Introduction

Thymus vulgaris L., Rosmarinus officinalis L. (Lamiaceae) and Ruta chalepensis L. (Rutaceae) are important medicinal plants, highly recommended due to a range of biological and therapeutic properties including anti-carcinogenic, anti-inflammatory, antimicrobial, antioxidant, antiseptic, hepatoprotective, anti-HIV, diuretic and expectorant (Dorman and Deans 2000; Ruberto and Baratta 2000). The detailed chemical composition of these medicinal plants is very complex. The most important classes of its volatile compounds are characterized, namely, by monoterpenoids, sesquiterpenoids and their oxygenated derivatives, carbonyl compounds, ketones, aldehydes and higher alcohols (Cao et al. 2006; Guedes de Pinho et al. 2009). The content of volatile metabolites may vary as a function of the origin of samples. In addition to volatile metabolites, another important class of medicinal plant constituents is represented by secondary metabolites such as polyphenols, including flavonoids, phenolic acids and their esters, terpenes alkaloids and others. Their chemical complexity may explain their wide range of biological and pharmacological activities.

The biosynthetic pathways of important plant volatile metabolites have been traced back up to intermediates of primary metabolism. It has been shown that carbohydrates, fatty acids and amino acids, represent the natural carbon pool for volatile metabolites, which can also be released from their polymers (Fig. 1 ). In plants two biosynthetic pathways are responsible for synthesis of isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP), universal precursors of all terpenoids, which are localized in different subcellular compartments. While DMAPP, formed in plastids, is used by isoprene synthase (ISPS) to form isoprene in some plants, IPP and DMAPP precursors are further condensed by prenyl diphosphate synthases in the respective compartments to form prenyl diphosphate intermediates serving as substrates for a large group of terpene synthases (TPS) enzymes, resulting in the final terpenoid compounds (Nagegowda 2010).

Fig. 1
figure 1

Schematic of biosynthetic pathways leading to the formation of plant volatile metabolites

Full size image

For thymol, the pathway formation proceeds from γ-terpinene via the aromatic compound, p-cymene, as an intermediate (Fig. 2 ). γ-Terpinene is most likely converted to thymol by the action of one or more cytochrome P450 oxidases, catalyzing a hydroxylation (Crocoll et al. 2010). α-Terpinyl cation is a key intermediate to form several terpenoids, such as thymol, eucalyptol, limonene, α-pinene, sabinene, among others.

Fig. 2
figure 2

Eucalyptol and thymol biosynthesis pathways

Full size image

Several researches reported thymol followed by carvacrol, eucalyptol, borneol and α-pinene, as the main volatiles of T. vulgaris (thyme, Lamiaceae) volatile composition. It has been reported that it possesses numerous biological activities including breaking up congestion, calming bronchial spasms, stimulating respiration, antimicrobial and antioxidant effects (Baranauskienė et al. 2003; Grigore et al. 2010), whereas R. officinalis (rosemary, Lamiaceae) is widely used in folk medicine in the treatments of variety of diseases such as external stimulant, relaxant for nervousness and muscle spasms (Barnes et al. 2007). Some volatile constituents, namely eucalyptol, camphor, borneol, bornyl acetate and α-pinene, could be responsible for these pharmacological properties. R. chalepensis (ruta, Rutaceae), is characterized mainly by aliphatic ketones being 2-decanone, 2-undecanone and 2-dodecanone the most dominant volatile constituents. It has been reported that ruta possesses numerous biological activities including emmenagogue, abortifacient, anti-helmintic, and spasmolytic (Mejri et al. 2010). The biosynthesis of methyl ketones has been hypothesized to derive from a variety of biological pathways such as fatty acid β-oxidation or aerobic alkene/alkane degradation. In plants, 3-keto fatty acids could themselves be derived from the hydrolysis of either 3-ketoacyl-acyl carrier proteins, intermediates in the fatty acid biosynthetic pathway of chloroplasts, or could be derived from 3-ketoacyl-CoA, which are intermediates in the degradation of fatty acids in the peroxisomes (Fig. 3 ) (Yu et al. 2010). Decarboxylation of 3-keto fatty acids could thus give rise to straight-chain methyl ketones such as those found in the ruta plants.

