Isolation of a New Phenolic Compound, 3, 5-Dihydroxy- 2-(Methoxycarbonylmethyl)phenyl 3, 4-Dihydroxybenzoate, from Leaves of Actinidia chinensis (Kiwifruit)
Kirstin V. Wurms
Janine M. Cooney
Extraction of leaf tissue from the golden–fleshed kiwifruit cultivar Actinidia chinensis `Hort16A` expressing genotype–resistance against the fungus Botrytis cinerea revealed the presence of a novel phenolic compound, 3,5–dihydroxy–2–(methoxycarbonyl-methyl)phenyl 3,4–dihydroxybenzoate (1). In addition, the known phenylpropanoids, caffeic acid (2) and quercetin (3) were amongst the principal compounds present in a fungi–toxic phenolics extract from `Hort16A` leaf tissue. TLC fungi–toxicity bioassays indicated that 100 Î¼g quantities of these pure compounds were not fungi–toxic, but caffeic acid is a central compound in the phenylpropanoid pathway and therefore might function indirectly in host resistance, by acting as a pre–cursor of fungi–toxic secondary metabolites. The other compounds could possibly act in a similar role.
Although many species of Actinidia exist in the wild, A. chinensis Planch. var. chinensis Hort16A, which was recently introduced onto the world market, is one of only two kiwifruit cultivars considered commercially important, the other being A. deliciosa (A. Chev.) C.F. Liang et A.R. Ferguson var. deliciosa Hayward. Grey mould, caused by Botrytis cinerea is a serious disease of Hayward kiwifruit, but studies have shown that leaves and fruit of Hort16A express greater constitutive resistance to this pathogen than Hayward and that resistance is associated with phenolic compounds (Wurms et al., 2003; Wurms, 2004). Given the recent introduction into commercial production of Hort16A, there are few detailed studies on the phenolics composition of the Hort16A cultivar, particularly in relation to disease resistance.
The present study was initiated to characterize some of these compounds and determine their fungi-toxicity to facilitate better understanding of resistance mechanisms against Botrytis cinerea in Hort16A. Here we present results from a bioassay directed fractionation study of leaf tissue from Actinidia chinensis Hort16A and describe the isolation and identification of a novel phenolic compound, 3,5-dihydroxy-2-(methoxycarbonylmethyl)phenyl 3,4-dihydroxybenzoate (1) along with several known phenolic compounds which together constituted the principal components of a fungi-toxic extract.
Materials and Methods
General Experimental Procedures
NMR spectra were determined on a Bruker Avance DRX-400 spectrometer at 400.13
(1H) and at 100.61 MHZ (13C). Homonuclear 1H
connectivities were determined by 2D double-quantum filtered COSY, long range
2D double-quantum filtered COSY and long range 1D TOCSY experiments. One-bond
heteronuclear 1H-13C connectivities were determined by
a 2D proton-detected HSQC experiment and two- and three-bond 1H-13C
connectivities were determined by a 2D proton-detected HMBC experiment. Chemical
shifts are reported relative to DMSO.
LC-MS employed a LCQ Deca ion trap mass spectrometer fitted with an APCI interface (ThermoQuest, Finnigan, San Jose, CA, USA) and coupled to a SurveyorTM HPLC and PDA detector. The analytical column employed was a Prodigy 5 μ ODS(3) 100Ε (Phenomenex, Torrance, CA, USA), 150 x 2 mm. A 0.2 μ in-line filter (Alltech, Deerfield, Illinois, USA) was installed before the analytical column and the temperature of the column oven was maintained at 35°C. Compounds were separated isocratically using 36:64 MeOH+0.1% FA (formic acid)-H2O. Flow rate was 250 μL min-1 and the injection volume 10 μL. UV-VIS detection was by absorbance at 200-600 nm. MS data was acquired in the negative mode for most experiments using a data-dependent LC-MSn method. The APCI vaporizer temperature, capillary temperature, capillary voltage, tube lens offset, sheath gas pressure and auxiliary gas were set at 450, 200°C, -12 and -20 V, 60 and 0 psi, respectively.
High-resolution mass spectra were determined in the FAB mode on a VG70S magnetic sector mass spectrometer.
