Essential Oils from Piper lhotzkyanum Kunth Leaves from Brazilian Atlantic Forest: Chemical Composition and Stability in Different Storage Conditions.

This work aimed to evaluate the impact of different storage conditions and light and temperature exposures on the visual aspect and chemical composition of the essential oil (EO) of Piper lhotzkyanum Kunth, obtained from leaves by hydrodistillation from a region of altitude. For this purpose, aliquots of the EO were stored for up to 90 days (a) under a refrigerator condition of 5 ± 3°C, (b) under a long-term (LT) condition of 30 ± 2°C and 75 ± 5% relative humidity (RH) and an accelerated condition (AS) of 40 ± 2°C and 75 ± 5% RH, and (c) in a photostability test achieved in amber and colorless glass vials. The changes were monitored on days 0 (control), 60, and 90 for the refrigerator, LT, and AS conditions. All EO chemical analyses were assessed by GC-FID and GC-MS for quantification and identification, respectively. It is reported, for the first time, that the EO of P. lhotzkyanum is rich in the sesquiterpenes β-elemene and α-zingiberene. No significant changes in the EO was observed, revealing a minimal impact of temperature on the sample at the different storage conditions. However, there was a change in the content of α-zingiberene to bicyclogermacrene after exposure to light. The visual appearance of the samples was altered for all test conditions except the refrigerator condition. These results can potentially contribute to the product development of a bioactive EO from leaves of P. lhotzkyanum, a sesquiterpene rich natural material.


Plant material
Fresh leaves of Piper lhotzkyanum Kunth Piperaceae were collected in winter season July, 2019 at Serra dos Órgãos National Park, near city of Teresópolis, Rio de Janeiro State, Brazil Elevation: 1,145 m; GPS location 12 11 45 S, 38 58 05 W . A voucher specimen was deposited at Herbarium of the Rio de Janeiro Botanical Garden, Rio de Janeiro, Brazil RB01426181 . This study was registered at Genetic Heritage Management Council under AE4E953 and Chico Mendes Institute for Biodiversity Conservation under 57296-1.

Essential oil extraction
Fresh leaves 2000 g from P. lhotzkyanum were cleaned and cut into small pieces and submitted to EO extraction by hydrodistillation using a Clevenger-type apparatus. The extraction was carried out for 3 h after the mixture reached boiling point. Traces of water were removed from the EO by treatment with anhydrous Na 2 SO 4 Sigma-Aldrich, Brazil . Yields were measured as weight of EO per weight of fresh leaves, in percentage w / w 10,11,15,36 . Analysis was carried out in triplicate. After extraction, samples of the EO were subjected to a photostability test with different vials, as well as stability test in three different storage conditions, in order to investigate the impact of temperature and light during a pre-established period of time.

Photostability study
In order to assess the stability to light, aliquots of EO 300 µL, triplicate were placed in amber PS1 and colorless PS2 2 mL glass vials, and instantly sealed tightly with a crimp cap containing a teflon septum crimp 2 mL, Agilent Technologies . A control EO sample control was storage under refrigeration 20 until the duration of the chromatographic analysis the same used for accelerated and long-term storage study . PS1 and PS2 were submitted to a photostability study using a chamber Ethik Technology 424-CF, São Paulo, Brazil according to ICH Q1B 37 , option 2 and Brazil, 2019. Samples were exposed to light providing an overall illumination of not less than 1.2 million lux hours and an integrated near ultraviolet energy. Aliquots of 10 µL of PS1, PS2 and control were diluted ten times in dichloromethane HPLC grade, Tedia, Brazil for gas chromatographic GC analysis after the exposure experiment.

