Mevalocidin: A Novel, Phloem Mobile Phytotoxin from Fusarium DA056446 and Rosellinia DA092917

Abstract

A multiyear effort to identify new natural products was built on a hypothesis that both phytotoxins from plant pathogens and antimicrobial compounds might demonstrate herbicidal activity. The discovery of one such compound, mevalocidin, is described in the current report. Mevalocidin was discovered from static cultures of two unrelated fungal isolates designated Rosellinia DA092917 and Fusarium DA056446. The chemical structure was confirmed by independent synthesis. Mevalocidin demonstrated broad spectrum post-emergence activity on grasses and broadleaves and produced a unique set of visual symptoms on treated plants suggesting a novel mode of action. Mevalocidin was rapidly absorbed in a representative grass and broadleaf plant. Translocation occurred from the treated leaf to other plant parts including roots confirming phloem as well as xylem mobility. By 24 hr after application, over 20 % had been redistributed through-out the plant. Mevalocidin is a unique phytotoxin based on its chemistry, with the uncommon attribute of demonstrating both xylem and phloem mobility in grass and broadleaf plants.

Introduction

Phytotoxins have been studied for their role in pathogen virulence and as starting points for herbicide discovery. The emergence of weed resistance to glyphosate has refocused attention on the need for new herbicides with new modes of action (Gerwick, 2010). Phytotoxins offer a wide range of unique chemistries with commercially unexploited modes of actions, and their strengths and limitations have been described previously (Duke and Dayan, 2011; Dayan et al., 2012). Particularly interesting because of their activity level and spectrum are the non-host specific phytotoxins, such as phaseolotoxin. That the modes of action of these are generally different from those of herbicides discovered in industrial screening programs can be attributed to their unique chemical backbones (Dayan et al., 2012), but may also involve other factors. For example, while natural products are typically oxygen and nitrogen rich and lack halogens (Henkel et al., 1999), many commercial herbicides also share these properties. Of the 291 commercial herbicides for which the modes of action are defined (Herbicide Resistance Action Committee Poster 2010), over 100 of them lack a halogen. Further, as all major modes of action of commercial insecticides are represented by known natural products (Geng et al., 2012), it is puzzling why the same parallel is not evident in herbicides. It seemed surprising to us that no phytotoxins have been described that target acetolactate synthase, one of the most sensitive and commercially significant herbicide modes of action in use today, and one that is inhibited by a number of different chemical classes that lack halogens. In addition to the differences in chemistry described above, the lack of significant overlap in modes of action of phytotoxins and commercial herbicides could reflect fundamental differences in selection, e.g., many phytotoxins also demonstrate strong antibiotic activity against one or more microorganisms (Demain and Fang, 2000). Our hypothesis for sourcing organisms to be screened in our program was that antibiotic activity may also result in phytotoxic activity, particularly for non-host specific toxins that are more associated with virulence than pathogenicity (Graniti, 1991). Competition also occurs among strains of a given pathogen rendering antibiotic production a potential selective advantage (Demain and Fang, 2000). Hence, we screened a wide-range of antibiotic-producing bacteria and fungi as well as fungal plant pathogens for phytotoxic activity over a multi-year period in search of herbicide leads.

Several of the more interesting compounds discovered from these sources in our program include macrocidin [1] (Graupner et al., 2003), cinnacidin [2] (Irvine et al., 2008), albucidin [3] (Hahn et al., 2009), and described in detail in this report, mevalocidin [4]. Some of the information presented here on mevalocidin appears in US patent US7393812 B2 (Gerwick et al., 2008) and an ACS abstract (Graupner et al., 2008).

The macrocidins are produced by Phoma macrostoma, a pathogen of Canada thistle (Graupner et al., 2003). The isolate was originally made from diseased plants, and releasing the isolated fungus back into the soil reestablished disease on healthy thistle plants. Symptoms on infected plants included intense bleaching and stunting of new growth. These symptoms could be reproduced upon treating plants with extracts of the fungus grown in liquid shake-flask culture, which enabled bioassay directed isolation. Several related factors were identified with the major component designated macrocidin A, the first of a novel family of cyclic tetramic acids. The macrocidins were quite active on other members of the Asteraceae besides thistle, but activity on other broadleaf weeds was more limited, with no significant activity on grasses observed. A synthetic effort was initiated but did not enhance the potency or spectrum. The use of Phoma macrostoma as a bioherbicide is under development for the control of broadleaf weeds in turf (Falk and Bailey, 2012).

