Plant Science 175 (2008) 818–825 Contents lists available at ScienceDirect Plant Science journal homepage: www.elsevier.com/locate/plantsci Phyllostictine A, a potential natural herbicide produced by Phyllosticta cirsii: In vitro production and toxicity Maria Chiara Zonno a,*, Maurizio Vurro a, Sergio Lucretti b, Anna Andolfi c, Carmen Perrone c, Antonio Evidente c a b c Institute of Sciences of Food Production, National Research Council, via Amendola 122/O, 70125 Bari, Italy Plant Genetics and Genomics Section, Biotechnologies, Agro-industries and Health Protection Department, ENEA Casaccia Research Centre, Via Anguillarese 301, 00123 Rome, Italy Department of Soil, Plant, Environmental and Animal Production Sciences, University of Naples Federico II, via Università 100, 80055 Portici, Italy A R T I C L E I N F O A B S T R A C T Article history: Received 7 May 2008 Received in revised form 31 July 2008 Accepted 5 August 2008 Available online 19 August 2008 Phyllostictine A is a powerful toxin produced by Phyllosticta cirsii, a potential mycoherbicide of Cirsium arvense. To support its potential use as a natural herbicide, toxin production has been studied using different media and cultural conditions. The toxin content in the crude extracts has been determined by using a HPLC method set up for this purpose. Furthermore, its phytotoxicity has been evaluated on tobacco protoplasts by flow cytometric analysis, and on C. arvense protoplasts, by fluorescence microscopy. The best cultural conditions found allowed to produce more than 28 mg ml 1 of toxin in culture filtrate. The pure metabolite proved to have rapid dose-dependant toxic effects on host and nohost plant protoplasts. ß 2008 Elsevier Ireland Ltd. All rights reserved. Keywords: Phyllostictine A Flow cytometry Cirsium arvense Nicotiana tabacum HPLC analytical method Protoplasts 1. Introduction Weed pathogens are considered interesting potential sources of novel natural herbicides. Recently Phyllosticta cirsii has been proposed as a potential mycoherbicide for the biological control of the noxious and widespread weed, Cirsium arvense, commonly known as Canada thistle [1]. Species belonging to the genus Phyllosticta are known to produce bioactive metabolites, including non-host phytotoxins, e.g.: phyllosinol, brefeldin and PM-toxin, isolated by cultures of Phyllosticta sp., P. maydis and P. medicaginis, respectively [2–4]. This led us to investigate the production of toxins by P. cirsii [5]. From the fungal liquid culture phyllostictine A (1, Fig. 1) was purified and chemically characterized as a new oxazatricycloalkenone. This toxin, isolated together with other 3 related metabolites (named phyllostictines B–D) (2–4, Fig. 1), in preliminary bioassays showed interesting biological properties. In particular, it proved to have a noteworthy phytotoxicity, no antifungal activity and an antibiotic and zootoxic activity at high concentrations. Considering its interesting and promising properties, phyllostictine A was proposed as a potential natural herbicide * Corresponding author. Tel.: +39 080 5929332; fax: +39 080 5929374. E-mail address: mariachiara.zonno@ispa.cnr.it (M.C. Zonno). 0168-9452/$ – see front matter ß 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.plantsci.2008.08.003 deserving further studies on its phytotoxicity. Three main objectives were defined for the present study: (a) finding the best conditions for phyllostictine A in vitro production, in order to obtain sufficient amounts of the toxin to perform the biological assays, and to evaluate the capability of the fungus to produce this metabolite in vitro; (b) setting up a rapid analytical method to determine the phyllostictine A content in the culture filtrates, in order to easily compare the influence of the different cultural conditions on the toxin production; (c) assaying the phytotoxic activity of phyllostictine A, to better evaluate its potential as natural herbicide. 