Sesquiterpene Lactone Composition of Wild and Cultivated

Across a broad range of Helianthus annuus, differences in composition of sesquiterpene lactones extracted from disc florets were found between wild an...
0 downloads 0 Views 469KB Size
Article pubs.acs.org/JAFC

Sesquiterpene Lactone Composition of Wild and Cultivated Sunflowers and Biological Activity against an Insect Pest Jarrad R. Prasifka,*,† Otmar Spring,§ Jürgen Conrad,Δ Leonard W. Cook,† Debra E. Palmquist,⊥ and Michael E. Foley† †

Northern Crop Science Laboratory, Agricultural Research Service, U.S. Department of Agriculture, Fargo, North Dakota 58102, United States § Institute of Botany and ΔInstitute of Organic Chemistry, University of Hohenheim, 70593 Stuttgart, Germany ⊥ Agricultural Research Service, U.S. Department of Agriculture, Peoria, Illinois 61604, United States S Supporting Information *

ABSTRACT: Sesquiterpene lactones in sunflowers, Helianthus spp., are important to interactions with pathogens, weeds, and insects. Across a broad range of Helianthus annuus, differences in composition of sesquiterpene lactones extracted from disc florets were found between wild and cultivated sunflowers and also between distinct groups of inbreds used to produce sunflower hybrids. Discriminant function analysis showed the presence and relative abundance of argophyllone B, niveusin B, and 15hydroxy-3-dehydrodesoxyfruticin were usually (75%) effective at classifying wild sunflowers, cultivated inbreds, and hybrids. Argophyllone B reduced the larval mass of the sunflower moth, Homeosoma electellum, by >30%, but only at a dose greater than that found in florets. Low doses of mixed extracts from cultivated florets produced a similar (≈40%) reduction in larval mass, suggesting combinations of sesquiterpene lactones act additively. Although the results support a role for sesquiterpene lactones in herbivore defense of cultivated sunflowers, additional information is needed to use these compounds purposefully in breeding. KEYWORDS: Helianthus annuus, Asteraceae, terpenoids, host plant resistance, Pyralidae, argophyllone B, argophyllin, argophyllone B tiglate



INTRODUCTION For sunflowers, Helianthus spp. (Asteraceae), sesquiterpene lactones are important to interactions with pathogens,1 weeds,2,3 and insects4 and have also been useful in chemosystematic investigations of sunflowers and related species.5 Although the capitate glandular trichomes that contain sesquiterpene lactones are present on both sunflower leaves and florets, the very high density of glandular trichomes on disc florets supports their role in defense against floret-feeding insects. For wild sunflower species, pure sesquiterpene lactones cause mortality or delayed development for larvae of the sunflower moth, Homeosoma electellum, a significant floret- and seed-feeding insect pest of sunflower in North America.4,6,7 Several sunflower terpenoids also inhibit feeding or growth of insects that are not strongly associated with sunflowers.6,8 However, cultivated sunflowers are considered deficient in glandular trichomes and associated defensive terpenoids.4,7 Closer examination of the sesquiterpene lactone composition in wild and cultivated Helianthus annuus appears justified for several reasons. First, in at least one instance, sesquiterpene lactones from cultivated sunflower show repellency against an occasional floret-feeding insect (western corn rootworm, Diabrotica virgifera virgifera).8 Second, contrary to previous assertions,4,7 florets of cultivated sunflowers often have equal or greater densities of capitate glandular trichomes relative to wild sunflowers.9 Finally, although some comparative data on sesquiterpene lactones in H. annuus exist,10,11 no broad comparison of sesquiterpene lactones in wild and cultivated specimens is available. Consequently, a cross-section of wild © XXXX American Chemical Society

and cultivated germplasm was used to determine whether significant differences in sesquiterpene lactone composition exist among different types (inbreds, hybrids, and wild accessions) of H. annuus. Additionally, assays with sunflower moth were used to assess insect development when larvae were exposed to mixed and pure extracts of sesquiterpene lactones found in cultivated sunflower.



