Eugenia neonitida Sobral and Eugenia rotundifolia Casar . ( Myrtaceae ) Essential Oils : Composition , Seasonality Influence , Antioxidant Activity and Leaf Histochemistry

Investigou-se a influência da sazonalidade na composição do óleo essencial de folhas frescas de Eugenia neonitida Sobral e Eugenia rotundifolia Casar. (Myrtaceae) coletadas trimestralmente, assim como a atividade antioxidante do óleo essencial de folhas secas e a histoquímica foliar. Os componentes majoritários do óleo essencial de E. neonitida foram biciclogermacreno, germacreno D e b-cariofileno e os de E. rotundifolia foram a-pineno, b-pineno e b-cariofileno. Os óleos essenciais de ambas as espécies apresentaram principalmente hidrocarbonetos e álcoois cíclicos, sendo o óleo essencial de E. neonitida composto principalmente por sesquiterpenos e o de E. rotundifolia por monoterpenos. A precipitação parece influenciar na composição dos óleos essenciais de ambas as espécies. O óleo essencial de folhas secas apresentou baixa atividade antioxidante. Os testes histoquímicos evidenciaram que as cavidades secretoras armazenam soluções complexas e contêm, além dos óleos essenciais, outras classes de lipídeos.


Introduction
Eugenia neonitida Sobral and Eugenia rotundifolia Casar.(Myrtaceae) are found only in Brazilian sandy coastal environments, 1 also referred to as "restingas", which are ecosystems constantly subjected to the impact of human occupation and to water and nutritional deficiencies due to sandy soil. 2 Eugenia neonitida is commonly known as "pitangão", while the common name for E. rotundifolia is "araponga".Both species produce berry fruits, and E. neonitida ones are eatable, extremely aromatic, and used for beverage preparation. 3They also possess numerous translucent dots in the whole plant defined as secretory cavities of the essential oils. 4The composition of Eugenia leaf essential oils has been studied and many authors have reported the occurrence of a-pinene, a-copaene, b-elemene, alloaromadendrene, d-cadinene, spathulenol, globulol, epiglobulol, b-caryophyllene and a-humulene.These last two components are known for their anti-inflammatory properties and were identified in the essential oils of more than 30 Eugenia species. 5,6However, neither E. neonitida nor E. rotundifolia essential oils composition was found in the literature.
This paper appraises the composition of fresh and dried leaves essential oils of E. neonitida and E. rotundifolia.In addition, variation in the amounts of fresh leaves essential oils major compounds was evaluated trimestrially seeking to correlate it with abiotic factors or with the plant's phenologic stage.The antioxidant activity of dried leaves essential oils and the leaves histochemistry were also reported.

Abiotic factors
Data referring to the abiotic factors medium temperature (°C) and precipitation (mm) in Rio de Janeiro City during 2009 were obtained from Instituto Brasileiro de Meteorologia (INMET). 7According to that, in Rio de Janeiro the medium temperature varied in the range of 22-29 °C and precipitation varied from 51 to 211 mm.Usually, tropical climate seasons are not well defined, occurring only periods of months colder and drier than others, as shown by Gaussen-Gagnouls ombrotermic diagram built with the climatological normals of medium temperature (°C) and precipitation (mm) during the period between 1961-1990. 7However, analysis of the Gaussen-Gagnouls ombrotermic diagram built with the same abiotic factors during 2009 has shown a variation of this pattern, indicating the occurrence of a dry season from May to August in Rio de Janeiro. 7ant material Eugenia neonitida and E. rotundifolia leaves were obtained from georeferenced individuals in a sandy coastal environment located in Rio de Janeiro City called "Restinga de Grumari" (23º02'94''S and 43º31'98''W).Plant samples were collected trimestrially in a way that two collections occurred in drier and colder months (May and August) and both the other collections were carried out in wetter and hotter months (February and November).During the collection of the plant material the phenologic stages (presence of leaves, flowers and fruits) were observed. 8oucher specimens were deposited at Universidade Federal do Estado do Rio de Janeiro Herbarium (HUNI646 and HUNI650).