Fig. 3
figure 3

Schematic biosynthesis of straight-chain methylketones

Full size image

Several extraction techniques, including hydrodistillation (Sourmaghi et al. 2014), microwave (Sourmaghi et al. 2014), supercritical fluid extraction (Venkatachallam et al. 2010), soxhlet extraction (Bajerová et al. 2008), and ultrasound-assisted extraction (Ince et al. 2014) have been reported adequate to extract volatile metabolites from natural products. However, it is well-known that these techniques presents some shortcomings such as long extraction time, are laborious procedures needs large volume of organic solvents and amounts of sample and present low extraction efficiency. To overcome these drawbacks, solid-phase microextraction (SPME), emerged as an attractive and simple extraction technique which offers important advantages over conventional solvent extraction procedures as it is rapid, easy-to-use, solvent free, sensitive and does not require any concentration step prior to analysis, preventing artifacts. Combined to gas chromatography (GC) and mass spectrometry analysis (MS) SPME has been successfully applied in a wide range of fields (Cao et al. 2006; Gianelli et al. 2002; Guedes de Pinho et al. 2009; Vázquez-Araújo et al. 2013).

Many of medicinal plants have been investigated through of the world due to their promising health benefits namely their potent antioxidant activity protecting humans from oxidative damage. As synthetic antioxidants have been reported to cause liver damage and carcinogenesis, the interest in using naturally derived antioxidants is strongly recommended avoiding the deleterious effects of synthetic antioxidants (Gouveia and Castilho 2013; Halvorsen et al. 2002). In this sense, several in vitro assays are available to measure the antioxidant activity of different medicinal plant extracts and authentic compounds. Each of these assays is based on one feature of the antioxidant activity, such as the ability to scavenge free radicals and inhibition of lipid peroxidation (LP) (Choi et al. 2002; Cvetković et al. 2015; Katalinic et al. 2006; Lin et al. 2009). The antioxidant activity of medicinal plants is a result of complex interaction between its bioactive components, such as terpenoids, phenolic compounds, flavonoids, among others (Halvorsen et al. 2002).

Aqueous herbal extracts have attracted attention since they can be consumed in a daily basis as decoctions. Many Lamiaceae and Rutaceae extracts are extensively used in traditional diet and popular medicine. Therefore, in the present work, the establishment of the volatile pattern of medicinal plants, T. vulgaris, R. officinalis and R. chalepensis, growing at Madeira Island (Portugal) was carried out combining the high throughput ability of HS-SPME technique with the high resolution and identification potential of GC–qMS. In order to improve the SPME performance, based on number of isolated metabolites, total peak area of identified compounds and % of RSD, the conducted research involved a previous optimization step of the most important extraction-influencing parameters, namely fiber coating, extraction time and extraction temperature. In addition, the total phenolic content (TPC) and the antioxidant activity of the target medicinal plants aqueous extracts were also evaluated and compared using previously developed assays based on Folin-Ciocalteu’s assay and on scavenging activity of the stable free radical DPPH● and thiobarbituric acid reactive substances (TBARS) assay, respectively.

Material and methods

Reagents and materials

All chemicals were of analytical quality. Sodium chloride (99.5 %) was supplied from Panreac (Spain, Barcelona). The authentic standards of volatile compounds (purity higher than 99 %) used for the identification of target compounds were purchased from Sigma-Aldrich (Madrid, Spain) and Fluka (Buchs, Switzerland). The individual stock solutions were prepared in ethanol at concentration of 1000 mg/L, and stored at 4 °C. The 2,2-diphenyl-1-picrylhydrazyl (≥ 85 %) was purchased from Fluka (Buchs, Switzerland) and the ultra-pure water from a Milli-Q system (Millipore, Bedford, USA). Helium of purity 5.0 (Air Liquid, Portugal) was used as the GC carrier gas. The glass vials, SPME fibers and SPME holder for manual sampling were purchased from Supelco (Bellenfonte, PA, USA). The retention index (RI) was calculated through the injection of a series of C8 to C20 straight-chain n-alkanes (concentration of 40 mg/L) in n-hexane supplied by Fluka (Buchs, Switzerland).