2. Material and methods 2.1. Optimizing fungal growth and phyllostictin A production A strain of Phyllosticta cirsii was supplied by Alexander Berestetskiy, Russian Research Institute of Plant Protection, Saint-Petersburg, Russia. It was stored in the mycological collection of Institute of Science of Food Production (ITEM N. 8964) in 20% glycerol at 80 8C. For inoculum production, it was transferred to potato-dextrose-agar (PDA) plates and grown at 25 8C under near UV lights for 2 weeks in order to obtain a sufficient production of picnidia. Plates were then used to prepare M.C. Zonno et al. / Plant Science 175 (2008) 818–825 819 Fig. 1. Structures of phyllosticnines A–D (1–4). conidial suspensions. In order to find the best conditions for fungal growth and toxin production, the fungus was grown in four mineral liquid media, namely Malt extract [6], Fries modified [7], M-1-D [8] and Kent-Strobel [9]. For each medium, both Erlenmeyer flasks (containing 400 ml of medium) and Roux bottles (containing 200 ml medium) were used, for shaken and static conditions, respectively. Aliquots of suspensions (approximately 106 conidia ml 1) were used to inoculate flasks and bottles (0.1 and 0.2 ml respectively) that were then incubated in shaken (100 rpm) or static conditions, respectively, at the dark and at 25 8C. Static cultures were harvested at week intervals, up to 9 weeks, whereas the shaken cultures were harvested at four-day intervals, up to 20 days. All the cultures were produced in duplicate. The fungus was also grown on M-1-D medium with reduced saccharose content (1/2, 1/5 or 1/10 of the standard recipe, i.e. 28.25 g l 1), or with glucose (28.25 g l 1) as carbon source, or even with the addition to the standard M-1-D medium [10] of minced fresh leaves of C. arvense (2 g l 1), collected from naturally infested fields near Bari. These inoculated media were incubated for 7 weeks in static condition at 25 8C at the dark. After incubation, the cultures were filtered on filter paper (Whatman N.4), mycelium fresh and dry weights were determined and pH of culture filtrates was measured. All culture filtrates were tested for the toxicity by using the leaf puncture assay described below. Aliquots (100 ml) were lyophilized and stored until the determination of the phyllostictin A content, as described below. 2.2. Leaf puncture bioassay To compare the biological activity of the culture filtrate to the phyllostictine A content, the toxicity of all the cultured filtrates obtained under various conditions and media was tested by a puncture assay on thistle leaves. Young leaves were cut from C. arvense plants grown in the greenhouse, obtained from seeds collected from wild plants in a field near Bari. Droplets (20 ml) of solutions were placed on punctured leaves that were kept in a moist chamber at room temperature for 72 h. The diameter of the necrotic area was measured. 2.3. HPLC analysis—toxin content The lyophilised fungal culture filtrates (corresponding to 50 ml for the static cultures and to 100 ml for the shaken cultures), obtained as reported above, were extracted tree times by CHCl3– iso-propanol (9:1, v/v) solution (15 and 30 ml for static and shaken 820 M.C. Zonno et al. / Plant Science 175 (2008) 818–825 cultures respectively). The organic phases were combined, filtrated on paper (Whatman N.4), and evaporated under reduced pressure. The samples were dissolved in methanol prior to their use. The phyllostictine A standard sample was purified and identified from P. cirsii culture filtrates as described in a previous work [5]. The HPLC calibration curve for quantitative phyllostictine A determination was performed with standard amounts of the pure toxin dissolved in methanol in the range between 0.05 and 20 mg ml 1 and each concentration was run in triplicate. HPLC linear regression curve (absolute amount against chromatographic peak area) for phyllostictine A was obtained as regression weighted lines calculated from nine multiple injections of the standard metabolite in the concentration range above indicated. For the determination of the phyllostictine A content in the fungal cultures, aliquots of the samples (20 ml) were injected for analysis, having the mobile phase consisting in acetonitrile and HPLC grade water (1:1, v/v), and with a flow rate of 1 ml min 1. Detection was performed at 263 nm, corresponding to the maximum phyllostictine A absorption [5]. Each sample analysis was in duplicate. The quantitative determination of