(4Z)-Lachnophyllum Lactone, an Acetylenic Furanone from Conyza bonariensis, Identified for the First Time with Allelopathic Activity against Cuscuta campestris

Abstract

Cuscuta species are obligate parasitic plants that infect the stems of a wide range of hosts including many crop and weed species causing severe agricultural problems. Using in vitro experiments to screen organic extracts prepared from fifteen autotrophic weed species found in agricultural fields infested with Cuscuta campestris, we have identified for the first time a strong phytotoxic activity in Conyza bonariensis extract against C. campestris. Additional pot experiments revealed that seven day-old Cuscuta seedlings had reduced capacity to coil and properly attach on Conyza plants, leading to reduced parasitic weed infection. Via activity-guided fractionation of Conyza extracts, we isolated and identified the acetylenic furanone (4Z)-lachnophyllum lactone as the major active component, with a concentration required to achieve reduction of 50% Cuscuta seedling growth (IC₅₀) of 24.8 µg/mL. The discovery of (4Z)-lachnophyllum lactone bioactivity could aid the development of efficient and sustainable management strategies for C. campestris, whose control is limited or non-existent.

1. Introduction

Approximately 1% of angiosperms, distributed among 28 dicotyledonous families, are parasitic on other plants [1,2]. Some of these parasitic plants are obligate parasites that have abandoned key mechanisms that allow plants to function autotrophically and therefore depend on their host plants for nutrient acquisition, growth and reproduction. Among them, about 170 species of dodders (Cuscuta spp., Convolvulaceae) thrive at the expense of other plants in tropical, subtropical and temperate regions [3,4]. Cuscuta plants have no roots nor leaves and their seedlings coil around the stems of other plants, forming infective haustoria that withdraw nutrients and water through vascular connections [5,6]. Among dodder species, Cuscuta campestris Yunck. is one of the most damaging species for agricultural production, for which control in the majority of crops is limited or non-existent [7]. On one side, the intimacy of connections between Cuscuta and its crop host renders the available selective herbicides ineffective, and on the other side, there is a lack of development of resistant varieties against Cuscuta infection for the majority of crops affected [7,8,9]. In addition, the persistent Cuscuta seedbank and broad host range in the agricultural fields, which includes many species of crops and weeds, make the use of rotation ineffective for its control.

Elucidation of novel structures and modes of action of natural compounds with allelopathic activity against parasitic weeds is an alternative solution to provide efficacy and sustainability in strategies for parasitic weed management [10,11,12]. Plants are a generous source of natural pesticides [13,14], but only a small fraction of plant metabolites has been screened for herbicidal activity [15]. From the screening of natural compounds produced by allelopathic plants, compounds with specific herbicidal activity against parasitic weeds have been previously discovered [10,11,16]. Conyza species (Asteraceae) are invasive weeds native to America, affecting more than 40 crops in 70 countries [17]. In Spain, three Conyza species, Conyza bonariensis (L.) Cronq., Conyza canadensis (L.) Cronq. and Conyza sumatrensis (Retz.) E. Walker, cause important problems in agricultural fields [18,19,20]. Allelopathy plays a part in their invasive success [21,22,23,24]. In this work, we used allelopathy assays to screen fifteen weedy species found in southern Spanish agricultural fields with soils infested with Cuscuta. This screening allowed us to identify for the first time the strong allelopathic activity of C. bonariensis dichloromethane extract against the growth of C. campestris. The bioactivity-guided purification of the Conyza extract led to the isolation of a main metabolite responsible for the phytotoxic activity. Using spectroscopic methods (essentially, ¹HNMR and ESI-MS), we identified this metabolite as (4Z)-lachnophyllum lactone, a phytotoxin with a potent activity against C. campestris growth never reported before.