Plant material

Fresh leaves, ca. 200 g, of T. vulgaris, R. officinalis (Lamiaceae) and R. chalepensis were collected randomly on September 2011, in Caniço [32°65′N, 16°85′W], Madeira Island. Medicinal plant extracts were prepared based on a standard protocol (Katalinic et al. 2006), with some modifications. 5 g of plant material were added with 200 mL of deionised water at 98 °C. The infusions were left at room temperature during 30 min. The medicinal plant extracts were filtered and their volatile profile and antioxidant activity was evaluated.

Optimization of SPME influencing parameters

Several experimental parameters affecting the SPME efficiency performance, such as fiber coating, extraction time and extraction temperature, were optimized using R. chalepensis aqueous extract.

A preliminary screening of six coating fibers with different polarities and retention capabilities, carboxen/poly(dimethylsiloxane) (CAR/PDMS, 75 μm), carbowax/divinylbenzene (CW/DVB, 65 μm), poly(acrylate) (PA, 85 μm), poly(dimethylsiloxane)/divinylbenzene (PDMS/DVB, 65 μm) and poly(dimethylsiloxane) (PDMS, 100 μm) and divinylbenzene/carboxen/poly(dimethylsiloxane) (DVB/CAR/PDMS, 50/30 μm), were considered to select the best coating in terms of extraction efficiency, number of isolated metabolites and repeatability. All fibers were thermally conditioned, as recommended by the manufacturer. Before the first daily extraction, the fibers were conditioned during 10 min at 240 °C. A blank test was performed to check possible carry-over. At least three replicates were performed for each fiber.

The extraction time and temperature are important parameters affecting the volatility and solubility of analytes and consequently the SPME performance due to their strong influence on vapor pressure and on equilibrium of volatile compounds in the HS of the sample. The temperature also influences the distribution coefficient between the sample and the HS and between the HS and the fiber. Therefore, these parameters were tested and optimized. The HS-SPME of the R. chalepensis aqueous extract was carried out at different exposition times, between 15 and 60 min using DVB/CAR/PDMS fiber at 40 °C under constant stirring (800 rpm). Secondly, in order to optimize the extraction temperature, 25, 40 and 60 °C using DVB/CAR/PDMS fiber for 45 min under constant stirring (800 rpm) were tested.

HS-SPME procedure

Aliquots of 10 mL of R. chalepensis aqueous extract were placed into a 20 mL glass vial, which corresponds to the ratio volume of the liquid phase and headspace volume (1/β) of 0.5. After, the addition of 2 g of sodium chloride (NaCl) and a stirring bar (2 mm × 0.5 mm, stirring at 800 rpm), the vial was closed, and placed in a thermostated bath adjusted to 40.0 ± 0.1 °C. The SPME fiber was manually inserted into the sample vial headspace for 45 min. After sampling, the SPME fiber was retracted and immediately inserted into the GC injection port for thermal desorption of volatile metabolites. A desorption time of 6 min at 240 °C was used in splitless mode. Before sampling each fiber was reconditioned for 5 min at 240 °C. All measurements were performed, at least, three replicates.

GC-qMS conditions

The desorbed volatiles were separated in an Agilent Technologies 6890 N Network gas chromatograph equipped with a BP-20 (polyethylene glycol) fused silica capillary column (30 m × 0.25 mm I.D.; film thickness, 0.25 μm) from Scientific Glass Engineering (Darmstadt, Germany) connected to an Agilent 5973 N quadrupole mass selective detector (Palo Alto, CA, USA). Helium (Air Liquid, Portugal) was used as the carrier gas at a flow rate of 1.1 mL/min (column-head pressure: 89,632 Pa). The injections were performed in splitless mode (6 min). The GC oven temperature was maintained at 50 °C for 1 min, then raised at 2.5 °C/min to 100 °C (2 min), then to 180 °C (1 min) at 2 °C/min, and finally increased to 220 °C at a rate of 15 °C/min and held isothermally for further 10 min. For the MS system, the temperatures of the transfer lines, quadrupole and ionization source were 250, 150 and 230 °C, respectively; electron impact mass spectra were recorded at 70 eV and the ionization current was about 30 μA. The acquisitions were performed in full scan mode (30–300 m/z).

