Analytica Chimica Acta 589 (2007) 120 124 Volatiles emission patterns of different plant organs and pollen of Citrus limon Guido Flamini, Marianna Tebano, Pier Luigi Cioni Dipartimento di Chimica Bioorganica e Biofarmacia, Via Bonanno 33, 56126 Pisa, Italy Received 24 November 2006; received in revised form 21 February 2007; accepted 21 February 2007 Available online 25 February 2007 In memory of Prof. Ivano Morelli (1940 2005). Abstract The volatiles emitted in vivo by different plant parts of Citrus limon (Rutaceae) have been identified by mean of head space-solid phase micro extraction (HS-SPME) and gas chromatography coupled with mass spectrometry (GC MS) analyses. In particular, the profiles of flower buds, mature flowers, petals, stamens, gynaecium, pericarp of unripe and ripe fruits, young and adult leaves and pollen have been examined. Furthermore, the essential oil obtained from expression of ripe pericarp was studied. Volatiles were produced in distinctive amounts by the different plant organs, creating an interesting contrast, particularly within the flower parts: the highest amount of limonene (62.5%) was emitted by gynaecium, followed by stamens (22.9%) and petals (3.1%). Pollen did not produce limonene at all. The same compound is contained in higher amounts in the young leaves than in old ones (65.3% versus 30.1%). A possible defensive role of limonene and other volatiles, mainly terpene aldehydes, produced by young leaves has been hypothesized. 2007 Elsevier B.V. All rights reserved. Keywords: Citrus limon; Rutaceae; Volatiles; Head space-solid phase micro extraction; Plant organs; Pollen 1. Introduction It is known that the color of a flower is the first and foremost cue for pollinator s attraction, but the scent of a flower also plays a major role in attracting pollinating insects [1,2]. Odors can act both at long distances as attraction cues and at short distances as orientation cues among different parts of the flower or among different flowers [1,3 5]. The use of headspace technique to sample volatiles in the air surrounding a whole plant or plant organ has permitted one to ascertain that olfactory floral stimuli differ not only between species, but also between different organs within a single flower [3,5 9]. Distinctive volatile compounds could allow insects both to recognize specific host plants and to assess the amount of rewards in a flower, since pollen, an important food source for many flower-feeding insects, also produces odors. In fact it is well documented that pollen of many species has distinctive fragrances as evaluated by the human Corresponding author. Tel.: +39 050 2219686; fax: +39 050 2219660. E-mail address: flamini@farm.unipi.it (G. Flamini). nose [10 12]. This situation has been confirmed by instrumental analyses (GC and GC MS) of the volatiles sampled from the air surrounding the pollen [3,6,7,13,14]. Probably pollen odors evolved as defense compounds against pollen-feeding animals, but when plants became dependent on animals for pollination, some attractive compounds have been included among pollen volatiles [1]. Actually plants must face two simultaneous contrasting pressures: the need to protect their pollen from non-pollinating insects and the need to advertise it as reward to pollinators. Studies of floral scents and of their patterns within a single flower are important to better understand the chemical bases of plant-animal relationships and pollination ecology. Furthermore, they may reveal new scented molecules that could be of value to both the food industry and perfumery. In the present paper the profiles of the volatiles emitted in vivo from whole fresh flowers (collected at different development stages), different flower parts and pollen of lemon (Citrus limon Ten., Rutaceae) have been analyzed by mean of solid phase micro-extraction (SPME) coupled with gas chromatography mass spectrometry (GC MS). Furthermore, 0003-2670/$ see front matter 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2007.02.053
G. Flamini et al. / Analytica Chimica Acta 589 (2007) 120 124 121 the volatiles emitted by the pericarp of ripe and unripe fruits, young and adult leaves and the essential oil obtained by expression of the ripe fruit pericarp have been studied. Finally, to evaluate whether the identified volatiles could be useful to discriminate among diverse volatile emission patterns of different organs, all the compounds were submitted to hierarchical cluster analysis (HCA). 