HPLC was performed using a Shimadzu system (Kyoto, Japan), equipped with a 150x4.6 mm Reversed-phase (RP) column (Prodigy 5μ ODS (3), Phenomenex, Torrance, California, USA), with detection at 280 nm and photodiode-array analysis at 200-360 nm.
Standards of known kiwifruit phenolic compounds (Webby, 1990; Webby et al., 1994; Dawes and Keene, 1999; Lahlou et al., 2001), purchased from Sigma-Aldrich (St. Louis, Missouri, United States) and Apin Chemicals Ltd (Abingdon, England), were analyzed by RP-HPLC and LC-MS and comprised 125 μg mL-1 MeOH of chlorogenic acid, hydroxybenzoic acid, caffeic acid, epi-catechin, catechin, ρ-coumaric acid, rutin, quercetin-3-glucoside, quercetin-3-rhamnoside, quercetin, kaempferol and rhamnetin and 100 μg mL-1 MeOH of 2α,3β,23-trihydroxy-12,20(30)-ursadien-28-oic acid, otherwise known as actinidic acid (which was a generous gift of Dr Nobuhiro Hirai, Kyoto University, Japan).
A TLC spot bioassay described earlier (Wurms et al., 2003) was used to determine anti-fungal activity of crude extracts, standards, purified compounds and chromatographic fractions and thereby to guide compound isolation. In a biological system, a compound is generally considered to be fungi-toxic, if growth inhibition in the TLC-bioassay occurs with <50 μg of pure product (Skipp and Bailey, 1977; Niemann, 1993; Fawe et al., 1998).
Randomly-selected healthy leaves from T-bar orchard-grown mature Hort16A
kiwifruit vines (Hamilton, New Zealand, 2002) were inoculated with 5 mm diam.
14-28 day old agar inoculum plugs (four plugs/leaf) of B. cinerea, which
were secured face-down on the adaxial leaf surface with a piece of Danco 130
all-weather transparent tape (Danco (NZ) Ltd., Wellington, New Zealand). Six
days after inoculation, leaf tissue was pooled from: (1) the diseased inoculation
sites and (2) the ring of symptom-free leaf tissue immediately surrounding the
diseased lesions. A total of 27.5 g of lyophilized leaf tissue (equivalent to
approximately 110 g FW) was used for extraction and isolation of the phenolic
Extraction and Isolation
The detailed extraction protocol is described by Wurms et al. (2003),
except that the initial extraction comprised 2x24 h extractions in 80% v/v MeOH
at 20 mL MeOH/g dry matter (DM). The acid-hydrolyzed phenolic extract (20 mL
total of 1.5 g DM/mL MeOH) from the above extraction procedure was diluted 1:1
with water and then fractionated on 10 g Reversed-phase (RP) Solid Phase Extraction
(SPE) cartridge (High Load C18, Alltech Associates, Inc., Deerfield, IL) using
a step-wise gradient of 0:100, 50:50, 80:20 and 100:0 MeOH-H2O, where
all solvents were acidified with 2.5% v/v glacial acetic acid to suppress compound
ionization. TLC fungi-toxicity bioassay screening indicated that only the acidified
50:50 MeOH-H2O fraction was markedly fungi-toxic and therefore of
interest for further purification. Between each subsequent purification step
described below, the fraction containing the compound(s) of interest was concentrated
by rotary evaporation until the point of dryness, followed by re-dissolution
in the smallest possible volume of 50:50 H2O-MeOH (both solvents
acidified with 2.5% v/v acetic acid). Resolution of the fungi-toxic fraction
by RP-HPLC using initial solvent conditions of 14.5:85.5 CH3CN-H2O+0.1%
TFA (trifluoroacetic acid) and a linear solvent gradient described previously
(Wurms et al., 2003), with a flow rate of 1 mL min-1, revealed
the presence of 5 major peaks eluting at 8.3, 20.9, 22.5, 23.8 and 29.1 min.
