Journal of Chemical Ecology, Vol. 19, No. 11, 1993
S E A S O N A L V A R I A T I O N OF E X U D A T E OF
Cistus ladanifer
NATIVIDAD
CHAVES,
and CARLOS
l'* JOSt~ CARLOS
ESCUDERO,
GUTII~RREZ-MERINO
1
2
IDepartamento de Ecologia Facultad de Ciencias Universidad de Extremadura, 06080-Badajoz, Spain 2Departamento de Bioqufmica y Biolog[a Molecular Facultad de Ciencias Universidad de Extremadura 06080-Badajoz, Spain (Received January 29, 1993; accepted June 28, 1993)
Abstract--The production of labdanum exudate by Cistus ladanifer L. is highly seasonal, reaching a maximum concentration during summer and a minimum concentration in winter. Because this exudate strongly absorbs in the wavelength range of 260-400 nm (the near-UV-visible range), it may be important biologically as an UV-visible filter. Separation of exudate components has been achieved by reverse-phase high-performance liquid chromatography (HPLC). The retention times of HPLC chromatograms and the spectral characteristics (absorption and fluorescence) of the exudate identify flavonoids as the most relevant chromophores regarding the potency of the exudate as a UV-visible filter. HPLC studies show that kaempferol-3(O)methyl, kaempferol-3,7-di(O)methyl, and apigenin-4'-(O)methyl are the most enriched flavonoids in the exudate. Other flavonoids [apigenin, apigenin7-(O)methyl, apigenin-7,4'-di(O)methyl, kaempferol-3,4'-di(O)methyl and kaempferol-3,7,4'-tri(O)methyl] are present in the exudate as minor components, e.g., each contributes by less than 10% to total flavonoids. The ratio of kaempferols to apigenins of the exudate also shows seasonal variation (maximum value in summer and minimum in spring). However, due to the similar absorption spectra of both groups of flavonoids, this has a minor influence on the exudate's ability to filter near-UV-visible radiation. Key Words--Flavonoids, seasonal variation, Cistus ladanifer L., UV filter, labdanum exudate. *To whom correspondence should be addressed. 2577 0098-0331/93/1100-2577507.00/09 1993PlenumPublishingCorporation
2578
CHAVES ET AL. INTRODUCTION
Plants have adapted to hostile environments by a variety of morphological and physiological mechanisms. Biochemical adaptations have resulted in a complex pool of secondary metabolites. In particular, the flavonoids have been suggested to function as natural antioxidants, specific enzyme inhibitors, regulators of plant growth and development, microbial and animal toxins, and near-UV light screens (Fimin et al., 1986; Harborne, 1986; Hedin et al., 1983; McClure, 1975; Peters et al., 1986; Tom~is-Llorente et al., 1990). Flavonoids strongly absorb UV-visible radiation between 260 and 380 nm, a wavelength range associated with enhanced mutagenesis, DNA thymine dimerization, and photolysis of NAD and NADP, all capable of causing cellular death (Swain, 1970). Prior studies have shown that plant tissues exposed to high UV irradiance accumulate large amounts of flavonoids, mostly in the epidermal cells and in their vacuoles (Caldwell, 1971; Les and Sheridan, 1990; Hahlbrook et al., t982; Knogge and Wessenb6ck, 1986; Matem et al., 1983; McClure, 1986; Schnabl et al., 1986; Swain, 1975). These studies lend support to the hypothesis that flavonoids serve as plant UV filters. Although flavonoids occur in epidermal cells of most plants, there can be variation in tissues within the same plant, or from plant to plant. These patterns have been attributed both to environmental and genetic bases (Bohm, 1987; Les and Sheridan, 1990; Roberts and Haynes, 1986; Swain, 1970). Swain (1970, 1975) suggested that flavonoids may have played a major role in the colonization of land by plants. We used the Mediterranean species Cistus ladanifer L. to test the hypothesis that flavonoids might function ecophysiologically as UV filters, because C. ladanifer is a pioneer species (Herrera, 1984; Nufiez, 1989) and is able to colonize land exposed to irradiance higher than 360 cal/m2/day (Matias, 1990). Leaves of this plant are also known to exude a sticky mixture containing flavonoids (J. De Pascual et al., 1974; Proksch and Gtilz, 1984), whose physiological relevance is not well understood. If the labdanum exudate of C. ladanifer is biologically important to protect this plant against the damaging effects of near-UV-visible solar radiation, it can be predicted that its secretion should be seasonal, attaining a maximum during summer. The aim of this study is to test this hypothesis. METHODS AND MATERIALS