2. Materials and methods The plant material has been obtained from plants grown in the Botanical Garden of the University of Pisa. Twelve different samples were prepared: Sample 1: flower buds; Sample 2: mature flowers; Sample 3: petals; Sample 4: stamens; Sample 5: pollen; Sample 6: gynaeceum Sample 7: pericarp of unripe fruits; Sample 8: pericarp of ripe fruits; Sample 9: young leaves; Sample 10: adult leaves; Sample 11: essential oil from expression of ripe pericarp. Samples 1 and 2 were prepared using five organs collected at the proper vegetative stage, cut few millimeters below the calyx and the ends were wrapped in aluminum foil to minimize water loss. They were introduced in a 10 ml septum-cap vial and allowed to equilibrate for 20 min at 25 C before sampling. Samples 3 5 were prepared using 10 20 pieces obtained from freshly opened flowers, avoiding contamination from other flower parts. They were introduced in a 4 ml septum-cap vial and allowed to equilibrate for 20 min at 25 C before sampling. Samples 6 and 7 were obtained from pieces of pericarp cut from ripe or unripe fruits and allowed to dry the margins before sampling. Than they were introduced in a 4 ml septum-cap vial and allowed to equilibrate as above. Samples 8 and 9 were made of five leaves collected at the proper vegetative stage, cut few millimeters below the calyx and the ends were wrapped in aluminum foil to minimize water loss. They were introduced in a 50 ml septum-cap vial and allowed to equilibrate as already described. Sample 10 consisted of 3 5 mg of pollen obtained by gentle tapping from flowers after anther dehiscence. It was allowed to equilibrate as described above. All the samples were sampled by means of the solid phase micro extraction technique (see below). The GC analyses were accomplished with a HP-5890 Series II instrument equipped with HP-WAX and HP-5 capillary columns (30 m 0.25 mm, 0.25 m film thickness), working with the following temperature program: 60 C for 10 min, ramp of 5 C min 1 up to 220 C; injector and detector temperatures 250 C; carrier gas nitrogen (2 ml min 1 ); detector dual FID; split ratio 1:30; injection of 0.5 l. The identification of the components was performed, for both the columns, by comparison of their retention times with those of pure authentic samples and by mean of their linear retention indices (l.r.i.) relative to the series of n-hydrocarbons. GC/EIMS analyses were performed with a Varian CP-3800 gas-chromatograph equipped with a DB-5 capillary column (30 m 0.25 mm; coating thickness 0.25 m) and a Varian Saturn 2000 ion trap mass detector. Analytical conditions: injector and transfer line temperatures 220 and 240 C, respectively; oven temperature programmed from 60 C to 240 C at 3 C min 1 ; carrier gas helium at 1 ml min 1 ; injection of 0.2 l (10% hexane solution); split ratio 1:30. Identification of the constituents was based on comparison of the retention times with those of authentic samples, comparing their linear retention indices relative to the series of n-hydrocarbons, and on computer matching against commercial (NIST 98 and ADAMS) and home-made library mass spectra built up from pure substances and components of known oils and MS literature data [15 20]. Moreover, the molecular weights of all the identified substances were confirmed by GC/CIMS, using MeOH as CI ionizing gas. SPME analyses were performed using a Supelco SPME device coated with polydimethylsiloxane (PDMS, 100 m). The fiber was pre-conditioned according to the manufacturer instructions. After the equilibration time, the fiber was exposed to the headspace for 20 min at room temperature. Once sampling was finished, the fiber was withdrawn into the needle and transferred to the injection port of the GC or GC/MS system, operating in the same conditions as above both for quantification and identification of the constituents, except that the splitless injection mode was used and the injector temperature was 250 C. All the analyses were performed at least in quadruplicate. The results were expressed as mean percentages obtained by FID peak-area normalization. In the absence of information about shape of the groupings and/or statistical distribution of values, Euclidean distances combined with unweighted pair group average linking were used for cluster analysis [21]. 3. Discussion The floral scent sampling techniques employed in previous studies have the drawbacks of requiring considerable amounts of flower material, especially when sampling volatiles from pollen (50 200 mg), and very long sampling times (24 48 h for pollen, or 8 27 h for flower parts), as well as posing possible risks of sample contamination and loss of volatiles during the concentration of the solvent volatile mixture in a water-bath in preparation for analysis by GC [3,6,7]. We have recently demonstrated that it is possible to improve the sampling method by application of the solid phase micro-extraction technique [22]. The results of all the samplings are reported in Table 1. Altogether, 65 compounds were identified, accounting from 87.7% to 99.8% of the whole volatiles. All the volatiles were mono- and sesquiterpenes, both hydrocarbons and oxygenated derivatives, together with some non-terpene compounds such as C 9 C 14 straight-chain aldehydes, and C 13 C 17 saturated and unsaturated aliphatic hydrocarbons.