Cinnacidin is a novel natural product produced by Nectria sp (Irvine et al., 2008). The partial 28 S large subunit ribosomal DNA sequencing most closely matched Nectria cinnabarina and sister taxon, Nectria pseudotrichia, both of which are known plant pathogens. Nectria cinnabarina causes “coral spot” or “Nectria twig disease” and infects a number of woody species (http://www.hedges.ie/Coral-Spot-Disease.html). Nectria pseudotrichia also is a pathogen of woody plants such as pears where it causes stem canker (http://www.cabdirect.org/abstracts/20053045060.html). What was particularly interesting to us was the strong structural resemblance of cinnacidin to both coronatine and jasmonic acid. Coronatine is a bacterial toxin produced by Pseudomonas syringae, where it is a well described virulence factor acting via the jasmonic acid signaling pathway (Nomura et al., 2005; Browse, 2009). Whether cinnacidin plays a similar role in the virulence of Nectria sp. is unknown.

Albucidin was identified in liquid shake flask cultures of Streptomyces albus subsp. chlorinus following bioassay directed isolation (Hahn et al., 2009). It is a new natural product but known by synthesis following synthetic modification of the natural product oxetanocin. Both materials have been reported as having anti-viral activity. Albucidin is a broad spectrum herbicide with activity following both pre-emergence and post-emergence applications. The symptoms of chlorosis and stunting are very slow to develop, and chlorosis is restricted to new growth. In some cases chlorosis did not progress to necrosis and plant death for >30 days. The mode of action is unknown and it is unclear if the anti-viral activity and herbicide activity are linked.

In the current paper, we describe the discovery and structure elucidation of mevalocidin, a broad spectrum phytotoxin produced by two unrelated fungi. We further describe the synthesis of this material and related compounds, the biological activity, and absorption and translocation properties in plants.

Methods and Materials

Identification

Strain DA056446 grew slowly with limited aerial mycelia and produced a dark reddish-orange pigment on Potato Dextrose Agar (PDA) and Malt Extract Agar (MEA). Microscopically, strain DA056446 produced branched hyphae and 4–5 cell macroconidia with a tapered apical cell at each end. This morphology is consistent with the fungi of the genus Fusarium. Strain DA092917 produced abundant white vegetative mycelium with black aerial mycelium on PDA and white mycelium and no aerial mycelium on Oatmeal Agar. No spores or conidia were evident on either media. DNA from strains DA056446 and DA092917 was submitted to Midi Labs (Newark, NJ, USA) for partial DNA sequencing of the large subunit ribosomal RNA gene. The DNA sequence was compared to the GenBank database using the BLAST algorithm (http://www.ncbi.nlm.nih.gov/BLAST/). The strongest BLAST matches for strain DA056446 were various Fusarium species. The strongest BLAST matches for strain DA092917 were various members of the ascomycetous fungal family Xylariaceae particularly the genus Rosellinia. The morphology of the cultured strain DA092917 is consistent with Rosellinia. The partial 28S sequences for DA056446 and DA092917 were deposited in GenBank under the accession numbers DA056446 and DA092917. The two strains were designated Fusarium sp. DA056446 and Rosellinia sp. DA092917, and have been deposited in the patent collection of the NRRL culture collection of the National Center for Agricultural Utilization Research (1815 N. University St., Peoria, IL, USA) under the accession numbers NRRL 30882 (DA056446), and NRRL 30883 (DA092917).

Fermentation

Fusarium sp. DA056446 was fermented on Malt Extract Agar (Becton Dickenson, Sparks, MD, USA) and Rosellinia sp. DA092917 was fermented on Oatmeal Agar (Becton Dickenson, Sparks, MD). Both strains were fermented at 25 °C for 14 days. Mature cultures were extracted by overlay of mycelia with 50:50 ethanol:water or 50:50 methanol:water for 1 h and filtration through a 0.2 μm sterile filter.

Sample Screening

Fungal samples DA056446 and DA092917 were received from MYCOsearch (now Mycosynthetix, Inc., 505 Meadowlands Drive, Suite 103, Hillsborough, NC, USA) in microtitre plate format under an Extract Supply Agreement with Dow AgroSciences. Each well contained 1.6 ml equivalent of dried test material. To each well was added 160 μl of methanol followed by sonication to create a soluble test sample. An 80 μl sample subsequently was transferred to a second shallow well plate (daughter plate) and 120 μl of 50 % methanol were added to create a 1:2.5 dilution. A 100 μl sample from the daughter plate was transferred to a 96 well “bubble tray” (38.5 × 32 × 2 cm). The samples in the bubble tray were air dried prior to testing.

