• Effects of two algal polysaccharides, laminarin and carrageenans, on defence responses and signalling in tobacco plants is presented. A possible role as defence elicitors is important in the context of the use of algal extracts as plant protectants.
• The effect of the extracts was assessed after infiltration of tobacco leaves, and compared to the effect of a known elicitor of Phytophthora parasitica var. nicotianae(Ppn).
• Of the two algal polysaccharides, only carrageenans efficiently induced signalling and defence gene expression in tobacco leaves, as observed with Ppn elicitor. λ-carrageenan, with its high sulphate content, proved the most active. Defence genes encoding sesquiterpene cylase, chitinase and proteinase inhibitor were induced locally, and the signalling pathways mediated by ethylene, jasmonic acid and salicylic acid, were triggered. Some effects lasted for at least a week.
• λ-Carrageenan can elicit an array of plant defence responses, possibly through an effect of its high sulphate content. This helps clarify the mechanism of plant protection by algal extracts.
Algae have long been used in coastal regions as soil fertilisers and advantageous effects of spraying seaweed extracts on crop plants have been reported. Thus, improved seed germination, higher yields, increased resistance to pests and diseases and longer shelf life of fruits were recorded upon treatment of various plants with seaweed extracts (Stephenson, 1966; Hankins & Hockey, 1990; Blunden, 1991; Jolivet et al., 1991).
Due to their effects as plant protectants, it has been proposed that algal extracts act as elicitors of plant defence responses. However, their mode of action is far from being understood, which limits their utilisation as beneficial products. Mechanisms by which plants respond to pathogens and their elicitors in order to limit penetration and growth into their tissues are well characterised. Elicitors act as signal molecules whose perception at the cell surface and subsequent transduction lead to the activation of plant defence genes. As a result various defence compounds are synthesised, most notably structural proteins which reinforce plant cell walls, enzymes involved in the synthesis of phytoalexins, hydrolases such as chitinases or glucanases, and enzyme inhibitors which counteract pathogen hydrolytic enzymes (Kombrink & Somssich, 1995). Elicitor effects are mediated by signalling pathways, among which salicylic acid (SA), jasmonic acid (JA) and ethylene either alone or in combination, play major roles in local and systemic induction of defence responses (Hammond-Kosack & Jones, 1996; Reymond & Farmer, 1998).
A variety of molecules, including oligo- and polysaccharides, peptides, proteins, and lipids can act as elicitors. Most of them derive from plant or pathogen cell surfaces (Hahn, 1996; Côtéet al., 1998). However, it has been reported that some polysaccharides purified from seaweeds, as well as derived oligosaccharides (Potin et al., 1999), also have the ability to trigger plant defence responses. Thus, the storage β-1,3-glucans of brown algae, known as laminarans and laminarin, are main components of several commercial seaweed liquid fertilisers. Laminarans induce the formation of antifungal compounds in alfalfa cotyledons (Kobayashi et al., 1993), and elicit D-glycanases (β-1,3-glucanase and α-amylase) in Rubus fruticosus cell suspension culture (Patier et al., 1993). Laminarin was shown to stimulate phytoalexin (glyceollin) accumulation in soybean cotyledons (Keen et al., 1983) and seedlings (Bonhoff & Grisebach, 1988), an effect which was accompanied by the subsequent localised resistance of soybean against Phytophthora megasperma f. sp. glycinea. Carrageenans are a family of partly sulphated linear galactans found in the cell walls of many red algae. They are composed of repeating dimers of an α-1,4-linked D-galactose (λ-carrageenan) or 3,6-anhydro-D-galactose residue (κ- or ι-carrageenan) and a β-1,3-linked D-galactose residue. κ-, ι-, and λ-carrageenan show increasing sulphate contents, c. 22%, 32% and 38% (w/w), respectively, in commercial preparations (De Ruiter & Rudolph, 1997). Patier et al. (1995) reported that κ-carrageenan elicits β-1,3-glucanase activity in Rubusfruticosus cell suspension cultures, oligo-κ-carrageenans being more efficient than the native polysaccharide.