2. Materials and Methods

2.1. General Experimental Procedures

¹H and ¹³C NMR spectra were recorded at 400 and 100 MHz, respectively, in CDCl₃ on a Bruker spectrometer (Karleshrue, Germany). The NOESY (nuclear overhauser enhancement spectroscopy) experiment was performed using standard Bruker microprograms. The same solvent was used as an internal standard. ESI mass spectra and liquid chromatography (LC)/MS analyses were performed using the LC/MS TOF system Agilent 6230B (Agilent Technologies, Milan, Italy), HPLC 1260 Infinity. The HPLC separations were performed with a Phenomenex (Bologna, Italy) LUNA (C18 (2) 5 μ 150 × 4.6 mm). Analytical and preparative thin-layer chromatography (TLC) was performed on silica gel (Kieselgel 60, F₂₅₄, 0.25 and 0.5 mm, respectively) plates (Merck, Darmstadt, Germany), and the compounds were visualized by exposure to UV light and/or iodine vapors or by spraying first with 10% H₂SO₄ in MeOH and then with 5% phosphomolybdic acid in EtOH, followed by heating at 110 °C for 10 min.

2.2. Plant Material

Seeds from fifteen weed species from Amaranthaceae (Amaranthus albus L. and Amaranthus retroflexus L.), Asteraceae (Conyza bonariensis (L.) Cronq.), Boraginaceae (Heliotropium europaeum L.), Brassicaceae (Capsella bursa-pastoris (L.) Medik. and Diplotaxis virgata (Cav.) DC.), Convolvulaceae (Convolvulus arvensis L.), Malvaceae (Malva sylvestris L.), Papaveraceae (Fumaria officinalis L.), Polygonaceae (Polygonum aviculare L.), Portulacaceae (Portulaca oleracea L.), Solanaceae (Datura stramonium L. and Solanum nigrum L.), Urticaceae (Urtica dioica L.) and Zygophyllaceae (Tribulus terrestris L.) were collected during the season of 2016–2017 from a buckwheat field at Institute for Sustainable Agriculture (IAS-CSIC, Córdoba, southern Spain). Weed seeds were surface sterilized with 0.5% (w/v) sodium hypochlorite and 0.02% (v/v) Tween 20 for 5 min, rinsed thoroughly with distilled water and dried in a laminar airflow cabinet. Then, weed seeds were sown in a greenhouse in 1 L pots containing sand and peat (1:1, v:v) and grown for 40 days (23/20 °C, 16/8 h day/night). Then, the stem of each weed plant was cut 2–3 cm above the soil surface, and the roots were carefully washed, dried with filter paper, immediately frozen and then maintained at −80 °C until lyophilization.

Seeds of Cuscuta (Cuscuta campestris Yunck.) were collected in July 2019 from mature Cuscuta plants parasitizing pea plants in fields of IAS-CSIC. Dry Cuscuta seeds were separated from capsules using a winnowing with a fan and sifting with a 0.6 mm mesh-size sieve. Cuscuta seeds were stored dry in the dark at room temperature until use for this work in 2022.

2.3. Plant Extraction for Screening of Allelopathy in Weed Species

About 6 g of lyophilized tissue of each weed species described in Section 2.2 were milled in a Warry Blender and the resulting powder was macerated overnight in 200 mL of a mixture of methanol–distilled water (1:1, v:v) under stirring in the dark at room temperature. Then, the suspension was centrifuged at 7000 rpm for 1 h, at 4 °C. The supernatant was extracted with dichloromethane (3 × 200 mL). For each weed species, the organic extracts were combined, dried with sodium sulfate, filtered and evaporated under reduced pressure.

