Essential Oil Variation in Brazilian Varronia curassavica Jacq. in Response to Drying and Edaphoclimatic Conditions

Varronia curassavica Jacq. (Boraginaceae) is a native species of the Atlantic Forest with medical importance. This study determined the essential oil variation of 16 populations of V. curassavica of restinga in Santa Catarina, Brazil, in response to drying and edaphoclimatic conditions. The populations supplied essential oil with a content between 0.27 to 1.15% in the summer and from 0.33 to 1.12% in the winter. 41 chemical compounds were identified in the summer and 40 in the winter. The compounds were grouped into 4 chemical classes in each station studied. The common chemical constituents found in the essential oil of all populations and in both seasons were α-thujene, α-pinene, sabinene, α-humulene, (E)-cariophylene, spatulenol, mircene, allo-aromadendrene, β-sesquifelandreno and α-zingiberene. Cluster analysis using the nearest neighbor method based on Euclidean distance grouped the 16 populations into 3 groups in the summer and 8 groups in the winter. As the habitats have distinct pedological characteristics, we identified that pH, organic matter, sum of bases and base saturation are associated with the synthesis of (E)-caryophyllene, α-humulene, and allo-aromadendrene from populations.


Introduction
The Atlantic Forest is one of 25 recognized biodiversity hotspots globally; it is home to more than 19,000 species, of which 35% are endemic (Oliveira et al., 2019;Souza et al., 2021). Despite the remarkable endemism levels that make the Atlantic Forest one of the most distinct regions in the Neotropics (Ribeiro et al., 2011;Souza et al., 2020), little is known about the aromatic plants' potential genetic resources in this biome. Among the botanical species that occur in the restinga-an ecosystem associated with the Atlantic Forest biome and established on sandy soils of marine origin-Varronia curassavica Jacq. (synonym = Cordia verbenacea DC.) (Boraginaceae) is considered one of the main sources of molecules used in the treatment of inflammation, rheumatism, and ulcers Roldão et al., 2008). This is due to the high diversity of secondary metabolites, specifically, the essential oils (EOs), which are synthesized and stored in the glandular trichomes present on the leaf surface (Feijó et al., 2014). In addition, the EO of this species stands out for being the first topical phytotherapeutic developed entirely in Brazil, with anti-inflammatory action (Nizio et al., 2015). With the commercial name of Acheflan ® and launched by the Aché Laboratory in 2011, this phytomedicine has achieved prescription leadership in the medicinal plant segment, with over 1 million units sold, representing USD 8.1 million and accounting for 10% of industry revenue (Oliveira, 2017). Besides, V. curassavica has recognized efficacy by the Brazilian Health Regulatory Agency and appears in different official lists of the Brazilian Ministry of Health (Oliveira, 2017).
Pre-clinical studies with histamine-induced edema assays in mice have attributed to the sesquiterpene α-humulene the role in the anti-inflammatory effect of the EO from V. curassavica  EO with the desired amount of the substance of interest. For example, the pharmaceutical industry demands that plants have a minimum content of 2.0% α-humulene to meet the quality standard (Magalhães, 2010).
Various environmental conditions influence the production of EOs; among them, the year-season stands out; when the season changes, plants perform physiological changes in their metabolism to amplify CO 2 uptake and water and nutrient cycling, thus reflecting in the increase or decrease in the content and/or relative percentages of the compounds present in EOs (Dehsheikh et al., 2019). Some authors attribute that temperature can alter EO production via activation of thermosensitive enzymes involved in the mevalonic acid pathway, precursors of terpenes (Burbott & Loomis, 1967;Rahimmalek & Goli, 2013;De Almeida et al., 2016). Additionally, solar radiation can influence EO production directly or indirectly through increased plant biomass (Burbott & Loomis, 1967). EOs production is influenced by various environmental conditions, of which the seasons of the year stand out. Season change entails physiological changes in plant metabolism to amplify CO 2 uptake and water and nutrient cycling, reflecting the increase or decrease in the content and/or relative percentages of compounds present in EOs (Dehsheikh et al., 2019). Some studies have ascribed that temperature can alter EO production via activation of thermosensitive enzymes involved in the mevalonate pathway, a precursor of terpenes (Burbott & Loomis, 1967;Rahimmalek & Goli, 2013;De Almeida et al., 2016). Additionally, solar radiation can influence EO production directly or indirectly through increased plant biomass (Burbott & Loomis, 1967).
Previous studies have reported that the geographic location of plants along with the soil and climate conditions serve as modulators in EO production (Rahimmalek et al., 2017;Marques et al., 2019); thus, the same species may show differential EO production depending on the environment in which it is established. Similarly, that different genotypes can result in differential essential oil production, possibly due to the behavior of floral visitors that can induce gene flow among them (Hoeltgebaum et al., 2018). Another factor is that different plant phenological stages contribute to marked differences in EOs productivity (Bouyahya et al., 2019). However, regardless of the phenological stage at which the plant is harvested, this procedure is usually performed when the plant has high water content, and drying is the most widely used process to ensure the quality and stability of EOs after harvest.
Understanding the genotype × environment interaction may provide new insights for selecting V. curassavica matrices with potential pharmaceutical use, which should be coordinated with sustainable use practices of the species. Although V. curassavica has a wide geographic distribution along the coastal zones of the Santa Catarina state in Brazil (Bayeux et al., 2002), to our knowledge, there is no information about the chemical compounds of EOs found in these populations. In addition, the diversity of EOs among plants collected in natural habitats enables determining the collecting seasons, as well as the ideal growing conditions for domestication and improvement of the plants. In this sense, our objective was to determine the chemical variation of EOs from V. curassavica specimens collected from 16 native restinga populations in Santa Catarina, in response to drying and edaphoclimatic conditions.