The aim of this work was to evaluate and compare the biological effects of the two algal polysaccharides, laminarin and carrageenans on defence responses and signalling in tobacco plants, a prerequisite to their use as plant protectants. An elicitor prepared from the tobacco pathogen Phytophthora parasitica var. nicotianae (Ppn), whose activity was previously characterised on tobacco cell suspension culture (Rickauer et al., 1990, 1997) was included in all bioassays in order to set up the conditions required for defence gene induction in whole tobacco plants.
Materials and methods
Nicotiana tabacum L., cv. 49–10 (Helgeson et al., 1972) was grown on vermiculite in a growth chamber at 80% hygrometry, with a photoperiod of 16 h light at 125 µmol m−2 s−1 and 25°C, and 8 h dark at 18°C.
Biological compounds and chemicals
The elicitor of Phytophthora parasitica var. nicotianane (Ppn), hereafter called Ppn elicitor, was prepared as described previously (Pélissier et al., 1986; Roux et al., 1994). Briefly, it consists of the dialysed ethanol-soluble fraction of the extract recovered upon autoclaving of the cell walls isolated from the mycelium.
Salicylic acid (SA), ortho-anisic acid (oANI), laminarin and carrageenans were purchased from Sigma (St. Quentin Fallavier, France). Laminarin originated from Laminaria digitata, κ-carrageenan from Eucheuma cottonii, ι-carrageenan from Eucheuma spinosa, and λ-carrageenan from Gigartina acicularis and Gigartina pistillata.
The compounds to be tested were solubilised in ultrapure water (MilliQ) before being infiltrated into the mesophyll of fully expanded leaves of 2-month-old plants with a syringe without needle. Routinely, 150 µl of solution were infiltrated until spreading of the solution in the leaf tissue covered an area of c. 5 cm2, and this area was delineated with a marker pen. Each plant received only one compound except in one experiment where all the compounds were infiltrated into distinct areas of the same leaf for comparison of the different polysaccharides. Accumulation of fluorescent compounds in the infiltrated leaves was followed by examination under UV light at 365 nm. As a function of time after infiltration, the tissues corresponding to the infiltrated area, the surrounding area (1 cm width beyond the marker line) or the rest of the leaf (> 2 cm away from the infiltrated area) were harvested, frozen in liquid nitrogen and stored at −80°C until use.
Cell death in tobacco leaf tissue was monitored by Evans blue staining as described by Baker and Mock (1994). Each assay was performed with four leaf disks (8 mm in diameter) punched out with a cork borer from each infiltrated area of either six plants (24 disks) for water or Ppn elicitor, or two plants (eight disks) for carrageenan or laminarin. Leaf disks were incubated for 30 min in 0.25% Evans blue on a rotary shaker, extensively washed, and ground in eppendorf tubes containing 0.5 ml 1% SDS. The tubes were vortexed, centrifuged and the absorbance of the supernatant was measured spectrophotometrically at 600 nm. The experiment was repeated twice independently and data represent mean values ± SE.
RNA extraction and analysis
Total RNA was isolated from samples using the ready-for-use Extract-All kit (Eurobio, Les Ulis, France). RNA concentrations were measured spectrophotometrically at 260 nm and samples (10 µg RNA) were denatured at 50°C for 1 h with glyoxal/dimethylsulfoxide prior to being subjected to electrophoresis in a 1.2% agarose gel in 10 mM sodium-phosphate buffer pH 7.0 according to Sambrook et al. (1989). The separated RNA were blotted overnight onto a neutral nylon membrane (Hybond N; Amersham, UK), and fixed at 80°C for 2 h. Equal loading of the gel was checked by visualisation under UV light (λ254nm) of ribosomal RNA on the membrane.
The membranes were prehybridised at 42°C for 2 h in ‘High SDS Buffer’ (Boehringer Mannheim, Germany) supplemented with 0.005% yeast RNA (Boehringer Mannheim), then hybridised overnight at 42°C in the same solution containing the radioactive probe. The probes used in this study were pTL-35 for tobacco lipoxygenase (LOX, Véronési et al., 1995), pBS-TEAS for tobacco 5-epi-aristolochene synthase (EAS, also called sesquiterpene cyclase, Facchini & Chappell, 1992), pCHN50 for tobacco chitinase (CHN, Shinshi et al., 1987) and pSSy20 for Nicotiana sylvestris Rubisco small subunit (RBCs, Pinck et al., 1986), kindly provided by the respective authors.
