ night, 16/8 h; temperature, 21 °C; relative humidity 75 %; photosynthetic photon flux density of ~280 μmol photons m−2 s−1.
Transmission electron microscopy Nodule tissues were viewed and photographed in a JEM– 1200 EM transmission electron microscope (JEOL Ltd., Tokyo, Japan) at 80 kV.
Plant material used in the study
Table 1 Lines SGE −
Phenotype
References
Wild-type
Kosterin and Rozov (1993); Tsyganov et al. (1998)
SGEFix -1 (sym40)
Hypertrophied infection droplet and infection thread formation Tsyganov et al. (1994, 1998); Voroshilova et al. (2009)
SGEFix−-2 (sym33)
Abnormal IT growth inside nodule, no bacterial release
Tsyganov et al. (1994, 1998); Voroshilova et al. (2009)
RBT3 (sym33, sym40) Abnormal IT growth inside nodule, no bacterial release
Tsyganov et al. (2011)
RBT4 (sym33, sym42) Abnormal IT growth inside nodule, no bacterial release
Tsyganov et al. (2014)
Finale
Wild-type
Engvild (1987); Novák et al. (1995); Morzhina et al. (2000)
RisFixV (sym42)
Early senescence of nodule, thickness of ITs
Engvild (1987); Novák et al. (1995); Morzhina et al. (2000); Tsyganov et al. (2001)
K.A. Ivanova et al.
Quantitative analysis of immunogold labelling For statistical analysis, at least 5 different samples of root nodules and at least 20 sectioned walls of infection threads were examined. Morphometrical data were obtained as described by Fernandez-García et al. (2009). Briefly, at least three areas of wall section for each infection thread were evaluated and the number of gold particles per unit area was calculated. The areas and the number of gold particles were measured using software ColorViewII AnalySIS ® (Olympus Soft Imaging Solutions GmbH, Münster, Germany). The data were presented as the number of gold particle/μm2. They were analysed by one-way ANOVA using the software SigmaStat for Windows vesrion 3.5 (Systat Software, Inc, San Jose, California, USA). Means were separated by the Tukey multiple range test (P≤0.001).
To detect callose deposits, sections were washed twice for 10 min with 0.1 M potassium phosphate buffer (pH 8.0). They were stained with 0.1 % Aniline Blue in a potassium phosphate buffer (pH 8.0) for 60 min, washed three times for 10 min in potassium phosphate buffer (Currier and Strugger 1995) and embedded in ProLong Gold® antifade reagent. Light and fluorescent microscopy Analysis was performed on a microscope Axio Imager.Z1 (Carl Zeiss GmbH, Jena, Germany) using bright field, differential interference contrast (DIC) and fluorescence microscopy (filter DAPI). Photos were taken using a digital video camera ColorViewII and software ColorViewII AnalySIS ® (Olympus Soft Imaging Solutions GmbH, Münster, Germany). RNA extraction from plant root nodules
Histochemical staining Nodules were placed into glass vials containing a fixative made of 3 % paraformaldehyde and 0.25 % glutaraldehyde, to which was added 0.3 % Triton X-100 and 0.3 % Tween 20 in diluted 3-fold stabilising buffer MTSB (50 mM PIPES, 5 mM EGTA, 5 mM MgSO4.7H2O, pH 6.9). They were vacuum-infiltrated in a desiccator three times for 5 min with 10-min intervals and then placed at 4 °C overnight. After a rinse in MTSB (three times × 20 min), nodules were dehydrated at room temperature in a series of increasing ethanol concentrations till 96 % (v/v). Samples were then stained with Toluidine Blue (0.1 % in 96 % ethanol) overnight at 4 °C, and rinsed with 96 % ethanol (2×30 min). Nodules were placed in a series of Stidman’s wax+ethanol mixture (from 10 to 100 % of Steedman’s wax (v/v)) at 37 °C (Vitha et al. 1997). Finally, they were embedded in pure wax (37 °C) using a hot plate Präzitherm PZ 28–1 (Harry Gestigkeit GmbH, Düsseldorf, Germany). Sectioning (10–15 μm) was performed at room temperature using a microtome HM360 (Microm International GmbH, Walldorf, Germany). Sections were placed on a slide and were de-waxed in 96 % ethanol twice for 40 min, in 70 % for 10 min, and in 40 % for 10 min. To detect suberin deposits, sections were washed with buffer 1× TBS (50 mM Tris-HCI, 150 mM NaCl, pH 7.4) two times for 10 min. They were stained with Toluidine Blue (0.5 % in 1× TBS buffer) for 45 min, washed in buffer TBS (three times for 10 min), post-stained with 0.1 % Neutral Red in 0.1 M K2PO4 (pH 6.5) for 1 min, washed three times with buffer TBS for 10 min, and then embedded in ProLong Gold® antifade reagent (Life Technologies, Grand Island, NY, USA). The Neutral Red technique appears to be specific for the hydrophobic/lipid domain of suberin (Lulai and Morgan 1992).