Lemna and Agrostis were used as test species. Lemna was grown to the 4-leaf stage in a growth chamber (16:8 hr, L:D photoperiod, 29 °C) prior to testing. The growth media consisted of 3.1 g/L Gamborg’s B-5 Basal salt mixture, 4.4 g/L Murahige and Skoog modified basal salt mixture (NH4-Free) and 10 g/L sucrose, pH adjusted to 5.5. Agrostis was added to the test as ungerminated seeds. Five ml of surface sterilized Agrostis seed were added to 1,500 ml of growth media and placed on a table top stirrer. An automated dispenser was used to transfer 4.0 ml of the media containing un-germinated seeds into each bubble tray. The Lemna was added to each test well at this point using a sterile loop. The bubble trays were maintained in a Conviron growth chamber (16:8 hr, L:D photoperiod, 19 °C) for 7 days after which visual observations were made on a 0 to 100 scale, where 0 represents no injury and 100 represents complete necrosis. Extracts of fungal samples DA056446 and DA092917 provided >50 % growth reduction of both test species.

Isolation

One liter of freeze dried methanol extracts of cultures DA056446 and DA092917 was received from MYCOsearch and partitioned between n-butanol and water (pH = 4.5). Testing of these fractions by foliar application on whole plants of Helianthus annuus (HELAN) (see below) indicated the highest activity in the aqueous fraction. The aqueous fraction was dried, redissolved in 20 % methanol, and fractionated on a preparative C-18 column (particle size, 10 μ; column size, 40 × 2 cm; mobile phase, 100 % 10 mM ammonium acetate to 50 % acetonitrile in 15 min; flow rate, 10 ml/min; uv detection at 210 nm). Activity was localized to two regions by bioassay on HELAN. The fraction containing the more polar active was dried and re-chromatographed on C-8 AQS (particle size, 5 μm; column size, 15 × 4.6 mm; mobile phase, 0.05 % aqueous TFA; flow rate, 1 ml/min; detection, uv at 210 nm) in 1 mg aliquots. The activity was localized to a peak with retention time of 6.0 min. The sample was collected, dried, and re-chromatographed under identical conditions to yield ~1 mg of clear solid. This was submitted directly for NMR and MS analysis. The fraction containing the later eluting active from preparative chromatography was only weakly active. Nevertheless, it was pursued by fractionation on a C-18 analytical BOS column (particle size, 5 μm; column size, 25 cm × 4.6 mm; mobile phase 5 % aqueous TFA (0.05 %) to 80 % acetonitrile in 25 min; flow rate, 1 ml/min; detection, uv at 210 nm). Bioassay of chromatography fractions indicated activity in two regions; surprisingly, one of the active regions corresponded to the polar active previously isolated. Apparently, the non-polar active was partially converted to the polar active during chromatography. Further efforts to isolate the non-polar active were suspended. As discussed subsequently, the active was shown by NMR to exist in both an open chain and lactone form, which explains the above chromatographic behavior.

Greenhouse testing of samples was completed by methods similar to those described previously (Schmitzer et al., 2000). Applications were made either to the foliage (post-emergence) or soil (pre-emergence). Seeds of sunflower (HELAN), cocklebur (Xanthium strumarium, XANST), crabgrass (Digitaria sanguinalis, DIGSA), morningglory (Ipomoea hederacea, IPOHE), velvetleaf (Abutilon theophrasti, ABUTH), pigweed (Amaranthus retroflexus, AMARE), barnyardgrass (Echinochloa crusgalli, ECHCG), giant foxtail (Setaria faberi, SETFA), wild oats (Avena fatua, AVEFA), and blackgrass (Alopecurus myosuroides, ALOMY) were planted in mineral soil (sandy clay loam, 51 % sand, 26 % silt, 23 % clay, 2.8 % O.M., pH 7.8) and Metro-mix (Bulk Sak, Inc., Malvern, AR, USA) for pre-emergence and post-emergence treatments, respectively. Pots receiving pre-emergence treatments were watered immediately following application. Prior to post-emergence treatment, plants were grown in the greenhouse (16:8 hr, L:D photoperiod, 27 °C) and thinned to a density of 2 to 15 plants/pot, depending on species. At the time of post-emergence applications, plants were 8–12-d-old, 3 to 10 cm in height, and in the 1- to 3-true-leaf stage. Natural light in the greenhouse was supplemented with metal halide lights that provided an average photosynthetic photon flux of 500 μmol/m²/s.