the metabolite was calculated interpolating the mean area of their chromatographic peaks with the data of the calibration curves. concentrations [12]. Furthermore, the herbicide Glyphosate (trade name: Glifone, 360 g l 1 of pure active compound) was tested at four concentrations, i.e.: the technical dose (t.d.: 500 l ha 1 corresponding to 4.2  10 3 M) and three decimal dilutions corresponding to 1/10, 1/100 and 1/1000 of the t.d. 2.7. Cirsium protoplasts Aliquots (1 ml) of protoplast suspensions containing the metabolites at the desired concentrations were placed in Eppendorf tubes and incubated for 1, 3 and 6 h at 27 8C at the dark. After incubation, 50 ml of the protoplast suspension was added with 5 ml of fluorescine diacetate (FDA: stock 1 mg ml 1) and further incubated for 5 min. Droplets (15 ml) of treated cell suspensions were transferred onto a microscope slide, and the viability of cells Cell viability was assessed by using an epifluorescence microscope (Leitz, Orthoplan). Cells with intact plasma membranes fluoresced bright green, while dead cells remained unstained or weakly yellow-green. Methanol solutions (1%) were also used as control to assess cellular toxicity of the solvent. All assays were repeated in triplicate. Toxicity was expressed as protoplast viability percentage as: Number of green fluorescent protoplasts/Total number of protoplasts  100 and each sample was compared to the untreated control. 2.4. Recovery studies 2.8. Tobacco protoplasts Recovery studies were performed using the filtrate obtained by the 2-week-old fungal culture on M-1-D medium. Pure phyllostictine A was added to the culture filtrate in the range between 0.3 and 2.0 mg l 1. The samples were managed as described above and the extracts analysed by the HPLC method in order to determine the recovery values. Three replicate injections were performed for each concentration. 2.5. Protoplast bioassays Tobacco leaf protoplast suspension (3.6 ml) were placed in sterile 5 ml polystyrene round-bottom tubes (Falcon 2254, Becton Dickinson, Franklin Lakes, NJ USA) and incubated as described before. Protoplast suspension (400 ml) was stained with 0.1 mg ml 1 FDA and incubated for 7 min in polystyrene tubes. The toxicity was evaluated by using a FACSCalibur flow cytometer (Beckton-Dickinson, USA), by recording the green fluorescein emission (FL1) collected with a 530/42 nm band pass filter and red chlorophyll auto-fluorescence (FL2) (670/13 nm band pass filter). Data were recorded and analysed using the software CellQUEST 3.1 (Becton Dickinson, USA). Viable leaf protoplasts were characterized by simultaneous emission of green and red fluorescence, FDA and chlorophyll emission, respectively. Five thousands events were collected for each analysis at a flow rate of 100 cells s 1. Assays were repeated three times and all concentrations were tested in triplicate. Toxicity was expressed as percentage of viable protoplasts and compared to the control. 2.5.1. Protoplasts extraction For the production of protoplasts, plants of C. arvense were grown in soil in plastic pots (12 cm diameter) by sowing seeds harvested in naturally infested fields near Bari, Italy. Nicotiana tabacum cv. Samsung plants were grown on Murashige and Skoog Basal Medium (Sigma) agar (1.2% of agar) medium in sterile Magenta boxes (Sigma, 100 ml medium box 1). Plants were grown with a photoperiod of 16 h light at 25 8C (lamps OSRAM HQIT 400 W/N/SI), for 1 month (thistle) or for 8 weeks (tobacco). Leaf tissues were cut with a razor blade into small strips (about 1mm wide) and placed in Petri dishes containing 10 ml of an enzymatic solution made of Cellulase ‘‘Onozuka’’ R-10 (Yakult, Honsha Co. Ltd, Nigashi Shimbashi, Tokyo, Japan) (0.2%, w/v) Driselase (0.025%, Sigma), Macerozyme R-10 (0.025%, Yakult Honsha Co. Ltd, Nigashi Shimbashi, Tokyo, Japan) and D-Mannitol (0.5 M, Sigma–Aldrich), CaCl2