Samples were ground into a fine powder in ceramic mortars in liquid nitrogen and transferred into 1.5-mL Eppendorf tubes. Total RNA was isolated from nodules (0.1–0.15 g) using PURE-ZOL reagent (Bio-Rad, Hercules, California, USA) according to the manufacturer’s protocol. Samples were collected from different plants in order to minimise the individual plant variation in gene expression. The integrity of the total RNA was checked by electrophoresis on a 1.5 % standard agarose gel, and its quantity as well as purity were determined using a microchip electrophoresis system for the study of nucleic acids MultiNA (Shimadzu Corporation, Kyoto, Japan). All RNA samples were then adjusted to the same concentration, measured and adjusted again in order to homogenise RNA input in the subsequent reverse transcription reaction. Reverse transcription-PCR (RT-PCR) cDNA was synthesised from 1.5 μg DNA-free RNA using 200 U RevertAid ™ M-MuLV Reverse Transcriptase, 4 μL 5× RT buffer, 1 μg oligo(dT)18 primer, 40 U RiboLock ™ RNase Inhibitor and 0.5 μL 25 mM dNTPs (MBI Fermentas, Lithuania) and DEPC water. Reverse transcription reactions were carried out in a volume of 20 μL and the resulting cDNAs were diluted five times for use in real-time PCR. Temperature conditions were 40 °C for 60 min and 70 °C for 10 min. All cDNA samples were stored at −20 °C until used for PCR. The reaction was carried out in an automated thermocycler C1000™ Thermal Cycler (Bio-Rad). Real-time PCR In the first stage of the study, four reference genes were chosen based on their previous use as internal controls in plant gene expression studies. Our results showed that glyceraldehyde 3-
Mis-perception of Rhizobium as a pathogen in pea mutants
phosphate dehydrogenase (GAPC1) was the most stably expressed gene. Primers were designed with the following criteria: Tm of 56 ±2 °C and PCR amplicon lengths of 60–400 bp, yielding primer sequences with lengths of 18–25 nucleotides and GC contents of 45–65 % (Table 2). Polymerase chain reactions were performed in a 96-well plate with a CFX96™ real-time PCR System combined with a C1000™ thermal cycler using SYBR Green (Bio-Rad) to monitor dsDNA synthesis. Reactions contained 5 μL 2x iQTM SYBR® Green Supermix, 3.4 μL H2O, 0.8 μL of cDNA and 0.4 μL of each gene-specific primer (10 μM) in a final volume of 10 μL. The following standard thermal profile was used: polymerase activation (95 °C for 3 min), amplification and quantification cycles repeated 40 times (95 °C for 30 s, 56 °C for 30 s, 72 °C for 40 s). The specificity of the amplicons was checked by electrophoresis in 2 % (w/v) agarose gel and a melting curve analysis performed by the PCR machine after 40 amplification cycles (55°–95 °C with one measurement of fluorescence every 0.5 °C). All investigated qPCR products showed only single peaks and there were no primer-dimer peaks or other artefacts. Negative control with a quantification cycle (Cq) less than 38 was not observed. Two biological repetitions were used for the measurement, and two technical replicates were analysed for each biological repetition. Data analysis Data were analysed using the CFX Manager™ software version 2.0. Primer efficiency values with an R2 value less than 0.997 were ignored. The expression levels of the gene of interest (GOI) relative to the GAPC1 were calculated for each cDNA sample using the PCR efficiency and the equation: relative ratio ¼ E
ΔC control sample –ΔC test sample T T
The values of samples were used in a Student’s t test and Dunnett’s criterion to calculate probabilities of distinct Table 2
induction or repression, and the average ratio of these values was used to determine the fold change in transcript level in mutant lines compared with parental lines at the different stages of nodule development. Student’s t test was used for pairwise comparisons and Dunnett’s test for comparison of many experimental groups with one control.