Specimens of C ladanifer were sampled at the Sierra de los Conejeros, located near Alburquerque in the county of Badajoz (Spain). This site is rich in C. ladanifer, and its proximity (40 km from Badajoz) facilitates collection and processing within 1 hr. This mountainous site is 340-360 m above sea level, with an average annual rainfall of 500 mm (Cabezas et al., 1983). The rainy
FLAVINOID SEASONAL VARIATION
2579
months are October, November, March, and April. Annual average minimum and maximum temperatures are 10.7 and 19.8~ respectively (Cabezas and Escudero, 1989). Sample Collection. We examined exudate production in four tissue types: (1) mature leaves (nascent in the spring and exude high levels of labdanum during the summer), (2) sticky senescent leaves, characteristically darker with a brown-red color, (3) nascent summer leaves, active in labdanum secretion, and (4) upper part of photosynthetic stems, in which exudate is also evident. Samples were standardized to 17 g of each tissue type (Chaves, 1991), and were collected during a nearly two-year interval (May 1990 until December 1991). Sample Handlings. The extracts of labdanum exudate were prepared as follows: Using forceps, 0.7-g pieces of ethanol-soaked Whatman 118 filter paper were wiped over the surface of freshly cut leaves and stems until all visible traces of exudate were removed. Usually this operation was repeated several times, with special care taken to avoid putting pressure on the biological samples during handling. Filter paper samples were then transferred to glass flasks, and l0 ml of absolute ethanol was added to extract the exudate from the paper. Under mild stirring the ethanol solution soon turned yellow-green. The flasks were then carefully sealed to avoid losses by evaporation, and stored at 4~ until use. Sample Analysis. Absorption spectra were recorded with a diode array Hewlett Packard 8451A, from 250 to 700 nm. The samples were diluted in absolute ethanol to ensure that the maximum absorbance was less than 1.0 optical density at the peak wavelength in the spectrophotometer cuvette. HPLC chromatograms were carried out in a LCD Analytical HPLC instrument equipped with an integrator, and with a Nucleosil 5 C~8 column of 150 x 4 mm. In each case, 20 #1 of diluted ethanol extract was injected. Initially all samples were analyzed at wavelengths of 260, 280, 300, and 345 nm. After noting only minor differences between detection at 260 and 280 nm and that detection at 300 nm yields a mixed result between 280 and 345 nm, we decided to complete the study with the detection set at 260 and 345 nm. Chemicals. The following chemicals were used as standards: cinnamic acid, hydrocinnamic acid, benzyl alcohol, dimethyl- and diethyl-phthalate, apigenin (all from Aldrich Chemical Co.); apigenin-7-(O)methyl, kaempferol-3,7di(O)methyl, kaempferol-3,7,4'-tri(O)methyl (all kindly supplied by Dr. J. Gonzalez-Urones, University of Salamanca, Spain); other apigenins and kaempferols [kaempferol-3(O)methyl, kaempferol-3,7-di(O)methyl, kaempferol-3,4'di(O)methyl, apigenin-4'(O)methyl, apigenin-4',7-di(O)methyl and apigenin-7(O)methyl] were kindly supplied by Drs. Thomas Vogt (Botanisches Institut, Universitfit K61n, Germany) and E. Wollenweber (Institut ftir Botanik, Technische Hochschule Darmstadt, Germany). All other chemicals were purchased from Merck, Darmstadt, Germany, or Carlo Erba, Milano, Italy.