Samples were dissolved in 4 ml of distilled water followed by 10 ml of formulation stock “B” (200 ml isopropanol, 20 ml crop oil concentrate, 0.4 g Triton X-155 (Sigma Chemical Co., St. Louis, MO, USA), and 78 ml water). This was followed by three 1:1 (v/v) serial dilutions using formulation stock “C” (100 ml isopropanol, 10 ml crop oil concentrate, 0.2 g Triton X-155, 390 ml water, 485 ml acetone, and 15 ml DMSO). Treatments were made at four different doses (4000, 2000, 1000, and 500 g/ha). For pre-emergence treatments, 1.75 ml of spray solution were applied to the soil surface in each pot using a glass Cornwall syringe fitted with a Teejet SS8001 E flat-fan nozzle (Spraying Systems Co., Wheaton, IL, USA). For post-emergence applications, the foliage of the test plants was sprayed using a DeVilbiss atomizer (DeVilbiss Health Care, Inc., Somerset, PA, USA) driven by compressed air at a pressure of 22 kPa. Untreated controls and controls of solvent alone were included for reference. Pots were placed in the greenhouse following compound application. Pots from post-emergence and pre-emergence treatments were sub-irrigated and top watered, respectively, as needed for the duration of the test. Visual injury ratings were taken 9 and 16 days after treatment for post-emergence and pre-emergence applications, respectively, on a 0 to 100 scale where 0 represents no injury and 100 represents complete necrosis.

Absorption, Translocation and Metabolism

¹⁴C-labeled racemic mevalocidin, with a specific activity of 23.93 mCi/mmol was prepared by the Specialty Synthesis Group within Dow AgroSciences. Corn (Zea mays, ZEAMX) and IPOHE were grown from seed in a soil-free mix (Metro-Mix 360®) for experiments involving uptake and transport. Both species were grown under greenhouse conditions with supplemental light as described above. Mevalocidin was applied to the cotyledons of 1-leaf stage IPOHE and the 2nd leaf of 2.5 leaf stage ZEAMX for uptake and translocation studies. The compound was applied as a 2 % emulsifiable concentrate (EC) with 159,906 dpm of herbicide in 2 μl of solution. Two μl of the solution were spotted on each plant in 0.5 μl aliquots. At harvest times of 0, 4, 26, and 48 hr, the application leaf was washed in 50 % methanol/water, followed by a wash with 100 % CH₂Cl₂. The treated leaf, untreated leaves, and the roots were oxidized to determine uptake and transport. The experiment was performed with 5 replicate plants at each time point, with one plant kept without washing or dissection to be viewed by phosphorimaging. Phosphorimages of whole plants treated with radiolabeled compound were generated using a Molecular Dynamics Model SF phosphorimager. The plants were pressed between absorbent paper, frozen at −80 °C, and freeze dried for 4 days in a Virtis Genesis Model 12ES freeze drier at −40 °C with 15 mm Hg vacuum. The pressed plants then were covered with 12 μm thick Mylar film, mounted into a Molecular Dynamics storage phosphor screen (35 × 43 cm), and developed for 7 days. Plants were processed using the optimum high cutoff setting to background setting to generate optimum contrast for each particular phosphorimage.

Structure Elucidation

Mass spectral characterization indicated a very strong -ve ESI signal correlating to a MW of 160. Proton and carbon NMR revealed 7 carbon atoms, and 9 non exchangable protons accounting for 93 amu. Four oxygen atoms and 3 exchangeable protons make up the rest of the molecular weight giving a formula of C₇H₁₂O₄, which requires two unsaturations. A carbon signal at 176 ppm is consistent with a carboxylic acid, accounting for two of the oxygen atoms, leaving two hydroxy groups. Proton NMR indicated that the molecule contained 3 methylene groups, one of which was split into a pair of doublets (2JHH = 15 Hz) — indicating a chiral center in close proximity — and a methyl group. The carbon spectrum indicated that the two signals at 5.1 ppm were due to a terminal olefin methylene (at 109 ppm), where the geminal coupling between the two protons was ca. 0 Hz. DEPT analysis revealed that the peaks of 109, 61, and 45 ppm correlated to the 3 methylenes, indicating an olefin, and an oxygen substituted methylene. A one bond C/H heteronuclear multiple-quantum correlation experiment (HMQC) showed that the coupled methylene correlated to the carbon signal at 45 ppm, indicating carbon attachments. The rest of the spectrum indicated another olefinic carbon at 152 ppm, the carboxylic acid, an aliphatic quaternary atom at 73 ppm, and the methyl at 27 ppm.

