An Allelochemical from Myrica gale with Strong Phytotoxic Activity against Highly Invasive Fallopia x bohemica Taxa

We report the identification of the allelochemical 3-(1-oxo-3-phenylpropyl)-1,1,5-trimethylcyclo-hexane-2,4,6-trione, known as myrigalone A, from the fruits and leaves of Myrica gale. The structure of the compound was confirmed by high-resolution techniques (UV, MS and NMR analysis). The compound is phytotoxic towards classical plant species used for allelochemical assays and also against Fallopia x bohemica, a highly invasive plant. Application of either powdered dry leaves or dry fruits of M. gale also showed in vitro phytotoxic activity. We hypothesize that M. gale could be used as a green allelopathic shield to control Fallopia x bohemica invasion, in addition to its potential use as an environmentally friendly herbicide.


Introduction
Biological invasions are recognized as a result of global change and are of growing interest in the biological sciences [1] because of their potential effects on biodiversity. Invaded ecosystems are generally considered as disrupted and the native species as strongly threatened. Many invasive plant species are known to affect plant and animal communities [2,3], ecosystem functioning [4][5][6], soil properties and nutrient fluxes [7,8] in their new environments. The impact of the invasion can be site-specific and is linked, among other factors, to soil properties in the novel habitat prior to the invasion. The effect of the invasion on the soil nutrients (increased or decreased) is dependent on the original soil chemical conditions, even if the plant can be invasive in a broad range of soil nutrient conditions [9].
Invasive knotweeds (species complex Fallopia) represent a particularly aggressive taxonomic group in Europe and Northern America [10][11][12]. The genus Fallopia (Polygonaceae) includes two herbaceous perennial Asian polyploid species (F. japonica and F. sachalinensis), that have now become widespread and highly invasive weeds. Recent hybridization events implicating F. japonica and F. sachalinensis in the invasion area have led to a new species being formed, F. x bohemica, or more precisely to a complex of polyploid hybrids [13].
Ecosystems invaded by Japanese knotweeds are generally considered as modified and the endemic biodiversity as threatened [14][15][16][17]. Many watercourses are now highly invaded, with large populations of Fallopia being found along riverbanks in countries such as the UK [18], the Czech Republic [19] and France (F. Piola personal observation and [20]). The invasive capacity of the Fallopia complex may be the result of exceptionally high growth rate, gigantism and extremely effective vegetative multiplication, as found for other invasive species [21]. The invasive potential of the Fallopia complex may also result from its more direct effects on the environment, through the production of substances toxic to organisms native to the invaded habitat. It is now known that chemical compounds possessing antifungal, antiherbivory and antimicrobial properties or with phytotoxic (allelopathic) effects confer some invasive plants benefits in new environment. Invasive plant strategies based on such compounds constitute the hypotheses named « Novel Weapons Hypothesis » or « Allelopathic Advantage against Resident Species (AARS) » [22][23][24][25]. Extracts of F. japonica and F. x bohemica also exhibit phytotoxic activities [26][27][28] and antifungal properties [29,30].
We hypothesize that the novel weapon theory could be applied to invasive species themselves: plants producing allelochemicals unknown to invasive plants could affect the growth of invasives and may provide a safe, novel method of biocontrol. The plant used as chemical weapon against the invasive species must not itself become an invasive plant in alien environments. Consequently, the "weapon plant" must have well-defined, specific and known ecological requirements restricting it from growing in most environments and ensuring it is unknown to the invasive plant. Myrica gale (Myricaceae) fulfils these conditions, growing only in acidic wet to flooded edaphic conditions [31]. This shrub is found in environments such as the wet and flooded area around lakes, along rivers or in peat bogs and has a wide distribution from Northern and Western Europe the American continent. The fruits and leaves of M. gale are covered in droplets made up of exuded secondary metabolites [32]. The composition of the fruit exudates is made of a high diversity of chemical structures, including essential oils and flavonoids from the class of dihydrochalcones and for which several biological activities have been reported [33][34][35]. Although phytotoxic activity has already been observed for other Myricaceae species [36], no allelopathic properties have ever been described for M. gale. Here, we investigated the phytotoxic activity of M. gale exudates against a variety of plants used in classical phytotoxic assays and against F. x bohemica. We also report the identification of the active compound responsible for phytotoxic activity. We finally discuss the implications of our results for the development of a successful biocontrol solution for Fallopia in France.

Allelopathic potential of M. gale fruit exudates against F. x bohemica
The methanol extract of M. gale fruit exudates inhibited the root and shoot growth of all the plant species tested, including F. x bohemica ( Figure 1). Root and shoot growth was inhibited for all the plant species when at least 5 mg of extract was applied, except for sorghum where root growth was significantly inhibited only when 10 mg of exudate was applied. Germination rates were also reduced by M. gale fruit exudate extracts (Table 1). Table 1. Effects of M. gale methanol fruit exudate extracts on the germination rates of sorghum, cress, mustard and F. x bohemica. Concentrations are the total quantities of dry exudate extracts added to each Petri dish. Thirty plants were used for each treatment.