2580
CrtAVESET AL. RESULTS
Spectral Properties of Labdanum Exudate. The labdanum exudate is a sticky mixture that is clearly noticeable on leaves of C. ladanifer from the end of May until the end of October. Figure 1 shows the absorption spectra of samples taken in May, August, and October, with strong absorption in the nearUV to visible wavelength range up to 400 nm and peaks at 260, 310, and 350 nm. On the basis of previous studies (J. De Pascual et al., 1974; Proksch and Gfilz, 1984), several flavonoids are likely components of the extracts of labdanum exudate, namely, apigenin-7-(O)methyl, kaempferol-3,7-di(O)methyl and kaempferol-3,7,4'-tri(O)methyl. This is further supported by the comparison of the absorption and fluorescence spectra of the exudate (Figure 1) with that of purified flavonoids (Figure 2). The mild treatment used to prepare the samples of exudate does not extensively damage the epidermal plant cells. Leaves treated to extract the labdanum exudate in situ in plants, under low sunlight irradiance, do not become senescent and slowly (in several days) recover their sticky appearance, and their ability to produce exudate remains largely unimpaired. Moreover, epicuticular flavonoids have been shown to be predominantly aglycones, as opposed to intracellular (vacuolar) flavonoids, which are highly glycosylated (Vogt et al., 1987a, b; Proksch and Gfilz, 1984; Wollenweber and Dietz, 1981). Flavonoid glycosides have been shown to be well resolved and separated from aglycone flavonoids by paper chromatography using H20 as eluant, due to the much lower Rf values of the glycones (Markham, 1982). Chromatograms run under these
\
240
Wavelength (nm)
450
280 340
400 460 520 580 Wavelength (nm)
640
FIG. 1. Absorption and fluorescence emission spectra of the exudate of Cistus ladanifer. The spectra shown correspond to samples taken on May 29 (broken line), on August 30 (solid line), and on October (bold solid line). (A) Absorption spectra in methanol. (B) Fluorescence emission spectra recorded with an excitation wavelength of 320 nm in ethanol (a.u. stands for arbitrary unit).
FLAVINOID
SEASONAL
258 1
VARIATION
B v
S E A S O N A L V A R I A T I O N OF E X U D A T E OF
Cistus ladanifer
NATIVIDAD
CHAVES,
and CARLOS
l'* JOSt~ CARLOS
ESCUDERO,
GUTII~RREZ-MERINO
1
2
IDepartamento de Ecologia Facultad de Ciencias Universidad de Extremadura, 06080-Badajoz, Spain 2Departamento de Bioqufmica y Biolog[a Molecular Facultad de Ciencias Universidad de Extremadura 06080-Badajoz, Spain (Received January 29, 1993; accepted June 28, 1993)
Abstract--The production of labdanum exudate by Cistus ladanifer L. is highly seasonal, reaching a maximum concentration during summer and a minimum concentration in winter. Because this exudate strongly absorbs in the wavelength range of 260-400 nm (the near-UV-visible range), it may be important biologically as an UV-visible filter. Separation of exudate components has been achieved by reverse-phase high-performance liquid chromatography (HPLC). The retention times of HPLC chromatograms and the spectral characteristics (absorption and fluorescence) of the exudate identify flavonoids as the most relevant chromophores regarding the potency of the exudate as a UV-visible filter. HPLC studies show that kaempferol-3(O)methyl, kaempferol-3,7-di(O)methyl, and apigenin-4'-(O)methyl are the most enriched flavonoids in the exudate. Other flavonoids [apigenin, apigenin7-(O)methyl, apigenin-7,4'-di(O)methyl, kaempferol-3,4'-di(O)methyl and kaempferol-3,7,4'-tri(O)methyl] are present in the exudate as minor components, e.g., each contributes by less than 10% to total flavonoids. The ratio of kaempferols to apigenins of the exudate also shows seasonal variation (maximum value in summer and minimum in spring). However, due to the similar absorption spectra of both groups of flavonoids, this has a minor influence on the exudate's ability to filter near-UV-visible radiation. Key Words--Flavonoids, seasonal variation, Cistus ladanifer L., UV filter, labdanum exudate. *To whom correspondence should be addressed. 2577 0098-0331/93/1100-2577507.00/09 1993PlenumPublishingCorporation