The 1-aminocyclopropane-1-carboxylic acid oxidase (ACO) and proteinase inhibitor (PI) probes were prepared by polymerase chain reaction (PCR) amplification using the cDNA of a mass excised cDNA lambda zap library (Véronési et al., 1995) as the template. The degenerated oligonucleotide primers retained for ACO were deduced from the alignment of ACO sequences of tomato (Holdsworth et al., 1987, genbank accession X04792), and Nicotiana glutinosa (Kim et al., 1998, genbank accessions U54565, U54566 and U62764). The coding strand primer, GATGCTTGTGA[GA]AA[CT]TGGGG, corresponding to nucleotide 147–166 of U54565, and the noncoding strand primer, TCAT[AT]GCTTCAAA[CT]CTTGG, corresponding to nucleotide 966–984 of U54565 were used to amplify a 838-bp ACO fragment. In the case of PI, the oligonucleotide primer corresponding to T3 promoter in pBluescript vector, ATTAACCCTCACTAAAGGGA, and AACCCTTGTCTGCGTTACAA, corresponding to nucleotides 950–970 of a tobacco type II PI sequence (Balandin et al., 1995, genbank accession Z29537) as noncoding strand primer, were used to amplify a 541-bp PI fragment. The nucleotide sequence of the ACO and PI fragments was checked after cloning of the PCR products into pGEM-T vector (Promega, Charbonnieres, France).
Radioactive labelling of the probes was performed with [α-32P]dCTP using the RadPrime DNA Labelling System (GibcoBRL, Life Technologies, Cergy Pontoise, France).
After hybridisation, the membranes were washed once in 2 × SSC (1 × SSC = 0.15 M sodium chloride, 0.015 M sodium citrate, pH 7.0), 0.1% SDS at room temperature for 5 min, and twice at 65°C for 30 min in the same solution, before being autoradiographed at −80°C for 1–3 d (BioMax MS films, Kodak).
Each experiment including sample preparation and RNA isolation was carried out two or three times independently.
Salicylic acid extraction and analysis
Salicylic acid was measured in the phenolic extracts of leaf tissues harvested at various times after elicitation according to a protocol adapted from Meuwly and Métraux (1993). Free and bound phenolics were extracted sequentially from frozen leaf material (usually 0.2–0.5 g), to which an internal standard (oANI) was added at the beginning of extraction. Briefly, methanolic extraction followed by organic phase partitioning against ethylacetate : cyclohexane (v : v) was carried out to obtain free phenolics. Acid hydrolysis was performed on the aqueous layer containing the bound phenolics and the hydrolysate was partitioned as above. Alternatively, total salicylic acid contents were measured; in this case, acid hydrolysis was achieved directly after methanolic extractions and the hydrolysate was then subjected to organic phase partitioning.
Analysis was performed with an HPLC apparatus (SP 8800, Spectra-Physics France, Les Ulis, France) equipped with a deactivated LC-ABZ reversed phase column (25 cm length × 4.6 mm diameter × 5 µm packing, Supelco SA, France) and a LC-ABZ guard column (2 cm × 4.6 mm × 5 µm packing, Supelco), equilibrated in 15% acetonitrile in 25 mM KH2PO4 buffer pH 2.6. Chromatography was achieved at 27°C with a 15–55% acetonitrile gradient in the same potassium phosphate buffer, at 1 ml min−1, for 18 min. The column was then washed with a 55–80% acetonitrile gradient in water for 1 min, 80% acetonitrile in water for 5 min, and re-equilibrated with 15% acetonitrile in 25 mM KH2PO4 buffer pH 2.6 for 5 min. SA and oANI were quantified with a Spectrofluorimeter (FP920 JASCO, Nantes, France) with wavelengths optimised for each compound: λex 305 nm, λem 365 nm for 15 min to detect oANI, then λex 305 nm, λem 407 nm to detect SA. Corrections for losses were made for each individual sample according to recoveries of the internal standard. Experiments were repeated twice and data represent mean values ± SE.