Method
The experiments were conducted under field and laboratory conditions, during summer (February 2015 and 2016) and winter (September 2015 and, to determine the influence of soil and climatic conditions on the OE content and chemical composition of V. curassavica

Characterization of the Collection Site
Detailed information about the collection site of the V. curassavica populations used in this study is detailed in Table 1. For the soil physicochemical analyses, six soil samples were collected at a depth of 20 cm at each location in April 2015. The samples from each location were pooled to form a composite sample and subsequently air-dried and sieved (1 mm). The fraction thinner than 1 mm was retained for physicochemical analyses. The potential of hydrogen (pH) was determined in a 1:1 soil-water volume ratio. Phosphorus (P) was extracted with Mehlich solution, and Aluminum (Al) was extracted with 1 mol L -1 potassium chloride (KCl). Organic matter (OM), base saturation (V), and the sum of bases (SB) were determined according to Silva (1999

Plant Material and Isolation of the Essential Oils (EOs)
Leaves from 16 populations of V. curassavica were collected along approximately 77 km of the Santa Catarina coast, between latitudes 26°22′22.2″S and 26°51′52.4″S, during the summer (February 2015 and 2016) and winter (September 2015 and. The populations were identified with codes according to the municipality where they are located (Table 1). In the Atlantic Forest biome, the vegetation cover is inserted in the restinga associated ecosystem, whose ecological nature is conditioned to the vegetation complex established on sandy soils of marine origin. Each population consisted of a minimum number of 20 individuals. All individuals were selected within a maximum distance of 30 cm per work unit and collected on a single day. These populations were also characterized according to the size of their range, average plant height, and the average height of the portion of branches with leaves (Table 1).
To obtain the EOs corresponding to the fresh leaves, samples of the collected populations were immediately taken to the laboratory and subjected to hydrodistillation in a Clevenger apparatus (Vidrolabor ® , São Paulo, Brazil) for 3 h. On the other hand, for the EO obtained from the dried leaves, the plant material was previously kept under a forced air circulation oven (Fanem ® , São Paulo, Brazil) at 40 °C until the mass remained constant before performing the hydrodistillation process mentioned above. The EO content was determined based on dry matter basis, with three replicates per treatment. Subsequently, the EOs obtained were separated with anhydrous sodium sulfate (Na 2 SO 4 ) and kept refrigerated at -4 °C in amber flasks until the chemical analyses were conducted.

Chemical Analysis of Essential Oils (EOs)
GC-MS was carried out in an Agilent 6890 gas chromatograph coupled to an Agilent 5973 N mass selective detector. The GC was fitted with an HP-5MS fused capillary column (30 m × 0.25 mm × 0.25 μm film thickness) coated with 5% phenyl-95% dimethylpolysiloxane stationary phase. Helium was used as the carrier gas at a flow rate of 1.0 mL/min. Temperature programming was set to 60-240 °C at the rate of 3 °C/min, heated to 240 °C, and held at this temperature for 10 min. The injector temperature was kept at 250 °C. Essential oil samples were diluted to a 1% solution in dichloromethane, and 1.0 μL of the solution was injected with a split ratio of 1:20. The mass detector was operated in electron ionization mode (70 eV) at a 3.15 scan/min rate and a scan range of 40-450 Da. The transfer line was maintained at 260 °C, the ion source at 230 °C, and the analyzer (quadrupole) at 150 °C. For the quantification, the EOs were injected in an Agilent 7890A gas chromatograph fitted with FID operated at 280 °C. Hydrogen was used as the carrier gas at a flow rate of 1.5 mL/min, using the same column and conditions described above.
The quantification of each constituent was estimated by electronic integration of the FID signal with the corresponding peak area, which was determined based on the average of three injections (Tables 3 and 4). The identification of the oil's components was carried out by comparison of the mass spectra with those from commercial libraries and also by their linear retention indexes, after the injection of a homologous series of alkanes (C 8 -C 26 ), under the same experimental conditions, compared to literature data (Adams, 2007).