Essential oil extraction, seasonal variation and statistical analysis
Fresh leaves collected from four of both species georeferenced individuals during the months of analysis were subjected to hydrodistillation for 3 h using a Clevenger apparatus.Air dried leaves collected in November from five individuals of both species were likewise subjected to hydrodistillation.The essential oils obtained from dried leaves were weighted and the extraction yields expressed in relation to dry leaf mass were calculated.
A linear correlation test was used in order to detect any influence of medium temperature and precipitation on the percentages of the major compounds present in both species essential oils from the February, May, August and November 2009 collections.Likewise, major compounds of the essential oils were correlated with the phenologic stages of the collected georeferenced individuals.A t student test was also used to detect seasonal variation in the essential oils major compounds in May (cold and dry month) and in November (hot and wet month).In this case, the results presented here are the means of four replicate experiments.Statistical analyses were conducted using the software GraphPad Instat.

Gas chromatography-mass spectrometry (GC-MS)
The leaf essential oils were analyzed by high resolution gas chromatography with a flame ionization detector (GC-FID), and high resolution gas chromatography-mass spectrometry (GC-MS).For the GC-FID analysis a Varian Star 3350 apparatus equipped with a silica capillary column Quadrex Inc. 007-1 (50 m × 0.25 mm id; 0.5 mm film thickness) was used.For the GC-MS analysis a Varian Star CP3800 apparatus equipped with a triple quadrupole trap detector (Varian Star 12000L) and a fused silica capillary column Varian Inc. VF5-ms (30 m × 0.25 mm id; 0.25 µm film thickness) operating in electron impact mode (70 eV) were used.The carrier gas was H 2 in GC-FID and He in GC-MS.Similar chromatographic conditions were used in both analyses: injector and detector temperature was 250 °C; the column temperature was programmed to hold isothermal at 40 °C for 2 min, then heating rate of 6 °C min -1 to 250 °C, and at the end hold isothermal for 5 min.Injection volume was 0.5 mL of essential oil solution in CH 2 Cl 2 , in splitless mode.
Compounds identification was accomplished by comparing the obtained mass spectra with reference data from National Institute of Standards and Technology database (NIST 05 2.0d), and also by comparing the experimental Kovat's retention index (IK C ) calculated from a C 7 -C 26 n-alkanes series with literature data. 9The chemical compound amounts were expressed as percentage, calculated from the normalized area data.

Antioxidant activity
The radical scavenging activity was measured according to 2,2-diphenyl-1-picrylhydrazyl (DPPH • ) method. 10,11The essential oils obtained from dried leaves of both species from November collection were diluted in ethanol to final concentrations of 1.000, 250, and 125 mg mL -1 .A 0.37 mL sample of 0.1 mmol L -1 DPPH • (Sigma, 90% purity) solution in ethanol was added to 0.93 mL aliquots of essential oil solutions.Ethanol (0.37 mL) plus essential oil solution (0.93 mL) was used as blank, while DPPH • solution (0.37 mL) plus ethanol (0.93 mL) was used as negative control.After 30 min of reaction in the dark and at room temperature the absorbance values (sample, blank and negative control) were measured at 517 nm.As positive control a 2,6-di-tert -butyl-4-methylphenol (BHT, Fluka, ≥ 99.0% purity) solution in the same concentration of the essential oil solutions was used.
The percentage of antioxidant activity (AA%) was calculated using the equation: AA% = 100 -{[(Abs sample -Abs blank ) × 100] / Abs control }.Each assay was repeated three times and results are means ± standard deviation.Efficient concentration values (EC 50 ) were calculated by linear regression of the data.

Histochemistry
Fresh leaves from both species were transversally hand sectioned and stained with 1% alcian blue and 1% basic fucsin in 50% ethanol.The following histochemical tests were also applied to transversal sections of fresh leaves: Sudan III and IV, 12 Sudan black B 13 and Sudan red B 14 for lipids; Nile blue sulphate for acid and neutral lipids; 15 2,4-dinitrophenylhydrazine for terpenoids with carbonyl group 16 and Oil Red O for rubber. 17The observations were captured in an Olympus BX41 light microscope.