Macroscopic changes and cell death of tobacco tissues challenged with Ppn elicitor and algal polymers
Infiltration of the mesophyll of tobacco leaves with Ppn elicitor at 30 µg ml−1 induced rapid macroscopic changes in the infiltrated area (Fig. 1). Four–five hours post infiltration (hpi) the infiltrated area (I) turned slightly bright on the abaxial face, started to desiccate 9–12 hpi, being fully desiccated by 24 hpi (Fig. 1c). The necrosis turned brown thereafter (E30, Fig. 1e) and remained strictly limited to the infiltrated tissues. However, UV light examination (λ365nm) revealed that both the necrotic area (I) and the surrounding tissues (II) displayed a blue autofluorescence indicative of the accumulation of phenolic compounds, while the rest of the leaf (III) remained unaffected (Fig. 1d).
As shown on Fig. 1(e,f), laminarin-infiltrated tissues (L) were similar to water-infiltrated control tissues (C, Fig. 1a,b,e,f) and remained symptomless even at 168 hpi, for doses ranging from 100 to 1000 µg ml−1. By contrast, λ-carrageenan (G) elicited macroscopic changes similar, although less severe, than those occurring upon infiltration with Ppn elicitor. Thus, the infiltrated area became shiny 4–5 hpi with 1000 µg ml−1λ-carrageenan, but this initial response was not followed by complete necrosis of the tissues. The infiltrated area appeared slightly chlorotic at 168 hpi, and showed a local accumulation of fluorescent compounds (Fig. 1e,f) in a dose-dependent manner, being visible at 100 µg ml−1. In some experiments, a few necrotic spots were visible at high λ-carrageenan concentration (1000 µg ml−1 solution) but they never covered the whole infiltrated area.
Infiltrations with ι-carrageenan and to a lesser extent κ-carrageenan at 1000 µg ml−1 resulted in a faint brightness of the infiltrated areas, accompanied by a slight local accumulation of fluorescent compounds (data not shown). After 168 h, the observed symptoms were similar to that observed with λ-carrageenan at 100 µg ml−1. Hence, only λ-carrageenan was retained for further work since it proved the most active in this assay.
The effect of Ppn elicitor and algal compounds on cell death was assessed by measuring Evans blue uptake in the infiltrated tissues (Fig. 2). Accumulation of the vital dye was strongly increased at 24 hpi in Ppn elicitor-infiltrated tissues as revealed by the increase in absorbance at 600 nm, indicating the occurrence of cell death at this time point. Tissues harvested 168 hpi could not be assessed in this test since the dye did not penetrate dry tissues. When dye uptake was measured in water-infiltrated areas and surrounding tissues, the values were similar to that of noninfiltrated tissues (data not shown). A slight increase in cell death was also recorded in response to infiltration with λ-carrageenan at 1000 µg ml−1, while laminarin had no effect.
Altogether, these results indicate that tobacco leaves are particularly receptive to Ppn elicitor and to λ-carrageenan but not to laminarin. The elicitor activity of the three compounds was subsequently investigated.
Effect on defence-related genes
Based on preliminary experiments performed with various doses of the compounds at different time points between 8 and 168 hpi, a comparative study of the ability of λ-carrageenan, laminarin and Ppn elicitor to elicit defence gene expression was then undertaken by northern blot analysis (Fig. 3) of the RNA isolated from the infiltrated area (zone I) and the surrounding tissues (zone II).
