Biotechnological potential of Cannabis sativa adventitious roots for producing immunomodulatory and anti-inflammatory bioactive compounds

Abstract

Cannabis sativa L. (C. sativa), commonly known as hemp, is widely recognized for its diverse range of bioactive compounds with therapeutic potential in medicinal, industrial, and nutritional applications. This study investigates the use of adventitious roots (ARs) derived from C. sativa as a scalable platform for producing bioactive metabolites with immunomodulatory and anti-inflammatory properties. We first isolated extracts from C. sativa ARs (CS-AR) using various solvents: methanol (MeOH-E), chloroform (CHCl₃-E), and hexane (Hexane-E), and explored their effects on dendritic cell (DC) maturation, a key process involved in immune responses. Notably, MeOH-E demonstrated strong anti-inflammatory effects without inducing cytotoxicity in DCs, distinguishing it from the other extracts. Metabolomic analysis of these extracts annotated the presence of cannabinoid derivatives and metabolites, including cannabinoid glycoside derivatives, cannabigerolic acid-O-acetate (CBGA-O-acetate), cannabidiol diacetate derivatives, and cannabidiol mono-acetate mono-benzoate. Among these, cannabinoid glycoside derivatives and CBGA-O-acetate were found to be present at higher levels in MeOH-E. Further investigation into the functional properties of MeOH-E revealed that it could suppress the expression of key surface molecules and antigen-presenting ability in mature DCs, alongside attenuating mitogen-activated protein kinase (MAPK) signaling pathways as well as nuclear factor kappa-B (NF-κB) signaling. Additionally, MeOH-E inhibited T cell proliferation and activation. These findings underscore the CS-AR system as a promising, reproducible biotechnological platform for producing therapeutic bioactive compounds for inflammatory diseases, with significant potential for application in the pharmaceutical and nutraceutical industries.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-16130-1.

Introduction

Cannabis sativa L. (C. sativa), commonly referred to as hemp, has been utilized for centuries in the medicinal, industrial, and nutritional domains, primarily owing to its broad spectrum of bioactive compounds^1,2. Its various plant parts including roots, stems, leaves, and flowers possess distinct biological properties, such as anti-inflammatory, antioxidant, and antimicrobial activities, thus, positioning C. sativa as a versatile resource in diverse applications^3,4. Among these, the leaves are particularly noteworthy for their pharmacological and therapeutic potential. Recent studies have underscored their relevance in the development of functional foods, pharmaceuticals, dietary supplements, and cosmetics, attributing such potential to their prominent antioxidant, anti-inflammatory, and hypoglycemic properties^5,6. These biological effects are driven by a rich diversity of secondary metabolites—including terpenoids, stilbenoids, alkaloids, cannabinoids, and flavonoids—present in the leaves^6–8. Of particular interest are the cannabinoids (e.g., cannabidiol) and flavonoids (e.g., cannflavins, luteolin, vitexin, genistein), which have been shown to exert anti-inflammatory effects via the inhibition of mitogen-activated protein kinase (MAPK) and nuclear factor-kappa B (NF-κB) signaling pathways—both key regulators of pro-inflammatory cytokine production^9,10. These pathways govern the expression of genes such as tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6), which play central roles in immune activation and chronic inflammation¹¹. By modulating these signaling cascades, C. sativa-derived compounds have demonstrated the ability to attenuate inflammation in various disease models, including inflammatory skin conditions, osteoarthritis, and edema^9,10,12. These findings collectively highlight the therapeutic potential of C. sativa compounds for targeting inflammation-related pathways.

Despite the growing body of evidence supporting their efficacy, substantial variability remains in the yield and biological activity of these compounds. Such inconsistency can be attributed to factors including extraction methods, plant origin, and environmental growing conditions, all of which hinder the reproducibility and standardization required for clinical or commercial applications^13,14. This underscores the need for alternative approaches that ensure consistent bioactive profiles and scalable production strategies for C. sativa-derived therapeutics.