122 G. Flamini et al. / Analytica Chimica Acta 589 (2007) 120 124 Table 1 In vivo volatile emission of different parts of Citrus limon Constituents a l.r.i. b Flower buds (1) Mature flowers (2) Petals (3) Stamens (4) Pollen (5) Gynaec. (6) Unripe peric. (7) Ripe peric. (8) Young leaves (9) Adult leaves (10) Express. ess. oil (11) -Thujene 932 c 0.2 0.2 0.2 0.4 -Pinene 942 2.4 0.5 tr d 0.5 0.9 1.0 1.6 1.0 0.6 1.9 Camphene 955 0.1 Sabinene 978 3.4 1.6 1.4 1.5 1.5 1.8 2.4 1.8 2.4 -Pinene 982 24.0 7.8 0.9 5.8 7.9 9.2 12.0 6.3 6.2 13.7 Myrcene 992 0.4 1.4 0.7 1.4 2.2 1.5 2.7 1.6 octanal 1002 tr (E)-3-Hexen-1-ol 1004 1.4 acetate -Phellandrene 1008 tr 3-Carene 1013 0.1 2.1 0.2 -Terpinene 1020 0.1 0.2 0.3 0.3 0.3 0.3 p-cymene 1028 0.2 0.4 0.4 0.1 0.2 0.1 Limonene 1033 38.9 44.3 3.1 22.9 62.5 65.3 68.3 65.3 30.1 62.5 1,8-Cineole 1035 6.7 10.2 23.1 1.7 tr (E)- -Ocimene 1051 7.0 5.4 3.9 3.1 3.1 0.1 0.1 3.0 5.4 tr -Terpinene 1064 6.0 6.7 0.7 4.5 9.2 11.5 11.4 0.3 0.3 11.6 cis-sabinene 1070 0.2 0.1 hydrate Terpinolene 1088 0.4 0.5 0.3 0.8 0.7 0.6 0.5 0.6 Linalool 1101 0.2 0.2 2.7 0.2 1.5 0.2 0.5 tr trans-sabinene 1103 0.2 0.3 hydrate Nonanal 1106 6.7 1.0 tr allo-ocimene 1131 tr 0.2 0.1 Limonene oxide 1136 0.2 Citronellal 1155 0.2 0.2 4-Terpineol 1179 0.1 -Terpineol 1192 tr tr 0.5 0.2 0.2 0.2 tr Decanal 1202 tr tr 1.5 tr Nerol 1228 0.1 1.3 0.2 0.1 1.3 Neral 1243 tr 0.2 tr 6.4 0.8 1.0 Geraniol 1256 0.4 0.2 0.1 0.7 Geranial 1272 0.2 tr 5.5 1.3 1.4 Indole 1289 4.8 2.5 tr 1.0 Tridecane 1300 0.1 0.2 Methyl 1338 0.7 8.4 0.2 anthranylate -Elemene 1340 0.3 0.2 0.1 0.8 Citronellyl acetate 1356 tr Eugenol 1358 tr Neryl acetate 1365 0.1 0.3 0.2 0.1 0.6 Geranyl acetate 1385 0.5 0.2 0.6 1-Tetradecene 1393 0.2 Tetradecane 1400 0.2 0.2 tr Italicene 1404 0.4 Dodecanal 1411 1.2 cis- - 1415 tr 0.1 0.1 Bergamotene -Caryophyllene 1419 11.3 9.5 2.5 14.5 3.7 1.1 0.1 2.3 25.1 0.3 trans- - 1436 1.2 1.4 0.2 3.2 0.7 0.2 2.8 0.2 Bergamotene (E)-Geranyl 1453 0.2 1.9 9.0 1.9 acetone -Humulene 1456 0.6 0.6 0.2 0.3 0.1 1.0 (E)- -Farnesene 1458 0.1 0.2 -Santalene 1463 0.1 (Z,E)- -Farnesene 1491 0.5 0.2 0.2 Valencene 1493 tr 0.1 Bicyclogermacrene 1496 1.3 0.9 0.2 4.4 0.5 0.1 3.5 tr Pentadecane 1500 0.9 4.2 2.0 2.2 (E,E)- -Farnesene 1507 0.3 9.1 2.9 tr 4.4 -Bisabolene 1509 1.6 1.1 0.2 5.2 1.0 0.4 0.1 1.8 0.3
G. Flamini et al. / Analytica Chimica Acta 589 (2007) 120 124 123 Table 1 (Continued ) Constituents a l.r.i. b Flower buds (1) Mature flowers (2) Petals (3) Stamens (4) Pollen (5) Gynaec. (6) Unripe peric. (7) Ripe peric. (8) Young leaves (9) Adult leaves (10) Express. ess. oil (11) trans-nerolidol 1565 4.5 27.3 14.2 30.7 tr Caryophyllene 1583 tr tr oxide Germacrene 1578 tr tr tr 0.7 D-4-ol Hexadecane 1600 1.8 Isolongifolan-7- ol 1621 13.6 6.9 10.0 1-Heptadecene 1673 0.3 (Z,E)- -Farnesol 1697 0.2 Heptadecane 1700 0.1 0.2 0.2 (E,E)- -Farnesol 1792 0.3 Total identified 99.0 96.5 93.6 95.2 87.7 99.5 97.8 98.9 99.5 97.0 99.8 a Percentages (by weight) obtained by FID peak-area normalization. b Linear retention indices (HP-5 column). c Not detected. d tr < 0.1%. Isolated petals (sample 3) were characterized by a low limonene content (3.1%) and oxygenated mono- and sesquiterpenes accounted for more than 60% of the total volatiles. In stamens (sample 4), where limonene reached 22.9%, oxygenated derivatives represented 46.7% of the whole volatiles. On the contrary, adult leaves (sample 10) contained about 30% of limonene and high amounts (38.6%) of sesquiterpene hydrocarbons, mainly -caryophyllene (25.1%), while oxygenated compounds just reached 10%. In the other samples, gynaecium, unripe and ripe pericarp and young leaves (samples 6 9), where