Volatile metabolites identification was based on (i) comparison of the GC retention times and mass spectra, with those, when available, of the authentic standard, (ii) comparison of all mass spectra with the data system library (NIST, 2005 software, Mass spectral Search Program V.2.od; NIST 2005, Washington, DC), and (iii) Kovats retention index (RI) value determined according to the van Den Dool (1963). For the determination of the RI, a C8-C20 n-alkanes series was used, and the values were compared, when available with values reported in the literature for similar chromatographic columns (Culleré et al. 2004; Galindo-Cuspinera et al. 2002; Perestrelo et al. 2011; Pontes et al. 2009).

Determination of TPC

TPC of the aqueous extracts of medicinal plants were determined by the Folin-Ciocalteu’s (FC) colorimetric method as described by Lim et al. (2007). Briefly, 0.6 mL of each extract and 3 mL of Folin-Ciocalteu reagent (previously diluted 10-fold with deionised water) was added and mixed thoroughly. Then, 2.4 mL of 7.5 % (w/v) of sodium carbonate was added to the mixture and mixed gently. The reaction mixture was kept in dark during 30 min, and its absorbance measured at 725 nm against water in a Perkin Elmer Lambda 25UV/Vis spectrophotometer. The estimation of phenolic compounds in the extracts was carried out in five replications and calculated by a calibration curve (linear dynamic range from 10 to 150 mg/L) obtained with gallic acid. Total phenolics were determined according the following equation A765 = 0.003GAE (mg/L) - 0.010 (r 2 = 0.999) and expressed as gallic acid equivalents of aqueous extract (mg (GAE)/L). All measurements were performed in triplicate.

Antioxidant activity

DPPH free radical-scavenging activity

The DPPH free radical-scavenging activity of the investigated medicinal plant aqueous extracts was determined as previously described by Lin et al. (2009), with some modifications. Briefly, an aliquot of 100 μL solution of the extracts at different concentrations was mixed with 1.0 mL of methanolic solution of DPPH (0.2 mM). The mixture was shaken vigorously and allowed to stand at room temperature for 30 min. Then the reduction of DPPH radical was determined by measuring the absorbance at 517 nm. Lower absorbance of the reaction mixture indicated higher free radical scavenging activity. The data presented are the average of five measurements given as mean ± standard deviation. The radical scavenging activity was calculated as percentage of DPPH discoloration of the extracts using the following formula and expressed as percent inhibition:

$$ \mathrm{Radical}\ \mathrm{scavenging}\ \mathrm{activity}\ \left(\%\right)=\left[\left({\mathrm{A}}_{\mathrm{control}}-{\mathrm{A}}_{\mathrm{plant}\ \mathrm{extract}}\right)/{\mathrm{A}}_{\mathrm{control}}\right]\times 100 $$

where: Acontrol is the absorbance of DPPH● solution and Aplant extract is the absorbance of the solution when the sample extract has been added at a particular level.

A gallic acid standard curve was obtained by using a gallic acid standard solution at various concentrations. The results obtained using following equation: ln (%ΔA517) = 0.760 ln(GAE(mg/L)) + 0.937; r 2 = 0.962, were also expressed as mg(GAE)/L. All determinations were carried out three times.

TBARS

Inhibition of lipid peroxidation was determined by measuring TBARS as previous described (Gonçalves et al., 2013). Briefly, aqueous extracts of target plants were incubated at 37 °C for 1 h with 30 μL of 0.1 M tris-HCl buffer (pH 7.4) containing 250 μM of freshly prepared FeSO4 and distilled water. The reaction was then stopped by adding 8.1 % SDS, acetic acid/HCl (pH 3.4) and 0.8 % TBA, and incubated at 100 °C for 1 h. The TBARS was measured by determining absorbance at 532 nm.

Results and discussion

SPME optimization

The effectiveness of SPME was optimized to obtain a broad spectrum of volatiles. In fact, fiber coatings, extraction time and extraction temperature are generally considered as the most important factors to achieve maximum extraction efficiency. The best HS-SPME/GC–qMS conditions were chosen based on intensity response (GC peak area), number of isolated and identified compounds, and relative standard deviation (RSD%). After the optimization step, the volatile pattern of T. vulgaris, R. officinalis and R. chalepensis, aqueous extracts was established.