One promising approach in this regard involves the use of adventitious roots (ARs), which represent a scalable and biotechnologically viable system for the stable production of plant secondary metabolites^15,16. ARs are roots that are induced in vitro from non-root tissues through exposure to specific plant growth regulators under controlled environmental conditions^17,18. Once induced, ARs activate root-specific metabolic programs leading to enhanced biosynthesis of root-associated compounds such as alkaloids and phenolics^19,20. Notably, ARs exhibit high genetic and metabolic stability, even when derived from differentiated donor tissues, thereby enabling the long-term and consistent production of target metabolites^21–23. Moreover, the use of controlled and sterile culture environments minimizes variability due to external factors such as pathogens, pesticides, and environmental fluctuations to ensure high-quality and reproducible metabolite yields^19,24. Given these advantages, ARs are increasingly recognized as valuable platforms for the production of medicinal plant-derived compounds, with significant potential for applications in the pharmaceutical, nutraceutical, and cosmetic industries^19,25,26.

In this study, we investigated the metabolomic profiles of crude extracts (CS-AR extracts) obtained from ARs induced from C. sativa leaf (CS-leaf) and evaluated whether these extracts exhibit bioactivity comparable to that of conventional CS-leaf extracts. Specifically, we focused on the immunomodulatory and anti-inflammatory potential of AR-derived extracts by examining their effects on the activation of dendritic cells (DCs) and T cells—key players in host immune defense and the pathogenesis of inflammatory diseases. Ultimately, this study proposes the AR system as a robust and reproducible platform for the production of bioactive compounds from C. sativa, thereby supporting its potential application as a natural therapeutic resource in the pharmaceutical, functional food, and cosmetic industries.

Results

Cytotoxicity differences in bone marrow-derived dendritic cells (BMDCs) treated with CS-AR extracts

To investigate the functional differences in the physiological activity of CS-AR extracts, we explored the potential cytotoxicity of CS-AR extracts (chloroform extracts; CHCl₃-E, hexane extracts; Hexane-E, and methanol extracts; MeOH-E) dissolved in DMSO. BMDCs were treated with various concentrations of CS-AR extracts for 18 h, followed by flow cytometric analysis using Annexin V and PI staining (Fig. 1A). The results showed that treatment with 10 µg/mL of CS-AR extracts, specifically CHCl₃-E and Hexane-E increased necrotic cell death (Annexin V⁻PI⁺ cells) and late apoptotic cell death (Annexin V⁺PI⁺ cells) compared to the control cells (DMSO-treated cells). In contrast, 10 µg/mL of MeOH-E did not induce such cytotoxic effects. Furthermore, when cells were treated with 50 and 100 µg/mL of CS-AR extracts, CHCl₃-E and Hexane-E treatments induced higher levels of necrotic cell death (Annexin V⁻PI⁺), late apoptotic cell death (Annexin V⁺PI⁺), and early apoptotic cell death (Annexin V⁺PI⁻) than control cells. MeOH-E did not cause cytotoxicity at 10 µg/mL (Fig. 1B).

Fig. 1.

Cytotoxicity of CS-AR extracts. BMDCs were treated with various concentrations of CS-AR extracts or 50 nM staurosporine (STS) for 18 h. (A) Representative results of Annexin V and PI staining analyzed via flow cytometry. (B) Bar graphs showing the analysis of necrotic cell death (Annexin V⁻PI⁺), late apoptotic cell death (Annexin V⁺PI⁺), and early apoptotic cell death (Annexin V⁺PI⁻). Data are presented as mean ± SD (n = 3 samples). **p < 0.01, ***p < 0.001. Cell death analyses were performed in three independent experiments. CHCl₃-E: chloroform extract, Hexane-E: hexane extract, MeOH-E: methanol extract.

