QUARTERLY Novel oligosaccharides isolated from Fusarium oxysporum L. rapidly induce PAL activity in Rubus cells �

Activation of the phenolic pathway is known to be part of a defense response against cell wall-derived elicitors from pathogens. Many examples of a defense response by increasing the synthesis of phenolic compound against the elicitor were demonstrated in the past, but the elicitor structure has so far been poorly characterized. Our results indicate that a disaccharide fraction containing the following structure: alpha-D-mannopyranosyl (1-->2)alpha/beta-D-glucopyranosyl and alpha-D-mannopyranosyl (1-->x) inositol, isolated from Fusarium oxysporum L., promotes rapid and transient phenylalanine ammonia lyase activity in Rubus fructicosus cells at nanomolar concentration. The disaccharides were isolated by size-exclusion chromatography directly from extracts obtained by alkaline treatment of F. oxysporum mycelium. Their structure was determined by 500-MHz-1H-NMR spectroscopy combined with methylation analysis and fast atom bombardment mass spectrometry.

production of lytic enzymes like pectinases, glucanases, xylanases (Benhamou et al., 1990;Alconada et al., 1995;Christakopoulos et al., 1996), and large amounts of biologically active metabolites such as carotenoids, bikaverins, mycotoxins, phytotropins, gibberellins and estrogens (Bruckner et al., 1989).Plants react against Fusarium penetration by accumulating callose or plant cell-wall components (Rodriguez-Galvez & Mendgen, 1995;El-Gendy et al., 2001), by increasing the steady-state mRNA level of phenylpropanoid pathway enzymes (Ni et al., 1996) or pathogenesis-related proteins (Casacuberta et al., 1992).Fusarium also produces acidic polysaccharides as components of the cell wall, or as exopolymers in culture medium.Among the carbohydrate components of fungal cell walls, b (1®3)[1®6] D-glucans have been described as inhibitors of virus infection of Nicotiana tabacum (Rouhier et al., 1995).The uronic acid-containing glycoprotein glycans of Fusarium sp.M7-1 are conjugated to serine and/or threonine residues (Iwahara et al., 1992;1995;Jikibara et al., 1992a;1992b), and they are involved as Fusarium allergens in the development of respiratory diseases (Backman et al., 1995).In addition, the treatment of suspended plant cells or protoplasts by the carbohydrate enriched fractions isolated from uronic acid-containing glycoproteins from Fusarium sp.M7-1 indicated an ability of fungal oligosaccharides to elicit early plant defense reactions (Nita-Lazar et al., 2000).To get insight into the mechanism of action of Fusarium carbohydrates which may be signaling molecules in plants, an attempt has been made to identify the structure of the minimum molecular size of O-glycans required for the expression of a biological activity.In this study, we report the isolation and characterization of disaccharides which promote elicitor activity.The carbohydrates were examined here in terms of their ability to trigger early reactions in Rubus fructicosus suspended cells.