In water-infiltrated control tissues, the level of transcripts was very low or undetectable, except in rare cases (Fig. 3a, 6 hpi; Fig. 3b, zone II) where a slight expression was observed, probably as a result of the mild wounding effect of infiltration. In response to Ppn elicitor, sesquiterpene cyclase (EAS), an enzyme of the sesquiterpenoid pathway leading to the synthesis of capsidiol, the main phytoalexin produced in tobacco, was early and transiently induced to high levels (Fig. 3a). Transcript accumulation was observed as soon as 6 hpi, reached a maximum at 12 hpi and then declined sharply. Basic chitinase (CHN), a member of the PR3 family, and proteinase inhibitor type II (PI) were also transiently induced in the infiltrated tissue, with maximal transcript accumulation around 18 hpi (Fig. 3a, zone I). From experiment to experiment, defence gene expression followed the same kinetics in response to Ppn elicitor, except that the level of transcripts was still high at 24 hpi in some experiments (Fig. 3b). In zone II tissues, a weaker and somewhat delayed expression of EAS, CHN and PI genes was recorded (Fig. 3a, zone II). Defence gene expression was generally low or undetectable in the rest of the leaf, i.e. zone III of Fig. 1(c,d) (data not shown).
In response to λ-carrageenan at 1000 µg ml−1 (G), relatively high transcript accumulation was observed in zone I for EAS, CHN and PI at 24 hpi (Fig. 3b, zone I). This level remained high for at least 1 wk (168 hpi) for CHN and PI. Laminarin (L) had, generally, a much weaker effect on EAS, CHN and PI than λ-carrageenan. Neither algal polymers induced defence gene expression in zone II tissues.
These results indicated that λ-carrageenan and Ppn elicitor are potent inducers of defence responses in tobacco leaves. An inverse relationship was generally found between the extent of defence gene expression and of Rubisco small subunit (RBCs) transcripts whose level was totally or severely decreased in response to Ppn elicitor and to λ-carrageenan (1000 µg ml−1) in zone I and to a lesser extent, in zone II tissues.
Effect on signalling pathways
In order to further characterize the elicitor activity of the three compounds under study, their effect on the signalling pathways mediated by jasmonic acid, ethylene and salicylic acid was investigated. Lipoxygenase (LOX) and ACC oxidase (ACO) whose gene expression leads to jasmonic acid and ethylene biosynthesis, were first analysed. Fig. 3(a) shows that LOX and ACO gene expression were highly and transiently induced within 24 hpi in response to Ppn elicitor in zone I. As compared to ACO, elicitation of LOX was somewhat delayed, being maximum at 18 hpi. However, as in the case of EAS, CHN and PI, the level of ACO and LOX transcripts was still high at 24 hpi in some experiments in zone I as well as in zone II (Fig. 3b), as particularly observable for LOX gene expression.
Neither genes, ACO or LOX, was induced in zone III tissues, even at 168 hpi (data not shown).
In response to the two algal polymers (Fig. 3b), ACO gene expression was induced locally (zone I), and to a lesser extent in the surrounding zone II tissues at 24 hpi with λ-carrageenan. LOX gene expression was mainly induced by λ-carrageenan in zone I.
The effect of the three compounds on free, conjugated and total salicylic acid (SA) production was then investigated in zone I tissues only. A time course analysis of water-infiltrated control tissues (C, Fig. 4a) showed that the amounts of either free or conjugated SA remained low and constant for the duration of the experiment, with amounts averaging 44 and 76 ng g−1 f. wt. In response to Ppn elicitor (E), SA increased sharply between 6 and 24 hpi, reaching amounts of 204 ng g−1 f. wt free SA and 438 ng g−1 f. wt conjugated SA at 18 and 24 hpi, respectively, similar to values reported in the literature in response to elicitors (Baillieul et al., 1995). Then, both levels started decreasing, being close to that obtained in water infiltrated tissues at 96 hpi.
λ-carrageenan (1000 µg ml−1) was revealed to be a very potent elicitor of SA (G, Fig. 4b). As compared to Ppn elicitor, its effect was lower at 18 hpi (339 ng g−1 f. wt total SA) but more prolonged, total SA reaching values of 585 and 944 ng g−1 f. wt at 96 and 168 hpi, respectively. Even at lower concentrations (100 µg ml−1), λ-carrageenan induced a four-fold increase in total SA at 168 hpi. In contrast, laminarin (L) did not induce SA accumulation, regardless of the concentrations that were retained (100 or 1000 µg ml−1).
