Introduction

Phytochemicals are naturally occurring compounds synthesized by plants in response to their surrounding environment for growth and survival1. These compounds accumulate in various parts of the plant, including the roots, stems, leaves, flowers, fruits, fruit peels, and seeds2. Beyond protecting plants from biotic and abiotic stresses that cause disease, phytochemicals also exhibit significant pharmaceutical potential, making them valuable for various applications. In particular, they serve as essential sources of bioactive compounds for pharmaceutical, cosmetic, and functional food industries3,4,5. Herbal supplements are widely recognized as a complementary approach to modern medicine and have demonstrated sufficient effectiveness in pharmaceutical applications6. Currently, environmental changes, pandemics, lifestyle shifts, and nutritional factors continue to impact human health. Consequently, herbal products have gained popularity due to their rich content of bioactive molecules that contribute to health benefits, particularly in enhancing the immune system6,7.

The human immune system is a vital and intricate system that defends against various microorganisms and also functions as an immunomodulator8. Different external factors, immunizations, and prior infections can trigger the immune response in the body. The immune response is mediated by specific components of the immune system, including cytokines, acute phase proteins, macrophages, monocytes, complement proteins, and neutrophils9. Additionally, the host’s immune system is capable of distinguishing between its own cells and proteins and foreign substances. A healthy immune system ensures the maintenance of homeostasis within the body7,8,9.

Immunomodulation amplifies any modification of the immune response involving activation, amplification and expression as well as suppression of a part or phase of the immune response triggered by an immunomodulator, which is an endogenous or exogenous factors to optimize the immune response8,10. In clinical practice, immunomodulator is classified into three categories including immunostimulants, immunosuppressants, and immunoadjuvants9,10,11. Phytochemicals act as an immunostimulants, which are a non-specific promoters that function through both innate and adaptive immune responses to be immunotherapeutic agents to increase the basic level of the immune response11. Thereby, finding plants containing bioactive compounds are challenging to know a novel source of plant that they are potential to be immunomodulators to boost or dysfunction the immune response. Additionally, there are some documents reporting plants derived bioactive compounds, functioning as immunomodulator. Resveratrol, a natural polyphenol extracted from Polygonum cuspidatum, can decrease the beta-amyloid accumulation in the hippocampus by lowering the interleukin‐1 beta (IL‐1β) in BV‐2 murine microglial cells to show activity on neuroprotection12. Curcumin‐a compound derived from Curcuma longa‐inhibits IL‐1, IL‐6 and tumor necrosis factor (TNF) and suppresses oxidative stress to prevent the progression of SARS‐CoV‐2 infection to pneumonia and acute respiratory distress syndrome (ARDS)13. In addition, Lupeol‐a pentacyclic triterpene, commonly found in edible fruit and vegetables. It can modulate IL‐2 expression to inhibit NF‐κB and Akt‐dependent signalling, which can lead to its several anticancer and anti‐inflammatory activities8. However, scientific information on the immunomodulatory activity of Barleria species remains limited.

Barleria species (Acanthaceae) have long been utilized as traditional medicines based on their pharmacognostic activities to treat toothache, cough, fever, bronchitis, diabetes, asthma, insect bites and inflammations14,15,16. There are approximately 300 Barleria species worldwide, with only eight species reported in Thailand: B. biloba J.B. Imlay, B. cristata L., B. longiflora L., B. lupulina Lindl., B. prionitis L., B. repens Nees, B. siamensis Craib, and B. strigosa Willd17. In addition to their use in traditional medicine, these plants also have economic value. Some species, such as B. lupulina, B. prionitis, and B. repens, are cultivated as ornamental plants17. Moreover, B. siamensis, one of these species, is endemic to Thailand and rare population. In the past couple of years in Thailand, there have been anecdotal reports on the consumption of boiled Barleria leaves, including B. lupulina, B. prionitis, and B. strigosa, as a potential treatment for cancer patients by local herbalists, with patients reportedly showing positive responses. However, numerous phytochemical constituents of Barleria species have been widely reported, including flavonoids, saponins, sterols, tannins, terpenoids, neohesperidoside, balarenone, barlerinoside, phenolics, steroids, phytosterols, iridoids, and lupulinoside15,16, which are associated with antioxidant, anti-inflammatory, and anticancer activities16,18. Nevertheless, the role of Barleria species and its related compounds in anticancer activity remains unclear, and there is no academic evidence to support this claim18. Based on the above context, investigating the phytochemical constituents and their biological activities, particularly in terms of the immunomodulatory function and anticancer potential of Barleria species, is essential. This research would help bridge existing knowledge gaps, validate anecdotal claims, and contribute to a more comprehensive understanding of these plants through scientific study.

Additionally, the interaction between phytochemicals and the monooxygenase enzyme family known as cytochrome P450 (CYP450) enzymes is noteworthy for studying their effects on drug metabolism, which influences bioavailability and therapeutic efficacy19,20. However, there is still a lack of data on the effects of Barleria extracts on cytochrome P450 activity. This study provides additional insights into the role of Barleria species in regulating CYP450 gene expression.

In this study, we report the detection of barlerin and verbascoside, both of which are present in six Barleria extracts, including B. cristata, B. lupulina, B. prionitis, B. repens, B. siamensis, and B. strigosa. Additionally, we investigate their biological activities, including toxicity, immunomodulatory effects, and their role as cytokine gene expression promoters to activate peripheral blood mononuclear cells (PBMCs) against prostate cancer (PC-3) cells, as well as their function in regulating CYP450 gene expression. These findings significantly enhance the scientific understanding of Barleria species in terms of bioactive compound profiling and biological activities, contributing to their potential applications in human health.

