Volatile Organic Compounds from Orchis Species Found in Basilicata (Southern Italy)

This study is part of a project devoted to determining the scent of all the orchid species present in Basilicata. All the analyses were performed by using the solid-phase microextraction technique coupled with gas chromatography-mass spectrometry. The scent of eight species belonging to the Orchis genus was investigated. In the case of O. anthropophora, caryophyllene, tetradecanal and hexadecanal were the main components of the aroma; in O. purpurea, 3,5-dimethoxytoluene and elemicin were found; in O. italica, caryophyllene and 4-(3-hydroxy-2-methoxyphenyl)butan-2-one were found; in O. pauciflora, linalool and 1,4-dimethoxybenzene were found; in O. mascula, linalool was found; in O. quadripunctata, pentaand heptadecane were found; in O. provincialis, β-farnesene and farnesal were found; and in O. pallens, curcumene was the main product.


Introduction
Within the families of flowering plants, the Orchidaceae family is one of largest, with more than 28,000 species. The scent of some orchids has a relevant importance in perfume industries. However, the emission of volatile organic compounds from an orchid can have a relevant role in the life of the plant considering the possible effect of these compounds in attracting pollinators, or in defense against pathogens. One third of all the orchid species are food-deceptive species because the flowers do not contain nectar and the volatile organic compounds emitted mimic the floral signal of rewarding plants to attract pollinators. Furthermore, many compounds show antimicrobial and antifungal activities [1]. Volatile organic compounds are mainly terpenes, phenylpropanoid derivatives and fatty acid derivatives.
The above-reported data show that very different results can be obtained by using different GC-MS analytical methods able to characterize the aroma components, showing that the use of a homogenous method can provide valuable information on the scent of these species. In this work, the same HS-SPME-GC-MS method was used in order to characterize the scent of eight species of the Orchis genus.      To prevent plant damage to the whole plant from a population in Basilicata, a large portion of soil all around the plant was removed from its habitat and placed in a greenhouse for a few days of acclimatization.

Experimental Section
Following this period, for three days, the plant was placed under a bell jar. In view of the fact that the investigated taxa are rare wild plants, in order to preserve the species, we chose to use a single plant for our analysis.

Analysis of Volatile Organic Compounds
SPME [4] analysis of eight different samples of Orchis was performed. This way, the identified plants were collected and inserted in a glass jar for 24 h where a fiber (DVB/CAR/PDMS) and SPME syringe were also present. After this time, the fiber was desorbed in a gas chromatographic apparatus equipped with a quadrupole mass spectrometer detector. A 50/30 µm DVB/CAR/PDMS module with a 1 cm fiber (57328-U, Supelco, Milan, Italy) was employed to determine VOCs. The SPME fiber was maintained in the bell jar for 24 h. The analytes were desorbed in the splitless injector at 250 • C for 2 min. Analyses were accomplished with an HP 6890 Plus gas chromatograph equipped with a Phenomenex Zebron ZB-5 MS capillary column (30 m × 0.25 mm i.d. × 0.25 µm FT) (Agilent, Milan, Italy). An HP 5973 mass selective detector (Agilent) was utilized with helium at 0.8 mL/min as the carrier gas. The analyses were performed by using a splitless injector. The splitless injector was maintained at 250 • C, and the detector at 230 • C. The oven was held at 40 • C for 2 min, then gradually warmed, 8 • C/min, up to 250 • C and held for 10 min. Tentative identification of aroma components was based on mass spectra and Wiley 11 and NIST 14 library comparison. A single VOC peak was considered as identified when its experimental spectrum matched with a score over 90% present in the library. All the analyses were performed in triplicate.
To avoid contamination on the sample due, for example, to volatile organic compounds emitted from the soil, analysis of Orchis anthropophora was conducted on a single flower without the presence of soil, showing that this type of contamination does not exist. Otherwise, all the analyses were carried out by inserting the flowering plant in a glass bell jar and isolating the plant from the soil.

Discussion
It is interesting to note the large differences between our reported results and those reported in the Introduction section. For O. anthropophora, two different analyses are available [16,20]. While in the work of Cozzolino [16], only hydrocarbons with an extremely high molecular weight were found, the other article [20] reported that β-caryophyllene was the main component. In our study, β-caryophyllene was present, but the main component was tetradecanal. In the case of O. purpurea, no other results on the composition of the scent are available. For O. italica, only an article published by Cozzolino is available [16], and, also in this case, only high-molecular weight hydrocarbons were found. In our experiment, on the contrary, β-caryophyllene was the main component of the scent.
The scent of O. pauciflora has been determined through headspace analysis, showing the presence of 2-methyl-6-methylene-3,7-octadiene-2-ol as the main component [15]. However, in our analysis, linalool and 1,4-dimethoxybenzene were found as the main components. O. mascula was the object of an intense study where several different analytical methods were used. This way, headspace analysis returned E-ocimene as the main component of the scent [13,14]. This result was confirmed by SPME analysis [15,17]. In our analysis, as it is evident considering Figure 4, linalool was the main component of the aroma. For O. quadripunctata, the work of Schiestl and Cozzolino found only hydrocarbons [16]. Only hydrocarbons were found in this work, but with a significant difference in the molecular weight of the detected compounds. In the case of O. provincialis, the work of Schiestl and Cozzolino determined only the presence of hydrocarbons [17], while the presence of relevant amounts of β-farnesene was determined in this study. Finally, while an SPME analysis of the scent of O. pallens found phenethyl alcohol, farnesene and farnesol [21], our analysis of the same species found β-curcumene as the main component.

Conclusions
This work shows the analysis of Orchis samples from Basilicata. The analyses were performed by using the same procedure and the same fiber in SPME-GC-MS, which allowed achieving a homogenous dataset. The analyses showed different scent compositions from those determined on samples deriving from different sites. These observed differences, when SPME of other headspace techniques is used, can depend both on the different absorption rates of the analytes on the fiber and on the variation in the scent due to natural adaptation of the plant to different environmental conditions, due, for example, to different pollination insects. A completely different consideration can be found in the work of Schiestl and Cozzolino, where a completely different analytical method was used (hexane extraction of labellum, and GC-MS analysis of the extracts). In their case, only hydrocarbons were determined. Probably, their analytical procedure was not the correct method for the determination of the orchid scent.