Treatment
Sorghum Cress and mustard were the most affected, with only 1% and 24% of germinated seeds, respectively, when 10 mg of fruit exudate was applied. By contrast, in the control assays 100% and 93% of seeds germinated for cress and mustard, respectively. Germination rates of sorghum and F. x bohemica were also slightly reduced by M. gale fruit exudates but to a lesser extent (Table1).
The taxa of Fallopia are highly invasive plants threatening numerous ecosystems in Europe and North America. There is currently no successful strategy to eliminate F. x bohemica and effective control methods are urgently needed. In this work we have hypothesized that one of the mechanisms by which a plant becomes invasive, the Novel Weapons Hypothesis, might be used against invaders themselves. We show that indeed M. gale fruit extracts exhibit phytotoxic activity against F. x bohemica.

Identification of an allelochemical from M. gale fruit exudates
An active substance (9 g) was purified from M. gale fruit exudates (350 g of fruits). Based on a literature comparison of its 1 H-, 13 C-NMR, and MS data, as well as its UV spectra, the molecule was identified as 3-(1-oxo-3-phenylpropyl)-1,1,5-trimethylcyclohexane-2,4,6-trione, also known as myrigalone A (MyA), a flavonoid from the dihydrochalcone class ( Figure 2). The compound has only been described from M. gale [34,37] but, to our knowledge, no phytotoxic activity has ever been reported for it. HPLC analysis revealed that MyA was also present in M. gale leaves (data not shown). We show that this compound is phytotoxic against plants generally used in allelochemical bioassays (cress, mustard and sorghum) as well as against F. x bohemica (Figure 3). This molecule is produced in a high yield in M. gale fruit exudates (40% of the total fruit exudates) and, to a lesser extent, in the leaves. Myrigalone A shares structural features with the natural βtriketone phytotoxin leptospermone, produced by the bottlebrush plant (Callistemon spp.). The βtriketone moiety is responsible for the phytotoxicity of the compound and its molecular target is p-hydroxyphenylpyruvate dioxygenase (HPPD), an enzyme involved in the biosynthesis of tocochromanols (tocopherols and tocotrienols) and prenylquinones [38][39][40]. Inhibition of HPPD causes photodynamic bleaching of the foliage. Triketones have mainly been described in Myrtaceae species such as Leptospermum, Eucalyptus, and Corymbia and commercial herbicides have been developed based on the triketone moiety (sulcotrione, mesotrione…).
The mechanisms by which the plant producing allelochemicals is protected from these compounds are still unclear. However it is thought that the site of production of the compounds, glandular trichomes and other glands, might compartmentalize the bioactive products and thus protecting the plant from autotoxic effects. Such glands can be found on M. gale leaves as well as in other plants producing allelochemicals such as Myrtaceae [32]. In M. gale fruits, myrigalone A is exuded outside the fruit, on the surface of the lignified drupe, possibly to protect the seed (unpublished data).

Phytotoxicity of whole fruits and leaves of M. gale
Entire dry fruits and powdered dry leaves of M. gale inhibit the root and shoot growth of all the tested plant species, similar to the inhibitory effects observed for methanol extracts (Figure 4). The results show root growth inhibition is observed in cress and mustard as soon as 100 mg of entire fruit or leaf powder is applied. For all the plants tested, root and shoot growths are inhibited when 200 mg of entire fruits and leaf powder is applied.
The results described in this paper show that both pure myrigalone A and fruit exudate methanol extract can inhibit the root and shoot growth of F. x bohemica. However, applying solvent-dissolved phytosanitary products in the field can be challenging and the toxicity of these solutions must be taken into account.
In order to avoid the utilization of solvents for applying allelochemicals, we showed that the use of M. gale entire fruits or leaf powder sprinkled was enough to demonstrate a phytotoxic effect on the plant species tested, including F. x bohemica. Our results strengthen the argument for the development of allelopathic shields against alien species. Indeed, the identification of endemic allelopathic plants in an environment may lead to use these plants as a source of allelopathic compounds unknown to invading species. Such plants could also be used as green allelopathic barriers to prevent the invasion of a threatening species within their own environment. Consequently, one "weapon plant" will not be adapted to each environment and as such, for every threatened environment, the identification and use of a suitable allelopathic plant must be done. We are currently testing this hypothesis using M. gale to control F. x bohemica species in Rhône-Alpes region of France. M. gale appears to be suitable for testing the concept of green allelopathic shields because it grows only around wet and flooded areas as well as river shores, sites that are currently heavily invaded by F. x bohemica in the Rhône-Alpes. We acknowledge however that the implementation from laboratory studies to field experiment is a difficult and critical step and identifying other potential "weapon plant" might be necessary to fully achieve protection against F x bohemica in this environment. Moreover, even if seed germination appears to be a mechanism of dissemination of F. x bohemica [41], its propagation is mainly vegetative through rhizome fragments. We are currently evaluating M. gale allelopathic properties on rhizome fragments regeneration. Other experiments are currently undertaken to evaluate in glasshouse and mesocosm conditions the possible use of M. gale as an allelopathic shield. These include the evaluation of M. gale effects on other plant species living in the same ecosystem, even if since they have coevolved together, no phytotoxicity is expected to be observed. Finally those mesocosm experiments will provide data about M. gale and F. x bohemica in situ competition along with possible phytotoxic activity of F. x bohemica against M. gale. If the results of those experiments show that M. gale fulfils the requirements to be an allelopathic barrier, then implementation on invaded sites in the Rhône-Alpes region to evaluate this biocontrol strategy will be performed.