Results Immunocytological localisation of unesterified pectin in IT wall in the symbiotic nodules of ineffective pea mutants The composition of the IT wall in 14-day-old pea wild-type and mutant nodules was studied using antibodies JIM5 specific for unesterified pectin. In nodules of wild-type SGE antibody JIM5 labelled uniformly the IT wall (Fig. 1a, b). Similar distribution was characteristic for wild-type Finale and mutant SGEFix−-1 (sym40) (data not shown). In nodules of the mutant line SGEFix−-2 (sym33) (Fig. 1c, d) and double mutant RBT3 (sym33, sym40) (Fig. 1e) the thickened ITs were extensively labelled by JIM5 in comparison with SGE (Table 3). In mutant RisFixV (sym42) JIM5 labelled the IT wall (Table 3) but label was unevenly distributed with many obvious clusters (Fig. 1g, h). In the double mutant RBT4 (sym33, sym42), the distribution of JIM5 labelling was similar with SGEFix−-2 (sym33) and double mutant RBT3 (sym33, sym40) (Fig. 1f; Table 3). The additional abnormality in JIM5 localisation was observed in RisFixV (sym42) mutant, when degrading bacteroids in zone IV were surrounded by deposition of unesterified pectins (Fig. 1i). Histochemical localisation of suberin and callose depositions in the symbiotic nodules of ineffective pea mutants Histological organisation in 21-day-old nodules of the parental pea line SGE was typical for wild-type nodules, including a meristem, a zone of infection and a nitrogen-fixation zone (Fig. 2a) (Tsyganov et al. 1998; Voroshilova et al. 2009).
Description of primer sequences for qPCR
Accession no.
Description
Primer sequence (5′–3′)
PCR product size (bp)
Reference
L07500.1
Glyceraldehyde-3-phosphate dehydrogenase (GapC1) Peroxidase 7RA84
222
AAGAACGACGAACTCACCG TTGGCACCACCCTTCAAATG 89 TGTTTGAATCAGATGCTGCATTG 140 CATTTGATTGAAGATGTTGTGCAA
188
Martin et al. (1993)
75
Pérez-de-Luque et al. (2006); Die et al. (2009)
a
Hypersensitive reaction 203 J
70
Die et al. (2007, 2009)
Z15128.1
ABA-responsive protein
TGTTTGAATCAGATGCTGCATTG CA TTTGATTGAAGATGTTGTGCAA 21 TGGGTGTCTTTGTTTTTGATGATGA 436 TATGGCCTTGATAAGTCCAGTTCCT
440
Iturriaga et al. (1994)
AF396465 TC107779
a
– Identifier in the TIGR M. truncatula Gene Index
390
K.A. Ivanova et al. Fig. 1 Immunogold labelling of ultrathin sections through infection zone of pea wild-type and mutant nodules. Infection thread wall (a– h) and pectin depositions around bacteroid (i) labelled by JIM5 antibody (arrow), which is specific for unesterified pectin. IT – infection thread, ITW – infection thread wall, ba – bacterium, Ba – bacteroid. a Wild-type SGE. b High magnification of the boxed area in (a). c Mutant SGEFix−-2 (sym33). d High magnification of the boxed area in (c). e Double mutant RBT3 (sym33, sym40). f Double mutant RBT4 (sym33, sym42). g, i Mutant RisFixV (sym42). h High magnification of the boxed area in (g). Bar= 500 nm (a, c, e, f, g, i), bar= 250 nm (b, d, h)
Additionally, in 42-day-old nodules a senescence zone was observed below the nitrogen-fixation zone (data not shown). Suberin depositions in the 21- and 42-day-old nodules were only identified in the nodule endodermis, which separates the inner from the outer layers of the nodule cortex in postmeristematic regions (Fig. 2b). Nodules of cv. Finale were similar to those of SGE with respect to suberin deposition in the nodule endodermis (data not shown). In 21- and 42-day-old white nodules of the mutant line SGEFix−-1 (sym40) (Fig. 2c, e) suberised cells were not observed in the central tissue. However, there were suberin depositions in cell walls of both the nodule endodermis and the vascular endodermis (Fig. 2f), and the nodule endodermis enclosed the tip of the nodule, leading to the isolation of the
entire nodule from the rest of the plant (Fig. 2d). An agedependent increase of suberin accumulation in the cell walls was not revealed. In 21-day-old white nodules of the mutant line SGEFix−-2 (sym33) (Fig. 3a) suberin deposits were observed in the walls of host cells that were traversed by ITs (Fig. 3b, d). Suberin also accumulated in the thickened walls of the ITs (Fig. 3c, d). The increase of suberisation was observed in 42-day-old nodules (data not shown). In 21-day-old nodules of the double mutant line RBT3 (sym33, sym40), suberin distribution was similar to that of the mutant line SGEFix−-2 (sym33) and it was also increased in 42-day-old nodules (data not shown). In 21-day-old nodules of the mutant line RisFixV (sym42) suberin deposits in
Mis-perception of Rhizobium as a pathogen in pea mutants Table 3 Distribution of gold particles in infection thread walls in pea wild-type and mutant lines Genotype
Mean value
SE
SGE SGEFix−-1 (sym40) SGEFix−-2 (sym33) RBT3 (sym33, sym40)
139a 112a 506b 569b
20.6 18.8 25.0 39.4
depositions were revealed in the central tissues of 21- and 42day-old nodules of the mutant line RisFixV (sym42) (Fig. 4c, d). Callose accumulated in the walls of mature ITs and in the cell walls (Fig. 4 e, f). Changes associated with age of the nodule were not found. In the other mutant lines, callose was not found at all (Fig. 4g, h) or was registered only in the form of small deposits in phloem elements in 21- and 42-day-old nodules (Fig. 4 i).