H2O (10 mM, Carlo Erba Analyticals) and MES (3 mM, Sigma–Aldrich). Suspensions were put on an orbital shaker overnight at 27 8C, then filtrated through a cell strainer (70 mm) and re-suspended twice in the holding buffer after centrifugation at 50  g for 10 min. Protoplasts were cultured in the buffer solution at a final concentration of 105 protoplasts ml 1 [11]. 2.8.1. Chemicals and equipments Analytical and HPLC grade solvents for chromatographic use were acquired from Carlo Erba (Milan, Italy). All other analytical grade chemicals were purchased from Merck (Darmstadt, Germany). Water was HPLC quality, purified in a Milli-Q system (Millipore, Bedford, MA, USA). The HPLC system employed was supplied by Shimatzu (Tokio, Japan) and consisted of a Series LC-10AdVP pump, FCV-10AlVP valves, SPD-10AVVP spectrophotometric detector and DGU-14A degasser. The HPLC separations were performed using a MachereyNagel (Duren, Germany) high-density reversed-phase Nucleosil 100-5 C18 HD column (250 mm  4.6 mm i.d.; 5 mm) provided with an in-line guard column from Alltech (Sedriano, Italy). 2.6. Protoplast bioassay 3. Results Protoplasts were exposed at 5 different concentrations of phyllostictine A, i.e.: 10 3, 5  10 4; 10 4; 5  10 5 and 10 5 M). For comparative studies, the fungal toxin fusaric acid (Sigma– Aldrich) dissolved in methanol (1%) was tested at the same 3.1. Fungal growth If the dry weight of the P. cirsii mycelium is considered, the best medium for the fungal growth proved to be the malt extract, both 821 M.C. Zonno et al. / Plant Science 175 (2008) 818–825 Table 1 Toxicity of the culture filtrates obtained by growing the fungus on different media in static conditions Weeks Media M-1-D 1 2 3 4 5 6 7 8 9 + +++ ++++ +++ +++ ++ Fries ++ + + + + Malt Kent-Strobel + + + + + + + + + + + The assay was performed on punctured thistle leaves. Scale of toxicity: , nontoxic; +, necrosis 1 mm; ++, 2–3 mm; +++, 3–5 mm; ++++, wider necrosis. Table 2 Toxicity of the culture filtrates obtained by growing the fungus on different media in shaken conditions Days Media M-1-D 4 8 12 16 20 Fries Malt Kent-Strobel + – – – – ++ +++ +++ + + + + + The assay was performed on punctured thistle leaves. Scale of toxicity: , nontoxic; +, necrosis 1 mm; ++, 2–3 mm; +++, 3–5 mm; ++++, wider necrosis. 3.2. Leaf puncture bioassay Fig. 2. Growth of Phyllosticta cirsii on different media in static conditions up to 9 weeks. Changes in fungal dry and fresh weights and pH of the culture filtrates. After 72 h incubation, the strongest toxicity was observed on leaves treated with M-1-D culture filtrates from 4 to 7 weeks (Table 1, static culture). A lower phytotoxic activity was caused by younger or older M-1-D filtrates, whereas an almost negligible or nil toxicity was observed all the way for the other three media used (Table 1). In shaken conditions, phytotoxic activity was also observed for the culture filtrate obtained growing P. cirsii on Fries medium, whereas negligible was the toxicity for the other three media used (Table 2). 3.3. Toxin quantification in shaken (data not shown) and static (Fig. 2) conditions. For this latter, both fresh and dry weights increased during the whole experiment, reaching more than 34 g and almost 3.5 g flask 1, respectively. Intermediate results were obtained using Fries or M1-D medium (Fig. 2), whereas the worst medium proved to be Kent-Strobel medium for both conditions, with an almost negligible growth. With regard to the pH of the culture filtrates, in the static conditions it was quite stable for M-1-D and Malt, ranging between 4.8 and 5.8, whereas it was more variable for the other two media. With regard to the method set up for the quantification of the phyllostictine A content in the culture filtrates, the characteristics of the calibration curves, the absolute amount range and the detection limits (LOD) of phyllostictine A are reported in Table 3. The linearity of the calibration curves in the reported interval is proven by the regression data. The best elution conditions to quantify phyllostictine A content in the extract from culture filtrates of P. cirsii were obtained on C18 reverse phase using isocratic elution with acetonitrile and HPLC grade water (1:1, v/v) with a flow rate of 1 ml min 1. HPLC Table 3 Analytical characteristics of calibration curvea for phyllostictine A tR (min) Range (mg) Slope Intercept S.D. (y%) r2 Number of data point Limit of detection (LOD) (ng) 13.8 0.05–20 9153.6 917.7 0.64 0.9977 27 4.8 tR: retention time. a Calculated in the form y = a + bx where y is chromatographic peak area and x is mg of toxin. 