Materials.
Suspension cultures of Rubus fructicosus, originally derived from cambial explant from twigs, were grown as described by Hustache et al. (1975).
Size-exclusion chromatography fractionation.Dried mycelium of Fusarium oxysporum L. (100 g) was extracted overnight with acetone (250 ml at 20°C) and the residual powder was subjected to alkaline treatment in 100 ml 0.1 M NaOH at 60°C for 2 h.The supernatant collected by centrifugation (15 000 ´g, 15 min) and neutralized to pH 7 with 50% acetic acid was stored overnight at 4°C.The resulting sample was centrifuged (15 000 ´g, 20 min at 20°C), and an aliquot (1 ml) of supernatant was applied to an HW 40F/50F column (50 cm ´2.5 cm; Interchim) and eluted with 0.1 M NaNO 3 at a flow rate of 2 ml/min.Detection was run both at 206 nm and 280 nm using a Beckman DU 640 spectrophotometer, and the refractive index was measured with a Waters 410 differential refractometer.The presence of carbohydrates or proteins in the isolated fractions was determined (independently of the HPLC detection system) according to Bradford (1976) andDubois (1956) Elicitation experiments.Rubus cells (4 10 6 ) were suspended for 0 to 30 h in 25 ml of a pH 4.8 buffer (25 mM Bis-Tris/HCl containing 1 mM KCl, 1 mM CaCl 2 , 0.06 M sucrose, 0.56 M mannitol supplemented or not with cycloheximide (1 mM)) on a roller mixer in the presence or absence (controls) of fungal elicitor.In each experimental set, the fungal elicitor isolated from Fusarium was added up to 0.2 mg/ml to the medium.The elicitation medium was removed by centrifugation (500 ´g, 5 min, 4°C), and the cells were submitted to PAL activity analysis.At least four repli-cates were monitored from three independent elicitation sets.PAL assay.The cells were subjected to PAL assays carried out according to the modified procedure of Hagendoorn et al. (1991).Briefly, the cells were sonicated on ice (70 W, 20 s) with a Vibra cell (Bioblock) in 0.1 M sodium borate buffer (pH 8.8) containing 2 mM dithiothreitol.The homogenate was centrifuged (5000 ´g, 15 min) and the supernatant was mixed with 100 mg of  for 10 min at 4°C.The mixture was centrifuged for 5 min at 5000 ´g and 0.05% (w/v) polyvinylpyrrolidone was added to the supernatant.The mixture was incubated for 10 min at 4°C and then centrifuged for 5 min at 5000 ´g.The supernatant was dialyzed and concentrated with an Ultrafree unit equipped with a 10 kDa molecular mass cut-off membrane (Ultrafree ÔMillipore, Bedford, MA, U.S.A.).The crude enzyme extracts were then subjected to PAL assays according to Zucker et al. (1965).The reaction mixtures containing the L-phenylalanine substrate (2.25 mM) and enzyme (final concentration 6 mg protein/ml) in 0.1 M sodium borate, pH 8.8, supplemented with 2 mM dithiothreitol were incubated at 40°C.Kinetics were developed at 290 nm for 30 min using a Beckman DU 640 spectrophotometer and kinetic curves were drawn and fitted with second-order polynomial regressions; the statistical significance was set at 5%.The initial velocity of PAL reaction (D A 290 min -1 ) was calculated from the regression equation using Excel software and PAL activity was expressed as mkat/kg protein (1 mkat representing the formation of 1 mmol of product per s).Enzyme activation was expressed by the R value, i.e. the ratio of fitted-curve slope obtained with elicited cells versus controls.
Controls.Kinetic curves were developed in parallel from elicited samples and from non-induced ones.Controls consisted of cells incubated in a buffer without the elicitor in the presence or absence of an effector, like cycloheximide.The viability of cells was con-trolled during each experimental set using Evan's Blue indicator.Blanks for enzyme assays were performed with boiled enzyme or in the absence of L-phenylalanine.

Elicitor activity
Alkaline extract (about 1 g) from the acetone powder was fractionated by size-exclusion chromatography into three fractions with different retention times (Fig. 1).About 400 mg of lyophilized powder of fraction I, 80 mg of fraction II and 100 mg of fraction III were obtained.Independently of the HPLC elution profiles protein and carbohydrate analysis showed that fractions I and II contained both carbohydrates and proteins, while fraction III contained carbohydrates only.The three fractions were used as elicitor to trigger mainly the PAL activity using Rubus cells in suspension for up to 30 min in at least 4 independent experiments.A study of the ability of the three fractions to elicit PAL activity in 30 min showed a significant activity of 150 mkat/kg protein for fraction III, a lower activity of 95 mkat/kg pro-tein for fraction I while fraction II had no activity (Fig. 2).On the basis of these results, we focused our interest on fraction III to characterize the PAL response and also to identify the signaling structure.
Three to six micrograms of fraction III was applied to Rubus suspended cells for up to 30 h in at least 4 independent experiments.The PAL response was followed as changes in absorbance at 290 nm of reaction mixtures containing 2.25 mM L-Phe and the enzyme extract prepared from treated cells or from controls.We verified that the viability of cells was not affected by the elicitor application since it remained as high as in controls (85-95%).The induced response was biphasic with respect to the kinetics shown in Fig. 3 (curve a).After 30 h the slope increased only slowly and after 35 h the presence of contaminations was observed.The first response peaked at 30 min with the activity of 150 mkat/kg protein, giving an R value of 18, and the second one reaching at 30 h the level of 191.6 mkat/kg protein, equivalent to the R value of 23.In contrast, PAL activity in ab-  sence of elicitor remained unchanged over 15 h (Fig. 3, curve b).In the presence of 1 mM cycloheximide, the early response maintained the R value of 17 while the long-term response was strongly attenuated since it decreased to the R value of 5 at 30 h (Fig. 3,  curve c).These data reveal that the treatment for up to 5 h affects the specific enzyme activity since it did not change significantly the total amount of extractable proteins.In a marked contrast, the long term-treatment promoted a large increase of the total extractable protein content, suggesting on involvement of a de novo synthesis of proteins in the delayed response (not shown).