Co-chromatography with authentic standards and resulting matches in retention
times, absorption spectra and mass spectral data indicated that the peaks at
8.3 and 29.1 min were caffeic acid (2) and quercetin (3), respectively. The
other peaks did not match any of the known standards and LC-MS analysis indicated
the presence of multiple compounds within each peak (data not shown). To further
separate these compounds, RP-HPLC was employed using an initial mobile phase
of 18:82 CH3CN-H2O+0.1% TFA and isocratic conditions for
10 min, followed by a linear gradient to 28:72 CH3CN-H2O+0.1%
TFA over 10 min, held for 10 min before resetting to the original conditions
over the next 5 min. Final separation by RP-HPLC was achieved isocratically
using 27:73 MeOH-H2O+0.1% TFA and yielded 2 mg of compound 1. This
compound, which eluted at 22.5 min in the original RP-HPLC gradient, was then
dried under a nitrogen stream, re-dissolved in 1 ml of DMSO and analyzed by
3,5-dihydroxy-2-(methoxycarbonylmethyl)phenyl 3,4-dihydroxybenzoate (1)
Beige needle-like crystals; UV (MeOH) λmax 210, 268, 298
nm; 1H NMR and 13C NMR (Table 1); HRFABMS
m/z 335.0760 (calcd for C16H15O8, [M+H]+,
335.0767); LC-MS m/z 333.0 (M-H)-; 301.0 (MS2 333); 165.1
(MS3 333>301); 137.2 (MS4 333>301>165); Rf
0.29 (Si gel 60 F234, 35:65 CH2Cl2:EtOAc);
100 μg of pure compound was not fungi-toxic in the TLC bioassay.
Caffeic acid (2) and quercetin (3) were identified by comparison of their
spectral data with literature values (Mabry et al., 1970; Harborne and
Mabry, 1982; Dey and Harborne, 1989) and unambiguously confirmed using RP-HPLC
and LC-MS co-chromatography experiments with authentic standards.
|| 1D and 2D NMR Spectral Data for Compound 1
Results and Discussion
A crude methanolic extract of 110 g fresh weight (FW) of A. chinensis Planch. var. chinensis Hort16A leaf tissue was fractionated by reversed-phase solid phase (RP-SPE) extraction and the fractions were bioassayed (using a TLC spot bioassay described earlier) (Wurms et al., 2003) for fungi-toxicity. Fungi-toxic activity was predominantly associated with the fourth RP-SPE fraction (RP-SPE 4). This fraction became the focus of all subsequent purification work. Three of the major phenolic constituents in RP-SPE 4 were identified: 3,5-dihydroxy-2-(methoxycarbonylmethyl)phenyl 3,4-dihydroxybenzoate (1) a novel compound which to our knowledge has never before been characterized and two known phenolics-a hydroxycinnamic acid, caffeic acid (2) and a flavonol, quercetin (3) (Fig. 1). Approximately 80% of the total quantity of compound 1 was found in RP-SPE 4 and 2 mg of pure material was recovered from this fraction (ca. 20% recovery of the total yield in RP-SPE 4), while 2 and 3 were present at ca. 139.0 and 973.4 μg g-1 FW leaf tissue, respectively.
Identification of 1 was achieved via a combination of mass spectral and one
and two-dimensional NMR studies. LC-MS using APCI ionization in negative ion
mode afforded a strong (M-H)- ion at m/z 333. Fragmentation of this
ion employing a data-dependent MSn experiment gave the MS2
daughter ion m/z 301, suggesting loss of a methoxyl group as a mole of methanol,
followed by the MS3 granddaughter ion m/z 165 and MS4
ion m/z 137 (Fig. 2). HRFABMS gave [M+H]+ m/z 335.0767
corresponding to a molecular formula of C16H14O8,
which was consistent with the results of 13C NMR and DEPT experiments
(Table 1). These experiments revealed 9 singlet resonances,
2 of which were consistent with chemical shifts expected for ester carbonyl
groups (171.2 and 163.6 ppm), 5 doublet resonances, a quartet resonance consistent
with an oxy methane group (51.3 ppm) and a triplet resonance (28.6 ppm). 1H
NMR, COSY and LRCOSY experiments showed 2 isolated aromatic spin systems comprising
2 doublets (7.52 ppm J = 2.0 Hz, 6.97 ppm J = 8.2 Hz) and a doublet of doublets
(7.50 ppm J = 8.2, 2.0 Hz) in 1 ring and 2 doublets (6.35 ppm J = 2.0 Hz, 6.16
ppm J = 2.0 Hz) in a second ring. Coupling constants and chemical shift values
suggested one aromatic ring with substituents at the 1, 3 and 4 positions, the
latter 2 oxygenated and a second ring with 4 substituents, 3 of which were oxygenated.