2.4. In Vitro Experiments for Screening of Allelopathy against Cuscuta Seedling Growth

Effects of dichloromethane extracts of fifteen weed species described in Section 2.2 were tested on growth of Cuscuta seedlings. To promote Cuscuta germination, the hard coat of Cuscuta seeds was eliminated by scarification with sulfuric acid during 45 min [25], followed by thorough rinses and air-dried. Then, five scarified Cuscuta seeds were manually placed using tweezers on 5 cm-diameter filter paper discs inside 5.5 cm-diameter Petri dishes. Stock solutions of each organic extract dissolved in methanol were diluted up to 100 µg weed extract/mL sterilized distilled water. The final concentration of methanol was 2%. Triplicate aliquots of 1 mL of each weed extract were applied to filter paper discs containing the scarified Cuscuta seeds. Triplicate aliquots of a treatment only containing 2% methanol and sterile distilled water was used as control. Treated Cuscuta seeds were incubated in the dark at 23 °C for 6 days. The seedling length was measured in each of the five Cuscuta seedlings for each of the three replicate filter paper discs per treatment. A second in vitro bioassay was performed to confirm the Conyza inhibitory activity identified in the first allelopathic screening. Conyza organic extract dissolved in methanol was applied at seven concentrations (100, 75, 50, 25, 10 and 5 µg Conyza extract/mL sterilized distilled water, maintaining the final concentration of methanol constant at 2%) to filter paper discs containing five scarified Cuscuta seeds as described before. Triplicate aliquots of a treatment only containing 2% methanol and sterile distilled water was used as control. After six days, Cuscuta seedling length was determined.

2.5. Pot Experiments for Validation of Conyza Allelopathic Activity

In a greenhouse, 40 pots containing sand and peat (1:1, v:v) were prepared for the validation of Conyza allelopathic activity against Cuscuta campestris. As a non-allelopathic control we used a subset of eight weed species from those that showed no allelopathic activity during the in vitro screening. Plants of Amaranthus albus, Amaranthus retroflexus, Diplotaxis virgata, Convolvulus arvensis, Conyza bonariensis, Malva sylvestris, Polygonum aviculare, Portulaca oleracea and Solanum nigrum were grown in pots at 23/20 °C, 16/8 h day/night. Each weed plant, at the stage of 4 leaves, was inoculated with pregerminated Cuscuta seeds. To promote Cuscuta germination, two days before inoculation, Cuscuta seeds were scarified with sulfuric acid for 45 min [25], rinsed thoroughly, and then spread in wet filter paper inside Petri dishes to allow their germination in the dark at 23 °C for 2 days. Then, nine pregerminated Cuscuta seedlings were manually placed using tweezers on the soil surface surrounding each weed plant at 1 cm distance from the weed stem. Seven days after germination, the Cuscuta seedlings were visually inspected and classified as either (i) unattached Cuscuta seedling or (ii) attached Cuscuta seedling. Fourteen days after inoculation, Cuscuta attached seedlings were classified as seedlings with adhesion disks (i) without posthaustorial growth, or (ii) with posthaustorial growth emerging at the Cuscuta–host interface.

2.6. Isolation and Identification of (4Z)-Lachnophyllum Lactone from Conyza Bonariensis Extracts

A measure of 27 g of Conyza bonariensis-lyophilized tissues obtained as described in Section 2.2 were extracted (1 × 150 mL) using a mixture of methanol–distilled water (1:1, v:v), 1% NaCl, under stirred conditions at room temperature for 24 h. The suspension was centrifuged, and the supernatant extracted using CH₂Cl₂ (3 × 150 mL). The residue (60 mg) of the organic extract, showing specific inhibitory activity against Cuscuta campestris, was purified by TLC eluted with EtOAc/n-hexane (6/4, v/v), yielding five homogeneous fractions which were screened for allelopathic activity against Cuscuta seedling growth as described in Section 2.7. The fraction with the strongest toxicity against Cuscuta was studied using spectroscopic methods (essentially ¹HNMR and ESI-MS).