General
Chromatographic analysis of extracts was achieved by HPLC on an Agilent 1200 series HPLC instrument equipped with a degasser (G132A), a quaternary pump module (G1311A), an automatic sampler (G1329A) and a DAD (DAD G1315B). Separations were carried out using a Kromasil RP18 column (250 × 4.6 mm, 5 µm, 100 Å) with a linear gradient of acetonitrile in water from 0% to 100% in 60 min supplemented with formic acid (0.4%). Chromatograms were recorded between 200 and 700 nm and a specific channel set at 280 nm was used to monitor chromatographic traces. Structural elucidation was achieved by UV spectroscopy (Agilent 8453 UV spectrophotometer) recorded between 220 and 500 nm, HPLC-MS and by 1 H-and 13 C-mono and bidimensional NMR (Bruker DRX 500). For MS analysis, the HPLC system described above was interfaced with an HP MSD 1100 series allowing the same chromatographic conditions as those used for HPLC-DAD analysis. The operating conditions of the mass spectrometer with an APCI interface were: gas temperature 330 °C at a flow rate of 9.0 L/min, nebulizer pressure 50 p.s.i, quadripole temperature 30 °C, capillary voltage 4,000 V and fragmentor 100. The full scan spectrum from m/z 100 to 900 in both positive and negative ion mode was recorded

Plant material
M. gale fruits and leaves were collected in December 2004 on the shore of Biscarosse Lake, Bordeaux, France. A voucher specimen was deposited at the Herbarium of the University Claude Bernard Lyon1, Villeurbanne, France, under the name 'Collection Piola' and collector number 3. Airdried fruits (500 g) were sonicated (Branson 2510, 40KHz) twice in methanol (2L) and evaporated, allowing us to obtain 50 g of dried residue. Leaves were dried and ground and compounds were extracted by sonication in methanol. The two extracts were then diluted in methanol for the further experiments. Fallopia x bohemica akenes were collected in January 2008 on stands growing along the river Dorlay in the Loire region of France. The akenes were collected randomly on several shoots of Fallopia, pooled and stored at 4 °C.

Phytotoxicity assays
Total fruit exudates and purified myrigalone A were first dissolved in methanol. Aliquots of the extracts were evaporated on filter paper (Whatman n°2) placed in a 9 cm-Petri dish to obtain final quantities of 1, 5 and 10 mg of dry extract per Petri dish. Three additional sheets of filter paper were placed underneath the sheet with the extracts and moistened with 8mL of water. Controls were obtained by placing methanol only on the paper sheet, which was then evaporated before adding water. Five, 10, 20 and five seeds of respectively sorghum (Sorghum saccharatum), mustard (Sinapis alba), cress (Lepidium sativum) and F. x bohemica were placed on the paper sheets in the Petri dishes and incubated for 72 h-10 days at 25 °C in darkness. After incubation, germination rate and shoot and root lengths were measured. Each bioassay was repeated three times. Significant differences between treatment and control plants were examined by Tukey's test (R software 2.9.0).
A second bioassay protocol was developed to evaluate the phytotoxicity of whole M. gale organs (fruits and leaves) without methanol extraction. The protocol was similar to that described above with the key difference that test seeds were sprinkled with entire fruits or dry leaf powder instead of having methanolic extracts applied to them. One hundred, 200 and 300 mg of fruits or leaves were sprinkled on each Petri dish. After 72 h of growth, shoot and root lengths were measured and compared to nonsprinkled control plants.

Extraction and isolation
Gel filtration on Sephadex LH-20 (600 × 45 mm, methanol) of fruit exudate extract (35 g) resulted in 10 major fractions that were assessed for phytotoxic activity as described above. One fraction was active and its content was analysed by HPLC. This fraction, weighing 8 g, was composed of a single molecule. Structural elucidation was achieved by UV spectroscopy, Mass Spectrometry and NMR analyses and comparison with the literature led to the identification of myrigalone A [35,37].

Conclusions
We have shown that M. gale fruits and leaves extracts exhibit phytotoxic activities against different plant species, including against the invasive species Fallopia. The phytotoxicity is due to the presence of 3-(1-oxo-3-phenylpropyl)-1,1,5-trimethylcyclohexane-2,4,6-trione (myrigalone A), a dihydrochalcone present in both the fruits and leaves. We hypothesize that M. gale could be used in the future as an allelopathic green shield to control invasion of F. x bohemica in some environments, in addition to being a potential source of environmentally friendly herbicides.