Composition and seasonal variation of fresh leaves essential oils
Altogether, 41 compounds were identified in E. neonitida fresh leaves essential oil, and 40 compounds in E. rotundifolia (Tables 1 and 2).Essential oils major compounds were bicyclogermacrene, germacrene D and b-caryophyllene in E. neonitida, and a-pinene, b-pinene and b-caryophyllene in E. rotundifolia.These compounds are commonly identified in Eugenia leaves essential oils. 18he essential oils of both species are rich in cyclic compounds, especially hydrocarbons.A few alcohols and one oxide derivative were also identified.The essential oils compounds were grouped into four major classes: hydrocarbon monoterpenes, oxygenated monoterpenes, hydrocarbon sesquiterpenes, oxygenated sesquiterpenes and others, category in which non terpenoid compounds were included.Eugenia neonitida essential oil was rich in hydrocarbon sesquiterpenes throughout the analysis (Figure 1A), while hydrocarbon monoterpenes were the major compounds in the E. rotundifolia essential oil (Figure 1B).
The seasonal variation in the six major compounds of the essential oils of both species was addressed through the analysis of their relative abundance and is represented in Figures 1 and 2. In E. neonitida a decrease was observed in the amount of bicyclogermacrene, globulol, viridiflorol and epi-globulol during the coldest and driest month (August), while the amount of germacrene D increased remarkably (Figure 1A).These five mentioned compounds plus b-caryophyllene added up more than 50% of E. neonitida essential oil (Table 1).In E. rotundifolia a decrease was observed in the amount of b-caryophyllene, spathulenol and a-cadinol accompanied by an increase in the amount of a-pinene and b-pinene in August (Figure 1B).These last remarked components plus limonene added up more than 60% of E. rotundifolia leaves essential oil (Table 2).
Besides the mentioned quantitative variation, qualitative differences were also observed in both species essential oils.For instance, in E. neonitida essential oil b-copaene was not identified in February and May, while a-pinene, cyperene, valencene, and ledol were not identified in August (Table 1).In E. rotundifolia essential oil, a-cubebene, a-gurjunene, b-copaene and spathulenol were not identified in August (Table 2).
Since small differences in the amounts of the major compound of both species essential oils were detected, these values were statistically correlated to the phenologic stages of the collected individuals and the abiotic factors medium temperature (°C) and precipitation (mm).In relation to phenologic stages, leaves were observed in both species during the four months of analysis.Flowers and fruits were observed only in E. neonitida in February and November.
Statistically significant correlation between the amounts of major components of E. neonitida and E. rotundifolia essential oils with either the phenologic stage or the medium temperature was not detected.Statistically significant positive correlations were detected only between the precipitation and the compounds biclyclogermacrene (r = 0.987; P = 0.013*) in E. neonitida essential oil (Figure 2A), and b-caryophyllene (r = 0.973; P=0.057*) in E. rotundifolia (Figure 2B).Additionally, one statistically significant negative correlation was detected also between the precipitation and the a-pinene in E. rotundifolia essential oil (r = -0.984;P = 0.016*) (Figure 2B).Thus, the phenologic stage and medium temperature do not seem to influence both species essential oils composition.Likewise, according to the t student test, statistically significant differences in the composition of essential oils of leaves collected during the coldestdriest and the hottest-wettest periods of the year were not detected.Similar observations were found by other researchers. 19ropical climate seasons are not well defined.In Rio de Janeiro, for instance, 2009 monthly medium temperature and precipitation varied only from 22 to 29 °C and from 51 to 211 mm, respectively, as mentioned earlier. 7This small variation does not establish the occurrence of spring, summer, autumn and fall, it only indicated the occurrence of a dry season from May to August.These abiotic factors are even more stable in sandy coastal environments such as "restinga de Grumari," characterized by high temperatures and low water availability, 2 which probably contributed for the small variation observed in the essential oils of both species.
However, statistically significant correlation was detected between precipitation and one out of six major compounds of E. neonitida essential oil and two out of six major compounds in the essential oil obtained from E. rotundifolia leaves.Even though only three statistically significant correlations were detected, a noteworthy decrease occurred in the percentages of viridiflorol, globulol and epi-globulol in E. neonitida essential oil and in the percentages of spathulenol and a-cadinol in E. rotundifolia essential oil in August, the month with the lowest precipitation level.[22]