2.7. Bioassays against Cuscuta Seedling Growth for Identification of (4Z)-Lachnophyllum Lactone Phytotoxic Activity

A third in vitro bioassay was used to guide the identification of the phytotoxic compound(s) during the fractioning of Conyza extract. Test fractions were dissolved in dimethyl sulfoxide and diluted up to 100 µg/mL sterilized distilled water. The final concentration of dimethyl sulfoxide was 2% in all treatments including the control. As described above, triplicate aliquots of 1 mL of each test fraction and control were applied to filter paper discs containing scarified Cuscuta seeds, and six days later, Cuscuta seedling length was determined. A subsequent screening was conducted to confirm the activity of (4Z)-lachnophyllum lactone and characterize its dose–response curve on Cuscuta campestris. Triplicate aliquots of 1 mL of (4Z)-lachnophyllum lactone dissolved in dimethyl sulfoxide was applied on Cuscuta scarified seeds at seven concentrations (100, 75, 50, 25, 10 and 5 µg/mL sterilized distilled water, maintaining the final dimethyl sulfoxide concentration constant at 2%). Triplicate aliquots of a treatment only containing 2% dimethyl sulfoxide and sterile distilled water was used as control. Cuscuta seedling length was determined six days later.

2.8. Statistical Analysis

All bioassays were performed using a completely randomized design. Cuscuta seedling length for each treatment was calculated relative to the Cuscuta seedling length of the corresponding control. Percentage data were approximated to normal frequency distribution by means of angular transformation (transformed value = 180/Π × arcsine [√(%/100)]) and subjected to analysis of variance (ANOVA) using SPSS software for Windows (SPSS Inc., Chicago, IL, USA). The significance of mean differences among treatments was evaluated by Tukey test. Null hypothesis was rejected at the level of 0.05.

3. Results and Discussion

A first in vitro screening was conducted in order to identify candidate weed species as sources of allelochemicals that could be used for the control of Cuscuta campestris. Dichloromethane extracts obtained from fifteen weed species were individually applied to Cuscuta seeds at a concentration of 100 µg weed extract/mL sterilized distilled water and levels of Cuscuta seedling growth rated in comparison with the control (Figure 1). This first study revealed significant differences in phytotoxicity against Cuscuta growth among weed extracts tested (ANOVA, p < 0.001) and allowed us to identify an exceptional phytotoxic activity in the dichloromethane extract prepared from Conyza bonariensis while the dichloromethane extract prepared from the rest of the weed species showed no or negligible phytotoxicity.

In a second in vitro study, a dose–response screening was conducted to validate the effect on Cuscuta growth induced by Conyza extract in comparison with the growth of Cuscuta when treated with the control. This second study confirmed the phytotoxic activity of Conyza against Cuscuta and revealed an average of 99.3 ± 0.4% and 66.8 ± 1.8% inhibition of Cuscuta seedling length when, respectively treated with Conyza extract at 100 and 75 µg/mL. Negligible phytotoxicity was observed when Cuscuta seeds were treated with lower concentrations (ranged from 50 to 5 µg/mL) of Conyza extract (Figure 2).

Species of Conyza are sources of abundant phytotoxic compounds such as catechol, gallic acid, syringic acid and vanillic acid [24]. In vitro phytotoxicity of the closely related species Conyza canadensis against Lactuca sativa and Agrostis stolonifera was previously identified at 1 mg/mL [26]. To the best of our knowledge, there are no previous reports on phytotoxicity from any Conyza species against Cuscuta seedlings. On the contrary, Gaertner [25] described Conyza canadensis as a weed which C. campestris has the capacity to infect. Parasitic weeds can display preferences to infect different hosts in a species-specific manner [27]. To further explore in vivo the allelopathic potential of C. bonariensis revealed by our in vitro screening, and also to confirm the differences between our results with C. bonariensis and those from Gaertner [25] with C. canadensis, we conducted a pot experiment to observe the interaction between our candidate allelopathic species C. bonariensis for Cuscuta control and Cuscuta plants. The interaction between C. bonariensis and Cuscuta was compared with the interaction between Cuscuta and a control group of eight non-allelopathic weed species whose dichloromethane extracts showed no phytotoxicity against Cuscuta in the first in vitro screening. Plants of Conyza and the eight control weeds were cultivated in a greenhouse and individually inoculated with pre-germinated Cuscuta seeds. Without host infection, Cuscuta seedling viability expires in 3–7 weeks depending on the photosynthetic capacity of the Cuscuta species considered [28]. In our work with the species Cuscuta campestris, unattached Cuscuta seedlings did not show visual evidence of photosynthetic activity and their viability expired in two weeks without attachment to a host plant. Therefore, we determined success of Cuscuta coiling on the host at seven days after Cuscuta germination (Figure 3) and the success of infection at fourteen days after Cuscuta germination (Figure 4).