Selection of the fiber coating

The selection of a suitable fiber coating is an important step in SPME optimization. The SPME efficiency depends greatly on the value of the distribution constant of analytes partitioned between the sample and fiber coating material. Thus, the performance of commercially available SPME fibers was tested and compared to evaluate the extraction efficiency of each coating towards volatile metabolites, using R. chalepensis aqueous extract as the matrix.

Figure 4 shows the results of the fiber selection carried out by HS-SPME/GC-qMS. Each fiber was exposed to the headspace for the same time at the same temperature. The results of the fiber screening revealed that DVB/CAR/PDMS coating showed a strong extraction affinity for the volatile metabolites, followed by PDMS/DVB, CAR/PDMS, PDMS, CW/DVB and PA. For all assays, the repeatability was lower than 20 %. Considering the number of identified compounds (Fig. 4 a), the best performance was achieved by DVB/CAR/PDMS (32 volatiles) followed by PDMS/DVB (25), PDMS (24), CAR/PDMS (21), CW/DVB (19) and PA (19). DVB/CAR/PDMS coating combines the absorption properties of the liquid polymer with the adsorption properties of porous particles, which contains macro (> 500 Å), meso (20–500 Å) and microporous (2–20 Å). The mutually synergetic effect of adsorption and absorption of the stationary phase explains its high retention capacity. Consequently, based on the results, DVB/CAR/PDMS fiber was selected for all further optimization steps, and subsequently for the analysis of the volatile metabolites in the investigated medicinal plants (Fig. 4 b).

Fig. 4
figure 4

Effect of the experimental conditions on the extraction performance of the HS-SPME procedure for volatile metabolites in R. chalepensis L. plant extract, including a Number of volatile compounds; b response areas of different terpenoids, carbonyl compounds, esters, alcohols, volatile phenols extracted with different fibers measured in the R. chalepensis L.; c Effect of extraction time (fiber: 65 μm DVB/CAR/PDMS; extraction temperature: 40 °C; stirring: 800 rpm; desorption conditions: 6 min at 240 °C); and d extraction temperature (fiber: 65 μm DVB/CAR/PDMS; extraction time: 45 min; stirring: 800 rpm; desorption conditions: 6 min at 240 °C). a.u. arbitrary unit. Error bars show the standard deviation of the mean (n = 3)

Full size image

Extraction time and extraction temperature

To optimize the extraction time, the DVB/CAR/PDMS fiber was exposed in the headspace at different times, between 15 and 60 min. The influence of the extraction time on the SPME efficiency is reported in Fig. 4 c. The volatile profile, in terms of total GC peak area, shows a slight increase between 30 and 60 min, but the best extraction efficiency was obtained for 45 and 60 min. A total of 34 volatile compounds have been identified at 45 min, while using an extraction time of 60, 30 and 15 min, the number of identified volatiles are 33, 26 and 24, respectively. Repeatability (RSD%) was lower than 20 % for all extraction times tested. At 45 and 60 min, similar total GC peak areas, RSD values and number of identified volatiles were obtained. Thus, in order to implement a more expeditious methodology, the lower extraction time was selected (45 min).

The extraction temperature presents several effects on extraction efficiency. In general, increasing extraction temperature can enhance the mass transfer of the analytes from bulk sample to the headspace and diffusion coefficients and Henry’s constants, while the time needed to reach equilibrium decreases (Lord and Pawliszyn 2000; Perestrelo et al. 2011). To assess the influence of temperature on SPME efficiency, different extraction temperatures ranging from 25 °C to 60 °C using DVB/CAR/PDMS fiber for 45 min were tested. To prevent analyte degradation, temperatures higher than 60 °C were not considered. Figure 4 d shows the effect of the extraction temperature on the total GC peak area. The results revealed that the total GC peak areas increase with temperature between 25 °C and 40 °C, while a slight decrease was observed at 60 °C. A total of 34 volatile compounds have been identified at 40 °C, while using an extraction temperature of 60 °C and 25 °C, the number of identified metabolites is slightly lower, 29 and 20, respectively. Therefore an extraction temperature of 40 °C was selected for the HS-SPME analysis of the target medicinal plants volatiles. Repeatability (RSD%) was lower than 20 % for all extraction temperatures tested.