Carbohydrate analysis
The 1 H spectrum of the oligosaccharide fraction III displayed 5 signals in the anomeric region at 5. 37, 5.18, 5.11, 4.95 and 4.65 p.p.m. (Fig. 4) and other very weak signals (not assigned) which represented no more than 1% of the total integrated resonance.
The doublets at 5.37 and 4.65 p.p.m., based on chemical shifts and coupling constants, were attributed to the aand b-anomeric protons of a reducing end glucose unit.The integration of these resonances yielded an anomeric equilibrium (%) a/b: 52/48.The 3 J 1,2 coupling constants displayed (about 2 Hz) by the two remaining signals, at 5.18 and 4.95 p.p.m., revealed the a-configuration of mannose.The a-configuration is supported by the internal H-1/H-2 NOE effect (not shown).
This result was confirmed by TOCSY, 1 H-1 H COSY and two-steps relayed COSY, spectra (Fig. 5) for the configurational assignments of the sugar units, by consideration of the vicinal ( 3 J) coupling constant value displayed in its J connectivities (Table 1).
The 1 H-13 C heteronuclear spectrum (HMQC) of the compound allowed the assignment of the 13 C chemical shift (Table 2).
The relatively low-field resonance of the proton of C-2 from a-glucopyranosyl and b-glucopyranosyl was caused by glycosylation and revealed the O-2 substitution.Moreover, the HMBC spectrum (Fig. 6) clearly showed a long-range correlation between the H-1 protons at 5.18 and 4.95 p.p.m. and the C-2 at 76.02 p.p.m. (a-glucopyranosyl) and 80.08 p.p.m. (b-glucopyranosyl), indicating that the distinct 1 H resonances displayed by the two a-mannopyranosyl units are a consequence of the proximity between the anomeric center of the reducing end residue and the glycosidic linkage.
The presence of a minor compound may be correlated with the weak signal observed at 5.11 p.p.m. in the anomeric region of the 1 H spectrum (9.2% of the total anomeric resonance).The FAB-MS spectrum of methylated fraction III indicated the existence of a minor mannopyranosyl-cyclichexitol, demonstrated by the presence of fragments at m/z 491 [M+Na] + and m/z 469 [M+H] + (not shown).The cells (4 ´10 6 ) were challenged with elicitor only (curve a) or with elicitor plus 1 mM cycloheximide (curve c).Curve b is the control response.The cells were challenged by 3-6 mg elicitor up to 30 h, and PAL extracts were then assayed at 290 nm.In ordinate: relative rate of reaction products expressed in mkat/kg protein or by the R value (activity in treated cells vs. controls).Experiments were carried out in triplicate.
The presence of a-mannopyranosyl-(1®x)inositol can be also correlated with the occurrence of the anomeric proton of D-manno- pyranosyl observed at 5.11 p.p.m.
These data unambiguously characterize the disaccharide: a-D-mannopyranosyl-(1®2)a/b-D-glucopyranosyl as the major compound (90.8% of fraction III) and suggest the presence of a-mannopyranosyl-(1®x)-inositol as a minor compound (9.2% of the fraction).The peaks at m/z 477 [M+Na] + and m/z 455 [M+H] + obtained by FAB-MS (not shown) additionally confirm the presence of these major compounds.