Cuscuta seedlings explore the environment searching for a host to which they can coil using a rotative movement guided by host-derived volatiles and far-red light [29,30,31]. Cuscuta seedlings also coil nonspecifically around inert objects, such as metal or plastic sticks they accidentally encounter during their rotative movement. Therefore, reduced coiling can be a sign of allelopathic activity. Figure 3 shows that the success of coiling of seven day-old Cuscuta seedlings was significantly affected by the weed species considered (ANOVA, p = 0.03), with the percent of Cuscuta seedlings that coiled and established proper contact with Conyza plants being significantly lower (34.2 ± 9.9%) than the percent of coiling around the stems of the eight non-allelopathic control weed species (percent of coiling ranged from 75.6 ± 12.4% in Amaranthus albus to 93.8 ± 3.8% in Portulaca oleracea).

Once Cuscuta coils around the stems of its hosts, tactile signals, light spectrum and phytohormones promote the development of an haustorium that enables infection [32,33,34,35]. The haustorium invades the host stem, connecting the host xylem to withdraw nutrients and water used by Cuscuta to develop posthaustorial stems [3]. In our pot experiment, the success of infection was observed as the percent of coiled Cuscuta seedlings that were able to develop posthaustorial stems from the site of attachment (Figure 4). There were not significant differences among the infection success of Cuscuta on stems of Conyza (57.5 ± 10.9%) and the infection success of Cuscuta on stems of the non-allelopathic weed species, which ranged from 36.1 ± 7.3% in A. albus to 88.9 ± 7.9% in Polygonum aviculare (data for the rest of species are not shown).

Despite the capacity of Cuscuta to infect plants of Amaranthus and P. oleracea observed in our work (Figure 3 and Figure 4) and by that from Orkić et al. [36], a previous work by Gaertner [25] described A. retroflexus and P. oleracea as weed species on which Cuscuta campestris would not be able to survive. On the contrary, Cuscuta campestris was reported to have high binding ability on Conyza canadensis by Orkić et al. [36], however, our work revealed that Cuscuta seedlings had reduced capacity to coil and properly attach to Conyza bonariensis plants, but those few Cuscuta seedlings able to attach on Conyza had the capacity to infect and grow for at least fourteen days. Orkić et al. [36] obtained the results through observations of field infections which could be influenced by a high Cuscuta density because these authors did not distinguish between success in coiling and success in infection as we did in our work (Figure 3 and Figure 4). In addition, field observations could not distinguish whether the infection was produced by either Cuscuta seedlings or by mature Cuscuta stems originated in nearby plants (capacity of infection could differ between primary infection of prehaustorial Cuscuta seedlings and secondary infection of mature posthaustorial Cuscuta stems). Our results indicate that Cuscuta seedlings had a reduced capacity to coil and attach on Conyza bonariensis plants (Figure 3D) in comparison with the non-allelopathic control weed species (Figure 3), but those Cuscuta seedlings that were able to properly attach to Conyza had the capacity to infect (Figure 4C), indicating that Conyza does not seem to impose resistance mechanisms against the invasion of the attached haustorium and subsequent parasitic growth of Cuscuta seedlings up to at least an age of 14 days old. Resistance to C. campestris haustorium invasion and subsequent parasitic growth have been previously described in other plant species [7,8].

To identify the compound(s) responsible for the allelopathic activity against Cuscuta in Conyza bonariensis dichloromethane extract, an increased amount of Conyza-lyophilized tissue was extracted. The resulting organic extract was subjected to fractionation using TLC as reported in the Materials and Methods Section, yielding five homogeneous fractions (CBA, CBB, CBC, CBD and CBE). Phytotoxicity screening revealed that, among the five fractions of the Conyza extract, the CBB fraction caused the strongest phytotoxicity in seedlings of Cuscuta (Figure 5). This phytotoxicity was observed as the abnormal growth of the Cuscuta seedling with a length reduction in comparison with the control seedlings.