||Chemical structures for compounds 1, 3,5-dihydroxy-2-(methoxycarbonylmethyl)-phenyl
3,4-dihydroxybenzoate, 2 caffeic acid and 3 quercetin
|| LC-MS fragmentation data for compound 1
|| Selected HMBC correlations for compound 1
|| Chemical structure of 2-O-(3,4-dihydroxy-benzoyl)-2,4,6-trihydroxyphenylacetic
This data was consistent with a penta-hydroxylated structure comprising A and
B rings as for quercetin (Fig. 1), but without a C ring. The
complete structure was finally elucidated using a combination of a long range
1D selective TOCSY experiment and HSQC and HMBC spectral data. In an experiment
with a 250 msec mixing time, irradiation of the CH2 group at 3.35
ppm showed weak but detectable correlations to the methoxy signal at 3.48 ppm
and the aryl B ring protons at 6.16 ppm (H-6) and 6.35 ppm (H-4).
Selected HMBC correlations are shown in Fig. 3 and are consistent
with the LC-MS fragmentation data previously described (Fig. 2).
The structure of 1 was thus determined to be 3,5-dihydroxy-2-(methoxycarbonylmethyl)phenyl
To our knowledge, compound 1 represents a novel member of the depside family of compounds. Depsides are defined as substances derived from 2 or more molecules of the same or different phenolcarboxylic acids by esterification of the carboxyl group of one molecule with the hydroxyl group of a second molecule (Coffey, 1976; McNaught and Wilkinson, 1997). Most natural depsides occur in lichens, where they are biosynthesized via predominantly the acetate pathway and to a lesser extent via the shikimic acid pathway (Coffey, 1976), but they have also been isolated from higher plants including a number of commonly known medicinal herbs from the Lamiaceae family, including peppermint (Mentha piperita (L.) Hudson), marjoram (Origanum majorana L.), sage (Salvia officinalis L.), thyme (Thymus vulgaris L.) lavender (Lavandula officinalis Chaix), rosemary (Rosmarinus officinalis L.) and hyssop (Hyssopus officinalis L.) (Zgórka and Glowniak, 2001). Depsides are known to be active ingredients in European and Chinese folk medicines owing to their anti-oxidant, anti-inflammatory, anti-microbial and anti-pyretic activities (Hidalgo et al., 1994; Neamati et al., 1997; Li, 1998; Esimone and Adikwu, 1999; Zgórka and Glowniak, 2001; Piovano et al., 2002). The most closely related previously-reported structure to compound 1 is the depside, 2-O-(3,4-dihydroxy-benzoyl)-2,4,6-trihydroxyphenylacetic acid, shown in Fig. 4, which was isolated from the herbaceous plant, Papaver rhoeas L. (Hillenbrand et al., 2004). The difference between the 2 compounds is that the carboxyl group attached to ring A in 2-O-(3,4-dihydroxy-benzoyl)-2,4,6-trihydroxyphenylacetic acid is replaced by a methyl ester in compound 1 (Figure 1 c.f. Figure 4). However it is extremely unlikely that compound 1 is a derivative of 2-O-(3,4-dihydroxy-benzoyl)-2,4,6-trihydroxyphenyl-acetic acid, created artifactually by extraction in methanol, since the initial stages of the extraction process involving methanol were identical for both compounds, i.e., in both cases powdered plant material was extracted in 80% methanol, followed by evaporative removal of the methanol under vacuum, leaving the aqueous residue, which was then partitioned against various solvents (Hillenbrand et al., 2004). Methanol is a commonly used solvent in the initial stages of extraction of phenolic compounds, including depsides (Elix et al., 1999; Zgórka and Glowniak, 2001; McNally et al., 2002; Hillenbrand et al., 2004) and methanolic degradation of phenolic compounds to form depsides does not occur spontaneously, but rather requires irradiation of the methanol solvent to generate reactive free radicals that initiate redox reactions of flavonoids leading to depside formation (Marfak et al., 2002, 2003). Moreover, based on the mass and spectral data for 2-O-(3,4-dihydroxy-benzoyl)-2,4,6-trihydroxyphenylacetic acid (Hillenbrand et al., 2004), this compound was not present as a major component in our fungi-toxic RP-SPE 4 phenolics fraction.