2578
CHAVES ET AL. INTRODUCTION
Plants have adapted to hostile environments by a variety of morphological and physiological mechanisms. Biochemical adaptations have resulted in a complex pool of secondary metabolites. In particular, the flavonoids have been suggested to function as natural antioxidants, specific enzyme inhibitors, regulators of plant growth and development, microbial and animal toxins, and near-UV light screens (Fimin et al., 1986; Harborne, 1986; Hedin et al., 1983; McClure, 1975; Peters et al., 1986; Tom~is-Llorente et al., 1990). Flavonoids strongly absorb UV-visible radiation between 260 and 380 nm, a wavelength range associated with enhanced mutagenesis, DNA thymine dimerization, and photolysis of NAD and NADP, all capable of causing cellular death (Swain, 1970). Prior studies have shown that plant tissues exposed to high UV irradiance accumulate large amounts of flavonoids, mostly in the epidermal cells and in their vacuoles (Caldwell, 1971; Les and Sheridan, 1990; Hahlbrook et al., t982; Knogge and Wessenb6ck, 1986; Matem et al., 1983; McClure, 1986; Schnabl et al., 1986; Swain, 1975). These studies lend support to the hypothesis that flavonoids serve as plant UV filters. Although flavonoids occur in epidermal cells of most plants, there can be variation in tissues within the same plant, or from plant to plant. These patterns have been attributed both to environmental and genetic bases (Bohm, 1987; Les and Sheridan, 1990; Roberts and Haynes, 1986; Swain, 1970). Swain (1970, 1975) suggested that flavonoids may have played a major role in the colonization of land by plants. We used the Mediterranean species Cistus ladanifer L. to test the hypothesis that flavonoids might function ecophysiologically as UV filters, because C. ladanifer is a pioneer species (Herrera, 1984; Nufiez, 1989) and is able to colonize land exposed to irradiance higher than 360 cal/m2/day (Matias, 1990). Leaves of this plant are also known to exude a sticky mixture containing flavonoids (J. De Pascual et al., 1974; Proksch and Gtilz, 1984), whose physiological relevance is not well understood. If the labdanum exudate of C. ladanifer is biologically important to protect this plant against the damaging effects of near-UV-visible solar radiation, it can be predicted that its secretion should be seasonal, attaining a maximum during summer. The aim of this study is to test this hypothesis. METHODS AND MATERIALS
Specimens of C ladanifer were sampled at the Sierra de los Conejeros, located near Alburquerque in the county of Badajoz (Spain). This site is rich in C. ladanifer, and its proximity (40 km from Badajoz) facilitates collection and processing within 1 hr. This mountainous site is 340-360 m above sea level, with an average annual rainfall of 500 mm (Cabezas et al., 1983). The rainy
FLAVINOID SEASONAL VARIATION
2579
months are October, November, March, and April. Annual average minimum and maximum temperatures are 10.7 and 19.8~ respectively (Cabezas and Escudero, 1989). Sample Collection. We examined exudate production in four tissue types: (1) mature leaves (nascent in the spring and exude high levels of labdanum during the summer), (2) sticky senescent leaves, characteristically darker with a brown-red color, (3) nascent summer leaves, active in labdanum secretion, and (4) upper part of photosynthetic stems, in which exudate is also evident. Samples were standardized to 17 g of each tissue type (Chaves, 1991), and were collected during a nearly two-year interval (May 1990 until December 1991). Sample Handlings. The extracts of labdanum exudate were prepared as follows: Using forceps, 0.7-g pieces of ethanol-soaked Whatman 118 filter paper were wiped over the surface of freshly cut leaves and stems until all visible traces of exudate were removed. Usually this operation was repeated several times, with special care taken to avoid putting pressure on the biological samples during handling. Filter