limonene was the main volatile (more than 62%), oxygenated derivatives were always detected in low amounts, 4.4%, 1.0%, 3.6% and 13.1%, respectively. In pollen (sample 5), limonene was absent and the main volatiles were sesquiterpenes; in this case, oxygenated derivatives constituted more than 55% of the spontaneous emission of volatiles. In many samples limonene was one of the main volatiles detected: in gynaecium, epicarp and young leaves it reached 62.5%, 65.3%, 68.3% and 65.3%, respectively. When leaves become older, their limonene content halves (30.1%). In flower buds, it was detected at 38.9% levels. This percentage increased to 44.3% with flower opening, but it seems that its emission was mainly due to stamens and pistil (22.9% and 62.5%, respectively), since petals emitted very low amounts of this monoterpene (3.1%). A similar trend can be noted for -pinene, sabinene and -pinene: their concentration decreased in the flowers with respect to the buds. On the contrary, 1,8-cineole and trans-nerolidol were not detected at all in the flower buds but formed significant parts of the natural emission of volatiles of the developed flowers (4.5% and 6.7%, respectively). Apart from petals, the main 1,8-cineole sources were stamens, whereas in the case of trans-nerolidol, this alcohol was also produced in high percentages by the pollen. Other flower volatiles did not undergo significant variations. Other interesting differences can be noted between young and old leaves. Apart from the above-cited halved limonene content of the adult leaves (65.3% versus 30.1%), other compounds were emitted in significant higher amounts by the young leaves, such as the isomeric aldehydes neral and geranial (6.4% and 5.5% versus 0.8% and 1.3%, respectively). The opposite trend was also noted: -caryophyllene emission passed from 2.3% of the young leaves to 25.1% of the older ones. (E,E)- -farnesene was not produced at all by young leaves, whereas it appeared in the volatile profile of the older ones (4.4%). Pollen was characterized by a very different emission profile with respect to all the other flower parts. The main volatile detected in pollen was trans-nerolidol (30.7%), followed by -caryophyllene, (E)-geranyl acetone (9.0%), isolongifolan-7- -ol, and nonanal (6.7%). This latter aldehyde was exclusive of pollen emission, and probably it represents one of those chemicals produced as defense compounds against pollen-feeding animals [1]. During fruit ripening, the volatiles of the epicarp did not undergo significant variations in their emission profile. Some minor volatiles disappeared from the ripe pericarp, such as linalool, sabinene hydrates, italicene, valencene and bisabolene. The expressed essential oil obtained from the ripe pericarp, showed a composition very similar to that of the volatiles emitted by this plant part. The main differences consisted in lesser limonene amount (62.5% versus 68.3%) and minor changes in the percentages of the aldehydes citronellal, neral and geranial and of the acetates of nerol and geraniol. Hierarchical cluster analysis (HCA) is a method in which samples are considered as lying in an n-dimensional space and distances between samples are calculated joining the objects with an agglomerative procedure [23,24]. The HCA (Fig. 1) confirmed the above observations. In particular, it confirmed that the emissions from the pericarp were very similar in ripe and unripe fruits and that the essential oil obtained by expression from this plant part had a comparable composition. This could indicate that the emitted volatiles originated from the glandular structures of the pericarp.
