Drying the leaves of Perilla frutescens increases their content of anticancer nutraceuticals

Abstract A regular intake of plant‐derived bioactive agents has gained popularity because of the health benefits. Fresh leafy greens, however, normally have a low concentration of such bioactive agents. In this study, we found that drying markedly affected the accumulation of secondary metabolites and that dried leaves of Perilla frutescens L. (perilla) contained more anticancer flavonoids than fresh leaves. Drying is a major method of food preparation, particularly for plant‐based foods, but the quality of the bioactive agents contained in the fresh and dried leaves of perilla has received only scant attention. Quantitative analysis of the concentrations of perillaldehyde, rosmarinic acid, apigenin, luteolin, 4‐hydroxyphenyllactic acid, and 4‐coumaric acid, some of which are known as nutraceuticals, revealed that the effect of drying significantly increased apigenin (28‐fold) and luteolin (86‐fold), but decreased rosmarinic acid in all leaf stages. We examined the positive effect on flavonoid levels on perilla leaves and confirmed that, by comparison with fresh perilla leaves, the dried leaves contained greater concentrations of anticancer flavonoids regardless of variety, form, or manner of cultivation. This indicates that drying can significantly increase the level of flavonoids in perilla leaves without a loss of flavor. Therefore, drying is a simple and effective method to improve the concentrations of bioactive agents, which increases the intake of beneficial substances derived from herbs and edible plants. This finding serves as a method for the supply of raw plant materials rich in bioactive agents that are suitable for labeling as edible nutraceuticals.

deliver a concentrated form of bioactive agents from a food, presented in a nonfood matrix, and used to enhance health in dosages that exceed those that could be obtained in normal foods. In his proposal, the culinary herbs, such as those used in designer foods, should not be recognized as nutraceuticals, diet supplements, or functional foods. Many bioactive agents, however, have been identified in the herbs and edible plants that are considered designer foods. We suppose that plant materials rich in bioactive agents should be recognized as nutraceuticals.
Perilla leaves contain many important bioactive agents and secondary metabolites such as perillaldehyde (PA), rosmarinic acid (RA), apigenin (AG), and luteolin (LT). PA is an original monoterpene and a major compound of the essential oil of perilla (Ito, Toyoda, & Honda, 1999). PA has a unique and faintly sweet flavor and imparts antimicrobial and antidepressant activities (Igarashi & Miyazaki, 2013;Ito, Nagai, Oikawa, Yamada, & Hanawa, 2011). The content of PA in perilla leaves for clinical use is defined in the Japanese pharmacopoeia (Ministry of Health, Labour, & Welfare of Japan, 2017). RA is a phenylpropanoid of polyphenols that is abundant in Lamiaceae's plants such as savory, thyme, oregano, mint, and lemon balm (Petersen, 2013;Vladimir-Knežević et al., 2014). RA possesses anti-allergic, anti-inflammatory, and antioxidant activities and has recently been identified as a potential treatment for Alzheimer's disease (Hamaguchi, Ono, Murase, & Yamada, 2009;Ono et al., 2012;Osakabe et al., 2004;Takano et al., 2004). The content of RA in perilla stems and fruits for clinical use is defined in the Chinese Pharmacopoeia (2015). AG and LT are major plant flavonoids and are widely found in many medicinal plants, vegetables, and fruits. AG and LT possess antioxidant, anti-inflammatory, and anticarcinogenic activities. AG has been reported for DNA protective activity against UV-B-induced DNA damage in skin cells and mice (Das, Das, Paul, Samadder, & Khuda-Bukhsh, 2013;George, Dellaire, & Rupasinghe, 2017). LT has been reported for its apoptosis potential in many types of cancer cells and in multidrug-resistant cancer cells by inducing cell-cycle arrest and apoptosis via intrinsic signaling pathways (George et al., 2017;Rao, Satelli, Moridani, Jenkins, & Rao, 2012).
Drying is a major method of food preparation, particularly for plant-based foods, but it also has a large impact on the secondary metabolism of harvested plants. When applied during the flowering stages of basil leaves (Mandoulakani, Eyvazpour, & Ghadimzadeh, 2017) and during the developmental stage in grape berries, drought stress results in increased levels of gene expression and compounds involved in the phenylpropanoid pathways (Savoi et al., 2016). Our hypothesis is that the secondary metabolisms of these plants, as well as their secondary metabolites, respond to the drying process in varied ways. Drying methods have been evaluated in various ways to minimize the loss of nutrition from fresh plant materials during processing. We realized, however, that few studies have investigated the effect that drying the leaves of perilla can exert on the content of potent bioactive agents that include PA, RA, AG, and LT, although these are known to be responsible for the plant's medicinal efficacy and for its ability to enhance health and fitness. As previously mentioned, the medicinal substances in perilla have the potential to prevent and treat disease, which has recently made them popular nutraceuticals in Japan. These facts prompted the questions of whether fresh or dried leaves would be more suitable as raw material for nutraceuticals in order to avoid the loss of secondary metabolites, as well as what would be the advantages of dried leaves other than long-term preservation, and, finally, how drying affects the content of particular substances in perilla leaves.
This article describes the effects that the drying of perilla leaves exert on the bioactive agents PA, RA, AG, and LT with a focus on their metabolic changes during each leaf stage. PA and other compounds in perilla essential oil are volatile and have a low boiling point, which means they are easily evaporated under conditions of reduced pressure or increased temperature. Therefore, drying was conducted in a laboratory under atmospheric pressure and air conditioning (<25°C) without light. Quantitative analyses of bioactive agents were conducted via high-performance liquid chromatography (HPLC) using a photodiode array (PDA) and liquid chromatographymass spectrometry (LC-MS). In addition, 4-hydroxyphenyllactic acid (4HL) and 4-coumaric acid (4CA) were also subjected to quantitative analysis in order to qualify the biosynthetic activity toward RA, AG, and LT. Herein, we describe the extent to which bioactive agents were changed in the perilla herb during a simple drying process and discuss how the drying may influence the secondary metabolism in each plant. This information should be valuable for researchers working in the development of plant-based foods and drugs.