Results

Untargeted metabolite investigation of Barleria extracts

The leaves of Barleria species, including B. cristata, B. lupulina, B. prionitis, B. repens, B. siamensis and B. strigosa (Fig. 1A), were extracted using 60% ethanol. Their derived crude extracts were screened to detect the untargeted metabolites, achieved by working with GC-MS. The derived untargeted metabolite data was further deconvoluted and then calculated to obtain the percentage of relative content of compound to identify the key active compound of each species. The targeted metabolites‐a part of a key active compound was quantified to attain the authenticity of concentration and amount of those metabolites, calculated based on a standard calibration equation (Fig. 1B) to ascribe a net volume of targeted metabolites in plant extract. To classify the untargeted metabolite of Barleria extracts based on the classic PCA method, the result demonstrates that the Barleria metabolite data of each species is pretty consistency with classification according to PCA score, PC1 (27.9%) and PC2 (24.1%), to evaluate the infallibility of the metabolite data and define the class separation. The B. prionitis and B. repens metabolite data were closely plotted (Fig. 1C). While the metabolite data of Barleria species was analyzed to cluster by hierarchical cluster analysis (HCA). The result indicated that there are two clades including a clade of B. prionitis and B. repens, and another clade containing B. cristata, B. lupulina, B. siamensis and B. strigosa, which is corresponded to derived PCA result (Fig. 1D). To search for the key compound of each Barleria species based on showing the highest percentage of relative content of those compounds, there are peculiarly containing key compounds in each species, referred to as indicating the highest percentage of relative content. Verbascoside (12%), cyclopenta[c]pyran‐4‐carboxylic acid, 1‐(beta‐D‐glucopyranosyloxy)‐1,4a,5,6,7,7a‐hexahydro‐5,6‐dihydroxy‐7‐methyl‐, 2‐(4‐hydroxyphenyl)ethyl ester (20.62%), methyl palmitoleate (15.29%), murolic acid (15.91%), olean‐12‐en‐28‐oic acid, 2,3,19‐trihydroxy‐, (2alpha,3beta,5xi,9xi,19alpha) (15.90%) and stigmasterol (24.37%) were represented to be a major key bioactive compound in B. cristata, B. lupulina, B. prionitis, B. repens, B. siamensis and B. strigosa extracts, respectively (Table S1). When the metabolite data of all Barleria species were intensively sorted to probe for common metabolites detected in all species, the result indicates that there are four common metabolites including γ‐tocopherol, barlerin, stigmasterol and verbascoside, found in all Barleria extracts (Fig. 1D and E). However, the barlerin and verbascoside showed high percentage relative content in all species more than the other two compounds. Hence, these compounds were selected for further study as a targeted metabolite. Additionally, the Barleria species also contain their own exclusive metabolite in a total of 25, 16, 10, 29, 20 and 4 metabolites in B. cristata, B. lupulina, B. prionitis, B. repens, B. siamensis and B. strigosa extracts, respectively (Fig. 1E; Table S1).

Incidentally, to categorize the Barleria metabolite into compound class level, the various metabolites of Barleria species belonged to flavonoid and terpenoid classes, and these compound classes highly distributed in B. cristata (Fig. 1F). Meanwhile, the anthraquinone and lignan glycoside class are merely abundant in B. cristata and B. siamensis, as well as long-chain fatty acid is mostly abundant in B. prionitis and B. repens. There are two species including B. siamensis and B. strigosa contained metabolite in xanthenes and xanthones class and only one species of Barleria as B. lupulina contained metabolite in diterpene glycosides class (Fig. 1F).

Barlerin and verbascoside as a targeted metabolite contained in Barleria extracts

After the untargeted metabolites in Barleria extracts were completely analyzed, barlerin and verbascoside were subsequently studied for targeted analysis. Their presence in the extracts was confirmed using authentic reference standards. To examine the barlerin and verbascoside in Barleria extracts, there were detecting both compounds in all extracts. The barlerin was detected at retention time (Rt) ranging between 17.54 and 18.03 min, while verbascoside showed a Rt signal at 22 min roughly (Fig. 1G). To quantify the actual amount of both compounds executing with their calibration equation. The derived standard calibration equations revealed their correlation factor (R2) as 0.9950 and 0.9931 in barlerin and verbascoside standard curve to show sufficiently confident stability of the equation to use for compound quantification. Remarkably, the B. siamensis contained high amounts of both compounds as 0.43 (for barlerin) and 1.02 (for verbascoside) mg/g dried leaf, while B. lupulina and B. prionitis showed low amount of barlerin as 0.03 mg/g dried leaf, in addition to B. prionitis and B. repens revealed low amount of verbascoside as 0.01 mg/g dried leaf (Fig. 1G).

Fig. 1
figure 1

Untargeted and targeted metabolite data of Barleria extracts. (A) the morphological traits of studied Barleria species; (B) a schematic to proceed metabolite analysis; (C) PCA plot based on untargeted metabolite data to classify Barleria metabolites of each species; (D) a hierarchical cluster analysis showing key compounds; (E) a venn diagram containing the number of common and exclusive metabolite of each Barleria species; (F) class of compound classification of the Barleria metabolites representing by Sanky diagram; (G) Barlerin and verbascoside detection and quantification in Barleria extracts showing concentration (mg/mL) and amounts of their compound in mg/g dried leaf of each species.

Full size image

Barleria extracts and their toxicity effect to normal and cancer cells

Normal human immune PBMCs, Primary Prostate Epithelium Cells: Normal, Human (HPrEC) and prostate cancer PC-3 cells were treated for 48 h with five working concentration levels of the Barleria extracts to evaluate the cytotoxic and genotoxic effects of the extracts against these cells, assessed using MTT and comet assays due to the Barleria extracts containing barlerin and verbascoside in different amount, these single compounds were employed to treat with the mentioned cells to investigate the toxicity activities of the compounds as well. The cytotoxic effect was compared to cisplatin‐chemotherapeutic drug, acted as a positive control. As results, some of the treatments did not show cytotoxic effect on the two normal cells but affected cancer cells, except for B. lupulina extract (Fig. 2A).

At concentrations of 50 and 100 µg/mL, B. lupulina extract significantly reduced the viability of PBMCs and HPrECs to below 50%, with calculated IC50 values of 69.94 µg/mL and 50.51 µg/mL, respectively. Although the extract also reduced the viability of PC-3 cells to 57.86 ± 1.41% at 100 µg/mL, it exhibited a relatively high IC50 value of 120.57 µg/mL, indicating lower potency and selectivity against cancer cells (Fig. 2B and C; Table S2 and S3). The results indicated that B. lupulina extract showed non-toxic activity in the cells (Fig. 2C). Interestingly, the B. cristata, B. siamensis, barlerin, and verbascoside revealed high percentage of cell viability beyond 65% in PBMCs and HPrEC with extremely high IC50 over the highest concentration used and effectively promoted the cytotoxic activity to PC‐3 with low IC50 values to indicate that the mentioned treatments show selective toxicity to cancer cells (Fig. 2C). Even though, the B. cristata extract represented non‐selective toxicity to PBMCs (Fig. 2C) because the IC50 ratio between PBMCs and PC‐3 lower than 3; however, the percentage cell viability in each concentration of PBMCs showed higher than PC‐3 (Fig. 2B; Table S2). Notably, B. strigosa extract exhibited strong cytotoxic activity against PC-3 cells, reducing their viability to below 20%. However, it also demonstrated significant cytotoxicity toward HPrEC cells at 100 µg/mL, with a cell viability of 45.89 ± 3.40% (Fig. 2B; Table S2). However, the extract revealed selective toxicity to PC‐3 cancer cells (Fig. 2C). All working concentration levels of B. prionitis and B. repens extracts revealed high percentage of cell viability beyond 80% in all studied cells (Fig. 2B and C; Table S2).

