How do essential oil composition and phenolic acid profile of Heracleum persicum fluctuate at different phenological stages?

Abstract Heracleum persicum, commonly named Persian hogweed, is a principal native medicinal plant in Iran. Collecting H. persicum at the most appropriate growing stage is the key factor to achieve the high phytochemical quality to meet consumer's needs. In the present experiment, the aerial parts of this plant were harvested at up to six different developmental stages during the growing season to determine the phytochemical profiles. Our results indicated that the highest essential oil content was obtained in the mid‐mature seed stage (3.5%). The most elevated extract content was recorded in the floral budding stage (10.4%). In the vegetative stage, limonene (18.1%), in floral budding stage, caryophyllene (14.1%), anethole (14.6%), and β‐bisabolene (12.7%), in the full flowering stage, myristicin (15.0%), and hexyl butyrate (9.1%), in the early development of seeds stage, hexyl butyrate (32.1%), and octyl acetate (11.7%), in the mid‐mature seeds stage hexyl butyrate (38.8%), octyl acetate (14.5%), in the late‐mature/ripe seeds stage, hexyl butyrate (23.6%), and octyl acetate (10.5%) are recorded as the main components. The highest phenolic acids content was obtained in the floral budding stage (287.40 mg/g dried extract). The analysis of phenolic acids demonstrated cinnamic acid (8.0–225.3 mg/g extract), p‐coumaric acid (1.7–39.2 mg/g extract), p‐hydroxybenzoic acid (0.8–16.8 mg/g extract), and ferulic acid (2.4–15.8 mg/g extract) as the main phenolic acids. Cinnamic acid was found as the major phenolic compound in the vegetative stage following by floral budding, the full flowering stage, the early development of seeds, and late‐mature/ripe seeds stages. P‐coumaric acid was the most abundant phenolic compounds in the mid‐mature seeds stage. In this regard, the harvest time of H. persicum aerial parts can be selected to achieve the highest secondary metabolites of interest. The results of this study can be used as a guideline for grower to obtain the highest possible amount of desirable metabolites, beneficial in both food and pharmaceutical industries as well as their undeniable economical benefits.


| INTRODUC TI ON
Until now, more than 125 species of the genus Heracleum have been discovered all around the world. Most species of this genus are distributed in Asia which among them ten perennial aromatic species grow in the flora of Iran. Heracleum persicum L., commonly known as Persian hogweed simply hogweed, or Golpar, is a polycarpic perennial herbaceous and a flowering shrub belongs to Apiaceae family.
It is originally native to humid mountainous areas of Iran, Iraq, and Turkey (Radjabian et al., 2013).
This plant is widely distributed and grown in Iran in regions with different ecological conditions. The best-growing condition for this plant is moist and fertile areas, especially in the mountains in the northern part of the country, with altitudes from 1,500 to more than 3,000 m above the sea level. (Hoseinifar et al., 2016;Radjabian et al., 2013).
H. persicum L. is a perennial herb that usually grows up to 1.5-2 m. This alternate-leaved plant has an anise-like smell with thick and hollow stem. The leaf blades are elongated, densely haired on the lower side, glabrous on top, and pinnate with blunt-toothed margins. The flowers with five petals and five stamens are small with pale white and lime-green color. The fruits are broadly obovate, with slightly ridged schizocarp (Majidi & Sadati Lamardi, 2018).
The major source of secondary metabolites in plants of the Apiaceae family is the leaves, stems, flowers, and especially fruits.
Secondary metabolite contents are mostly allocated to the essential oils and phenolic compounds, which are more important for industrial use, than other phytochemical compounds. Other favorable compounds with antioxidant properties are mainly used in the food industry, while essential oil is edible and has pharmaceutical properties (Tajkarimi et al., 2010). Medicinal plants with antioxidant properties can replace synthetic antioxidants, because they are less costly and more environmentally friendly. In addition to the above-mentioned advantages, dietary phenolic antioxidants of H. persicum play important roles in delaying the development of chronic diseases such as cardiovascular diseases, inflammatory bowel syndrome, and Alzheimer's diseases (Goleniowski et al., 2013;Kaurinovic & Vastag, 2019). Studies have shown that Heracleum species are rich in phenolic compounds and exhibit high biological activities (Firuzi et al., 2010) making them worthy medicinal plants to study.
According to previous researches, phenological and harvest stages are bare bone essential factors that influence phenolic compounds contents as well as their biological activities in different organs of medicinal plants (Esmaeili et al., 2018;Grevsen et al., 2009;Hazrati et al., 2019). Choosing the best-growing stage will help us to harvest the higher amount of essential oil yield as well as other beneficial compounds. Moreover, the amount of phenolic compounds in medicinal plants is affected by genetic variation among different species, even within the same species and also by the maturity of plant organs at harvest time (Alirezalu et al., 2018;Deveci et al., 2018;Hazrati et al., 2019;Li et al., 2020;Wang et al., 2009).
However, the proper time of harvest to achieve the maximum beneficial compounds of H. persicum remains unknown and needs more research work. Until now, there is limited information on phenolic compounds of H. persicum in different organs and different growth stages. Thus, it is essential to specify more precisely the time when H. persicum plant should be harvested to attain the highest possible quality and quantity of beneficial contents. In this study, we aimed to investigate the distribution pattern of phenolic compounds and essential oil in the H. persicum. We also tried to find the best phenological stage that has the high concentration of these components in the aerial parts of H. persicum that were collected from the Iranian population at different developmental stages. To the extent of our knowledge, this is the first report aimed to evaluate the phytochemical composition of different growing stages of H. persicum, during its biological cycle, which results will be useful to find the best harvest time to reach the high beneficial yield with less harm to the environment.