| Cultivation and sampling
Green perilla (P. frutescens var. crispa f. viridis, Takii & Co., Ltd., Kyoto, Japan) seeds were sown on soils generally used for the growth of vegetables (Kanuma Kosan Co. Ltd., Tochigi, Japan) in plastic pots placed in a cultivation room. The light intensity was set at approximately 80 µmol m −2 s −1 (the unit of µmol m −2 s −1 represents photosynthetic photon flux density) with a photoperiod of 16 hr per day provided by cool white fluorescent lamps. Water was provided once per week and was supplemented with a 1/200 dilution of HYPONeX (HYPONeX JAPAN Co. Ltd., Osaka, Japan) once every two weeks. Air temperature was set at 23°C/18°C during light/dark periods. After germination, 10 seedlings were cultivated for 12 weeks and grew to plants that each had eight pairs of leaves. Among them, four pairs of true leaves (the 4th-7th) from the middle regions of each plant were selected as samples and subjected to analytical experiments. All leaf samples were harvested at the same time with great care to avoid any wounding of the leaves, and the fresh weight of each leaf was immediately measured. The two leaves of a single pair will be the same age and at the same stage of growth, which means the appearances of plant growth and metabolism will be approximate. Then, one leaf from each pair was dried and the other remained fresh to allow for accurate comparisons.

| Leaf stage
With the exception of cotyledons, the seven pairs of true leaves were numbered from the ground to the top as the first-to-seventh stages of leaf development (leaf stage). Cotyledons were eliminated from the experiments because they differed in their plant growth stages. Four pairs of true leaves (the 4th-7th) from the middle regions of each plant were selected as samples and subjected to analytical experiments. One leaf of each pair of true leaves was used as a fresh leaf sample for analysis and the other leaf was used for the preparation of a dried leaf sample.

| Fresh leaf samples
A fresh leaf was placed into a polypropylene tube with a metalcorn, and the tube was then frozen in liquid nitrogen and ground using a Multibeads Shocker (Yasui Kikai Co. Ltd., Osaka, Japan). The metalcorns were removed, and the crushed leaf samples were stored at −80°C until they were subjected to analysis along with the dried leaf samples.

| Dried leaf samples
Dried leaf samples consisted of the whole leaf. Drying was accomplished by spreading the leaves in darkness under atmospheric pressure at ambient temperature that ranged from 18 to 23°C for a period of three weeks. Subsequently, the weight of the dried leaves was measured. Dried leaf samples were pulverized using the µT-01 Beads Crusher (Taitec, Saitama, Japan), and the steel beads were removed before the next extraction step.