Composition and antioxidant activity of dried leaves essential oils
The essential oils obtained from dried leaves collected from E. neonitida and E. rotundifolia in November are very similar to the essential oils obtained from fresh leaves.Most of their major compounds are also the major compounds in the essential oils obtained from fresh leaves (Tables 1 and 2).Bicyclogermacrene, b-caryophyllene and germacrene D are also the major compounds in E. neonitida dried leaves essential oil (Table 1).Likewise, a-pinene, b-pinene and b-caryophyllene are the major compounds in E. rotundifolia dried leaves essential oil (Table 2).Eugenia neonitida and E. rotundifolia dried leaves essential oils yields were 0.1% and 0.4%, respectively (Tables 1 and 2).
However, some qualitative and quantitative differences were observed when comparing both species fresh and dried leaves essential oils.In E. neonitida, for instance, aromadendrene and b-bisabolol were observed only in fresh leaves essential oil, while globulol, viridiflorol and a-cadinol percentages in dried leaves essential oil were inferior when compared to fresh leaves essential oil (Table 1).In E. rotundifolia, terpinolene and b-gurjunene were detected only in fresh leaves essential oil, while b-bisabolene and germacrene B percentages in dried leaves essential oil were superior when compared to fresh leaves essential oil (Table 2).
Another noteworthy comment is the occurrence of components such as terpinolene (E.neonitida) and cyperene (E.rotundifolia) only in the essential oils obtained from dried leaves, what may be explained by the superior number of individuals from which leaves were collected for essential oil extraction.
Eugenia neonitida and E. rotundifolia dried leaves essential oils exhibited weak antioxidant activities according to the DPPH • assay.EC 50 values were 8.6 mg mL -1 for E. neonitida and 5.8 mg mL -1 for E. rotundifolia.Because of the similarities in the composition of E. neonitida, E. rotundifolia, and other Eugenia species essential oils, it is possible to predict that the antioxidant activity of the last ones will also be weak.The exception is E. caryophyllata, commonly known as clove, whose essential oil exhibits strong antioxidant activity (EC 50 = 0.2 mg mL -1 ) due to the presence of eugenol. 5,6,23

Histochemistry
The presence of lipids in secretory cavities and parenchyma cells of both species was detected by Sudan III, IV, black B, and red B. Nile blue sulphate indicated the presence of neutral lipids (such as hydrocarbons) also in both species parenchyma cells and secretory cavities.It also detected the presence of acid lipids (such as like fatty acids or phospholipids) only in E. rotundifolia secretory cavities.Rubber was detected by Oil Red O in both species secretory cavities and parenchyma cells, indicating once more the complexity of these structures content.Finally, 2,4-dinitrophenylhydrazine test used to detect the presence of terpenoids with carbonyl group did not react with the secretory cavities content, certifying the GC-FID and GC-MS analyses.Vol.22, No. 8, 2011

Conclusions
The major compounds in the essential oils of fresh leaves were bicyclogermacrene, germacrene D and b-caryophyllene, in E. neonitida, and a-pinene, b-pinene and b-caryophyllene in E. rotundifolia.The essential oils of both species are mainly constituted by cyclic hydrocarbons and cyclic alcohols, but also possess cyclic oxides and non terpenoid aliphatic compounds.Eugenia neonitida essential oil is rich in sesquiterpenes, while E. rotundifolia's is rich in monoterpenes.Despite the small quantitative and qualitative variations in both species essential oils, statistically significant correlation with either phenologic stages or medium temperature and essential oil major components was not detected.Statistically significant seasonal difference in these last was also not detected.Precipitation, however, seems to be related to the percentages of some essential oils major components.The essential oils of both species dried leaves are quite similar to the essential oils obtained from fresh leaves and exhibited weak antioxidant activity.The leaves' histochemistry certified the GC-FID and GC-MS analyses and also attested the complexity of the solutions stored in secretory cavities and parenchyma cells: besides hydrocarbon and alcohol lipids, they may possess fatty acids, phospholipids and rubber.

Figure 1 .
Figure 1.Seasonal variation in relative percentages of Eugenia neonitida (A) and Eugenia rotundifolia (B) essential oil major compounds and chemical classes during the analysis.

Figure 2 .
Figure 2. Seasonal variation in relative percentages of Eugenia neonitida (A) and Eugenia rotundifolia (B) essential oil major compounds during the analysis and the abiotic factors medium temperature and precipitation. 7

Table 1 .
Seasonal variation in chemical composition of the essential oil of Eugenia neonitida fresh leaves, chemical composition of dried leaves essential oil, and calculated Kovat's retention index (KI C ). Entries are normalized area values

Table 2 .
Seasonal variation in chemical composition of the essential oil of Eugenia rotundifolia fresh leaves, chemical composition of dried leaves essential oil, and calculated Kovat's retention index (KI C ). Entries are normalized area values