paper samples were then transferred to glass flasks, and l0 ml of absolute ethanol was added to extract the exudate from the paper. Under mild stirring the ethanol solution soon turned yellow-green. The flasks were then carefully sealed to avoid losses by evaporation, and stored at 4~ until use. Sample Analysis. Absorption spectra were recorded with a diode array Hewlett Packard 8451A, from 250 to 700 nm. The samples were diluted in absolute ethanol to ensure that the maximum absorbance was less than 1.0 optical density at the peak wavelength in the spectrophotometer cuvette. HPLC chromatograms were carried out in a LCD Analytical HPLC instrument equipped with an integrator, and with a Nucleosil 5 C~8 column of 150 x 4 mm. In each case, 20 #1 of diluted ethanol extract was injected. Initially all samples were analyzed at wavelengths of 260, 280, 300, and 345 nm. After noting only minor differences between detection at 260 and 280 nm and that detection at 300 nm yields a mixed result between 280 and 345 nm, we decided to complete the study with the detection set at 260 and 345 nm. Chemicals. The following chemicals were used as standards: cinnamic acid, hydrocinnamic acid, benzyl alcohol, dimethyl- and diethyl-phthalate, apigenin (all from Aldrich Chemical Co.); apigenin-7-(O)methyl, kaempferol-3,7di(O)methyl, kaempferol-3,7,4'-tri(O)methyl (all kindly supplied by Dr. J. Gonzalez-Urones, University of Salamanca, Spain); other apigenins and kaempferols [kaempferol-3(O)methyl, kaempferol-3,7-di(O)methyl, kaempferol-3,4'di(O)methyl, apigenin-4'(O)methyl, apigenin-4',7-di(O)methyl and apigenin-7(O)methyl] were kindly supplied by Drs. Thomas Vogt (Botanisches Institut, Universitfit K61n, Germany) and E. Wollenweber (Institut ftir Botanik, Technische Hochschule Darmstadt, Germany). All other chemicals were purchased from Merck, Darmstadt, Germany, or Carlo Erba, Milano, Italy.
2580
CrtAVESET AL. RESULTS
Spectral Properties of Labdanum Exudate. The labdanum exudate is a sticky mixture that is clearly noticeable on leaves of C. ladanifer from the end of May until the end of October. Figure 1 shows the absorption spectra of samples taken in May, August, and October, with strong absorption in the nearUV to visible wavelength range up to 400 nm and peaks at 260, 310, and 350 nm. On the basis of previous studies (J. De Pascual et al., 1974; Proksch and Gfilz, 1984), several flavonoids are likely components of the extracts of labdanum exudate, namely, apigenin-7-(O)methyl, kaempferol-3,7-di(O)methyl and kaempferol-3,7,4'-tri(O)methyl. This is further supported by the comparison of the absorption and fluorescence spectra of the exudate (Figure 1) with that of purified flavonoids (Figure 2). The mild treatment used to prepare the samples of exudate does not extensively damage the epidermal plant cells. Leaves treated to extract the labdanum exudate in situ in plants, under low sunlight irradiance, do not become senescent and slowly (in several days) recover their sticky appearance, and their ability to produce exudate remains largely unimpaired. Moreover, epicuticular flavonoids have been shown to be predominantly aglycones, as opposed to intracellular (vacuolar) flavonoids, which are highly glycosylated (Vogt et al., 1987a, b; Proksch and Gfilz, 1984; Wollenweber and Dietz, 1981). Flavonoid glycosides have been shown to be well resolved and separated from aglycone flavonoids by paper chromatography using H20 as eluant, due to the much lower Rf values of the glycones (Markham, 1982). Chromatograms run under these
\
240
Wavelength (nm)
450
280 340
400 460 520 580 Wavelength (nm)
640
FIG. 1. Absorption and fluorescence emission spectra of the exudate of Cistus ladanifer. The spectra shown correspond to samples taken on May 29 (broken line), on August 30 (solid line), and on October (bold solid line). (A) Absorption spectra in methanol. (B) Fluorescence emission spectra recorded with an excitation wavelength of 320 nm in ethanol (a.u. stands for arbitrary unit).
FLAVINOID
SEASONAL
258 1
VARIATION
B v
