Chemical composition and antigenotoxic properties of Lippia alba essential oils

The present work evaluated the chemical composition and the DNA protective effect of the essential oils (EOs) from Lippia alba against bleomycin-induced genotoxicity. EO constituents were determined by Gas Chromatography/Mass Spectrometric (GC-MS) analysis. The major compounds encountered being citral (33% geranial and 25% neral), geraniol (7%) and trans-β-caryophyllene (7%) for L. alba specimen COL512077, and carvone (38%), limonene (33%) and bicyclosesquiphellandrene (8%) for the other, COL512078. The genotoxicity and antigenotoxicity of EO and the compounds citral, carvone and limonene, were assayed using the SOS Chromotest in Escherichia coli. The EOs were not genotoxic in the SOS chromotest, but one of the major compound (limonene) showed genotoxicity at doses between 97 and 1549 mM. Both EOs protected bacterial cells against bleomycin-induced genotoxicity. Antigenotoxicity in the two L. alba chemotypes was related to the major compounds, citral and carvone, respectively. The results were discussed in relation to the chemopreventive potential of L. alba EOs and its major compounds.


Introduction
Lippia alba (Mill.) N.E. Brown (Verbenaceae), an aromatic shrub reaching 1.7 m high, is distributed throughout the Caribbean, South and Central America and Tropical Africa. The species is mainly used in folk medicine against digestive and respiratory ailments, but also as a sedative, analgesic, anti-inflammatory, antipyretic and antihypertensive remedy (Pascual et al., 2001a;Hennebelle et al., 2008a). In Colombia it is popularly known as "Orégano de cerro" (Hill oregano), "Pronto alivio" (ready-relief) and "Curatodo" (all-round cure) depending on the region (Stashenko et al., 2003).
The species L. alba is characterized by variability in the chemical composition of the essential oils, depending on the origin of plant material, as well as the stage of the plant and the part selected for distillation of the oil (Zoghbi et al., 1998). Various chemotypes have been proposed (Hennebelle et al., 2006;Oliveira et al., 2006). Based on both the composition and the possible common biosynthetic pathways among the different oils, the existence of at least seven has been indicated (Hennebelle et al., 2008a). These are: chemotype I (citral, linalool and b-caryophyllene, as the main constituents), chemotype II (tagetenone), chemotype III (limonene and carvone or related monoterpenic ketones), chemotype IV (myrcene), chemotype V (g-terpinene), chemotype VI (camphor -1,8-cineole) and chemotype VII (estragole). In Colombia, L. alba chemotypes I and III, and a combined (I/III) form, not previously reported, have been found.
After determining the EOs composition of the two L. alba specimens by GC-MS analysis, their specific antigenotoxic activity against the clastogenic mutagen, bleomycin, was evaluated by using the SOS Chromotest (Quillardet et al., 1982). The antigenotoxic properties of the major EO constituents (citral, carvone and limonene) were also studied and their activity compared with the antigenotoxic standard compound Trolox. Our work provides new insights into chemoprevention by L. alba EO major compounds.

Plant material
L. alba plants were collected from the experimental gardens at CENIVAM Agroindustrial Pilot Complex, located at the Universidad Industrial de Santander campus (Bucaramanga, Colombia). Plant growing conditions were as indicated by Stashenko et al., (2008). Taxonomic identification was undertaken by Dr. José Luis Fernández Alonso (National University, Bogotá, Colombia). The two L. alba specimens (COL512077 and COL512078) were stored at the Colombian National Herbarium.

EO extraction and chromatographic analysis
Fresh leaves and flowers from L. alba plants were used for EO extraction using the microwave-assisted hydrodistillation method, as described by Stashenko et al., (2004). Briefly, a Clevenger-type hydro-distillation apparatus was placed inside a domestic microwave oven (LG, 1100 W, 2.45 GHz) with a side orifice, through which an external glass condenser linked the 2 l-round flask with the plant material (ca. 300 g) and water (ca. 0.5 l) inside the oven. The oven was operated for 40 min (4 x 10 min) at full power, which caused water to boil vigorously and reflux. Essential oil was decanted from the condensate, and then dried with anhydrous sodium sulfate. For chromatographic analysis, neat essential oil (50 mL) and n-tetradecane (0.5 mL) were dissolved in 1 mL of dichloromethane (Chromatography-grade reagent, Merck, Darmstadt, Germany). EO compound identification was based on chromatographic/spectroscopic analysis, as previously indicated by Vicuña et al., (2010).