| Plant material
The current study took place in the spring and summer of 2018 in the city of Sarab, East Azerbaijan Province, Iran (Location: 37°56′27″N and 47°32′12″E; altitude: 1,750 m). The first sampling started 30 days after the onset of the vegetative growth, in spring. All samplings were randomly selected from ten H. persicum plants. The harvest was done at different developmental stages (Figure 1). Each sampling was repeated three times, and sample's weight was 1 kg per replication. The samples were dried in a shade at room temperature (25°C).

| Extraction of essential oils
Applying hydro-distillation for three hours, the essential oil was extracted from 100 g dried samples in 600 ml of distilled water in a 2-L flask using Clevenger apparatus extraction technique in three replications (British Pharmacopoeia, 1988). The obtained essential oil was dried over anhydrous Na 2 SO 4 and kept at −20°C until further analysis.

| Preparation of the extracts
In order to prepare the extracts, dried powdered of different samples of H. persicum were used. About 30 ml of the methanol was added to 2 g of samples and put in an ultrasonic bath (Frequency, 100 kHz; power intensity, 160 W; temperature, 35°C) for 30 min.
Then, the extract was filtered, evaporated, and stored at -20°C until further analysis.

| Extraction of phenolic acids
The phenolic acid extraction was performed using our previous work method with some modification (Hazrati et al., 2019). Briefly, 10 ml of 80% ethanol was added to 1.0 g of the dried powdered plant, vigorously shaked, and centrifuged at a speed of 12,000 rpm for 10 min. The collected supernatant was evaporated and stored at -20°C for further analysis.

| Analysis of phenolic acids
In order to analyze the phenolic acid compounds, HPLC (Waters 2695, USA) system equipped with a diode-array detector, a 20 µl loop, and an ODS column (250 mm × 0.46 mm, 5 µm) was operated.

F I G U R E 1 H. persicum at different phenological stages
The reverse-phase separation was done with gradient elution solvent A [methanol TFA (99.9:0.1, v/v)] and B [water TFA (99.9:0.1,v/v)] at a flow rate of 0.5 ml/min with the following elution gradient: 20% A, at 0 min; 30% A, (from 0 to10 min); 60% A, (from 10 to 30 min); 80% A, (30 to40 min); 100% A, (40 to 45 min); 20% A, (from 45 to 52 min); isocratic, 6 min. The phenolic compounds were detected at 254 nm, 275 nm, and 320 nm, and identified based on the retention time and spike method. Finally, the standard external method was applied to quantify the studied phenolic acids. The results were expressed as mg per g of the extract weight.

| Gas Chromatography (GC)
Essential oils analysis was done using a gas chromatography device (model: Agilent 7890 A G). The separation was carried out using the Column MS HP-5 (30 m × 0.25 mm, 0.25 µm). The oven temperature program was set at 50°C (3 min) to 260°C with a ramp-up of 3°C/min and then held for 5 min. The detector and the injector temperature were adjusted at 270 and 240°C, respectively. Nitrogen was used as carrier gas at a rate of 1 ml/min.