Compounds 1-3 are unlikely to be anti-fungal in planta, since 100 μg quantities of the pure compounds (equivalent to>0.5g FW of leaf tissue) were not fungi-toxic in the TLC fungi-toxicity bioassay, whereas replicates of RP-SPE 4 extract (equivalent to>0.5g FW of leaf tissue) produced fungal growth inhibition zones with a mean diameter of 17 mm. However, compounds 1-3 accounted for 3 of 5 of the major components in the fungi-toxic RP-SPE 4 phenolics fraction. Wurms et al. (2003) demonstrated that the degree of anti-fungal activity of Hort16A leaf extracts (as quantified in the fungitoxicity bioassay by diameter of fungal growth inhibition zones) was correlated positively with concentrations of three phenolic compounds-caffeic acid and two other compounds of unknown identity. A comparison of spectral and mass data and retention times indicates that one of these previously observed compounds can now be characterized as novel compound 1 described in the current study. Although not inherently fungi-toxic, these phenolics may contribute indirectly to defense in a number of ways: by acting as precursors or co-factors of directly anti-fungal compounds, as signal molecules, as elicitors of other plant defenses, or as inhibitors of key pathogenesis enzymes. For example, caffeic acid is central to plant secondary metabolism because it acts as a building block in the phenylpropanoid pathway (Dey and Harbourne, 1989; Harbourne, 1998), while quercetin is one of the most ubiquitous flavonols and functions ascribed to flavonols include UV and disease protection, pollination attractants, enzyme inhibitors, antioxidants, plant hormone controllers and allelopathic agents (Dey and Harbourne, 1989). Indeed, many consider the indirect contributions of various endogenous compounds to host resistance to be more important than direct effects (Kuc, 1995).
Future research will investigate the applicability of compound 1 as a marker of host resistance for screening novel germplasm arising from breeding trials.
We thank Dr. A. Wilkins, Dr. R. Thomson and Dr. L. Main, University of Waikato, for helpful NMR discussions and advice on nomenclature. Thanks also to Mr. D. Jensen and Mr. J. Allan, HortResearch, for excellent technical assistance with LC-MS and FAB mass spec. analyses; Dr. D. Rowan and Dr. D. Lauren, HortResearch for critiquing the manuscript; and the Agricultural and Marketing Research and Development Trust (AGMARDT) of New Zealand for funding this study.
Dawes, H.M. and J.B. Keene, 1999. Phenolic composition of kiwifruit juice. J. Agric. Food Chem., 47: 2398-2403.
PubMed | Direct Link |
Elix, J.A., C.E. Barclay, J.H. Wardlaw, A.W. Archer, S.H. Yu and G. Kantvilas, 1999. Four new β-orcinol meta-depsides from Pertusaria and Siphula lichens. Aust. J. Chem., 52: 837-840.
Esimone, C.O. and M.U. Adikwu, 1999. Antimicrobial activity and cytotoxicity of Ramalina farinacea. Fitoterapia, 70: 428-431.
CrossRef | Direct Link |
Fawe, A., M. Abou-Zaid, J.G. Menzies and R.R. Belanger, 1998. Silicon-mediated accumulation of flavonoid phytoalexins in cucumber. Biochem. Cell Biol., 88: 396-401.
CrossRef | Direct Link |
Harborne, J.B. and T.J. Mabry, 1982. The Flavonoids: Advances in Research. Chapman and Hall, London, ISBN: 0-412-22480-1, pp: 24-132.
Harborne, J.B., 1998. Phytochemical Methods: A Guide to Modern Techniques of Plant Analysis. 3rd Edn., Chapman and Hall, London, Pages: 302.