822 M.C. Zonno et al. / Plant Science 175 (2008) 818–825 Fig. 3. HPLC profile of phyllostictine A standard peak (1) (A) and CHCl3–iso-propanol extract of P. cirsii culture filtrate (B). chromatographic profiles of the standard phyllostictine A (A) and of the CHCl3–iso-propanol extract of P. cirsii culture filtrate (B) are reported in Fig. 3. In samples the chromatographic peak (1) of phyllostictine A was identified by the coincidence of the retention time with that of the standard, which was eluted at 13.8 min with a high reproducibility (0.50 min). For all the samples, the matrix substances absorbing at 263 nm were eluted within the first 6 min. This finding, together with the high similarity of the features of the metabolite peak in samples with those of standard, allow to indicate that no peaks of other substances overlap those investigated. Injecting phyllostictine A in amounts higher than the upper limit of the indicated ranges it was still possible to have a linear response. Amounts below the lower limit were still detectable but with a low reproducibility of peak integration, which highly increased the standard deviation of the regression curves. As shown in Table 4, the content of phyllostictine A in the culture filtrates obtained using different cultural conditions ranged between 0.02 and 28.8 mg l 1. Phyllostictine A was produced with a linear increase since the first week (2.2 mg l 1) until the fourth, when it reached the maximum (28.8 mg l 1). After that, the production linearly decreased until the seventh week (15.2 mg l 1). In Table 4, the fungal growth kinetic data are shown up to 9 weeks in static condition onto M-1-D medium. A similar trend was obtained also for the other two media, Fries and Malt, but the production of phyllostictine A was much lower than M-1-D. It linearly increased since the first week (0.08 and 0.04 mg l 1, for Fries and Malt, respectively), reached the maximum at the fourth week (0.59 and 0.96 mg l 1, for Fries and Malt, respectively), and decreased linearly until the ninth week (0.22 and 0.02 mg l 1, for Fries and Malt, respectively). Onto KentStrobel medium the fungus did not produce the toxin in any of the conditions used (data not shown). Under shaken conditions on M-1-D, Fries and Malt media the fungus produced phyllostictine A only after 4 days and in very low amounts, and only in the latter medium (1.46 mg l 1), whereas no production was detected for longer incubation times, irrespective the medium used (data not shown). The production of phyllostictine A using glucose as carbon source gave a toxin yield (7.9 mg l 1) lower than saccharose. The production of phyllostictine A strongly decreased when using lower amounts of saccharose in the medium, lowering to 4.83, 1.75 Table 4 Production of phyllostictine A by Phyllosticta cirsii grown on different mediaa in static conditions up to 9 weeks Weeks 1 2 3 4 5 6 7 8 9 a Toxin content (mg l 1 ) M-1-D Fries Malt 2.12 3.59 13.72 28.83 23.69 21.93 18.18 16.91 15.23 0.04 0.05 0.08 0.96 0.52 0.22 0.11 0.06 0.02 0.08 0.15 0.35 0.59 0.38 0.36 0.30 0.26 0.22 M-1-D, fries, and malt are different media. M.C. Zonno et al. / Plant Science 175 (2008) 818–825 and 0.05 mg l 1 for the production in M-1-D medium containing 1/ 2, 1/5 and 1/10 of the standard recipe saccharose content, respectively (data not shown). A slight increase in phyllostictine A production was obtained in 7-week-old cultures by the addition of C. arvense leaves (21.9 mg l 1 compared to 19.1 mg l 1 of the standard culture). 3.4. Recovery studies The recovery of phyllostictine A added at various amounts to the culture filtrate ranged between 94  3.5 and 97  2.1% (data not shown). These results also indicated that the extraction by CHCl3:isopropanol (9:1, v/v) was a satisfying procedure for quantitative analysis of the metabolite in the culture filtrates. 