Volatile metabolomic profile of aqueous plant extracts

The optimized HS-SPME/GC–qMS methodology was applied to establish the volatile profile of T. vulgaris, R. officinalis and R. chalepensis aqueous extracts. The typical total ion current (TIC) chromatogram profiles of the target medicinal plants are shown in Fig. 5 . The percentage of relative peak area (RPA, %) of each volatile metabolite is summarized in Table 1 .

Fig. 5
figure 5

TIC chromatograms of volatile metabolites from medicinal plant aqueous extracts obtained by HS-SPME/GC–qMS methodology, using DVB/CAR/PDMS fiber at 40 °C for 45 min. (Attribution of the peak number shown in Table 1). a.u. arbitrary unit

Full size image
Table 1 Comparison of the volatile metabolomic pattern of T. vulgaris, R. officinalis and R. chalepensis using HS-SPME/GC-qMS methodology
Full size table

As shown in Table 1 , a total of 99 volatile metabolites were identified, including 53 terpenoids, 19 carbonyl compounds, 7 esters, 6 higher alcohols, and 14 miscellaneous. These volatile metabolites were identified by their mass spectrum, retention index (using C8-C20 n-alkanes series), and further confirmed by authentic standard, when available.

As observed, the aqueous extracts of Lamiaceae medicinal plants, T. vulgaris and R. officinalis, had in common 23 volatile metabolites, mainly terpenoids such as α-thujene, camphene, α-phellandrene, β-myrcene, limonene, eucalyptol, (Z)-ocimene, γ-terpinene, (E)-ocimene, p-cymene, isopulegol, dihydrocarvone, borneol, β-citronellol, p-cymen-8-ol, eugenol and thymol. R. chalepensis and R. officinalis plant extracts shared 12 volatile metabolites, like (E)-linalool oxide, dihydrocarvone, carvone, β-eudesmol, (E)-isoeugenol, bornyl acetate, nonyl acetate, methyl benzoate, 1,4-bis(1-methylethyl)-benzene, 2-decanone, (E)-2-octenal and 1-octen-3-one. Only dihydrocarvone was simultaneously identified in all three medicinal plant aqueous extracts. Figure 6 presents the distribution of chemical groups by medicinal plants investigated.

Fig. 6
figure 6

Distribution of chemical groups by medicinal plant extracts. CC - carbonyl compounds

Full size image

The terpenoids were the most predominant chemical group in T. vulgaris and R. officinalis aqueous extracts accounting, 83 and 69 %, respectively, of the total volatile pattern, whereas for R. chalepensis 95 % of total volatile profile was composed by carbonyl compounds, namely aldehydes. The main volatile metabolites identified in T. vulgaris aqueous extracts were thymol (67 %), isopulegol (6 %) and α-terpinene (2 %), whereas in R. officinalis, eucalyptol (40 %), bornyl acetate (10 %) and borneol (7 %), were the most dominant. R. chalepensis was characterized by the high levels of 2-undecanone (53 %), (E)-2-octenal (28 %) and 2-nonanone (10 %). The carbonyl compounds contribution to the R. chalepensis total volatile profile was almost 11 and 3 times higher, respectively, than the content found in T. vulgaris and R. officinalis aqueous extracts.

These results are in accordance with previously published data on T. vulgaris (Baranauskienė et al. 2003; Grigore et al. 2010) and R. officinalis (Barnes et al. 2007; Jiang et al. 2011) medicinal plant extracts. Terpenoids are important for plant survival and also possess medicinal properties that are beneficial to humans, such as anti-carcinogenic, antimalarial, anti-ulcer, hepaticidal, antimicrobial or diuretic activity (Aharoni et al. 2005). Similar results were also found by Mejri et al. (2010) for R. chalepensis volatile profile.

TPC and antioxidant activity

The TPC and the antioxidant activity of the studied medicinal plant extracts are illustrated in Fig. 7 . The TPC is significantly different among the studied medicinal plant extracts. The highest values of total phenolics were obtained for R. chalepensis (57 mg (GAE)/L) aqueous extract, followed by R. officinalis (26 mg (GAE)/L) and finally T. vulgaris (14.9 mg (GAE)/L)) (Fig. 7 ). The difference of TPC was significant, up to 4 fold, and the phenolics contribution to quality and nutritional value in terms of, flavor, aroma and also antioxidant potential providing health-beneficial effects are well-known (Vaya et al. 1997).