| Extraction
Extraction was conducted according to the Japanese Pharmacopoeia (Ministry of Health, Labour, & Welfare of Japan, 2017) and a method described in a previous article (Lu et al., 2017) with modifications.
A sample, ca 20 mg of a dry leaf and ca 150 mg of a fresh leaf, was weighed accurately and transferred to a 1.5-ml tube. Methanol (1 ml) was added, mixed for 10 min at 2,000 rpm and 15°C using an Eppendorf ThermoMixer (Hamburg, Germany), and centrifuged for 5 min. To the residue, methanol (1 ml × 2) was added, and the same extract manner was performed twice. The extracts (about 3 ml) were combined and transferred to a 5-ml volumetric flask and diluted with methanol to a 5 ml total volume. The solution was filtered through a 0.22-µm nylon syringe filter (Shimadzu GLC Ltd., Tokyo, Japan) to prepare the samples for HPLC or LC-MS.

| PA and RA concentration
High-performance liquid chromatography analysis of PA and RA was conducted according to methods described in previous literature with modifications (Jirovský et al., 2007;Lu et al., 2017;Natsume, Muto, Fukuda, Tokunaga, & Osakabe, 2006;Öztürk, Duru, İnce, Harmandar, & Topҫti, 2010;Sevindik et al., 2015;Vladimir-Knežević et al., 2014). HPLC was performed on a Shimadzu LC- The PA or RA content per fresh leaf weight (hereafter, PA or RA concentration) was estimated by dividing the PA or RA content in the samples by the weight of the fresh sample. Therefore, for dried samples, the PA or RA concentrations were estimated by dividing the PA or RA content in dried samples by the weight of the fresh sample before drying.

| AG, LT, 4HL, and 4CA concentrations
Liquid chromatography-mass spectrometry analysis of AG and LT was conducted according to a method described in previous literature (Lu, Takagaki, Yamori, & Kagawa, 2018) with modifications. A LCMS-2020 mass spectrometer (MS) equipped with an electrospray ionization (ESI) source operating in negative mode was used for identification and quantification of AG, LT, 4HL, and 4CA by chromatographic data processed using LabSolutions software (Shimadzu, Kyoto, Japan

| Statistical analysis
Data were subjected to independent sample testing using SPSS 25.0 software (IBM Japan, Tokyo, Japan). The Levene's test was conducted for equality of variances. Significant differences between the means of fresh and dried leaves were calculated for each leaf stage using Student's t test (p < 0.05). The test was conducted to determine the significance of drying. All experiments were repeated twice for five replications.
TA B L E 1 Concentrations of compounds extracted from fresh and dried leaves of perilla  For each leaf stage, significant differences between fresh and dried leaves are indicated by *p < 0.05, **p < 0.01, ***p < 0.001, according to Student's t test

| Total concentrations of PA, RA, 4HL, 4CA, AG, and LT
The data reported in this study are based on a wet basis wherein concentrations (µg g −1 /mg g −1 ) are expressed as the content per unit of fresh (pre-dry) leaf weight. The fresh leave samples contained a higher water content at 74 ± 1.7% (the mean ± standard error, n = 20). The concentration of each dried leaf sample was estimated by dividing the content in the dried leaf sample by the weight of the pre-dry leaf sample that was measured before drying. Table 1 shows the changes in the concentrations of PA, RA, 4HL, 4CA, AG, and LT in every leaf sample to allow comparisons between the fresh and dried leaves. The concentration of PA slightly decreased after drying (−6%), but the reduction was not significant. The concentration of RA largely decreased after drying (−81%, p < 0.001).
The concentration of 4CA, however, showed a marked increase after drying (fourfold, p < 0.001). The concentration of AG, however, was significantly increased after drying (16-fold, p < 0.001). The effect on LT was also significant with a 109-fold increase in concentration in the dried leaves (p < 0.001).