MATERIALS AND METHODS

Plant Materials. Samples of sesquiterpene lactones for identification via 1H NMR (>1 mg) were obtained by collecting large amounts (>100 g) of H. annuus florets from blooming heads of one cultivated sunflower hybrid and one wild population growing west of Fargo, ND, USA, during 2011. These bulk samples were stored frozen (−20 °C) until needed and then dried under a fume hood prior to sesquiterpene lactone extraction. Bulk samples were used only for identification of individual compounds and were not included in subsequent quantification with smaller samples, in part because light intensity (and possibly other environmental factors) influences the amount of sesquiterpene lactones in H. annuus glandular trichomes.12 To represent the diversity across wild and cultivated H. annuus, 60 different germplasm entries were planted in Fargo, ND, USA, in 2012. Wild material comprised five entries each from North Dakota, Kansas, and Texas. This latitudinal cross section includes much of the native range of H. annuus and reflects the fact that sesquiterpene lactone composition has previously been observed to differ between northern Received: January 20, 2015 Revised: March 13, 2015 Accepted: April 8, 2015

A

DOI: 10.1021/acs.jafc.5b00362 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry and southern populations of Helianthus maximiliani.13 For cultivated material, 30 publicly available inbreds were used, 15 each of maintainer (female) and restorer (male) sunflower lines. Although female lines used to produce hybrids differ from the maintainer lines by having male-sterile cytoplasm (CMS), they are considered equivalent in their nuclear genetic material. Furthermore, the use of CMS lines is likely problematic for sampling terpenoids in floral glandular trichomes, because cytoplasmic male sterility in sunflowers often produces significant deformities of florets. Finally, a group of 15 commercial sunflower hybrids was included. Commercial lines included five entries each from three commercial seed companies; the hybrids are not specifically identified to maintain anonymity of companies donating seed. However, the identities of the other 45 entries are available in Table 2. Seed of the identified entries may be obtained either from the USDA National Plant Germplasm System (for all wild entries) or the USDA Sunflower and Plant Biology Research Unit (for inbred lines) in Fargo, ND, USA. Because pollinators may remove glandular trichomes from blooming sunflowers,14 florets collected in 2012 were sampled just prior to anthesis. For each entry, five florets were carefully removed from each of three plants. After the corollas were removed from the florets (exposing the glandular trichomes), each group of five florets was stored and allowed to dry under refrigeration until analysis. Sesquiterpene Lactone Identification and Quantification. To purify sesquiterpene lactones, air-dried wild or cultivated florets (2−6 g) were vortexed in dichloromethane (DCM, 10 mL/g). The solvent was filtered through a cotton plug and evaporated under a stream of nitrogen at 30 °C. Dried residue was dissolved in 1.0 mL of methanol (MeOH), vortexed, and centrifuged at 13000 rpm for 2 min. Distilled water (H2O) was added to the supernatant to result in a MeOH/H2O ratio of 25:75. Aliquots (1.0 mL) were injected onto a semipreparative HPLC column (10 × 250 mm Hypersil ODS, 5 μm) and eluted with acetonitrile (ACN)/H2O (25:75) at 5 mL/min at 40 °C. Peaks were detected by UV at 225 nm. Individual peak fractions were collected and placed in a −80 °C freezer until lyophilized. Compound identification was performed by means of NMR spectroscopic experiments and data comparison with authentic references. 1H NMR measurements were conducted in a Varian Unity Inova spectrometer at 500 MHz. 1H chemical shifts were referenced to residual solvent signals at δH/C 7.27 (CDCl3) or 3.35 (CD3OD), respectively. The following compounds were identified (numbered according to HPLC fractionation in order of decreasing polarity) according to signal identity with data previously reported in references noted: 1, argophyllin B, and 3, argophyllin A15 2, 1,2-anhydrido-4,5-dihydroniveusin A16 4, 15-hydroxy-3-dehydrodesoxyfruticin1 6, argophyllone B17 7, niveusin B, and 8, niveusin C18 9, haageanolide19 10, xanthomicrol (5, 4′-dihydroxy-6,7,8-trimethoxyflavone)20 An additional compound was identical in most 1H NMR signals with argophyllone B, except for the side-chain signals (H-3′, H-4′, H5′), which indicated an ester residue at C-8 with tiglic acid instead of angelic acid. This structure appears to be previously undescribed and is described here as argophyllone B tiglate (5): 1H NMR (500 MHz, CDCl3) δ 6.83 (1H q, H-3′), 6.36 (1H d, J = 1.0 Hz, H-13a), 5.83 (1H dd, J = 1.4 Hz, H-13b), 5.19 (1H t, J = 3.5 Hz, H-8), 4.60 (1H dd, J = 11.5, 3.1 Hz, H-6), 3.88 (1H dd, J = 9.5, 8.7 Hz, H-15a), 3.70 (1H dd, J = 9.5, 5.0 Hz, H-15b), 3.39 (1H dd, J = 9.6, 5.6 Hz, H-1), 3.30 (1H dd, J = 14.5, 9.6 Hz, H-2a), 3.21 (1H m, H-4), 3.06 (1H bs, H-7), 2.75 (1H dd, J = 15.2, 3.4 Hz, H-9a), 2.37 (1H dd, J = 14.4, 9.8 Hz, H-5a), 2.25 (1H dd, H-2b) 1.92 (1H m, H-5b), 1.80 (3H dq, H-4′), 1.79 (3H s, H-5′), 1.40 (3H s, H-14), 1.37 (1 H dd, J = 15.0, 3.2 Hz, H-9b). Relative quantification of sesquiterpene lactones was accomplished by analyzing three replicate samples from each of the 60 germplasm entries. For each sample, five air-dried florets were vortexed in 0.5 mL of DCM. After 5 min, the samples were vortexed again, and the solvent was pipetted into a microcentrifuge tube and evaporated under a stream of nitrogen at 30 °C. The residue was dissolved by vortexing in