Genotoxicity assay
The SOS Chromotest, as indicated by Quillardet et al., (1982), was used for genotoxicity assaying. Briefly, overnight-cultures were grown in fresh LB medium (indicated above) until reaching an optical density of OD 600nm = 0.4. They were then diluted 10-fold in doublestrength LB medium, and mixed (v/v) with a specific substance for identification (EO, citral, carvone and limonene). Pure EOs (density of 900 mg/mL determined with a BRAND picnometer, Wertheim, Germany) were diluted in distilled water by vigorously stirring to a concentration ranging between 1.7 and 450.0 mg/mL, this including the antioxidant dose as previously indicated (Stashenko et al., 2004). Negative (distilled water) and positive (1 mg/mL of bleomycin) controls were always included in each assay. Cells were exposed to substances during 30 min at 8°C, and then cultured during 2 h at 37°C. The assays for b-galactosidase and alkaline phosphatase activities were according to Vicuña et al. (2010).
The genotoxicity criterion applied was the Induction Factor (IF), which, by representing fold induction of the sulA gene in each treatment (EO, mutagen, etc), could be considered as an indirect measure of induced primary DNA damage. The IF was calculated as: IF = (b-galactosidase/alkaline phosphatase) t / (b-galactosidase/alkaline phosphatase) nt , where t and nt are the treated and non-treated cells, respectively.

Antigenotoxicity assay
Antigenotoxicity was assayed using the co-incubation procedure, as indicated by Fuentes et al., (2006). Although the procedure was basically the same as that of the genotoxicity protocol, the cells were simultaneously cotreated with different concentrations of the tested substances (EO, citral, carvone and limonene) and the mutagen (1 mg/mL of bleomycin). Antigenotoxicity, i.e., the DNAprotective capacity of the tested substance, was measured as a significant reduction in IF in the combined treatments (substance + bleomycin), and expressed as a percentage of genotoxicity inhibition: here IF co is the SOS induction factor in co-treated cells (substance + bleomycin), IF basal the basal SOS induction factor, and IF bleo the SOS induction factor in bleomycintreated cells.

Statistical analysis
The average values of alkaline phosphatase and IF and the corresponding standard errors were calculated. Normality of the data was tested using the Kolmogorov-Smirnov test. Variance homogeneity and analysis of variance (ANOVA) tests were also conducted. Mean values were compared using Student's t-test. Product-moment (Pearson) correlation analysis was applied for examining dose-response relationships in genotoxicity studies. In all statistical analyses, p < 0.05 was considered significant. The STATISTICA software package (Version 6.0, StatSoft Inc (2003), Tulsa, OK, USA) was used for all analyses.

Genotoxic and antigenotoxic effects of L. alba EOs
The genotoxicity of L. alba EO was assayed before the antigenotoxic effect was investigated. Oils did not increased the IF values in PQ37 Escherichia coli strain indicating that they do not induce the SOS response in E. coli cells (Table 2). Interestingly, a stimulating effect on protein synthesis, measured as alkaline phosphatase activity, was observed with increased EO concentration in the case of Lippia alba essential oils 481  Table 1.
López et al. No., Order of elution is given in DB-5MScolumn, I K , Values of retention index (Kovats, 1965) calculated from a minimum of three independent chromatograms.
citral chemotype. Since this did not occur with a water soluble EO fraction (data not shown), apolar compounds in the EO mix are possibly involved. The antigenotoxic properties of L. alba EO are shown in Table 3. As previously indicated (Vicuña et al., 2010), a dose of 1 mg/mL bleomycin was used for antigenotoxicity assaying. EOs produced a significant decrease in bleomycin-induced genotoxicity (IF values) at doses between 28.1 and 450 mg/mL, though insignificant at those lower. Complete inhibition occurred with both citral and carvone/limonene chemotypes at doses higher than 56.2 mg/mL.