Furthermore, the half-maximum lethal dose (LD50) values were evaluated based on IC50 values to classify the class of hazardous level. Almost all the treatments indicated that they are class III slightly or class IV non‐hazardous to normal cells for oral but indicating class II moderately hazardous to cancer cell. For cisplatin, its LD50 value referred in class II moderately hazardous to all types of studied cells (Table S3).

As a cytotoxic result mentioned previously, the following experiments were conducted at IC50 values for each cell. In case of the IC50 values are superior to the highest working concentration used in the experiment, the highest working concentration at 100 µg/mL was selected to use for the following experiments. To study the effect of treatments inducing DNA damage on cells preformed using comet assay after 48 h treatment, the results demonstrated that some of the treatments including B. cristata, B. siamensis, B. strigosa, barlerin and verbascoside showed no genotoxic activity to PBMCs and HPrEC cells but they can induce the DNA damage significantly in PC-3 by revealing evident DNA fragments and high olive tail moment (OTM) values (Fig. 2D). Besides, the concentration at IC50 values of B. lupulina extract and cisplatin induced DNA damage in PBMCs, HPrEC and PC‐3 significantly. The B. prionitis and B. repens extracts have no genotoxic activity in all types of studied cells indicated by OTM values not different to the control (Fig. 2D; Table S4).

Fig. 2
figure 2

The cytotoxicity and genotoxicity results of the treatments including Barleria extracts, barlerin, verbascoside and cisplatin in PBMCs, HPrECs and PC-3 obtained after 48 h of treatment. (A) a heatmap expressing percentage of cell viability of each studied cell normalized with DMSO as a control; n = 3; (B) the plotting of percentage of cell viability versus concentration treated to show the cell viability in each concentration; n = 3; (C) the selective toxicity plotting calculated by IC50 of normal cell versus IC50 of cancer cell; (D) the comet images (200x) of studied cells to reveal their DNA damage by showing OTM values after 48 h treatment, referring to UV treatment‐a positive control in this experiment; n = 150.

Full size image

The effect of Barleria extracts on programmed cell death and cell cycle distribution

Using the same conditions as for genotoxicity testing, PBMCs, HPrEC and PC-3 cells were harvested after 48 h treatment to evaluate necrotic and apoptotic cell death as well as cell cycle distribution by flow cytometry. Almost all of the treatments can induce apoptosis and necrosis in PC‐3 cancer cells with higher percentages compared to normal cells. Notably, treatment with B. siamensis extract at IC50 concentration (29.97 µg/mL) revealed a high percentage of apoptosis at 64% in PC‐3 and is remarkably 2.4 times higher than cisplatin (Fig. 3A; Table S5). In a similar fashion to cytotoxicity and genotoxicity experiments, B. lupulina treatment did not show selectivity towards cancer cells, inducing apoptosis in PBMCs (28%), HPrEC (22%) and PC‐3 (14.2%) for B. lupulina. Additionally, treatment with a concentration at the IC50 (81.85 µg/mL) of B. strigosa extract induced apoptosis in HPrEC (13%). As expected, cisplatin, induced significant cell death in normal cells as well as cancer cells, with a high percentage of apoptosis of 55% in PBMCs, 27% in HPrEC, and 24.85% in PC‐3, respectively (Fig. 3A).

Characterization of cell cycle distribution following treatment with Barleria extracts containing barlerin and verbascoside, showed a similar selective targeting of cancer cells and not normal cells. Almost all the treatments arrested the PC-3 cell cycle in G1‐phase and suppressed progression through S‐phase for DNA synthesis and G2‐M phase for cell division, but did not affect PBMCs and HPrEC. However, treatment with B. prionitis and B. repens extracts did not affect cancer cells. B. lupulina extract and cisplatin arrested all types of studied cells in G1‐phase (Fig. 3B; Table S6).

Fig. 3
figure 3

The programmed cell death evaluation and cell cycle distribution analysis after 48 h treatment in PBMCs, HPrEC and PC-3; n = 3. (A) The two‐dimensional dot plots of single cell plotting between PI and FITC to identify necrosis, late apoptosis, live and early apoptosis suffused in the quadrant Q1, Q2, Q3 and Q4, respectively, the percentages of alive, apoptotic and necrotic cell in each treatment were represented in bar chart; (B) The histogram of cell cycle showing the DNA content peak in subG1, G1, S, and G2‐M phases and their bar charts containing the percentages of DNA content in each treatment.

Full size image

Co-culture toxicity activities of Barleria extracts activated PBMCs against PC‐3 as an immunomodulatory study

To study an immunomodulatory function stimulated by the studied treatments and according to toxicity results of the treatments, the B. cristata, B. siamensis, B. strigosa, barlerin and verbascoside at the concentration of 50 µg/mL are strong candidates to be further selected for immunomodulatory study based on they did not show toxic effects to PBMCs assuredly. Even through the B. cristata extract is clustered as non-selective toxicity to PBMCs (Fig. 2C) but at the mentioned concentration employed, shows high percentage of cell viability (Table S2). The experiment was designed and described in Fig. 4A, in addition to that, the PBMCs were treated with the treatments and DMSO as a control. Furthermore, the tubes of culture media separately contained treatment and DMSO were prepared concurrently to act as vehicle control to confirm that the treatments suspended in the culture media did not interfere with the experiment. The treated PBMCs and culture media containing treatment were incubated for 2 conditions at 24 and 48 h to establish the activated PBMCs and further study the cytotoxicity of activated PBMCs against PC‐3 for 24 and 48 h co‐culture (Fig. 4A). The wells lacked PC‐3 in culture, contained culture media with DMSO and activated PBMCs with DMSO, were used as a blank to confirm that the suspending activated PBMCs were completely removed out of the well to prevent experimental interference (Fig. 4A). The results exposed that activated PBMCs for 24 h did not show cytotoxic activity against PC‐3 in both 24 and 48 h co‐culture and the suspending activated PBMCs were completely discarded by showing similar absorbance values in blank wells (Table 1). Interestingly, the activated PBMCs for 48 h disclosed the cytotoxic activity against PC‐3 significantly and this activity was not detected in treatment suspended in culture media conditions, indicating that the suspending treatment in culture media to generate activated PBMCs did not impair the experiment (Fig. 4B; Table 1). The activated PBMCs with B. siamensis extract were remarkably exposed that the percentage of PC‐3 cell viability was decreased significantly, 61.30% for 24 h co‐culture and 30.96% for 48 h co‐culture. Furthermore, the activated PBMCs with B. cristata, B. strigosa, barlerin and verbascoside inhibited the growth of PC‐3 for 48 h co‐culture with the percentage of cell viability of 41.25%, 72.69%, 62.37% and 71.17%, respectively (Fig. 4B; Table S7). Likewise, the activated PBMCs with B. siamensis can induce the DNA damage indicated by long tail DNA migration (Fig. 4C; Table S7), apoptosis with a high percentage of 21.2% (Fig. 4D; Table S7) and arrest the cell cycle in G1‐phase showing by low DNA content in S and G2‐M phase (Fig. 4E; Table S7) in PC‐3 for 48 h co‐culture significantly, correlated with cytotoxic result harmoniously. These results showed that the candidate treatments can activate PBMCs to suppress PC‐3 proliferation and those treatments act as an immunomodulator to immune cell.