| GC-MS chromatography
In addition to GC, a gas chromatography device connected to a mass spectrometer (Model: Agilent 7890 A G Chromatograph and Agilent 5975 c Mass), called GC-MS chromatography, was used to analyze the essential oil. In this experiment, GC-MS chromatography was equipped with HP-5 column (30 m × 0.25 mm, 0.25 µm). According to the planned program, the oven temperature remained constant at 50°C for 3 min and remained constant for 5 min after increasing the temperature to 260°C with a ramp-up of 3°C/min. Here again, nitrogen was utilized as carrier gas at the rate of 1 ml/min. The analysis was performed with scan time (30 m/z), range of analysis (600 m/z), ionization 0/6 s, 70 Electron volt and solvent evaporation rate of 2 min. The compounds were then identified by index (ki) and Wiley Library and Nist11.

| Statistical analysis
In order to determine the relationship between different phytochemical compounds, variance analysis, comparison of means, principal component analysis (PCA), and cluster analysis were done using the statistical package SAS 9.4. All determinations were conducted in triplicate, and the results were calculated as mean value ± standard error (SE). The variations (standard error of means, SE) and the significances of treatment effects (F-test) were calculated and tested using the General Linear Models procedure of SAS. The mean values of treatments were separated by Tukey's test (p < .05).

| Content of essential oil and extracts
Medicinal plants have different capacities to produce essential oil at different phenological stages (Li et al., 2020). To achieve their goal, breeders must consider the proper harvest time to reach the best yield for their target purposes, such as pharmaceutical, food, and cosmetic industries applications. For this reason, the essential oil content was studied in different phenological growth stages of H. persicum. Statistical analysis of this experiment showed that the essential oil content was significantly different in various phenological stages of this plant ( Figure 2). The results indicated that the obtained essential oil yield in the vegetative stages, floral budding stage, full flowering stage, early development of seeds stage, mid-mature seeds stage and, final stage, or late-mature/ripe seeds stages were 0.45, 0.72, 0.50, 1.00, 3.50, and 2.20%, respectively.
According to these data, the greatest yield was observed at the middevelopment stage of the seeds, while the minimum yield was asso- and Echinophora tenuifolia (Şanli et al., 2016), followed the same pattern and produced different essential oil content at various growth stages. This can be due to different physiological activities of these plants at various stages of development and its interaction with environment condition. These plants had low essential oil content in early growth stages, while the highest essential oil content was obtained at seed setting and seed phase. Less moisture content and lower activity of some essential enzymes which are necessary for biosynthesis of specific compounds at seeding stage lead to less essential F I G U R E 2 Changes in essential oil content (% w/w) of H. persicum at different phenological stages. Essential oil content with different subscripts was significantly different at p < .05 (Tukey's test) Essential oil content (%) oil production at early stages of growth (Hazrati et al., 2019;Şanli et al., 2016;Soltani Howyzeh et al., 2018). Furthermore, previous studies showed essential oil content was not the same in different plant organs. Results of the study on H. persicum showed 0.4 to 5.2% difference in essential oil content of various organs such as stem and seeds (Sefidkon et al., 2004). In our experiment, we observed a low content of the essential oil in vegetative and flowering stages and high essential oil content in seeds at setting stage. According to this, we can conclude that possibly, H. persicum plants spend most of their produced photosynthetic materials to create vegetative organs instead of synthesis of useful biological active compounds at the early stages of its growth in comparison to later stages.
The study of essential oil constituents indicated that α-pinene and β-pinene were predominant in the early growth stages. According to the results (Table 1), the highest α-pinene value was observed at the full flowering stage (4.9%), but we did not see the same results at the immature seed stage. Moreover, β-pinene was detected in the vegetative stage and at the time of flower opening. With entering into the seed set stage, values of these two compounds decreased significantly.
Limonene was observed only in the vegetative and floral budding stages, and the highest value was 18.1% that obtained in the vegetative stage. In line with these results, other researchers reported the existence of this compound in the vegetative organs of H. persicum (Sefidkon et al., 2004), while it was rarely seen in the seeds. In the few cases that limonene was observed in seeds, the value was not significant.