Fig. 7
figure 7

TPC and DPPH radical-scavenging activity determined in investigated medicinal plants

Full size image

DPPH method is based on the reduction of alcoholic DPPH solutions in the presence of a hydrogen donating antioxidant. The antioxidant activity of T. Vulgaris, R. officinalis, and R. chalepensis aqueous extracts was evaluated through scavenging activity of the stable free radical DPPH●. The antioxidant activity ranged from 22 to 35 mg (GAE)/L, and the difference was not very large, up to 2 fold. The highest value was obtained for T. vulgaris and R. officinalis and the lowest for R. chalepensis plant extract (Fig. 7 ).

Furthermore, as the volatile profile of these medicinal plant extracts was established by HS-SPME/GC–qMS methodology, this allowed us to try a possible correlation between chemical composition and their antioxidant activity. A possible explanation for the low antioxidant activity of R. chalepensis plant extract could be attributed to their high content of carbonyl compounds and low levels of terpenoids. A previous study that screened the antioxidant activity of 98 pure oil components belonging to different chemical groups, observed that carbonyl compounds (e.g. 2-undecanone) exhibited a low scavenging activity, if any antioxidant effect, whereas terpenoids, such as monoterpenes hydrocarbons and phenols showed a significant protective effect with several variants due to their different functional groups (Ruberto and Baratta 2000). Moreover, the high antioxidant activity of T. vulgaris and R. officinalis could be explained by the presence of some oxygenated monoterpenes, such as thymol, eucalyptol, borneol, β-citronellol, geraniol and eugenol, with high antioxidant activity. From these, thymol is in fact responsible for the antioxidant activity of many essential oils (Ruberto and Baratta 2000), so these medicinal plants can be explored as a natural source of antioxidant compounds for possible applications in nutraceutical industries. The antioxidant activity of the aqueous extracts was also assessed by TBARS inhibition assay. In this assay, the greatest effectiveness was shown by R. officinalis (0.002 μg/mL), slightly lower for T. vulgaris (0.003 μg/mL), and the lowest for the R. chalepensis oil (0.11 μg/mL). Compared to positive control (gallic acid) the aqueous extracts exhibited quite similar antioxidant activity.

Conclusions

HS-SPME combined with GC–qMS provides a suitable tool and a reliable alternative technique to establish the volatile metabolomic pattern of medicinal plant aqueous extracts. Since SPME technique is very sensitive to experimental conditions, some parameters were optimized in order to achieve maximum extraction efficiency. According to the data obtained, DVB/CAR/PDMS fiber in addition to an extraction time and temperature of 45 min and 40 °C, respectively, were found to be the optimum conditions to ensure the best extraction efficiency. Sixty-six, 40, and 31 volatile metabolites were identified in T. vulgaris, R. officinalis, and R. chalepensis aqueous extracts, belonging to different chemical families namely monoterpenes, sesquiterpens, alcohols, carbonyl compounds and esters. Terpenoids are the dominant chemical group in T. vulgaris and R. officinalis aqueous extracts, whereas in R. chalepensis aqueous extracts carbonyl compounds is the main chemical group. Thymol (67 %), eucalyptol (40 %) and 2-undecanone (53 %) were the main volatile metabolites detected in T. vulgaris, R. officinalis and R. chalepensis aqueous extract, respectively.

The obtained results indicate that among the studied medicinal plants T. vulgaris and R. officinalis possess stronger protective effects than R. chalepensis. These finding could be attributed to the high terpenoid content, mainly monoterpenes hydrocarbons (e.g. β-myrcene, γ-terpinene) and oxygenated monoterpenes (e.g. thymol, eucalyptol), which are responsible for several biological activitiesThe investigated medicinal plant extracts, especially thyme and rosemary, can probably be used to prevent several human diseases. In addition to their well-known traditional use in food and cosmetics, the great potential of tested medicinal plants, encourage further investigation.