Anti-inflammatory activity of CS-AR extracts

We evaluated the anti-inflammatory effects of CS-AR extracts on DCs, independent of their cytotoxicity. DCs were treated with LPS to induce inflammation. Subsequently, we analyzed the expression levels of various inflammatory cytokines by treating DCs with CS-AR extracts alone. In addition, we evaluated their anti-inflammatory activity by co-stimulating DCs with LPS and CS-AR extracts. We found that treatment with CS-AR extracts (10, 50, 100 µg/mL) alone did not alter the expression levels of pro-inflammatory cytokines (TNF-α, IL-6, IL-1β, IL-12p70) compared to that in control cells (DMSO-treated cells) (Fig. 2A, B, C). However, when the cells were co-treated with LPS and 10 µg/mL of CS-AR extracts (Fig. 2A), the MeOH-E significantly inhibited the production of TNF-α, IL-6, and IL-12p70 compared to LPS-treated cells. In contrast, CHCl₃-E and Hexane-E only suppressed IL-12p70 production, and CHCl₃-E increased IL-1β production compared to LPS-treated cells. Furthermore, when DCs were co-treated with LPS and 50 µg/mL of CS-AR extracts (Fig. 2B), MeOH-E exhibited similar inhibitory effects on TNF-α, IL-6, and IL-12p70 expressions as observed with 10 µg/mL of the extract. However, it did not suppress IL-1β production. Meanwhile, CHCl₃-E and Hexane-E treatments reduced IL-6 and IL-12p70 levels but increased the IL-1β level. Finally, when cells were co-treated with LPS and 100 µg/mL of CS-AR extracts (Fig. 2C), MeOH-E showed an increasing inhibition of pro-inflammatory cytokine secretion (TNF-α, IL-6, IL-1β, IL-12p70) with increasing concentrations. Interestingly, treatment with 100 µg/mL of CHCl₃-E and Hexane-E inhibited IL-6 and IL-12p70 production and further suppressed TNF-α production. However, no inhibitory effect on IL-1β production was observed. These results suggest that MeOH-E is not cytotoxic to DCs and can elicit a strong anti-inflammatory response in mature DCs.

Fig. 2.

Anti-inflammatory activity of CS-AR extracts. BMDCs were stimulated with CS-AR extracts (10, 50, or 100 µg/mL) in the absence or presence of lipopolysaccharide (LPS) for 18 h. (A) Cytokine levels (TNF-α, IL-6, IL-12p70, IL-1β) in culture supernatants from BMDCs treated with 10 µg/mL CS-AR extracts. (B) Cytokine levels in culture supernatants from BMDCs treated with 50 µg/mL CS-AR extracts. (C) Cytokine levels in culture supernatants from BMDCs treated with 100 µg/mL CS-AR extracts. All experiments were performed in triplicate, yielding consistent results. Data shown represent the most typical findings. Data are presented as mean ± SD (n = 3). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001. CHCl₃-E: chloroform extract, Hexane-E: hexane extract, MeOH-E: methanol extract.

Multivariate analysis and annotation of discriminative chemical markers in CS-AR extracts

Based on our findings, we hypothesized that the differences in functionality and cytotoxicity among CS-AR extracts might be attributed to variations in their metabolite profiles. To investigate this, chemical composition of the three extracts was profiled using ultra-performance liquid chromatography–quadrupole-time-of-flight mass spectrometry (UPLC-Q-TOF MS) operated in both negative and positive electrospray ionization (ESI) modes (Fig. 3A, B). To visualize and discriminate extract-specific metabolites reflecting the compositional differences among the three CS-AR extracts, principal component analysis (PCA) was performed, including both score plots and loading plots. In the negative ESI mode, the PCA score plot revealed that principal component 1 (PC1) and principal component 2 (PC2) accounted for 60.5% and 23.3% of the total variance, respectively (Fig. 3C). Samples were clearly grouped into three distinct clusters, each corresponding to a specific extract, with each data point representing an individual replicate. The corresponding loading plot highlighted discriminative metabolites responsible for the separation among extract groups (Fig. 3D). Tentative annotation of the chemical markers was performed by comparing the proposed molecular formulas—derived from accurate mass and MS/MS fragmentation ions—with existing databases and published literature (Table 1 and Supplementary Fig. 1). Specifically, cannabinoid glycoside derivatives, cannabigerolic acid-O-acetate (CBGA-O-acetate; 2), 1-monooleoylglycerol, and an ion with m/z 605.4054 contributed significantly to group differentiation in the negative mode. In the positive ESI mode, PC1 and PC2 explained 66.0% and 18.5% of the variance, respectively (Fig. 3E), again yielding three well-separated clusters corresponding to the different extraction solvents. The PCA loading plot (Fig. 3F) indicated that cannabinoid glycoside derivatives, cannabidiol diacetate derivatives, cannabidiol mono-acetate mono-benzoate, and a compound with m/z 753.5727 served as discriminative chemical markers in the positive mode. A total of eight marker compounds were selected across both ionization modes as key discriminants among the CS-AR extracts, and their relative ion intensities were compared (Fig. 4 and Supplementary Table 1). Among the eight markers, cannabinoid glycoside derivative, CBGA-O-acetate, and cannabinoid glycoside derivative exhibited higher intensities in MeOH-E compared to the other extracts, and were least abundant in the Hexane-E. In contrast, cannabigerol, cannabidiol diacetate derivatives, and cannabidiol mono-acetate mono-benzoate were most abundant in Hexane-E. The ion at m/z 605.4054 showed high intensity only in CHCl₃-E, while the compound at m/z 753.5727 was more abundant in CHCl₃-E and Hexane-E than in MeOH-E.