Table 1 The percentages of cell viability of PC-3 after 48 h co‐culture system with activated PBMCs at 24 and 48 h with treatments at the concentration of 50 µg/mL.
Full size table

Barleria extracts stimulated cytokine genes expression in PBMCs

To insist that the candidate treatments including three Barleria extracts and single compounds (barlerin and verbascoside) acting as an immunomodulator to PBMCs, the result was supported by studies of cytokine gene expression, aimed at IL-2, IL-10, IL-12, IL-15, IL-21 and interferon (IFN)-γ normalized with β‐actin as a reference gene. The PBMCs were treated with the treatment including B. cristata, B. siamensis, B. strigosa, barlerin and verbascoside at the concentration of 50 µg/mL for 48 h to obtain the result. Noticeably, some of the treatments including B. cristata, B. siamensis, barlerin and verbascoside stimulated all cytokine gene expression, except for B. strigosa that merely promoted the gene overexpression of IL-2, IL-10 and IL-12 with percentage relative expression of 739.67%, 497.34% and 586.51% as well as slightly suppressed the expression of IFN-γ with percentage relative expression as 92.79%. In addition, the B. siamensis extract tremendously stimulated the cytokine genes to overexpress with a high percentage gene relative expression ranging between 370.41 and 865.10% and higher than other treatments (Fig. 4F; Table 2). These results related to co‐culture toxicity activity indicating that the activated PBMCs with B. siamensis extract for 48 h persuaded the perdition of PC‐3 with the lowest percentage of cell viability (Fig. 4B) and insist that the mentioned treatments act as an immunomodulator, functioned by their bioactive compounds.

Table 2 The percentages of relative cytokine gene expression in PBMCs after treatments at the concentration of 50 µg/mL for 48 h, normalized by β-actin.
Full size table

To confirm that the lymphocytes (T cell), helper T cell (Th cell) and cytotoxic T cells (Tc cell), and natural killer (NK) cells are comprised in studied PBMCs population to expose the relevance of the cytokine gene expression results. These cells were sorted by flow cytometry stained with their specific antibodies including CD3, CD4, CD8a and CD56 for T, Th, Tc and NK cells sorted in PBMCs population after treatment for 48 h (Fig. 4G). The result disclosed that the PMBCs employed in the experiment contain all types of mentioned cells by detecting their cell surface marker. The number of sorted T cell or CD3+CD56− cell, Th cell or CD4+ cell, Tc cell or CD8+ cell and NK cell or CD56+CD3− cell in each treatment was not much different from the control and indicated that the treatments did not show cytotoxic activity in the PBMCs population obviously (Fig. 4H; Table S8).

Fig. 4
figure 4

The results of Barleria extracts, barlerin and verbascoside activate PBMCs against PC-3 implicated to cytokine genes expression. (A) a schematic of activated PBMCs establishing and cytotoxic testing procedure using MTT assay; (B) a heatmap representing percentage of cell viability obtained after 24 and 48 h co‐culture; n = 3; (C) the genotoxic result showing comet image at 200x to reveal DNA damage of PC‐3 cell in each condition after 48 h co‐culture; n = 150; (D) The dot plot of single cell to indicate apoptotic PC‐3 cell and comparing bar chart of the percentage of apoptosis in each condition after 48 h co‐culture; n = 3; (E) the histogram of PC‐3 cell cycle after 48 h co‐culture showing DNA content in peak in subG1, G1, S, and G2‐M phases and comparing bar chart of percentage DNA content; n = 3; (F) the bubble plot to show percentage relative cytokine gene expression in PBMCs normalized by β‐actin as a reference gene; n = 3; (G) the procedure of PBMCs sorting using Flow cytometric assay; (H) The population of CD3+CD56− T cell, CD4+ T cell, CD8+ T cell and CD56+CD3− NK cell in studied PBMCs after 48 h treatment; n = 3.

Full size image

The effect of Barleria extracts on non-tumoral human hepatocyte (THLE-3) cytotoxic activity and cytochrome P450 genes expression

The candidate treatments (at concentration of 50 µg/mL) including three Barleria extracts, and two single compounds (barlerin and verbascoside) were further employed to evaluate their cytotoxic activity and CYP450 genes including CYP1A2, CYP2A6, CYP3A4, CYP2D6 and CYP2E1 expression. After 48 h treatment, the results of cytotoxic effect indicated that the B. strigosa extract inhibited the growth of THLE-3 significantly with a low percentage cell viability of 58.16% (Fig. 5A). Moreover, the extract also caused DNA damage indicated by long tail DNA migrating out of the nucleus (Fig. 5B), induced apoptosis and necrosis with a high percentage of 15.2% and 22.35%, respectively (Fig. 5C) as well as arrested the THLE‐3 cell in G1‐phase showing by low DNA content in S and G2‐M phase when compared to the control (Fig. 5D) as the same result as cisplatin. These results described that the B. strigosa extract showed toxic effects on normal hepatocytes but not detected in other four treatments including B. cristata, B. siamensis, barlerin and verbascoside (Fig. 5A‐D; Table S9).

To describe the effect of candidate treatments on CYP450 gene expression, the B. siamensis extract can suppress the expression of all studied CYP450 genes with a low percentage of relative expression for CYP1A2, CYP2A6, CYP3A4, CYP2D6 and CYP2E1 genes as 59.19%, 72.01%, 46.64%, 27.76% and 59.96%, respectively (Fig. 5E; Table 3). All treatment can suppress CYP2D6 gene expression significantly revealing a low percentage of relative expression, especially the lowest in B. siamensis extract as mentioned previously. In addition, the B. strigosa extract moderately induced the expression of CYP2A6 and CYP2E1 genes significantly with a high percentage of relative expression of 133.75% and 158.81%, respectively (Fig. 5E; Table 3).