0.2 mL of MeOH. After centrifuging at 13000 rpm for 2 min, the sample was transferred to an autosampler vial for HPLC analysis. Before injection, 0.8 mL of water containing 0.010 mg/mL 2,5dimethylphenol (2,5-DMP) internal standard was added to the sample vial and mixed using the autosampler. From the 1.0 mL mixed in each sample vial, 0.2 mL was injected onto an analytical column (4.6 × 250 mm Hypersil ODS, 5 μm) and eluted with ACN/H2O (25:75) at 1 mL/min at 40 °C. Peaks were monitored at 225, 254, and 265 nm. A representative chromatogram showing UV absorbance at 225 nm for a wild H. annuus is included as Figure S1 of the Supporting Information. Insect Assay. An assay with sunflower moth larvae was used to compare larval development upon exposure to mixed and pure extracts of sesquiterpene lactones found in H. annuus. Previous tests of biological activity of sesquiterpene lactones and diterpene acids have incorporated a known percentage of tested compounds into artificial diet on a dry mass basis.4,7 However, because glandular trichomes are found on the outer surface of florets, applying sesquiterpene lactones onto the surface of an artificial diet may more closely resemble how sunflower moth larvae are normally exposed to the contents of glandular trichomes. Florets from wild (PI 435430) and cultivated (HA 89) entries were collected from sunflower heads protected from pollinators by enclosure in mesh bags. Mixed extracts and pure sesquiterpene lactones were obtained from known numbers of florets as previously described. Each mixed extract or pure compound was dissolved in 600 μL of methanol, after which a 25 μL aliquot was applied to each of 24 wells containing 1.5 mL of a wheat germ-based artificial diet.21 Treatments included mixed extracts of PI 435430 or HA 89, pure argophyllone B (6) or 15-hydroxy-3-dehydrodesoxyfruticin (4), and methanol or untreated controls. On the basis of scanning electron micrographs, the surface area of the wells used for the insect assay is approximately 60 times the area over which glands are present on a single H. annuus floret tip. Consequently, expressing the amount of sesquiterpene lactone in diet assays as floret-equivalents (60 floretequivalents = estimated natural exposure) is an area-to-area conversion that incorporates the natural variation in H. annuus glandular trichome number.9 At least 30 min after application of treatments, two newly hatched sunflower moth larvae were placed into each well, and clear adhesive lids were used to contain the larvae. After 7 days, each well was scored for larval mortality; for all larvae still alive at 7 days, mass was also measured. Data Analysis. Analyses were performed using JMP and SAS software.22,23 Because relationships between peak area (UV absorption) and concentration (or mass) are not established for any of the compounds, mean peak areas (a relative measure) for sesquiterpene lactones were used for statistical analyses. To assess possible relationships between the seven major sesquiterpene lactones, simple and partial correlations (accounting for glandular trichome density; from ref 9) were calculated. To determine whether significant differences in sesquiterpene lactone composition exist among H. annuus germplasm types, quadratic discriminant function analysis was used, which allowed for differing covariance among germplasm types. Variables for inclusion on the discriminant function analysis were chosen by forward stepwise selection (P < 0.05 for entry). To test whether mixed or pure sesquiterpene lactone extracts influenced larval mass at 7 days, a single-factor analysis of variance was used for the seven treatments (five extracts and two controls); infestation date (occurring on two consecutive days) was used as a random blocking effect. Exclusion of outliers (mass more than three deviations above mean of treatment;