The molecule was assembled by analysis of a single HMBC experiment (optimized for 8 Hz coupling). The methyl protons correlated to the quaternary olefinic carbon, the quaternary aliphatic carbon atom, and the methylene at 45 ppm. The lack of coupling and the chemical shift of this methyl singlet indicate that it was connected to the quaternary aliphatic carbon atom, which had a hydroxy substituent on it. Cross peaks arising from the protons connected to the carbon at 45 ppm confirmed this, and also indicated that the carboxylic acid was connected to this methylene. The other methylene carbon at 63 ppm — indicating the attachment of the second hydroxy group — exhibited cross-peaks to both of the olefinic carbon atoms. These correlations were consistent with the structure of mevalocidin (Table 1).

Table 1 Proton and carbon NMR data for mevalocidin

Closer evaluation of the proton spectrum indicated a second component in the sample, at about 5 % of the total sample amount. The acid as drawn may be expected to readily form the lactone, thus mevalocidin exists in equilibrium between the two forms (Scheme 1). The lack of an HMBC cross-peak between the methylene singlet and the carbonyl oxygen, indicates that in aqueous solution at neutral pH (the conditions in which the NMR spectra were acquired), the open chain form is the preferred conformation.

Scheme 1

Equilibrium of acid and lactone forms of mevalocidin

Synthesis of Mevalocidin

The synthesis is depicted in Scheme 2. The Baylis-Hillman reaction of methyl vinyl ketone, with formaldehyde in the presence of 1,4-diazabicyclo[2.2.2]octane (Dabco™) at 0° provided the known 3-(hydroxymethyl)but-3-en-2-one (5) in 49 % yield after vacuum distillation. Reaction of this with 2 equivalents of lithio tert-butyl acetate gave the tert-butyl ester of the racemic natural product (6), in 77 % yield after chromatography. Deprotection of the carboxylate with trifluoroacetic acid gave the lactone (7, 77 %), which could be hydrolyzed with aqueous 1 N NaOH to produce the sodium salt of racemic mevalocidin (8, 100 %). Alternatively, treatment of the tert-butyl ester (6) with 1 N NaOH in methanol gave direct access to 8 quantitatively. All synthetic details are provided in Online Resource 1.

Scheme 2

Synthesis of the sodium salt of mevalocidin

Preparation of Enantiomers of Mevalocidin

Preparation of the individual enantiomers was accomplished readily by the preparation of the Moshers ester of the racemic mixture (9), and separation on a chiral HPLC column (9a and 9b). Subsequent treatment of the individual diastereomers with KOH in methanol led to complete deprotection, yielding the individual enantiomers of the potassium salts of mevalocidin (+4 and −4) (Scheme 3). Synthetic details are provided in Online Resource 1.

Scheme 3

Preparation of the individual enantiomers of mevalocidin

Results and Discussion

Mevalocidin demonstrated herbicidal activity on both broadleaf and grass species (Table 2). Activity was stronger post-emergence than pre-emergence, with greater than 50 % injury observed on all species post-emergence at 4 kg/ha after 16 days. Post-emergence symptoms included stunting, meristematic inhibition, anthocyanin accumulation, weak formative effects, and other symptoms consistent with metabolic perturbation (Fig. 1). These visual symptoms were unlike those of any known commercial herbicide or phytotoxin known by the authors, and suggested a novel mode of action. An interesting aspect of the phytotoxicity caused by mevalocidin was the slow onset of symptoms. Much of the meristematic inhibition and secondary necrosis observed post-emergence did not appear until 7–9 days after application. It also should be noted that lethality was observed on all species in Table 2 at 4 kg/ha after 21 days (data not shown). Mevalocidin was only weakly active pre-emergence (Table 2). With the exception of sunflower, mevalocidin did not cause greater than 40 % injury to any species at 4 kg/ha, and was inactive at rates less than 2 kg/ha. Symptoms pre-emergence included reduced growth and necrosis of some meristems. In contrast to the post-emergence activity which increased in severity with time, pre-emergence activity diminished with time. Much of the initial stunting and necrosis observed pre-emergence at 7 days after application was absent by 14 days after application. This suggests instability in the soil but has not been further pursued.