RBT4 (sym33, sym42) Finale RisFixV (sym42)
575b 157 a 231c
35.7 22.4 23.7
Real-time PCR for 7RA84 (peroxidase), HSR203J (a marker of hypersensitive reaction) and ABR17 (PR10 protein)
Results are presented as the number of gold particles/μm2 Mean value±SE (n=20-25) are shown. Means without a common letter are significantly different by the Tukey multiple range test (P≤0.001)
the thickened walls of ITs were not observed. Suberin was found only in the nodule endodermis, as in the nodules of the wild-type Finale (data not shown). Callose distribution was also analysed in order to investigate the composition of thickened cell walls surrounding mature ITs in greenish nodules of the mutant line RisFixV (sym42). In contrast to wild-type Finale (Fig. 4 a, b), callose Fig. 2 Histochemical localization of suberin deposition in the 21-day-old nodules of the wild-type SGE (a, b) and mutant SGEFix−-1 (sym40) (c–f). Sections stained with Neutral Red. I – nodule meristem, II – zone of infection, III – nitrogen-fixation zone, indicated where they are identified, arrows indicates nodule endodermis, arrowheads – vascular endodermis, * – vascular bundles, Bar=200 μm. a, c, e DIC (differential interference contrast), b, d, f fluorescence microscopy. a–d Sagittal sections, e, f transverse sections made in the middle of nodule, in the region corresponding to infection zone in wild-type
The expression of three genes known to be involved in host defence responses was monitored by real-time PCR: 7RA84 (peroxidase), HSR203J (a marker of hypersensitive reaction) and ABR17 (PR10 protein). For the gene 7RA84, the maximum level of expression was observed at 14 DAI in wild-type and in all mutant nodules examined (Fig. 5a). However, in the mutant lines SGEFix−-1 (sym40) and SGEFix−-2 (sym33), the levels of expression were significantly higher (at least 2.3fold) than in wild type SGE for each day of analysis (Fig. 5a). In nodules of wild-type Finale and the mutant line RisFixV (sym42), the levels of 7RA84 expression were similar
K.A. Ivanova et al. Fig. 3 Histochemical localisation of suberin depositions in a 21-day-old nodule of the mutant SGEFix−-2 (sym33). Sections stained with Neutral Red. I – nodule meristem, arrowheads mark the infection threads, arrows indicate suberised walls of infection threads. Note suberised cell walls. Bar=200 μm (a, b), bar=20 μm (c, d). a, c DIC (differential interference contrast), b, d fluorescence microscopy. a, b Sagittal sections
at 14 and 28 DAI but significantly different (5-fold higher) at 42 DAI. The level of expression of the gene HSR203J in nodules of the mutant line SGEFix−-1 (sym40) was significantly higher (at least 2.5-fold) than in nodules of wild-type SGE at all stages analysed (Fig. 5b). In the mutant line SGEFix−-2 (sym33), the level of expression was significantly increased at 28 and 42 DAI in comparison with wild-type SGE (at least 2.5-fold). In the mutant line RisFixV (sym42), the timedependent increase in the level of expression HSR203J in nodules was observed and the levels of expression were significantly higher (at least 8.1-fold) than in wild-type Finale at 28 and 42 DAI (Fig. 5b). The level of expression of the gene ABR17 was significantly higher in the mutant lines SGEFix−-1 (sym40) than in nodules of wild-type SGE at 14 and 28 DAI. The highest level of expression was at 28 DAI in this mutant line; this was also true in the line SGEFix−-2 (sym33) in contrast to the wild-type SGE (14.2- and 5-fold respectively) in which some decrease in the level of expression was observed at this stage (Fig. 5c). Interestingly, the level of expression of this gene in mutant line SGEFix−-2 (sym33) was significantly lower than with wildtype SGE (3.6-fold) at 42 DAI. In the mutant line RisFixV (sym42), the time-dependent increase in the level of expression of the gene ABR17 was observed and the levels of expression were significantly higher than in wild-type Finale at 28 and 42 DAI (6- and 3.8-fold respectively) (Fig. 5c).