Genotoxic and antigenotoxic effects of L. alba EO major constituents
Genotoxicity of the major EO constituents (citral, carvone and limonene) was also assayed (Table 4)
Citral induced a significant reduction in bleomycininduced genotoxicity at a dose of 182 mM. Percentages of genotoxicity inhibition (% GI) increased with citral doses suggesting a direct mode of action for antigenotoxicity of this compound mixture and supporting the results observed with the EO. Carvone and limonene were also antigenotoxic. S(+)-carvone was significantly active only from a dose of 798 mM on, as also very similarly its isomer (R(-)-carvone) (data not shown). Limonene was antigenotoxic from a relatively lower dose (97 mM) on, although GI percentages were always consistently lower than those observed with citral (Table 4). Thus, citral was considered of higher antigenotoxic potential.
Data on antigenotoxicity of positive standard Trolox were also presented for comparison with citral, carvone and limonene. Assayed doses were determined experimentally, since no previous reports on this standard compound and SOS Chromotest were available in the literature. Trolox produced a significant decrease in bleomycin-induced genotoxicity from a dose of 586 mM, onwards, thus comparatively nearly 90, 398 and 1548 times lower than those of citral, carvone and limonene, respectively.
Apparently, this is the first report on the genotoxic and antigenotoxic properties of L. alba EOs. Under the experimental conditions assayed here (absence of exogenous metabolic activation), the L. alba chemotypes (citral and carvone/limonene) did not induce DNA primary damage in the SOS Chromotest. In addition, EO major constituents as citral and carvone were not genotoxic in the SOS Chromotest. This was in accordance with previous studies using SOS Chromotest, Salmonella/microsome and Drosophila melanogaster SMART assays (Franzios et al., 1997;Gomes-Carneiro et al., 1998;Stammati et al., 1999). For limonene, IF increased at doses between 194 and 774 mM, thereby contrasting with the results obtained with EO of the carvone/limonene chemotype. This limonene-effect was possibly masked in the EO by interaction with other constituents, perhaps even carvone itself. Nevertheless, this presumption needs to be tested. A previous study (Vukovic-Gacic et al., 2006) indicated non-mutagenic effects for limonene using Salmonella/microsome assay. As the results so far have been inconclusive, harmonized studies on the genotoxicity of these compounds are now underway in our laboratory.
The antigenotoxic potential of L. alba EO was also shown. Although both the citral and carvone/limonene chemotypes were antigenotoxic against the clastogen bleomycin, citral appears as the most promising source of chemopreventive compounds, apparent by the antigenotoxicity observed in the major constituents (Table 4). The order of antigenotoxic activity for these compounds was found to be citral > carvone > limonene, indicating that citral was the most active compound. Although the chemopreventive properties of L. alba terpenoids, as carvone, geraniol, limonene and perillyl alcohol, have already been well-documented (He et al., 1997;Crowell, 1999;Uedo et al., 1999;de Carvalho and da Fonseca, 2006;Paduch et al., 2007;Patil et al., 2009;Rabi and Bishayee, 2009), little is really known as regards citral. Connor (1991) was the first to indicate citral chemopreventive potentiality against skin chemical carcinogenesis in mice. Further experimental evidence lent supported that citral has an ability to suppress oxidative stress, possibly through the induction of endogenous antioxidant proteins, such as phase II xenobiotic metabolizing enzymes, as well as glutatione S-transferase (Nakamura et al., 2003). In addition, it has been recently demonstrated that citral strongly inhibited the CYP2B60 hydroxylase activity (Seo et al., 2008) involved, not only in xenobiotic activation of a wide variety of pro-mutagens, but also in the synthesis of Aflatoxin B 1 mycotoxin (Shukla et al., 2009), involved in gastric carcinogenesis. The present work provides new insights into citral and carvone chemoprevention. Since bleomycin genotoxicity involves the generation of radicals in the DNA molecule, which thus induce DNAstrand breakages (Claussen and Long, 1999), it can be expected that the antigenotoxic effect of citral and carvone against bleomycin occurs through radical scavenging mechanisms within the molecule. In fact, Stashenko et al. (2004) have previously demonstrated antioxidant properties for the L. alba carvone/limonene chemotype.
In conclusion, this study showed the antigenotoxic properties of L. alba EO, citral, carvone and limonene against the drug bleomycin, lending support to the potential of the oils and compounds in chemoprevention and cancer therapy. Since the role of chemopreventive agents in the etiology of cancer is very complex, and involves several modes of action, and our results concern only in vitro experiments with a bacterial assay, additional animal and human studies involving different endpoints should be addressed in order to clarify the antimutagenic potential of L. alba EOs and their major constituents. In addition, harmonized studies on the genotoxicity of citral, carvone and limonene, using a battery of in vivo assays that evaluate different levels of DNA damage expression, are required, prior to the practical use of these compounds in chemoprevention.