124 G. Flamini et al. / Analytica Chimica Acta 589 (2007) 120 124 References Fig. 1. Hierarchical cluster analysis of the different samples of lemon. Another similarity was found between stamens/pollen and petals. Perhaps pollen can contaminate the petals of mature flowers after anther dehiscence. It is interesting to note the very different emission pattern of male and female parts that could indicate an orientation cue for pollinators adopted by the flower. A further difference was confirmed between old and young leaves. Perhaps this behavior can be related to the production of defense metabolites by the more vulnerable young leaf. Limonene and monoterpene aldehydes (i.e. neral and geranial), detected in considerable higher concentration in young leaves, are in fact well known repellents against phytophagous [25 29]. 4. Conclusions The results evidenced the presence of hydrocarbon monoterpenes in the volatiles of lemon, and in particular the high content of limonene that characterize most of the plant parts of this species. This study confirms that the SPME technique permits a detailed analysis of the volatiles emitted from various plant parts and establish their different composition with respect to the essential oil contained in the glandular structures of the pericarp. These results can be useful to better understand pollination chemistry and other animal plant relationships. [1] H.E.M. Dobson, G. Bergström, Pl. Syst. Evol. 222 (2000) 63. [2] H.E.M. Dobson, in: E.A. Bernays (Ed.), Insect Plant Interactions, CRC Press, Boca Raton, 1994, p. 47. [3] H.E.M. Dobson, G. Bergström, I. Groth, Israel J. Bot. 39 (1990) 143. [4] N.H. Williams, in: C.E. Jones, R.J. Little (Eds.), Handbook of Experimental Pollination Biology, Academic Press, New York, 1983, p. 50. [5] J.T. Knudsen, L. Tollsten, Pl. Syst. Evol. 177 (1991) 81. [6] H.E.M. Dobson, J. Bergström, G. Bergström, I. Groth, Phytochemistry 12 (1987) 3171. [7] G. Bergström, H.E.M. Dobson, I. Groth, Pl. Syst. Evol. 195 (1995) 221. [8] H.S. McTavish, N.W. Davies, R.C. Menary, Ann. Bot. 86 (2000) 347. [9] L.N. Fernando, I.U. Grün, Flavour Fragr. J. 16 (2001) 289. [10] K. von Frisch, Zool. Jahrb. Abt. Allg. Zool. Physiol. 40 (1923) 1. [11] A. von Aufsess, Z. Vergl. Physiol. 43 (1960) 469. [12] S.L. Buchmann, in: C.E. Jones, R.J. Little (Eds.), Handbook of Experimental Pollination Biology, Academic Press, New York, 1983, p. 73. [13] H.E.M. Dobson, G. Bergström, I. Groth, Am. J. Bot. 83 (1996) 877. [14] G. Flamini, P.L. Cioni, I. Morelli, J. Agric. Food Chem. 51 (2003) 2267. [15] E. Stenhagen, S. Abrahamsson, F.W. McLafferty, Registry of Mass Spectral Data, J. Wiley & Sons, New York, 1974. [16] Y. Massada, Analysis of Essential Oils by Gas Chromatography and Mass Spectrometry, J. Wiley & Sons, New York, 1976. [17] W. Jennings, T. Shibamoto, Qualitative Analysis of Flavor and Fragrance Volatiles by Glass Capillary Chromatography, Academic Press, New York, 1980. [18] N.W. Davies, J. Chromatogr. 503 (1990) 1. [19] R.P. Adams, Identification of essential oil components by gas chromatography/mass spectroscopy, Allured, Carol Stream, 1995. [20] A.A. Swigar, R.M. Silverstein, Monoterpenes, Aldrich Chem. Comp., Milwaukee, 1981. [21] P.J. Dunlop, C.M. Bignell, D.B. Hibbert, Aust. J. Bot. 45 (1997) 1. [22] G. Flamini, P.L. Cioni, I. Morelli, J. Chromatogr. A 998 (2003) 229. [23] J. Davis, Statistics and Data Analysis in Geology, Wiley, New York, 1986. [24] B.S. Everitt, Cluster Analysis, Heineman, London, 1980. [25] L. Li, Y. Zhao, B.C. McCaig, B.A. Wingerd, J. Wang, M.E. Whalon, E. Pichersky, G.A. Howe, Plant Cell 16 (2004) 126. [26] K. Ament, R. Merjin, M.W. Sabelis, M.A. Haring, R.C. Schuurink, Plant Physiol. 135 (2004) 2025. [27] B. Bojana, E. Knez, V. Skerlavaj, Microsph. Microcaps. Lipos. 6 (2003) 271. [28] A.O. Oyedele, A.A. Gbolade, M.B. Sosan, F.B. Adewoyin, O.L. Soyelu, O.O. Orafidiya, Phytomedicine 9 (2002) 259. [29] P.H. Vartak, V.B. Tungikar, R.N. Sharma, J. Commun. Dis. 26 (1994) 156.