F I G U R E 3
Changes in extract content (% w/w) of H. persicum at different phenological stages. Essential oil content with different letters was significantly different at p < .05 (Tukey's test) Carene 1,027 No.  Among all the compounds, hexyl 2-methylbutyrate, octyl isobutyrate, and 4-methylpentyl isobutyrate contents were increased with increase in plant age, so that at the end of the phonological stages, they were the dominant compounds.

RI
The essential oils were classified according to their chemical formula to eight groups (Figure 4). Based on the results, the main percentage of the essential oil was aliphatic esters, which value was 3.4% in the vegetative stage and finally hit 83.7% in the immature seeds stage or in the mid-mature seeds stage. Therefore, by increas- In the present study, most of these compounds were observed in the flowering stage and prior to it, and when the plant entered into the seed set stage, the value of these compounds was significantly reduced (Figure 4). The most important compounds in this group were limonene, β-pinene, α-pinene, and γ-terpinene. Other studies also confirmed these compounds as predominant compounds in the H. persicum (Mojab & Nickavar, 2003;Radjabian et al., 2014;Sefidkon et al., 2004). Similar to the present study, other researchers reported high amount of these compounds in flowers and leaves, in comparison to seeds (Mojab & Nickavar, 2003;Sefidkon et al., 2004;). Other spatulenol, and β-bisabolene which also reported as predominant compounds in the H. persicum (Firuzi et al., 2010;Moshafi et al., 2009) in other studies.
The other compound evaluated in this experiment was phenylpropenes, which amount at the full flowering stage was 23.1%, and after passing this phenological stage it was significantly reduced ( Figure 4). Anethole and myristicin were two important compounds related to phenylpropenes group. In a study on H. persicum plants, researchers found three phenylpropenes compounds (anethole, myristicin, and estragole) that were usually predominant in the vegetative and flower organs that match the results of the present study (Firuzi et al., 2010;Radjabian et al., 2013;Sefidkon et al., 2004).

Changes in essential oil compounds in different phenological
stages of H. persicum growth are probably due to the fact that the production of essential oil and aromatic compounds are under the control of interactions between physiological, biochemical, and metabolic mechanisms which they depend on age and growth stages of plant. Besides, these changes related to terpene biosynthesis as well as its accumulation in the secretory organs (Németh, 2005). The change in the essential oil ingredients is also influenced by factors such as the age and the development stage of medicinal plants.

| Phenolic acids in different phenological stages
Phenolic compounds such as phenolic acids are one of the most critical compounds in medicinal plants and they have paramount importance due to their high biological activity and function as antioxidants, anti-inflammatory, anticancer, and anti-Alzheimer (Goleniowski et al., 2013;Saibabu et al., 2015).
In this study, we investigated the phenolic acids content in different phenological stages of H. persicum. According to the obtained results, there were 8, 9, 9, 10, ,8 and 7 different phenolic compounds at the vegetative stages, floral budding stage, full flowering stage, early development of seeds stage, mid-mature seeds stage and, final stage or late-mature/ripe seeds stage, respectively ( Table 2).
The methanolic extract gathered from the floral budding stage had the highest amount of phenolic acids (287.6 mg/g dried extract) following by early seed development (259.7 mg/g dried extract), full flowering stage (153.9 mg/g dried extract), vegetative stage (72.5 dried mg/g extract), and mid-mature seed stage (46.2 mg/g dried extract). On the other hand, the mature seed extract contained the minimum amount of phenolic acids (24.8 mg/g dried extract) ( 6.3 ± 0.6c 0.7 ± 0.0b
Cinnamic acid was recorded as the predominant compounds of phenolic acid in all growth stages. However, its maximum value (225.3 mg/g extract) was in the early stages of flowering, and its lowest value (8.0 mg/g extract) was in the mid-mature stage.
Cinnamic acid is a natural aromatic phenolic acid whose long-term consumption is associated with low toxicity to humans and is used in flavorings, artificial color, and some specific medications. The frequent use of cinnamic acid is as a precursor for the methyl cinnamate, ethyl cinnamate, and benzyl cinnamate production which have perfume industrial use. It is also a precursor for the artificial sweetener aspartame. Cinnamic acid is one of the phenolic acids with several biological and medicinal properties, as well as other economical and industrial values (Bradley et al., 2015;Goleniowski et al., 2013). According to our results, H. persicum is considered as one of the plants with a rich source of cinnamic acid, especially in the floral budding stage and early development of seeds stage; however, as the plant entered the mature seed stage, we observed the significant reduction in its value.
P-coumaric acid was another beneficial predominant phenolic acid we explored in this experiment. P-coumaric acid amount was increased during the plant growth to 39.2 mg/g at the full flowering stage and then decreased to the lowest amount in the seed maturing stage (1. 7 mg/g extract). Studies showed p-coumaric acid has antioxidant properties, which can reduce the stomach cancer by suppressing nitrous amines, and has anti-tumor and anti-mutagenesis activities (Goleniowski et al., 2013;Taofiq et al., 2017).
Ferulic acid is one of the other critical phenolic compounds that researchers have been proven its antioxidant properties (Goleniowski et al., 2013). This compound was found at all growth Rosmarinic acid is an ester of caffeic acid called 3,4-dihydroxy phenyl lactic acid (Goleniowski et al., 2013;Lamaison et al., 1991). This compound has many pharmaceutical properties such as antimicrobial, anti-rheumatism, and anticancer and was present in all phenological stages of the H. persicum. The highest (13.3 mg/g extract) and the lowest amounts (3.6 mg/g extract) of this compound were obtained in the vegetative stage and the immature seed stage, respectively.
Therefore, it can be concluded that H. persicum can be considered as a plant rich in phenolic acids, which content was different in different phenological stages. Phenolic compounds were at the lowest rate in the floral budding stage while the maximum content was recorded at the seed mature stage. An increase in phenolic acid levels during flowering stage recommended a higher expression level of the phenylalanine lyase enzyme (Andreotti et al., 2006;Feduraev et al., 2019), and it is also an indication of enzyme activity reduction with plant maturation. These changes that occur in the process of primary metabolites adsorption are a result of starch synthesis in the middle stages of seed maturation, and this phenomenon can affect the biosynthesis of phenolic acids (Ma et al., 2016).