Hidalgo, M.E., E. Fernandez, W. Quilhot and E. Lissi, 1994. Antioxidant activity of depsides and depsidones. Phytochemistry, 37: 1585-1587.
PubMed | Direct Link |
Hillenbrand, M., J. Zapp and H. Becker, 2004. Depsides from the petals of Papaver rhoeas. Planta Med., 70: 380-382.
PubMed | Direct Link |
Kuc, J., 1995. Phytoalexins, stress metabolism and disease resistance in plants. Annu. Rev. Phytopathol., 33: 275-297.
PubMed | Direct Link |
Lahlou, E.H., N. Hirai, T. Kamo, M. Tsuda and H. Ohigashi, 2001. Actinidic acid, a new triterpene phytoalexin from unripe kiwifruit. Biosci. Biotech. Biochem., 65: 480-483.
Direct Link |
Li, L.N., 1998. Biologically active components from traditional Chinese medicines. Pure Applied Chem., 70: 547-554.
Mabry, T.J., K.R. Markham and H.B. Thomas, 1970. The Systematic Identification of Flavonoids. Springer Verlag Heidelberg, New York.
Marfak, A., P. Trouillas, D.P. Allais, Y. Champavier, C.A. Calliste and J.L. Duroux, 2002. Radiolysis of quercetin in methanol solution: Observation of depside formation. J. Agric. Food Chem., 50: 4827-4833.
CrossRef | Direct Link |
Marfak, A., P. Trouillas, D.P. Allais, Y. Champavier, C.A. Calliste and J.L. Duroux, 2003. Radiolysis of kaempferol in water/methanol mixtures. Evaluation of antioxidant activity of kaempferol and products formed. J. Agric. Food Chem., 51: 1270-1277.
PubMed | Direct Link |
McNally, D.J., K. Wurms, C. Labbe and R.R. Belanger, 2002. A standardized methodology for the study of induced glycosylated plant phenolics. Can. J. Plant. Pathol., 24: 429-436.
Direct Link |
McNaught, A.D. and A. Wilkinson, 1997. IUPAC Compendium of Chemical Terminology. 2 Edn., Blackwell Science Ltd., Oxford, UK., ISBN: 0865426848.
Neamati, N., H. Hong, A. Mazumder, S. Wang and S. Sunder et al., 1997. Depsides and depsidones as inhibitors of HIV-1 integrase: Discovery of novel inhibitors through 3D database searching. J. Med. Chem., 40: 942-951.
PubMed | Direct Link |
Niemann, G.J., 1993. The anthranilamide phytoalexins of the caryophyllaceae and related compounds. Phytochemistry, 34: 319-328.
Direct Link |
Piovano, M., J.A. Garbarino, F.A. Giannini, E.R. Correche and G. Feresin et al., 2002. Evaluation of antifungal and antibacterial activities of aromatic metabolites from lichens. Bol. Soc. Chil. Quim., 47: 235-240.
Skipp, R.A. and J.A. Bailey, 1977. The fungitoxicity of iso-flavonoid phytoalexins measured using different types of bioassay. Physiol. Plant Pathol., 11: 101-112.
Webby, R.F., 1990. Flavonoid complement of cultivars of Actinidia deliciosa var. deliciosa, kiwifruit. N. Z. J. Crop Hortic. Sci., 18: 1-4.
Webby, R.F., R.D. Wilson and A.R. Ferguson, 1994. Leaf flavonoids of Actinidia. Biochem. Syst. Ecol., 22: 277-286.
Direct Link |
Wurms, K.V., 2004. The incidence of Botrytis cinerea and expression of putative host defences in green- and golden-fleshed kiwifruit of differing harvest maturity. N. Z. Plant Prot., 57: 125-129.
Wurms, K.V., M.P.George and D.R. Lauren, 2003. Involvement of phenolic compounds in host resistance against Botrytis cinerea in leaves of the two commercially important kiwifruit (Actinidia chinensis and A. deliciosa) cultivars. N.Z. J. Crop Hortic. Sci., 31: 221-233.
Direct Link |
Zgorka, G. and K. Glowniak, 2001. Variation of free phenolic acids in medicinal plants belonging to the Lamiaceae family. J. Pharm. Biomed. Anal., 26: 79-87.
Direct Link |