Discussion Pectin is one of the most abundant components of a cell wall and different modifications of pectin occur during nodule
development as well as during pathogenic infection. In legumes, a pectate lyase was recently identified that is inducible by NFs and apparently involved in the penetration of rhizobia in IT (Xie et al. 2012). The involvement of pectin in defence responses probably involves the production of oligogalacturonides (OGs), which are the result of degradation of homogalacturonan by the action of polygalacturonases and pectate lyases, produced by both plants and microbes. OGs trigger calcium-mediating ROS production and activation of defence genes (Ferrari et al. 2013). In the legume-Rhizobium symbiosis, OGs inhibit flavonoid-induced nod gene expression, which could be important for fine-tuning of the NF concentration during IT growth (Moscatiello et al. 2012). However, pectin may be involved not only in signalling during defence responses, but it may also participate together with other components in cell wall modification. Structural reinforcement of cell walls is one of the most prominent manifestations of defence responses triggered by pathogens (Hammond-Kosack and Jones 1996). Different components like lignin, suberin, callose and pectin have been shown to be involved in wall cross-linking (Matern et al. 1995). The modification of pectin, caused by the activity of pectin methylesterases dramatically alters the physical properties of demethylesterified forms, causing an increase in cell wall rigidity and the concomitant formation of gel-structures to support damaged walls (Pelloux et al. 2007; Wolf and Greiner 2012; Bethke et al. 2014). In this study, the high labelling of the IT wall by JIM5, which labels unesterified pectins, was observed for single and double mutants carrying a mutation in the gene sym33. For mutant RisFixV (sym42) JIM5 labelling was unevenly dispersed in the IT wall with some clusters of intense labelling. The other striking feature of the RisFixV (sym42) was the presence of JIM5 label around degrading
Mis-perception of Rhizobium as a pathogen in pea mutants Fig. 4 Histochemical localisation of callose depositions in the 21-day-old nodules of the wild type Finale (a, b), mutant RisFixV (sym42) (c–f) and mutant SGEFix−-2 (sym33) (g–i). Sections stained with Aniline Blue. I – nodule meristem, II – zone of infection, III – nitrogen-fixation zone, IV – senescence zone indicated where they are identified, ic – infected cell. Arrows indicate callose depositions around infection threads, arrowheads in cell wall, * in vascular bundles. h Note autofluorescence of suberised cell walls and walls of infection threads in mutant SGEFix−-2 (sym33). Bar=200 μm (a–d, g, h), bar=20 μm (e, f), bar=50 μm (i). a, c, e, g DIC (differential interference contrast), b, d, f, h, i fluorescence microscopy. a–d, g, h Sagittal sections
bacteroids, indicating that unesterified pectin encapsulated ineffective bacteroids. To our knowledge, this phenotype has never been observed previously in pea wild-type nodules (Rae et al. 1992) or in mutants; however, the reason for this observed encapsulation of degrading bacteroids is still unknown. Suberin deposits in the nodule endodermis of wild-type SGE and Finale nodules are physiologically important for nodule protection from excessive dehydration, salt stress and the penetration of pathogens (Schreiber et al. 1999). The intense level of suberisation observed in nodules of the mutant line SGEFix−-2 (sym33) is an example of a plant defence response which prevents the release of bacteria into the plant cytoplasm and their subsequent differentiation into bacteroids.
These host defence responses indicate that the mutation in the pea gene sym33 leads to the perception of nodule bacteria not like microsymbionts, but as pathogens. Previously, it was shown that the accumulation of polyphenolic substances in walls of thickened ITs is typical for M. truncatula mutant TE7 in the gene Mtsym1 (Benaben et al. 1995). Recently, it was shown that Sym33 and MtSym1 are orthologues, encoding a protein IPD3, which interacts with calcium-dependent and calmodulin-dependent kinase CCaMK, activating the expression of a nodule-specific remorin, which is necessary for the growth and formation of ITs and symbiosomes (Ovchinnikova et al. 2011). In L. japonicus, the ortholog of Sym33 and MtIPD is LjCYCLOPS, which encodes a transcriptional factor that
K.A. Ivanova et al.
Fig. 5 Gene expression dynamics: 7RA84 (peroxidase), Hsr203J (a marker of hypersensitivity reactions), ABR17 (PR10 protein) in nodules of pea wild types SGE, Finale and corresponding mutants. * – significant differences from the wild-types of corresponding mutants within one time-period (t test, p