Table 3 The percentages of relative cytochrome P450 gene expression in THLE-3 after treatments at the concentration of 50 µg/mL for 48 h, normalized by β-actin.
Full size table
Fig. 5
figure 5

Toxicity results of Barleria extracts, barlerin and verbascoside on THLE-3 after 48 h treatment. (A) the comparison of percentage cell viability in each condition normalized by DMSO showing low percentage cell viability in B. strigosa extract and cisplatin; n = 3; (B) the comet images indicating DNA damage by showing their OTM values in each condition; n = 150; (C) the dot plot of single cell to indicate apoptosis and comparing bar chart of the percentage of apoptosis in each condition; n = 3; (D) the histogram of cell cycle showing DNA content in peak in subG1, G1, S, and G2‐M phases and comparing bar chart of percentage DNA content; n = 3; (E) the bubble plot to show percentage relative cytochrome P450 gene expression in THLE‐3 normalized by β‐actin as a reference gene; n = 3; (F) a summarized schematic presenting the biological activities of B. siamensis extract containing barlerin and verbascoside acting as an immunomodulator to activate PBMCs against PC‐3 and the effect to cytochrome P450 gene expression in THLE‐3.

Full size image

Discussion and conclusion

Barleria extracts contain numerous pharmacologically active compounds, notably γ-tocopherol, barlerin, stigmasterol, and verbascoside, which are distributed across all studied species and act as common metabolites. These compounds were successfully identified using the GC‐MS technique, a powerful analytical tool for detecting small volatile metabolites (< 650 Da), including small acids, alcohols, hydroxyl acids, amino acids, sugars, fatty acids, sterols, catecholamines, drugs, and toxins21,22,23. Based on the GC‐MS-derived percentage relative content, two common metabolites, barlerin and verbascoside, were detected at relatively high levels, with barlerin exceeding 5% and verbascoside above 10%. Given their significant presence, these two metabolites were selected for further study as key targeted compounds of the Barleria species. To quantify barlerin and verbascoside, HPLC analysis was performed, confirming the presence of these compounds in all Barleria extracts. Notably, B. siamensis extract exhibited the highest concentrations of both metabolites. Consequently, barlerin and verbascoside were chosen for further biological activity studies to evaluate the potency of individual compounds compared to Barleria extracts containing both, thereby providing a comprehensive understanding of their biological effects. The two compounds were targeted because they have previously been found in Barleria plants18, possess anti-inflammatory and antioxidant properties18, and were detected in all the plants studied. Therefore, it is hypothesized that plants containing these two compounds may exhibit immunomodulatory properties.

Regarding toxicity results, several treatments, including B. cristata, B. siamensis, B. strigosa, barlerin, and verbascoside, exhibited potent selective cytotoxic activity against cancer cells while sparing normal human cells (Fig. 2C). These treatments demonstrated high cell viability in normal cells, such as PBMCs and HPrEC, while significantly reducing cancer cell viability (Fig. 2A, B). Additionally, they induced DNA damage (Fig. 2D), apoptosis, and cell cycle arrest in cancer cells (Fig. 3A‐B). Based on this evidence, barlerin and verbascoside, as individual compounds, exhibited strong anti-proliferative effects on cancer cells. This finding suggests that plant species containing high levels of these bioactive compounds, such as B. cristata, B. siamensis, and B. strigosa, possess superior anticancer activity compared to other species. Notably, previous studies have confirmed the pharmacological properties of barlerin and verbascoside, highlighting their anticancer, cellular chemoprotective, and anti-inflammatory activities while preserving normal cell viability18,24,25,26. To clarify why the other Barleria species—B. lupulina, B. prionitis, and B. repens—were not selected for further investigation, our results revealed the following: B. lupulina extract exhibited non-selective cytotoxicity, with IC50 values of 69.94 µg/mL in PBMCs and 50.51 µg/mL in HPrECs (Fig. 2C), indicating substantial toxicity to normal cells. Furthermore, at these concentrations, the extract significantly induced DNA damage, triggered apoptosis, and caused cell cycle arrest in normal cells (Figs. 2D and 3A and B), thereby disqualifying it from further immunological or mechanistic studies due to safety concerns. In contrast, B. prionitis and B. repens extracts exhibited minimal biological activity, correlating with their low levels of key bioactive metabolites—barlerin and verbascoside—and a higher abundance of long-chain fatty acids, which are not known to contribute significantly to immunomodulatory or anticancer functions (Fig. 1F; Table S1). As such, these species were excluded from further study to maintain focus on more promising candidates.

When comparing the toxic effects of Barleria extracts, barlerin, verbascoside, and cisplatin–a widely used chemotherapeutic drug for treating various human cancers, including head and neck, lung, bladder, testicular, and ovarian cancers–the results demonstrated their differential impact on cellular viability. Cisplatin is known to induce unrepairable DNA lesions, leading to permanent proliferative arrest and apoptosis via the mitochondrial pathway27,28,29. However, while cisplatin exhibited potent cytotoxic effects on cancer cells, it also caused significant toxicity to normal cells–an adverse effect not observed with the Barleria extracts or the studied compounds. Furthermore, based on their LD50 values (Table S3), the toxicity classification indicated that Barleria extracts, and the two bioactive compounds were categorized as Class III (slightly hazardous) and Class IV (non-hazardous) for normal cells, while exhibiting Class II (moderately hazardous) toxicity towards cancer cells. In contrast, cisplatin was classified as Class II for all studied cell types, reflecting its lack of selectivity between cancerous and normal cells. Notably, some Barleria extracts, particularly B. cristata, B. siamensis, and B. strigosa, demonstrated greater anticancer potential than cisplatin by exhibiting selective toxicity towards cancer cells while maintaining minimal toxicity to normal cells.