823 3.5. Thistle protoplasts assay In the assay on thistle protoplasts, at the highest concentration assayed (10 3 M) phyllostictine A was strongly effective already 1 h after its application, being able to kill almost the totality of protoplasts (Fig. 4). At 5  10 4 M the toxicity was time dependant, being mortality around 50% 1 h after the treatment and reaching 100% after 6 h. At lower concentrations the toxicity of phyllostictine was low even after 6 h, being almost completely inactive at 10 5 M. With regard to fusaric acid used for comparison, it proved to have a time-dependant toxicity at the highest concentration tested (10 3 M) (Fig. 4), causing around 60% mortality 1 h after the application, reaching 100% about after 6 h. The other treatments were sensibly less toxic, and not dependant by the time of exposure. In case of glyphosate, it proved to be highly toxic at the t.d. causing the total mortality of thistle protoplasts 3 h after the treatment, whereas at 1.10 and 1:100 dilutions it caused around 60% protoplast mortality after 3 h of exposure to the toxin. At the lower dose (1/1000 t.d.) it caused a noteworthy (60%) protoplast mortality only 6 h after the treatment. 3.6. Tobacco protoplasts assay In Fig. 5, fluorescence emissions from tobacco protoplasts treated with phyllostictine A and analysed by flow cytometry, are shown. Damaged protoplasts showed less and less fluorescent Fig. 4. Toxicity of phyllostictine A (upper) fusaric acid (middle) and glyphosate (lower) at different concentration assayed on thistle protoplasts (% of viability compared to the control). Error bars indicate standard deviations of the measurements for the three replications. Fig. 5. Citogram (dot-plot) of tobacco protoplast vitality exposed to phyllostictine A at 10 3 M (upper) and to methanol (1%) as control (lower) for 240 min. Quadrants: UR + LR = Live protoplasts UL + LL = Debris and dead cells. 824 M.C. Zonno et al. / Plant Science 175 (2008) 818–825 emissions corresponding cell membrane damage (green fluorescence after FDA staining) and subsequent chloroplast releasing into the medium (red fluorescence from chlorophyll) At the highest concentration tested (10 3 M) phyllostictine A acted very quickly. In fact, already 1 h after the treatment the toxin caused the death of almost all the protoplasts (Fig. 6). At 5  10 4 M the toxin was still very fast acting, causing around 80% mortality already 1 h after the treatment. At 10 4 M the toxicity was time dependant, being considerable (around 60%) after 3 h of exposure. At the lowest concentration tested (10 5 M) there was no detectable toxic activity (Fig. 6). With regard to fusaric acid, it proved to be highly Fig. 6. Tobacco protoplast viability (% of the control) after treatment with phyllostictine A (upper), fusaric acid (middle) and glyphosate (lower) at different concentration. Error bars indicate standard deviations of the measurements for the three replications. toxic al 10 3 and 5  10 4 M, causing almost 100% mortality after 6 h, whereas at 10 5 M around 60–70% protoplasts remained alive after 6 h of exposure. Glyphosate was highly toxic assayed at t.d., whereas proved to be much less toxic at the reduced doses assayed (Fig. 6). 4. Discussion One of the problems in studying and using bioactive metabolites produced by plant pathogens is that they are often produced and released in the culture medium in very low amounts, making the purification and identification very difficult to be performed. Beside the natural capability of the microbes to biosynthesize the compounds, the cultural conditions can strongly influence their production in vitro. In our study, the use of M-1-D medium, with a chemically defined composition, and the growth of the fungus in static conditions for 4 weeks proved to be the most effective cultural condition, allowing a production higher than 28 mg l 1. Our data showed that toxin production was not related to the fungal growth. In fact, on malt extract medium the fungus grew very well, producing abundant mycelium, but toxin production was very low. When growing the fungus for more than 4 weeks we observed a progressive reduction in the toxin content, which is probably due to the instability of phyllostictine (Table 4). The toxicity of the culture filtrates seems to be correlated to the toxin content, being the most phytotoxic culture filtrates (i.e.: static M-1-D cultures between 4 and 7 weeks) (Table 1) also those with the highest toxin content (Table 4). Anyway, the presence of other known (e.g. phyllostictine