Table 2 Phytotoxicity of mevalocidin to several plant species in the greenhouse

Fig. 1

Visual phenotype of plants treated with mevalocidin. Xanthium strumarium (XANST) and Digitaria sanguinalis (DIGSA) were treated post-emergence with 2 Kg/ha of racemic mevalocidin 16 days prior to photo. Note chlorosis and necrosis of apical tissues in XANST and intense anthocyanin production in DIGSA

Mevalocidin has a chiral center, and it was not possible to assign absolute stereochemistry based on the amount of sample isolated from the original source. Both enantiomers were subsequently obtained by preparative chromatography of their Mosher’s esters and tested as their K⁺ salts following ester hydrolysis. Activity was exclusively in the (+) enantiomer when tested post-emergence on a range of species including ABUTH, IPOHE, ECHCG, and SETFA. A comparison of activity of the lactone and open chain forms of mevalocidin on these same species over the rate range of 0.5–4.0 kg/ha indicated the former was 0.25–0.5X the activity of the later. No significant difference was observed in the activity levels of the Na+ and K+ salt forms of mevalocidin on ABUTH, IPOHE, ECHCG, or SETFA tested at these same rates.

When ¹⁴C-mevalocidin was applied to leaves, it was rapidly absorbed by ZEAMX and IPOHE (Fig. 2). At 4 hr after application, 75 % of the radioactivity was absorbed by IPOHE and 91 % by ZEAMX, with 5 % transported out of the treated leaf of both species. By 26 hr, 92 % of the radioactivity was absorbed by IPOHE and 95 % by ZEAMX, with 24 % to 28 % being transported out of the treated leaves of these species. Hence, mevalocidin demonstrated rapid leaf penetration and redistribution in both a representative broadleaf and grass species. While emulsifiable concentrate formulations are generally quite effective for facilitating leaf absorption of herbicides and often lead to the greatest levels of plant injury among formulation types (Grabowski and Hopen, 1984), rarely are absorption levels as high as observed in the current study. Mevalocidin that moved out of the treated leaf in IPOHE and ZEAMX was broadly distributed throughout the aerial portions of the plants, with material also moving to the roots, especially in IPOHE (Fig. 3). The data in Figs. 2 and 3 indicate mevalocidin is phloem and xylem mobile, which is a highly desirable attribute for herbicides, and a relatively rare attribute for known phytotoxins. Phloem mobility is important to enhancing phytotoxicity by enabling the movement of herbicide from the point of contact to the growing meristems (Lichtner, 2000).

Fig. 2

Phosphorimages showing time-course distribution of 14C radiolabel following application of 14C-mevalocidin in Zea mays and Ipomoea hederacea

Fig. 3

Time course of uptake and translocation of 14C radiolabel following application of 14C-mevalocidin in Zea mays and Ipomoea hederacea

While the importance of phloem mobility to herbicide performance is well understood, the significance of phloem mobility to the role of phytotoxins in the infection and virulence process is not. Several non-host specific toxins including, for example, macrocidin A and coronatine have demonstrated phloem mobility in our bioassays (unpublished). By activating the jasmonic acid induced defenses against necrotrophic pathogens, coronatine enhances plant susceptibility to Pseudomonas syringae (Browse, 2009). Strains lacking coronatine production are less virulent (Demain and Fang, 2000), but what is not known is whether a phytotoxin with, for example, xylem mobility only, instead of phloem mobility, would be equally effective. A wide range of chemically diverse materials with differing physical properties may act as suppressors of plant defenses (Metraux et al., 2009).

Both of the mevalocidin producers in the current paper are soil organisms from quite different genera, but both genera are well known plant pathogens. While it is possible that the phytotoxicity of mevalocidin has no role in the virulence of these organisms and could be the result of antimicrobial or other selection pressures, no such additional activity has yet been identified. More than 150 phytotoxins are known and typically demonstrate both antibiotic activity against microorganisms and plant phytotoxic activity (Demain and Fang, 2000). While the evidence is compelling that these metabolites serve survival functions in the producing organisms, much research remains in defining how. Mevalocidin represents a unique phytotoxin because of its broad spectrum activity and phloem mobility and is being further evaluated as a potential bioherbicide.