Remarkably, at the concentration of 50 µg/mL of B. cristata, B. siamensis and B. strigosa, can activate the PBMCs representing an immune cell to promote the cytokine gene expression against PC-3 cancer cells growing based on their bioactive compounds, especially barlerin and verbascoside. Barlerin and verbascoside have been reported in some Barleria species, albeit infrequently, including B. acanthoides, B. cristata, B. dinteri, B. lupulina, B. prionitis, B. strigosa, and B. trispinosa18. These compounds exhibit various biological activities, including antimicrobial, neuroprotective, chemoprotective, anti‐inflammatory, antioxidative, and antitumor properties30,31,32,33,34. Notably, this research highlights a novel function of Barleria extracts containing barlerin and verbascoside, specifically their role in immunomodulatory activity and anticancer progression. As a consequence, B. siamensis extract, which demonstrated high potency in activating PBMCs against PC‐3 cells and promoting cytokine gene overexpression, may contain higher amounts of barlerin and verbascoside than the other two species. However, PBMCs activated with either barlerin or verbascoside alone appeared to be less effective in inhibiting cancer cell growth and upregulating cytokine gene expression when compared to the B. siamensis extract, which contains both compounds and may exert a synergistic effect. Furthermore, PBMCs activated with B. siamensis extract were able to attack PC-3 cancer cells after 24 h of co-culture. These results may be related to the concentration of barlerin and verbascoside, as B. siamensis extract potentially contains higher amounts of these compounds compared to other species, as previously mentioned. However, although IL-12 has been reported to amplify IFN-γ production35, it was not detected under the B. strigosa treatment condition. Even though the extract induced IL-12 gene overexpression, it did not correlate with IFN-γ expression, as indicated by its low relative expression percentage. This result suggests that certain compounds present in B. strigosa extract may interfere with this activity. Cytokines play a crucial role in mediating interactions between immune and non‐immune cells within the tumor microenvironment (TME)35. The cytokine genes analyzed in this study, including IL-2, IL-10, IL-12, IL-15, IL-21, and IFN-γ, have been implicated in promoting cytotoxic activity and enhancing anticancer immunity, primarily through their expression in T cells and NK cells35,36,37,38,39.

To elucidate the signaling pathways related to the expression of the studied cytokine genes, some studies have ambiguously suggested that IL-2, IL-10, IL-12, IL-15, IL-21, and IFN-γ function as pleiotropic cytokines with both anti-inflammatory and immunostimulatory properties. In cancer, the expression of these cytokine genes may exert either pro- or anti-tumor effects on CD4⁺, CD8⁺, and/or CD56⁺CD3⁻ cells40. Furthermore, they may be involved in the downregulation of key signaling pathways, including phosphatidylinositol 3ʹ-kinase (PI3K)-AKT, mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK), NF-κB, and signal transducer and activator of transcription 3 (STAT3), thereby impairing cancer cell proliferation and stemness35.

Additionally, the interaction between Barleria extracts and CYP450 genes was investigated to assess their effects on drug metabolism. The CYP1, CYP2, and CYP3 gene families are primarily responsible for clinical drug metabolism. These genes include CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP2E1, CYP3A4, and CYP3A5, which are predominantly expressed in the human liver41. According to the results, B. siamensis extract significantly suppressed the expression of the studied CYP450 genes, suggesting that its bioactive compounds may undergo slight degradation, thereby enhancing their biological activity. Furthermore, these findings provide compelling evidence that B. siamensis extract exhibits greater biological potential than the other studied species. However, the induction or suppression of CYP450 gene expression is critically linked to the bioactivation of xenobiotics, making it a profound and clinically significant factor19,20,41.

Barleria extracts containing barlerin and verbascoside as pharmacologically active compounds, particularly B. cristata, B. siamensis, and B. strigosa, exhibited significant cytotoxic activities by impairing proliferation, inducing DNA damage, enhancing apoptosis, and arresting the cell cycle in PC-3 cells. However, these effects were not observed in PBMCs and HPrEC, except for B. strigosa extract, which demonstrated a slight toxic effect on HPrEC and a moderate toxic effect on THLE-3 cells. These findings suggest that B. strigosa extract may not be entirely safe for use. Notably, B. cristata and B. siamensis extracts displayed promising immunomodulatory activity by activating PBMCs as immune cells against PC-3 cells, with B. siamensis showing significant activation after 24 h of co-culture and both B. cristata and B. siamensis exhibiting activation after 48 h of co-culture. Additionally, these extracts promoted cytokine gene expression, contributing to anticancer progression. These results indicate that the extracts can target cancer cells both directly and indirectly through PBMC activation (Fig. 5F). Furthermore, B. siamensis extract demonstrated a desirable outcome by significantly suppressing the expression of the studied CYP450 genes, a result not prominently observed in the other extracts, thereby highlighting its potential bioavailability and therapeutic effects. Additionally, the findings clearly indicate that B. siamensis extract, which contains high amounts of both barlerin and verbascoside, exhibits greater biological potency compared to single-compound treatments. This observation may suggest a synergistic interaction between the two compounds, potentially enhancing their overall pharmacological efficacy.

In conclusion, B. siamensis extract demonstrated outstanding efficiency in immunomodulatory function and CYP450 gene suppression, contributing to anticancer progression as supported by multiple lines of experimental evidence. Therefore, our discovery highlights this plant species as a novel and promising natural source for potential applications in human health. Moreover, as B. siamensis is a rare and endemic species in Thailand, its promotion could support both conservation efforts and economic potential. However, this research serves as a preliminary investigation to demonstrate the bioactive potential of various Barleria species, providing sufficient evidence to identify candidates for further study. Nonetheless, future studies should address current limitations by incorporating cytokine protein quantification, evaluating the potential synergistic effects between barlerin and verbascoside, and conducting in vivo experiments to deepen the mechanistic understanding of their immunomodulatory and anticancer activities.

Materials and methods

Plant materials

The mature leaves of six Barleria species were completely collected at the natural population and ex-situ sites distributed in different regions of Thailand. The four Barleria species, including B. cristata, B. lupulina, B. siamensis and B. strigosa were obtained from Chiang Mai province, while the B. prionitis was collected from Khon Kaen province and B. repens was collected from Nakhon Ratchasima province, respectively. The plants were identified by a proficient botanist, Professor Arunrat Chaveerach, Ph.D. coupled with using taxonomic classification proposed by Shendage and Yadav42. The experimentation on plants is in accordance with relevant guideline and legislation. In addition, the voucher specimens were kept at the Department of Biology, Faculty of Science, Khon Kaen University, under the collection number A. Chaveerach 1095‐1099. To preserve the quality of Barleria leaves for phytochemical study, the leaves were rinsed in tap water and dried at room temperature protected from sunlight. After the leaves were completely dried, they were packed in plastic bags, tapped completely up and stored at room temperature.

Phytochemical extraction and preparation

The dried leaves were finely ground using an electric blender, and 25 g of fine leaves were macerated in 250 mL of 60% ethanol and incubated at 37 °C, 300 rpm in a rocking incubator for 72 h protected from light. After that, the extract was filtered through Whatman® qualitative paper, grade 1. Ten milliliters were aliquoted to a conical tube to screen the phytochemical constituent of the extract. For the remaining amount, the extract was transferred to a 50 mL conical tube to submit into a speed vacuum concentrator to completely evaporate the extraction solution for utilization to study their biological activities. All extracts were stored at -20 °C for further study. The % yield of each crude extract was calculated from [weight of extract after solvent elimination (g)/weight of plant sample (g)] × 100 to analyze the initial concentration of the extracts43.