B, C and D already described) [5] or novel phytotoxins cannot be excluded. This possibility is much more evident in the case of the shaken conditions, in which a quite interesting phytotoxicity was observed for the cultures on Fries medium, that contains low or nil amounts of phyllostictine A. The influence of the cultural conditions and medium types on toxin production is well known from many plant pathogens. For example, the production of the phytotoxic b-nitropropionic acid by Septoria cirsii was increased by the addition of thistle leaves to the M-1-D medium [13]. Ascochyta sonchii, a leaf pathogen of Sonchus arvensis, produce a large amount of ascosonchine (8 mg l 1) when grown on M-1-D static culture for 8 weeks, but less than 1 mg l 1 when grown in shaken culture [14]. Ascochyta rabiei produces solanapyrones A,B,C only if Czapek-Dox medium was added with extract of host plant seeds [15,16]. Moreover, in shaken culture A. rabiei produces only 0.1–10% of the solanapyrones produced in static culture [17]. The HPLC method for the qualitative and quantitative analysis of phyllostictine A proved to be efficient, rapid and sensitive. This method proved to be very useful for the identification of the best cultural conditions for toxin production, and could also be used for comparing the ability of different fungal strains to produce phyllostictine A. Furthermore, it could be a valuable tool for mycoherbicide studies. One of the main problems in using fungi as mycoherbicides is the assessment of the strain virulence, which is highly time and space consuming. If there were a positive correlation between strain virulence and production of toxins, it could be easier to select the most aggressive strains simply selecting the best toxin producers. Furthermore the method could also be useful to evaluate the toxin yield in the large scale productions. Compared to the use of whole plants, bioassays performed on isolated protoplasts can offer some advantages, e.g. the use of lower amounts of toxins, a shorter time for the analysis and clearer host responses at the cellular level. M.C. Zonno et al. / Plant Science 175 (2008) 818–825 Flow cytometry can produce information about protoplast size, chlorophyll auto-fluorescence and fluorescence due uptake and subsequent esterase cleavage of FDA into the cytoplasm, therefore producing the strong green fluorescent compound fluorescein [18]. Normally, this fluorochrome remains into the protoplasm since its a polar compound and cannot cross freely the cell membrane. If the membrane loses somehow its integrity fluorescein can diffuse out of the cell which then is no more fluorescent. Flow cytometry generates data which categorize intact protoplasts emitting both red chlorophyll auto-fluorescence and green fluorescein and cellular debris and released organelles. Analysis of the auto-fluorescence of unstained tobacco protoplasts produce histograms containing debris and dead cells (LL and LR) and red emitting mesophyll protoplasts (UL); after staining with FDA another population was apparent (UR). The latter population was made of active protoplasts exhibiting both high chlorophyll auto-fluorescence level and fluorescine fluorescence from FDA. Only protoplasts were analyzed whereas chloroplasts and debris were excluded from the analysis. Viable protoplasts had bright fluorescence during flow cytometric analysis, whereas dead protoplasts were nonfluorescent (Fig. 5). Auto-fluorescence of the protoplasts not treated with phyllostictin A was used as experimental control to monitor the integrity of protoplast preparations combined with FDA staining. In our bioassays on protoplasts, at the highest concentration used phyllostictine proved to act faster then glyphosate at the t.d. (Fig. 4). This strong activity, together with the appearance of necrosis when applied to leaves, could support the idea of using the toxin as a herbicide for foliar application and tissue desiccation. Compared to FA, which is a well known and powerful toxin, phyllostictine proved to be much more potent (Fig. 4). Acknowledgements This work was carried out within the project ‘‘Enhancement and Exploitation of Soil Biocontrol Agents for Bio-Constraint Management in