Data Analyses
The EOs and α-humulene contents were submitted to the Shapiro-Wilk test to analyze the normality of the residuals and homogeneity of the variances. All data were subjected to analysis of variance (ANOVA), and the means were compared by the Scott-Knott method at 5% error probability. For the multivariate analysis, the Euclidean distance was estimated using the 'dist' function; the HCA was performed by the Unweighted Pair Group with Arithmetic Mean (UPGMA) method through the 'hclust' function, and the PCA was performed by the 'princomp' function. All of the functions belong to the 'stats' package. All data were analyzed in the "R" statistical software version 2.15.1 (R Development Core Team 2012).
In the winter season, we found eight clusters, using the Euclidean distance of 12 units as a dissimilarity measure ( Figure 1B). Cluster 1 was formed by the populations SF1, NA2, BS3, BV1, and PE1, characterized by α-thujene (3.1-19.3%), α-pinene (7.8-18.2%), α-zingiberene (13.7-38.8%), and β-sesquiphellandrene (10.  Vol. 13, No. 8; throughout the year, except in summer, when the plants were in the vegetative phase, as previously described (Brandão et al., 2015). This probably contributed to an increase in EO content in winter compared to the oil content extracted in summer in the fresh leaves of the BV2 (1.1%) and PE3 (0.9%) populations. The high rate of EO biosynthesis during the reproductive period in these populations may be due to the activation of enzymes required for the biosynthesis of certain compounds. The higher level of EO accumulated during flowering is a net result of the anabolic and catabolic processes during the several stages of flower development (Dubey & Luthra, 2001).
This research showed that the EOs chemical composition varied among the studied populations, highlighting the chemical diversity found in the same species. The common compounds in all populations were α-thujene, α-pinene, sabinene, α-humulene, (E)-caryophyllene, spathulenol, myrcene, allo-aromadendrene, β-sesquiphellandrene, and α-zingiberene, but with quantitative variations among them. A previous study on native populations of V. curassavica collected in Sergipe, Brazil, identified 53 compounds, of which 18 were also found in this study, except for α-thujene, myrcene, β-cubebene, γ-muurolene, (E)-nerolidol, and β-copaen-4 α-ol (Nizio et al., 2015). On the other hand, these compounds were obtained in V. curassavica genotypes cultivated in São Paulo, Brazil (Marques et al., 2019). These findings allow us to suggest that the diversity of compounds in the EO of V. curassavica may be influenced by geographic location compared to other studies (Santos et al., 2006;Nizio et al., 2015;Marques et al., 2019), as well as, the ability of the terpene synthase enzymes to convert the acyclic prenyl diphosphates and squalene into a multitude of cyclic and acyclic forms (Degenhardt et al., 2009). This property is found in almost half of the known monoterpenes and sesquiterpenes and may be attributed to the fact that the various reactive carbocationic intermediates can be stabilized in more than one way (Degenhardt et al., 2009). In this respect, we also observe that the chemical compounds' greatest diversity belongs to the classes of hydrocarbon sesquiterpenes and oxygenated sesquiterpenes compared to monoterpenes. Sesquiterpenes synthesized in the cytosol from farnesyl diphosphate by sesquiterpene synthases are structurally more diverse than monoterpenes due to the increased number of different cyclizations possible with five additional carbon atoms (Bohlmann et al., 1998;Nagegowda, 2010).
O α-humulene, an economically valuable constituent in the EO of V. curassavica, is a monocyclic sesquiterpene produced by terpene synthase enzymes using farnesyl pyrophosphate (Bohlmann et al., 1998). Its synthesis is related to the formation of CO 2 and acquisition of photosynthesis intermediates (Dehsheikh et al., 2019). Variations in the content of α-humulene (0.3-6.8%) found in this work may result from the influence of genetic load and environmental conditions. The α-humulene content was higher in the PE3 population (2.0-6.8%), regardless of season and leaf drying. A previous study found higher α-humulene content (31.6 %) in V. curassavica under field conditions with water and nutrition supply to the plants (Queiroz et al., 2020). In contrast, in our study, V. curassavica populations were established in sandy leached, and nutrient-poor soils, and plants were also exposed to high salinity, solar radiation, constant winds, and high soil temperatures. We also observed that all populations in summer, after the leaf drying process, maintained high α-humulene contents (0.7 to 6.8%). Also, we found that γ-muurolene, α-cubebene, and β-copaen-4 α-ol were detected only after the leaves were dried. Similarly, Amaral et al. (2017) found an increase in sesquiterpene molecules after leaf drying when studying tree species from the Atlantic Forest. These authors attributed that such changes in the EOs chemical composition are due to the higher stability of sesquiterpenes compared to monoterpenes and oxidation processes during drying.