Pre-processing of phytochemical Preparation for untargeted and targeted metabolite analysis

Ten milliliters of aliquoted extract were concentrated using a speed vacuum concentrator and redissolved with 2 mL of ethanol. After completely re-dissolved, the extract was fractionated with 2 mL of combined solvents between petroleum ether and diethyl ether (2:1 v/v). After fractionating, the clear portion was collected and filtrated through a 0.45 μm syringe filter into a new tube for untargeted and targeted metabolite analysis.

Untargeted metabolite screening using gas Chromatography-Mass spectrometry (GC-MS)

The conical vial containing Barleria extracts was loaded into the GC-MS Agilent 5977B GC/MSD (Agilent Technologies, Inc., Santa Clara, CA, USA) equipped with an HP–5MS UI column (30 m × 250 μm × 0.5 μm). The oven temperature was programmed from 50 °C (hold for 3 min), raised to 280 °C (10 °C/min for 10 min), and the final temperature was 300 °C (10 °C/min for 10 min). One microliter was injected with split mode (10:1) at an injection temperature of 250 °C, carried out by helium as a carrier gas at the flow rate of 1 mL/min. Additionally, the MSD transfer line temperature was constantly set at 280 °C. In contrast, the MS quadrupole and source temperature were programmed at 150 and 230 °C, and the mass scan ranged from 40 to 600 m/z44,45,46. The compound identification was analyzed using Agilent MassHunter Qualitative Analysis 10.0, authenticating the mass spectra with the W11N17 library. After compound identification, the area of each compound was applied to calculate % relative content to represent the key metabolite–derived Barleria extracts. To achieve a chemical classification of compounds, the data was sorted by using ClassyFire software47,48,49.

Barlerin and verbascoside detection by reversed-phase high performance liquid chromatography (HPLC)

Barlerin and verbascoside-the targeted metabolites were quantified by using reverse phase HPLC. The conical vial containing Barleria extracts was subjected to HPLC (LC‐20AD; Shimadzu, Kyoto, Japan) equipped with a quaternary pump, PAD (SPD‐M20A) detector, and a C18 (Inertsil ODS‐3) analytical column (4.6 mm × 250 mm, 5‐micron particle size). The chromatographic condition of each compound is shown in the following Table 4:

Table 4 The chromatographic condition to use for Barlerin and Verbascoside detection.
Full size table

The 1 mg of barlerin (Cas no. 57420-46‐9, PhytoLab, Germany) and verbascoside (Cas no. 61276‐17‐3, PhytoLab, Germany) standards were dissolved in their eluent buffer A/B (90:10 v/v) and 2‐fold serially diluted for 5 concentration levels to create the standard calibration equation for compound quantification.

Phytochemical preparation for biological activity studies

Ten milligrams of the Barleria extracts were completely dissolved in 1 mL of DMSO and then filtered through a 0.45 μm syringe filter. The filtrated extract was 10-fold diluted in sterile 1x PBS to keep at a stock concentration of 1,000 µg/mL. The stock concentration was diluted in culture media to obtain a first concentration of 100 µg/mL, containing 1% DMSO and 2–fold serially diluted to prepare a working concentration as 100, 50, 25, 12.50, and 6.25 µg/mL to be applied for toxicity testing and other biological activity studies. Furthermore, the 1% DMSO was used as a control of the experiment.

Human peripheral blood mononuclear cells (PBMCs) isolation and preparation

PBMCs isolation from the buffy coat, obtained from the blood bank of the Srinagarind Hospital, Faculty of Medicine, Khon Kaen University, Thailand, was compiled following the protocol of Freshney50 by using cold FicollPaqueTM Plus (Cytiva, USA). The isolated PBMCs, with a viability at least 98% as determined by hemacytometer counting following the protocol of Louis and Siegel51, were suspended in modified RPMI-1640 medium supplemented with L‐glutamine, 10% FBS, 100 µg/mL streptomycin, and 100 U/mL penicillin, at a concentration of 1 × 106 cells/mL to be used for an experiment, according to Arimilli et al.52.

Primary prostate epithelium cells: normal, human (HPrEC), prostate cancer cells (PC-3) and non‐tumoral human hepatocyte (THLE‐3) cell lines culture and preparation

HPrEC (ATCC, PCS-440‐010), PC‐3 (ATCC, CRL‐1435), and THLE‐3 (ATCC, CRL‐3583) cell lines were commercially purchased from the American Type Culture Collection (ATCC). For HPrEC, the cells were completely cultured in Prostate Epithelial Cell Basal Medium (ATCC, PCS‐440‐030), supplemented with Prostate Epithelial Cell Growth Kit (ATCC, PCS‐440‐040), 100 µg/mL streptomycin, and 100 U/mL penicillin, while PC‐3 cells were cultured in complete DMEM/F12 1:1 medium (HyClone, USA) with 10% FBS and 100 µg/mL streptomycin, and 100 U/mL penicillin. In addition, THLE‐3 was cultured in BEBM supplemented with a BEGM Bullet kit (Lonza, CC3170). The cells were incubated at 37 °C in a 5% CO2 atmosphere. The culture medium was changed every 3‐4 days. The cells were dissociated using 1 mL of 0.05% trypsin‐EDTA when having at least 70% density.

Cell viability assessment using (3-(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide) tetrazolium reduction (MTT) assay

After cell preparation, the PBMCs, HPrEC, PC-3 and THLE‐3 cells were seeded in 96 well cell culture plate at a concentration of 1 × 106, 5 × 104, 2 × 104 and 5 × 104 cells/mL for 100 µL/well and cultured for 24 h before treatment. Then, the cells were treated with 10 µL of plant extracts in a corresponding well and then incubated for 48 h. The MTT assay was compiled by following the steps by Freshney50 and Sudmoon et al.53. Briefly, after incubating the period, the culture media were carefully removed and replaced with 10 µL of 50 mg/mL MTT solution, and then the plates were further incubated for 4 h. After adding 100 µL of DMSO, the plates were read at the 570 nm wavelength to detect formazan solution intensity using a fluorescence microplate reader (SpectraMax M5 series, Molecular Devices). In addition, the cisplatin (Fresenius Kabi, USA) was prepared to be a positive control of the experiment. The % cell viability was manually calculated, normalizing with DMSO as a control, and further used to analyze IC50 values by using Quest Graph™ IC50 Calculator, http://www.aatbio.com/tools/ic50%calculator.html54. To indicate the selective toxicity, the IC50 value derived from normal cell were divided by the IC50 value of cancer. The ratio was cut‐off at 3 times to define selective toxicity to the cancer cell if those values are over 3 times55. In addition, LD50 values were estimated based on IC50 values to predict the toxicity levels of the extracts, following the classification system described by Walum56. According to this system, compounds are categorized as follows: Class I – strongly hazardous (LD50 < 100 mg/kg), Class II – moderately hazardous (LD50 = 100–500 mg/kg), Class III – slightly hazardous (LD50 = 500–1000 mg/kg), Class IV – non-toxic (LD50 > 1000 mg/kg).