Crops’’ (contract no. FOOD-CT-2003-001687), which was financially supported by the European Commission within the 6th FP of RTD, Thematic Priority 5—Food Quality and Safety. The research was also in part supported by a grant from Regione Campania L.R. 5/02. Contribution DISSPAPA N. 158. The authors thank Dr. Elisabetta Sbisà, Dr. Apollonia Tullo and Dr. Beatriz 825 Navarro, Institute of Biomedical Technologies, CNR, Bari, Italy, for the assistance in managing cell bioassays. References [1] A.O. Berestetskij, T.Y. Gagkaeva, Ph.B. Gannibal, E. Gasich, O.V. Kungurtseva, G.V. Mitina, O.S. Yuzikhin, I.V. Bilder, M.M. Levitin, Evaluation of fungal pathogens for biocontrol of Cirsium arvense, in: Proceedings 13th European Weed Research Society Symposium, Bari, Italy, June 19–23, 2005 (Abstract 7). [2] S. Sakamura, H. Niki, Y. Obata, R. Sakai, T. Matsumoto, Isolation and structure of phytotoxic compounds produced by Phyllosticta species, Agric. Biol. Chem. 33 (1969) 698–703. [3] C. Comstock, C.A. Martinson, B.G. Gengenbach, Host specificity of a toxin from Phyllosticta maydis for Texas cytoplasmically male-sterile maize, Phytopathology 63 (1973) 1357–1361. [4] I.D. Entwistle, C.C. Howard, R.A.W. Johnstone, Fungal metabolites. II: Isolation of brefeldin A from Phyllosticta medicaginis, Phytochemistry 3 (1974) 173–174. [5] A. Evidente, A. Cimmino, A. Andolfi, M. Vurro, M.C. Zonno, C.L. Cantrell, A. Motta, Phyllostictine A–D, oxazatricycloalkenones produced by Phyllosticta cirsii, potential mycoherbicide of Cirsium arvense, Tetrahedron 64 (2008) 1612–1619. [6] N. Zelikovitch, Z. Eyal, Y. Kashman, Isolation, purification and biological activity of an inhibitor from Septoria tritici, Phytopathology 82 (1992) 275–278. [7] J.F. Bouquet, H. Belhomme de Franqueville, A. Kollman, R. Fritz, Action de la septorine, phytotoxine synthétisée par Septoria nodorum, sur la phosphorylation oxydative dans les mitochondries isolées de Coléoptiles de Blé, Can. J. Bot. 73 (1980) 2575–2580. [8] F. Pinkerton, G.A. Strobel, Serinol as an activator of toxin production in attenuate cultures of Helminthosporium sacchari, Proc. Natl. Acad. Sci. U.S.A. 73 (1976) 4007– 4011. [9] S.S. Kent, G.A. Strobel, Phytotoxin from Septoria nodorum, Trans. Br. Mycol. Soc. 67 (1976) 355–358. [10] D.J. Robeson, G.A. Strobel, The influence of plant extracts on phytotoxin production and growth rate of Alternaria helianthi, J. Phytopathol. 117 (1986) 265–269. [11] L. Crepy, L.M.G. Barros, V.R.N. Valente, Callus production from leaf protoplasts of various cultivars of bean (Phaseolus vulgaris L.), Plant Cell Reports 5 (1986) 124–126. [12] C. Altomare, G. Perrone, M.C. Zonno, E. Evidente, R. Pengue, F. Fanti, L. Polonelli, Biological characterization of fusapyrone and deoxyfusapyrone, two bioactive secondary metabolites of Fusarium semitectum, J. Nat. Prod. 63 (2000) 1131–1135. [13] J. Hershenhorn, M. Vurro, M.C. Zonno, A. Stierle, G. Strobel, Septoria cirsii, a potential biocontrol agent of Canada thistle and its phytotoxin-b-nitropropionic acid, Plant Sci. 94 (1993) 227–234. [14] A. Evidente, A. Berestetskiy, A. Andolfi, A. Cimmino, M.C. Zonno, M. Vurro, Relation between in vitro production of ascosonchine and virulence of strains of the potential mycoherbicide Ascochyta sonchi: a method for its quantification in complex samples, Phytochem. Anal. 17 (2006) 357–364. [15] S.S. Alam, J.N. Bilton, A.M.Z. Slawin, D.J. Wiliams, R.N. Sheppard, R.N. Strange, Chickpea blight: production of the phytotoxins solanapyrones A and C by Ascochyta rabiei, Phytochemistry 28 (1989) 2627–2630. [16] Y.M. Chen, R.N. Strange, Synthesis of solanapyrones phytotoxins by Ascochyta rabiei in response to metal cations and development of a defined medium for toxin production, Plant Pathol. 40 (1991) 401–407. [17] B. Höhl, C. Weidermann, U. Höhl, W. Barz, Isolation of solanapyrones A, B and C from culture filtrates and spore germination fluids of Ascochyta rabiei and aspects of phytotoxin action, J. Phytopathol. 132 (1991) 193–206. [18] D. Berglund, S. Strobel, F. Sugawara, G.A. Strobel, Flow cytometry a method for assaying the biological activity of phytotoxins, Plant Sci. 56 (1988) 183–188. All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.