We found intrapopulation variability of the chemical compounds in the EOs of V. curassavica. This variability plays an important role in understanding natural populations, as they outline conservation and genetic improvement strategies. Thus, we distributed the plants into groups regardless of leaf drying. As a result, HCA grouped the 16 populations into three groups in summer and eight in winter. PCA identified that (E)-caryophyllene, α-humulene, and allo-aromadendrene were strongly associated with populations PE1, PE3, BS2, PI2, BS4, and NA1 in summer. These populations were strongly related to climatic conditions in summer, such as Tp and Pp. The seasonal dynamics of the municipalities where the collections were made are quite similar, with hot and rainy summer (average values Tp of 24.5 °C; Pp of 240.1 mm) and mild and dry winter (average values Tp of 17.3 °C, Pp of 95.3 mm). However, with increased sunlight in summer, photosynthesis tends to increase, and consequently, high levels of energy are available for plant growth and development (Rezaei et al., 2019), resulting in the balance of available energy being directed toward secondary metabolite production. Similarly, the higher Tp recorded in summer provides considerable effects on substrate concentrations because of its effect on modifying day length (Burbott & Loomis, 1967).
The water availability is known to increase the production of terpenes (Maatallah et al., 2016) due to biosynthetic reactions occurring in an aqueous medium. Similar results were reported by Boira and Blanquer jas.ccsenet.org Journal of Agricultural Science Vol. 13, No. 8; (1998), who revealed a positive relationship of sesquiterpenes, such as β-caryophyllene and caryophyllene oxide, when Tp and Pp increased. On the other hand, we observed that Tp and Pp in winter are related to sabinene and α-thujene, although their relationships are less evident. These monoterpenes are inverse to (E)-caryophyllene, α-humulene, and allo-aromadendrene in winter. This negative relationship between monoterpenes and sesquiterpenes can be interpreted as competition between two pathways for the same precursor (Ghaffari et al., 2018;Ghaffari et al., 2019). Thus, we can suggest that isopentenyl pyrophosphate fluxes are dominant to the plastid (site of monoterpene synthesis) under winter conditions, whereas this flux tends toward the cytosol (site of sesquiterpene synthesis) in summer (McCaskill & Croteau, 1994).
Another factor considered in our work was the soil in which the plants grow. This is one of the main aspects that differentiated our research, contrasting with V. curassavica plants from different geographical origins and propagated under ideal growing conditions (Santos et al., 2006;Nizio et al., 2015;Marques et al., 2019). In this study, most of the collection sites belong to the order of arid soils. This order is typically defined by saline or alkaline soils with very little OM, characteristic of arid regions (Dewan & Famouri, 1964). Most of the collection sites belong to the order of arid soils. This order is typically defined by saline or alkaline soils with very little OM, characteristic of arid regions (Dewan & Famouri, 1964). As the habitats have distinct pedological characteristics, we identified that pH, OM, S, and V are associated with the synthesis of (E)-caryophyllene, α-humulene, and allo-aromadendrene from populations PE3, BV3, BS4, PI2, and BS2, acting inversely with Al in the soil. The positive effects of these chemical compounds associated with pH affect plant nutrient availability and natural soil fertility (OM, S, and V). On the other hand, by acting inversely with Al, they decrease the toxic effects of this element on plants. Al impairs the synthesis of energy in the plant due to the inhibition of P uptake and transport and the ATPase enzyme activity (Ahn et al., 2001;Abichequer et al., 2003). Considering the need for energy in the form of ATP and other P-dependent enzymes in the synthesis of EOs (Loomis & Corteau, 1972), the limitation of this element may be determinant in the increase of (E)-caryophyllene, α-humulene and allo-aromadendrene contents under conditions of Al saturation in the soil. Thus, the chemical variation of EOs observed in V. curassavica populations is due to environmental conditions, as well as their interactions.

Conclusion
The PI1 and PE1 populations exhibit the highest EO content, while PE3 has the highest α-hulene content regardless of drought and season. HCA demonstrates differences in the chemical profile of populations from different locations. PCA corroborates these findings and shows that (E)-caryophyllene, α-humulene and allo-aromadendreno are related to soil and climate conditions (Tp, Pp, OM, SB, V and S) of PE1, PE3, BS2, Populations PI2, BS4 and NA1 in summer.