| Local concentrations of PA, RA, 4HL, 4CA, AG, and LT based on leaf stage
The

| States of variety, form, cultivation, and condition
In order to generalize these results concerning AG and LT accumulation, we shopped markets to collect commercially available perilla leaves of several varieties from different methods of cultivation to conduct further surveys. These additional results appear in Table 2. The perilla materials listed in Table 2 were composed of five varieties, one green form and four red forms, and were cultivated in four manners that differed from those described in Section 2.1.1 (Supporting information Table   S1). These had been grown in indoor cultivation systems or outdoor at various farms. The fresh leaves were obtained 3-5 days after harvest and were then analyzed. The fresh leaves were dried by repeating the methods described in 2.1.4. The concentration of AG increased significantly following drying (54-fold, p < 0.001), and the concentration of LT had increased in the dried leaves (37-fold, p < 0.001; Table 2).

| D ISCUSS I ON
Studies on a molecular basis have promoted an understanding of the biosynthetic pathways of RA, AG, and LT. The enzymes involved in the biosynthesis of RA from amino acids were unraveled in suspension cultures of Anchusa officinalis (Boraginaceae) and Coleus blumei (Lamiaceae) by Petersen and Simmonds, (2003) and Petersen (2013).
Flavonoid biosynthesis, which includes the core pathways, has been studied in different plants (Bashandy et al., 2015;Dao, Linthorst, & Verpoorte, 2011). Moreover, the entire transcriptome map of perilla leaves has been reported and the expression of genes involved in the biosynthetic pathways of flavonoids and phenylpropanoids has been clarified (Fukushima, Nakamura, Suzuki, Saito, & Yamazaki, 2015).
Thus, the proposed biosynthetic pathways in perilla are shown in
As shown in Table 1 (Okuda, Hatano, Isao, & Nishibe, 1986). In the present study, however, the concentration of PA was not significantly reduced after drying, as shown in Table 1 and in Figure 1, so that this level of heat is apparently suitable for chemicals such as PA that are sensitive to air, light, and temperature. Therefore, the significant decrease in RA accumulation following the drying process was due to plant metabolism rather than to decomposition by chemical reaction.
The reduction in RA accumulation in cells meant that RA consumption had overcome RA formation during drying. Otherwise, the RA level in the dried leaves would have remained at the same level of RA in the fresh leaves. We theorized that perilla had used RA as a protectant against the stresses and damages of the drying process (Döring & Petersen, 2014).  Table 1 and the flavone majority had switched from AG to LT. The ratio of the two compounds (AG/LT) in fresh leaves was 3.2 ± 0.58, but the ratio was switched to 0.42 ± 0.039 in dried leaves with a significant decrease in the amounts (p < 0.001). LT is formed from either an equivalent of AG or its corresponding precursor flavanone, naringenin, by hydroxylation at the 3′ position of the aromatic ring by the enzyme F3′H ( Figure 2). This switch in flavone accumulation implied that the molecular conversion into LT by F3′H was enhanced as the activity of F3′H was improved during drying.
To support the observation related to the activation of the biosynthetic pathway, we attempted to determine the expression levels of the enzymes in the pathway by measuring the RNA transcript levels. However, we observed a decomposition of the RNAs in the dried leaf cells and could not compare the expression levels between fresh and dried leaves. Therefore, the enzyme activity could not be accurately determined through the expression of proteins extracted from dried leaf cells because of damage sustained during drying.
We compared the data for AG and LT accumulation shown in Tables

| CON CLUS IONS
A comparison of the concentrations of PA, RA, AG, and LT in fresh and dried leaves revealed that after drying, AG and LT were increased 28-and 86-fold, respectively, with a 94% decrease in RA.
PA was not changed significantly. This proved that drying has a significant impact on phenylpropanoid and flavonoid biosynthesis and accumulation in perilla leaves. We also confirmed that the drying could promote AG and LT accumulation in perilla leaves regardless of variety, form, leaf stage, or manner of cultivation. This finding underscores the use of drying as a method for food processing that can provide raw materials rich in anticancer flavones that are suitable for use as nutraceuticals that contain P. frutescens.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.

DATA S H A R I N G A N D DATA ACCE SS I B I LIT Y
The data that support the findings of this study are openly available in the online version of this article at https://doi.org/10.1002/ fsn3.993

E TH I C A L R E V I E W
This study does not involve any human or animal testing.

I N FO R M E D CO N S E NT
Written informed consent was obtained from all study participants.