Accelerated and long-term storage study
Samples were placed in amber glass vials crimp 2 mL, Agilent Technologies, USA and submitted to a stability study using a climatic chamber BE-8209 SERIES Weiss, Brazil , according to Brazil, 2019 38 . Container closure system proposed for marketing, which according to ICH Q1A R2 39 is classified as impermeable that provided a permanent barrier to passage of moisture or solvent. Samples were exposed to 40 2 / 75 5 relative humidity RH for accelerated stability AS ; 30 2 / 75 5 RH for long-term study LT and 5 3 in refrigerator Consul CRA 30FB, Brazil for cold stability CT Tests were carried out in triplicate LT1 / AS1 60 days; LT2 / AS2 90 days; Control / CT1 / CT2 0, 60 and 90 days, respectively . LT, AS, CT and control aliquots of 10 µL were diluted ten times in dichloromethane HPLC grade, Tedia, Brazil for GC analysis done in triplicate.

Essential oils analysis
All samples of the EOs were injected 1 µL splitless 10, 15, 40 for chemical identification by gas chromatography coupled to mass spectrometry GC-MS and for quantification by gas chromatography coupled to flame ionization detection GC-FID .
GC-MS analysis for chemical identification was performed using a gas chromatograph 6890 GC coupled to a mass spectrometer Agilent MS 5973N Hewlett-Packard, Brazil , operating at 70 eV of ionization energy in positive mode and mass range m/z 40 -600 atomic mass unit u . GC conditions were a capillary column HP-5MS 30 m 0.25 mm i.d. 0.25 µm, film thickness , temperature programing from 60 to 240 with increasing of 3 /min, using helium 99.999 as carrier gas at a constant flow rate of 1.0 mL/min. Injector and detector was set at 270 ; and transfer line were set at 280 .
GC-FID analysis was performed for quantitative acquisition, using a gas chromatograph HP-Agilent 6890 Hewlett-Packard, Brazil equipped with a capillary column HP-5MS 30 m 0.25 mm i.d. 0.25 µm, film thickness , temperature programming from 60 to 240 , with an increase of 3 /min, using hydrogen as carrier gas at a constant flow rate of 1.0 mL/min. Injector and detector temperatures were set at 270 . Retention indices RI as well as peak quantification were achieved based on the GC-FID results. Relative quantities of individual components were calculated based on the peak areas of the GC without correction of the FID response factor.
Compounds were identified by comparing the mass spectra fragmentation pattern with those from literature records libraries from National Institute of Standards and Technology -NIST, 2010 and Wiley7n as well as from calculated retention index RI based on a homologous series of n-alkanes C 8 -C 28 , Sigma-Aldrich, Brazil 41 .

Statistical analysis
The results are noted as mean standard deviation M SD of three independent measures. The data were processed using the Tukey test ANOVA with 5 significance level p 0.05 for an adequate comparison of the obtained averages. Venn diagram was used for qualitative analyzes with Venny 2.0 software StartSoft Inc., Tulsa, USA 42 . Pearson s correlation for compounds with a concentration above 3 of essential oils was calculated using the Statistic 10 software.

Essential oil chemical characterization without stress
testing Essential oils from fresh leaves of P. lhotzkyanum control showed a citric yellow color in a vibrant tone and a characteristic citrus spicy aroma impression. The EO showed a yield of 0.60 0.23 w/w . It was possible to identify 64 compounds, representing a total quantification of 92.61 of the volatile mixture. The EO was characterized by a high relative percentage of hydrocarbons sesqui-terpenes 62.80 and monoterpenes 21.37 Table 1 . The major sesquiterpene hydrocarbons that were identified were β-elemene 21.96 , α-zingiberene 13.75 , and E-caryophyllene 6.07 . The hydrocarbon monoterpenes α-pinene 5.42 and β-pinene 8.98 were detected in great amounts. It is interesting to note that compounds with relative percentage less than 1 add up to a total of 48 compounds and correspond to 12.79 of the volatile mixture. Therefore, these results illustrate the chemical complexity of the P. lhotzkyanum EO.