DNA damage evaluation via comet assay

To detect DNA damage, the comet assay was conducted following the method described by Singh et al.57. The concentration at IC50, or a maximum working concentration, in case no IC50 value was detected, was applied as treatment. Olive tail moment (OTM) value-the relative amount of DNA in the tail of the comet multiplied by the median migration distance was analyzed by CASP software version 1.2.3 (CASPlab, Wroclaw, Poland) based on 150 single cells58.

PBMCs and PC-3 co‐culture system

In the experiment, the PC-3 was seeded in a 96‐well cell culture plate at the concentration of 2 × 104 cells/mL and cultured at 37 °C in 5% CO2 for 24 h to allow their adherence. Meanwhile, the PBMCs were seeded in a 1.5 mL tube at the concentration of 1 × 106 cells/mL and then incubated at 37 °C in 5% CO2 for 4 h. After that, the 50 µg/mL of plant extracts or other compounds were added into the tube containing PBMCs and incubated in 2 conditions, including 24 h and 48 h, to create activated PBMCs. Additionally, the activated PBMCs with DMSO and culture media containing 50 µg/mL of the extracts or other compounds were applied to be a control, prepared simultaneously. After the incubating period, 50 µL of activated PBMCs were added to the corresponding 96‐well containing PC‐3 and then incubated in 2 conditions, including 24 h and 48 h. After the co‐culture incubation period, the culture media and suspended cells were discarded, and then the well was washed twice with 100 µL sterile 1x PBS, followed by 10 µL of MTT solution, and incubated in the plate for 4 h to evaluate % cell viability by MTT assay.

Programmed cell death and cell cycle arrest evaluation by flow cytometric assay

After the treatment period, to detect programmed cell death, the cells were gently resuspended in 100 µL of 1x annexin-binding buffer and dyed with 5 µL of FITC Annexin V as well as 1 µL of the 100 µg/mL PI working solution (Invitrogen, Waltham, USA). After the incubation for 15 min, the 400 µL of 1x annexin‐binding buffer was added and mixed gently. The stained cells were analyzed by measuring the fluorescence emission at 530 nm and > 575 nm59. The data were plotted in two‐dimensional dot plots between the stained cells with PI and FITC annexin V. The plots can be divided into four regions corresponding to (1) viable cells that are negative to both probes; (2) apoptotic cells that are PI negative and annexin positive; (3) late apoptotic cells that are PI and annexin positive; and (4) necrotic cells that are PI positive and annexin negative43.

For cell cycle arrest evaluation, the harvested cells were fixed in 70% ethyl alcohol at 4 °C for 30 min and washed twice with 1x PBS. After that, 0.5 mL of PI/RNase staining solution (Invitrogen, Waltham, USA) was added and incubated at 37 °C for 30 min. The cell cycle distribution was analyzed under a 532 nm excitation with a 585/42 bandpass condition60. The G0/G1 and G2/M phase histogram peaks were separated by the S phase distribution.

Real-time quantitative polymerase chain reaction (RT‐qPCR) analysis

The cells were treated with treatments at the concentration of 50 µg/mL for 48 h. After cell harvesting, the RNA extraction was completed following the manufacturer’s protocol using the RNeasy Mini Kit (Qiagen, Hilden, Germany). The concentration of total RNA was quantified with a nanodrop spectrophotometer (DeNovix, Wilmington, Delaware, USA). Subsequently, cDNA synthesis was performed using the RevertAid First Strand cDNA Synthesis Kit (ThermoFisher Scientific, Waltham, USA) according to the manufacturer’s protocol. The cDNA was kept at -20 °C until use.

The qRT-PCR assay was performed using CFX Duet Real‐Time PCR system equipped with CFX Maestro Software 2.3 (BIO‐RAD, California, USA). Amplifications were carried out with the final reaction solution as 20 µL containing 10 µL, 2x SensiFAST SYBR® No‐ROX Mix, 6 µL of first stranded cDNA (diluted 1:10), 0.8 µL of each specific primer (10 µM) and 2.4 µL nuclease-free water. PCR conditions slightly modified from Freitas et al.61, were programmed following pre‐denaturation at 95 °C for 3 min followed by 55 cycles of 95 °C for 5 s, 60 °C for 10 s, and 72 °C for 20 s, for denaturation, annealing and extension, respectively. The melting curve analysis was set to verify the specificity of each pair of primers, following 95 °C for 30 s, 60 °C for 2 min and a continuous raise in temperature to 95 °C at 0.11 °C/s ramp rate. The cycle of threshold (Ct) in each treatment was applied to calculate ΔCt for relative gene expression calculation. The expression of target genes was normalized by the β‐actin expression value. The primer sequences used in this study were presented in Table S10. All values were expressed as the mean ± S.D. The relative gene expression value was derived using the 2−ΔΔCt method62,63, to expose the percentage of relative gene expression.

Flow cytometric assay for PBMCs phenotype sorting

After treating PBMCs with the plant extracts or used compounds at the concentration of 50 µg/mL for 48 h. The phenotypic analyses of PBMCs were performed by their cell surface staining with linage-specific fluorochrome‐labeled monoclonal antibodies including CD3 (OKT3), conjugated with eFluor™ 450 (eBioscience, 48‐0037‐42), CD4 (RPA‐T4), conjugated with FITC (eBioscience, 11‐0049‐42), CD8a (RPA‐T8), conjugated with PE (eBioscience, 12‐0088‐42) and CD56 (CMSSB), conjugated with APC (eBioscience, 17‐0567‐42), described by Maecker et al.64.

Statistical and data analysis

All values were expressed as the mean ± standard deviation (S.D.). For statistical significance, the non-parametrical Kruskal‐Walli’s test was selected to compare data within a group containing a single variant. In addition, the non‐parametrical Mann‐Whitney U test was employed to test the difference in data between groups using GraphPad Prism version 8.0.2 (GraphPad Software, Boston, USA). The probability value of p < 0.05 was considered to denote the statistically significant differences. The principal component analysis (PCA) and hierarchical cluster analysis of compounds were performed online using MetaboAnalyst 6.0, www.metaboanalyst.ca/home.xhtml65. The Venn diagram was created using InteractiVenn, www.interactivenn.net66 as well was Sankey diagram was constructed using SankeyMATIC, https://sankeymatic.com/67. All schematics presented in this research were illustrated using licensing BioRender, https://www.biorender.com/.