| Cluster analysis and principal component analysis (PCA)
Another goal of this study was to monitor the differences and similarities between different phenological stages in order to find the consequences of different harvesting time on identified phytochemical compounds in H. persicum. The results of the main components analyses are shown in Table 3. Based on these results, three components had highest eigenvalue, reporting 89.45% of the total variance. The relative variance for the first, second, and third components was 55.52%, 23.28%, and 10.65%, respectively. In the first component, the compounds of butyl butyrate, carene, 4-methylpentyl isobutyrate, hexyl butyrate, cis-5-octen-1-ol, octyl acetate, hexyl 2-methylbutyrate, octyl isobutyrate, β-bisabolene, 1-tetradecicanol, p-hydroxybenzoic acid, rosmarinic acid, and essential oil content had the highest loading factor. Differently, in the second and third components, m-coumaric acid and gallic acid compounds had the most top loading factor, respectively (Table 3).
Since 78.80% of the total variance was allocated to the first and second components, they were used for the biplot diagram obtained from the PCA. Different phenological stages were divided into three distinct groups which shown in the Figure 5. The biplot analysis showed the vegetative stage (S1), based on the first and second components alone and with high distance with other phenological stages.
In a study to investigate the phenological effects on phytochemical compounds at different populations of the Tithonia diversifolia, the cluster and principal component analyses were used and the results showed that phytochemical compounds were differed in various phenological stages (Pretti et al., 2018). ter analyses can be used to determine the optimum phenological stage and to achieve the best yield of phenolic acids and essential oil compounds from plants (Pretti et al., 2018). This research illustrated in the vegetative phase and in the beginning of the generative phase; the rate of phenolic acids was high. The current research confirmed that the different phenological stages had affected the phenolic acids contents and essential oil compounds of H. persicum. However, the mechanisms in which these changes have been made will require further studies. be concluded that to reach the maximum essential oil content and composition of aliphatic esters compounds, harvest at the seed set stage, especially in the mid-mature seeds stage, is desirable. In addition, harvesting the H. persicum at the flowering stage, especially at the floral budding stage, leads to maximum content of extract yield and phenolic acid compounds.

| E THI C AL RE VIE W
This study does not involve any human or animal testing.

| INFORMED CON S ENT
Written informed consent was obtained from all study participants.

ACK N OWLED G M ENTS
The support of university of Azarbaijan Shahid Madani is greatly appreciated.

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Research data are not shared.