Photostability study
The EO that was subjected to the photostability test had its color changed when compared to the EO with no treatment control citric yellow to colorless, respectively . The samples subjected to light in amber PS1 and colorless PS2 2 mL glass vials contained the major compound β-elemene 21.81 and 24.45 , respectively Table 1 . Therefore, no differences in relation to the control EO p 0.05 were observed. However, a large decrease in the α-zingiberene relative percentage for PS1 and PS2 0.27 and 0.19 , respectively and an increase in bicyclogermacrene for PS1 and PS2 10.77 and 12.48 , respectively were detected. In fact, α-zingiberene decreased from about 13 in the control EO to less than 1 after light exposure same for PS1 and PS2 . In addition, the samples in the colorless vials PS2 showed a 1.71 higher bicyclogermacrene relative percentage compared to the samples in the amber vials PS1 p 0.05 . It was possible to register a slight change in the relative percentage among oxygenated terpenoids decrease and hydrocarbon terpenoids increase between the samples in the amber PS1 and colorless PS2 vials.

Accelerated AS and Long-Term Storage LT
The EO in the different storage conditions i.e., longterm and accelerated stability for 60 days LT1 and AS1 and 90 days LT2 and AS2 , respectively, had its color slightly changed compared to the control EO. In the accelerated condition, an opaque yellow color was registered. The EO in storage at 5 for 60 days CT1 and 90 days CT2 showed no color alteration compared to the control EO. The EO samples in different analyzed conditions presented β-elemene as a major compound. The main components of the EO, which showed small variations influenced by heat, were sesquiterpene and monoterpene hydrocarbons in all analyzed conditions. In the different storage assays, there were no significant variations in chemical composition nor quantities p 0.05 . These results showed that there is a low influence of heat on the EO from P. lhotzkyanum Table 1 .
A Pearson correlation was carried out in order to obtain information on the relationship between the time factor and the content of the main compounds, which are pre-     sented in Table 2. It was possible to evaluate trends in each condition for which the samples were subjected. A significant negative correlation between the content of bicycloelemene 0.9985; p 0.035 with time to LT conditions was detected. An inversely proportional relationship can also be observed for this same condition for the compounds β-elemene 0.9953; p 0.062 and E-caryophyllene 0.9820; p 0.121 , but without statistical significance. These results were not observed for high AS and low CT temperatures.

Venn s Diagram and Qualitative Comparison
The Venn s diagram Figs. 1A-1C allows us to demonstrate and export any individual or multiple sub-sets as separate entity lists. Figure 1A shows the results obtained qualitatively over the period of 60 days at 5 , 30 , and 40 CT1, LT1, and AS1, respectively . Fifty compounds common to the three experiments and to the control day 0 were registered, which represent 60.2 of the total of compounds. Seven 7 , fifteen 15 , and five 5 different compounds of the control were identified in samples subjected to 5 , 30 , and 40 , respectively. We emphasize that at 30 LT1 , the presence of eight 8 unidentified compounds NI was registered, which were not common in the other samples. It was possible to identify common compounds in CT1, AS1, and LT1, which were 1,8-cineole, α-thujene, and α-cubebene. Figure 1B shows the comparison between CT2, LT2, and AS2 and control day 0 . It was observed that there was a decrease in compound similarity between samples 47 compounds, representing 53.4 . It was possible to iden-  This impact was confirmed by the absence of formation of compounds in the sample at low temperatures CT2 . In addition, qualitative variations after 90 days, although not relevant, were more evident than the quantitative data obtained for the accelerated and long-term storage experiments. Figure 1C shows the comparison of light exposed samples, PS1 amber vials , PS2 colorless vials , and control. The qualitative agreement values between the samples were 66.3 , representing 44 compounds. The light had a significant qualitative influence on the quality of the EO. This statement is evidenced by the strong agreement between the amber vials and the control 13.7 , 10 compounds compared to the colorless vials PS2 . Three 3 compounds were not registered in PS2, when compared to PS1 versus control. In other words, only 6 six compounds were common to PS1 and control 13.7 . This result suggests that there is a relationship between light exposure and P. lhotzkyanum EO stability.