A carbohydrate signaling molecule
Fusarium oxysporum like many fungi contain in its cell wall various compounds such as proteins, glycoproteins (Jikibara et al., 1992a), lipids (Iwahara et al., 1996) and free oligosaccharides (Bruneteau, 1992).Previous extensive structural studies carried out on the carbohydrate Fusarium cell wall components indicated the presence of b-D-(1®3) [1®6] linked glucose-and uronic acid-contain-ing glycoproteins (Iwahara et al., 1995;Jikibara et al., 1992a;1992c) (Jikibara et al., 1992b;Iwahara et al., 1995) including F. oxysporum.The oligosaccharides from Fusarium were mainly composed of mannose residues with an a(1®2) linkage, but neither linkages with glucose nor inositol have been detected up to now.To our knowledge, the structure a-D-mannopyranosyl-(1®2)-a/b-Dglucopyranosyl derived from our biological source is reported here for the first time.This structure is identical to that of a disaccharide methylglycoside previously described by Shashkov et al. (1993) and synthesized with the aim of developing a computerized approach for the structural elucidation of regular branched polysaccharides.
Oligosaccharides have been often described as active molecules in plant defense reactions.It is known that some oligosaccharides originating from fungal cell wall components 630 M. Nita-Lazar and others 2004 elicit phytoalexin accumulation and lignin or callose formation in plants (Kauss et al., 1989;Lesney, 1989).Furthermore, the system recognition of an oligosaccharide signal is highly sensitive and selective (Prome, 1996).The present study reveals the induction of PAL in Rubus cells within a few minutes after the application of the disaccharide elicitor.These very rapid responses are transient and they are maintained in the presence of cycloheximide.This suggests an early signal transduction cascade, which may be associated with plasma-membranes as suggested by Nurnberger et al. (1997), and provides evidence for post-translational regulation of the enzyme.In addition, changes in PAL activity in the presence of different elicitors were compared.The oligosaccharide fraction originating from Fusarium sp.M7-1 glycoproteins exhibits an R value of 108 (Nita-Lazar et al., 2002).According to our observations when cells instead of protoplasts were monitored, the detected response was attenuated by 35-40% but this can not be the only explanation of the difference in the PAL response between the data presented here and those from  Nita-Lazar et al. (2002).A different Fusarium species was previously used, but the main difference came from the fact that the previous elicitor fraction was a complex mixture of glycans of different length and carbohydrate units in which it was very difficult to identify the active compound and the specific elicitor response.In the present work we made important steps forward in addressing this question.
A delayed PAL response which decreased in the presence of cycloheximide, and therefore involved late changes in PAL mRNA translation levels, has also been detected.PAL elevation, as an indicator of plant resistance, has often been reported to occur at the transcriptional level.It is worth noting that the level of PAL mRNA does not always increase proportionally to the measured enzyme activity (Lee et al., 1992), thus suggesting that PAL can also be affected at posttranscriptional stages (Shaw et al., 1990).Up to now, enzymes of the phenol pathway have often been described as playing a central role in the orchestration of hypersensitive responses in plants, but they are shown here for the first time to be dependent on the application of a well-identified elicitor from Fusarium.
We thank Prof. Jean Montreuil and Dr. Gerard Strecker for fruitful discussions.

Figure 2 .
Figure 2. PAL responses in Rubus cells.The cells (4 ´10 6 ) were challenged for 30 min with 3-6 mg elicitor I, II, III.Plot C represents non-induced cells (control).In ordinate: relative rates of reaction products expressed in mkat/kg protein.I, II and III: fractions I, II, III, respectively.Experiments were carried out in triplicate and the standard deviation was between ±1.5 to 2.7% of each plot.

Figure 3 .
Figure 3. Time-course for PAL response in Rubus cells.
FAB-MS spectra were obtained on both an MS 50 KRATOS-AEI instrument (Manchester, U.K.) and a Nermag R 1010C mass spectrometer (Model 2000, Nermag, Rueil-Malmaison, France) equipped with an M Scan Wallis-type gun (8 kV, 20 mA).The samples were first dissolved in a glycerol matrix and submitted to Xe (9 kV) bombardment.