The aim of this work was to investigate the effect of two algal polysaccharides, carrageenans and laminarin, on plant gene expression related to defence, signalling, and primary metabolism. Carrageenans are sulphated linear galactans with varying amounts of sulphate half esters, while laminarin is a glucan mainly composed of β-1,3-linked D-glucose residues. So far, only very few studies showing that these polymers induce plant defence have been reported in the scientific literature; and except elicitation of D-glycanase activities and phytoalexin accumulation in cell suspension cultures, detached cotyledons, and soybean seedlings by κ-carrageenan and laminarans, an extended survey of their effects is still lacking. In the present study, their action on tobacco leaves was compared with the activity of an elicitor of Phytophthora parasitica var. nicotianae (Ppn). Since this elicitor was only known for its effects on tobacco cell suspension cultures (Rickauer et al., 1990; Rickauer et al., 1997), it was first checked that it was equally active upon infiltration of tobacco leaves. The three classes of target genes that were retained encode: defence molecules with antagonistic effects on microbial growth, structure, and pathogenicity; key enzymes of the signalling pathways leading to ethylene and jasmonic acid metabolism; and a protein typical of the primary metabolism. In addition, salicylic acid, another signalling molecule, was also highly increased in response to Ppn elicitor.
Of the two algal polymers, only carrageenans efficiently induced signalling and defence in tobacco plants. Among carrageenans, λ-carrageenan proved the most active, which suggests a relationship with the higher level of sulphation of this polymer as compared with ι- or κ-carrageenan. The elicitor activity of λ-carrageenan was of the same order of magnitude as the activity of Ppn elicitor in the infiltrated zone I tissues. The transient expression of signalling genes does not preclude that derived signals (i.e. ethylene and jasmonic acid) are produced for a longer time as shown for salicylic acid, whose levels exhibited a 7.3-fold increase over the control at 168 hpi. This would be consistent with the more prolonged expression of defence markers, which was particularly observed in response to λ-carrageenan. It is interesting to note that, although being as active as Ppn elicitor, λ-carrageenan had milder side effects, particularly on necrosis and on RBCs gene expression which was retained as a marker of the primary metabolism. This indicates that signalling and defence are not proportional to the extent of cell death.
λ-carrageenan elicited signalling and defence in a dose dependent manner, being already active at 100 µg ml−1 on SA production and fluorescent compounds accumulation. However, the concentrations that were used (100 and 1000 µg ml−1) are still high in comparison with Ppn elicitor (30 µg ml−1) or to purified elicitor molecules used in other studies (Keller et al., 1996; Villalba Mateos et al., 1997). Patier et al. (1995) reported that oligo-κ-carrageenans were more active as elicitors of laminarase than the polymer itself on Rubus fruticosus cell suspension culture. Since the λ-carrageenan that was used in our work was of commercial origin, one may assume that its activity would increase upon fractionation and purification of derived oligomers. Similarly one might expect oligoglucans to be more active than laminarin, as already reported by Kobayashi et al. (1993) and Patier et al. (1993) who showed that elicitation of phytoalexins and D-glycanases is much higher in response to the oligosaccharides.
The mechanisms by which plant cells perceive carrageenans, laminarin and oligosaccharides derived from these polymers are not fully understood. The isolation of a soybean receptor for β-glucan elicitors from Phytophthora megasperma f. sp. glycinea suggests that signalling in response to these and other glycan elicitors might be initiated by receptors on the plant plasma membrane (Umemoto et al., 1997). The recent report that carrageenans act as signalling molecules in algae–algae interactions lends support to this hypothesis (Bouarab et al., 1999).
In conclusion, this study provides the most complete evidence that λ-carrageenan elicits an extended array of defence-related genes in tobacco plants, without affecting too much the primary metabolism. Such knowledge is prerequisite to their potential use as plant protectants.
We are grateful to SECMA Biotechnologies Marines (Pontrieux, France) and to the French Ministry of Education and Research for the financial support of Laurence Mercier (CIFRE fellowship no. 96639).