Discussion
P. lhotzkyanum is a Brazilian medicinal plant used for many purposes. Likewise, many other species from the genus Piper is employed worldwide for medicinal and industrial applications 43 . The EO components have been obtained from uniform sample preparations from different vegetative parts and, by careful GC-MS and GC-FID analysis, it is possible to chemically identify as well as infer the relative percentage of each compound in the volatile mixture. Since EO chemical composition is made up of reactive compounds, such as unsaturated terpenoids, many factors can influence its quality. As such, knowledge of the influence of storage conditions on EO quality can guarantee commercialization of bioactive EO and its practical use 33 .
The EO from leaves of P. lhotzkyanum showed a chemical complexity 64 compounds , as expected for Piperaceae 10,11,15,27 . Its composition in the altitude region 1,145 m revealed a high sesquiterpenes β-elemene 21 and α-zingiberene 13 content and, for the first time, a chemical composition with these major components was found for this plant. P. lhotzkyanum EO from this high altitude differed from samples collected from a low-altitude region 763 m in São Paulo -SP Brazil 8 , which showed bicyclogermacrene 21 and caryophyllene oxide 14 as major constituents. This difference in EO composition may be attributed to environmental such as more exposure to sunlight and genetic factors 9,10,44,45 . We also believe that an assessment of the impact of seasonality and circadian rhythm on EO of these species is necessary 11 .
The main compound of P. lhotzkyanum from high altitudes, β-elemene Fig. 2 , is an asset of interest to the pharmaceutical and chemical industry. Studies with this sesquiterpene alone, or in mixture, afforded promising results as an anticancer agent with regulatory effects on the immune system, as well as possessing antimicrobial and anti-inflammatory 46 49 properties.
The visual aspect of EO revealed slight changes in color in the presence of temperature changes but great changes under the influence of light. In the EO extracted from marjoram Majorana hortensis Moench , whose composition was evaluated after one year of storage in a dark place, no significant changes in its color and characteristic aroma were observed 21 . In a study with seven EO subjected to possible worst-case conditions performed with Litsea cubeba Pers. and thyme Thymus vulgaris L. and stored under light and heat for 12 weeks, a discoloration process was observed 50,51 . Regarding light irradiation photostability test and the relative percentage of terpenoids in samples in amber PS1 and colorless PS2 vials, it was noted that both changed in a similar way. Compared to the control, the α-zingiberene content decreased dramatically, while the bicyclogermacrene relative percentage increased after light exposure about 13 . We suggest the conversion of α-zingiberene to bicyclogermacrene by C-C double bond rearrangement and C-6-ring breakage leading to a C-11-ring C-3-ring Fig. 2 . This chemical conversion is possible by light catalysis that generates free radicals due to the conjugated C-C double bond in α-zingiberene. The β-elemene content did not change under light exposure, possibly due to the absence of a conjugated C-C double bond in the sesquiterpene. There are few reports in the literature of variations of this range for sesquiterpenes 21 . In fact, UV and visible light are considered to accelerate selfoxidation processes, triggering the hydrogen abstraction that results in the formation of alkyl radicals 33 , which corroborates with our proposed theory. Some studies show changes in both the presence and in the absence of light, but in the presence of light, these changes may occur in a more pronounced way 23,52 . EO components are, generally, easily converted by oxidation, isomerization, cyclization, or dehydrogenization reactions due to the structural relationship within the same chemical group 20,33 . However, some authors point out that the variation between the oxygenated terpenoid components is more conducive to these recurrent changes to light irradiation compared to saturated hydrocarbons 33 .
In terms of unidentified compounds NI , an increase in these compounds after light exposure, both in amber and colorless vials, in relation to the control Venn diagram, Fig.  1C have been measured. Similar results were reported for the marjoram EO maintained in the presence and absence of light irradiation 19 . These NI compounds certainly arise by degradation after light exposure.