The investigation of the active fraction CBB, by the study of the ¹H NMR and ESI-MS spectra, revealed that it consisted in a pure compound, which was identified as (4Z)-lachnophyllum lactone, the (Z)-5-(hex-2-yn-1-ylidene)furan-2(5H)-one (Figure 6, R_f = 0.76, 5.10 mg). Its structure was confirmed by comparison of the ¹H-NMR data with those reported in the literature [26,37,38]. The configuration of the double bond was deduced from the presence of coupling between H-5 with H-3 and H-2 in the NOESY spectrum (Figure S1). In addition, the chemical shifts of H-5 (δ = 5.33) and C-5 (δ = 94.5) were very similar to those previously reported for lachnophyllum lactone and other natural furanones, with an α Z-disubstituted vinyl group, substantially differing from those having a E-vinyl group [38,39,40,41]. This structure was confirmed by the data of its ESI-MS spectrum which showed the sodiated adduct [2M + Na]⁺ and protonated [2M + H]⁺ dimers, and protonated [M + H]⁺ ions at m/z 347, 325 and 163, respectively. This lactone with unspecified configuration was previously reported from different plant species [26,37,38,42]. The ¹H NMR data of the (Z) and (E) isomers of the acetylenic lactone were reported when both the compounds were isolated from Baccharis paniculata. A clear upfield shift of proton H-5 was observed for the Z-isomer [38].

A subsequent dose–response screening was conducted to validate the phytotoxicity of (4Z)-lachnophyllum lactone, confirming the inhibitory activity of Cuscuta seedling growth at concentrations ranged from 100 to 10 µg/mL (Figure 7). The concentration required to achieve reduction of 50% Cuscuta seedling growth (IC₅₀) was observed at 24.8 µg/mL.

Previously, (4Z)-lachnophyllum lactone isolated from Conyza canadensis showed phytotoxic activity against Lactuca sativa, Agrostis stolonifera and Lemna paucicostata [26]. Furthermore, (4Z)-lachnophyllum lactone has reported fungitoxic activity against the fungi causing postharvest diseases in strawberry, i.e., Colletotrichum acutatum, C. gloeosporioides and C. fragariae [26], and causing postharvest diseases in citrus, i.e., Penicillium digitatum [43]. Fungitoxic activity against Pyricularia oryzae was identified in lachnophyllum lactone isolated from Erigeron apiculatus [44]. In addition, a repellent activity against Monotonda neritoides was identified in lachnophyllum lactone with unspecified configuration isolated from Erigeron sumatrensis [37]. These biological activities could be related to the presence in the structure of (4Z)-lachnophyllum lactone of an α,β-unsaturated carbonyl group, a known structural feature involved in nucleophilic Michael addition reaction mechanism frequently reported for bioactive natural compounds [40,45]. However, further studies are needed to elucidate the specific mode of action of this acetylenic furanone on Cuscuta development identified in this work.

4. Conclusions

From the allelopathy screening of dichloromethane extracts obtained from fifteen weed species, we identified that Conyza bonariensis extract causes strong phytotoxicity against Cuscuta campestris, a parasitic weed that causes worldwide agricultural problems and for which control is limited or non-existent. Sources of allelopathic activity have been previously identified in autotrophic weeds for control of parasitic weed species, such as Orobanche and Phelipanche species [46,47], however, to the best of our knowledge, this is the first report of such type of allelopathy screening of weed species against Cuscuta species, resulting in the identification of Conyza bonariensis as a source of compounds that can lead to the development of new bioherbicides. The bioactivity-guided fractionation of Conyza extract lead us to the isolation of the acetylenic furanone (4Z)-lachnophyllum lactone as the responsible compound for the allelopathic action of C. bonariensis against C. campestris.