The storage of the EO under different conditions was followed by no relevant changes in the compounds content. In fact, structural changes depend on the reactivity of each compound 21 . The relative percentage of β-elemene, E-caryophyllene, α-zingiberene, bicyclogermacrene, and αand β-pinene did not change over 90 days p 0.05 . Some oxygenated compounds such as trans-linalool oxide, elemol, 1,8-cineole, and α-terpineol did not change significantly in their relative percentages between the different samples exposed to heat and the control. The terpinen-4-ol content decreased in the presence of heat, except in the first 60 days in the refrigerator. In a study related to the drying of EO of laurel Laurus nobilis L. , the results indicated, in general, that there was little change in the volatile concentrations between fresh seasoning and bay leaves that were air-dried at room temperature or oven-dried at 45 . Concentrations of certain oxygenated terpenes, such as 1,8-cineole, linalool, and geraniol slightly decreased 53 .
The presence of caryophyllene oxide in all samples, a stable by-product that results from the oxidation of the caryophyllene sesquiterpene, can often be found in stored EO 54,55 . However, this oxygenated sesquiterpene is usually found in Piper EO 11,26 . Literature records showed that linalool and limonene oxidize when exposed to air. These compounds are not allergenic, but the result of their oxidation can cause contact dermatitis 56 . The presence of epoxides and oxides such as allo-aromadendrene epoxide and trans-linalool oxide, both in trace amounts less than 0.3 , identified in the P. lhotzkyanum EO after 90 days of storage, may also be indicative of the initiation of an oxidative process. However, the percentage content of these two compounds is too small, which suggests no changes in the EO biological effect and no potential harmful effects on human health. This hypothesis must be investigated.
One factor that can be observed through the Venn s diagram, which shows the partitioning of the variation Figs. 1A-1C , was the higher prevalence of compounds that are common to different samples stored in the heat patterns when compared to the control. This represents an important value of the total peak area in the samples CT1, CT2, AS1, AS2, LT1, and LT2 studied. Qualitatively, this may be an indication of the stability of the P. lhotzkyanum EO. For example, a study with the marjoram EO Majorana hortensis Moench. , Stintzing 2011, 2012 19, 51 reported that the EO of different plant species responds differently; while the EO of thyme does not change much, the EO of rosemary ends up being more susceptible to simulated daylight that promptly leads to a change in chemical composition. Mehdizadeh et al. 2017 34 evaluated the effect of storage conditions on the chemical composition of the EO of Cuminum cyminum L. during six months of storage. The best results, in terms of chemical stability of the major constituents, were obtained at 20 and 5 .
The data in the Venn diagram Figs. 1A and 1B showed that, in the refrigerator 5 2 , less compounds were formed, followed next by the irradiated samples. Therefore, we propose that these samples would be more stable in qualitative terms, when compared to storage in heat or exposed to light. This result agrees with the findings by Mehdizadeh et al. 2017 34 .

Conclusion
For the first time, the EO from leaves of P. lhotzkyanum showed to be rich in β-elemene and α-zingiberene. The stability study under different storage conditions demonstrated a low qualitative and quantitative chemical variation, allowing us to conclude that, after 90 days of storage, the EO is stable, mainly at low temperatures 5 . Therefore, it was demonstrated that heat has a low influence on the EO from P. lhotzkyanum. Also, we demonstrated significant qualitative and quantitative changes of the EO when submitted to a photostability test. It is interesting to note the chemical conversion of α-zingiberene to bicyclogermacrene when the EO was exposed to light, as well as the chemical stability of the main EO compound β-elemene under this condition. Our findings can contribute to the development of products based on the bioactive EO from leaves of P. lhotzkyanum.