Fig. 3.

Comparative LC-MS profiling and multivariate statistical analysis of CS-AR extracts. Base peak intensity (BPI) chromatograms of different CS-AR solvent fractions acquired in (A) negative and (B) positive ionization modes, respectively. (C) Principal component analysis (PCA) score plot and (D) loading plot of metabolites (n = 9) from different CS-AR extracts analyzed in negative ionization mode. (E) PCA score plot and (F) loading plot of metabolites (n = 9) from different CS-AR extracts analyzed in positive ionization mode. Red dotted circles in the loading plots indicate extract-specific tentatively selected discriminative metabolites. CHCl₃-E: chloroform extract, Hexane-E: hexane extract, MeOH-E: methanol extract.

Table 1.

Tentatively annotation of selected chemical markers using UPLC-Q-TOF-MS analysis.

No.	RT, min	Molecular formula	Detected ion [M-H]	Calculated ion [M-H]	Error (ppm)	MS/MS fragment ions	Tentative annotation	Ref.
1	8.31 (Neg)	C23H44O19	623.2397	623.2399	0.2	460, 445, 297	Cannabinoid glycoside derivatives	27
2	13.63 (Neg)	C18H34O4	313.2377	313.2379	−3.8	295, 255, 183	1-Monooleoylglycerol	28
3	16.21 (Neg)	C23H34O7	421.2260	421.2226	6.2	397, 353, 325	Cannabigerolic acid-O-acetate	29
4	18.93 (Neg)	C35H58O8	605.4054	605.4053	0.5	531, 513, 355	605.4054
5	8.30 (Pos)	C41H38O6	625.2573	625.2590	−2.7	462, 351, 325	Cannabinoid glycoside derivatives	27
6	15.81 (Pos)	C31H50O4	485.3646	485.3631	2.7	425, 407, 295	Cannabidiol diacetate derivatives	30
7	17.82 (Pos)	C33H52O5	527.3752	527.3736	2.5	467, 415, 184	Cannabidiol mono-acetate mono-benzoate	30
8	19.13 (Pos)	-	758.5727	-	-	-	758.5727

UPLC-Q-TOF-MS, ultra-high-performance liquid chromatography quadrupole time-of-flight mass spectrometry; No., number; RT, retention time; CHCl₃-E, chloroform extracts; Neg, negative ESI mode; Pos, positive ESI mode; Ref, reference.

Fig. 4.

Analysis of the relative intensities of tentatively identified chemical markers across CS-AR extracts obtained in negative (A) and positive (B) ionization modes. Data are presented as mean ± SD (n = 9). ***p < 0.001. CHCl₃-E: chloroform extract, Hexane-E: hexane extract, MeOH-E: methanol extract.

Inhibitory effect of CS-AR MeOH-E on DC maturation

As MeOH-E exhibited a strong anti-inflammatory activity on DC without cytotoxicity, we investigated whether it possessed additional immunological characteristics. We examined its effects on DC maturation by evaluating the expression levels of surface markers. MeOH-E inhibited the expression of surface markers (CD80, CD86, MHC-II) induced by LPS stimulation in a concentration-dependent manner. In contrast, cells treated with MeOH-E alone showed no changes in the expression of these surface markers compared to control cells (DMSO-treated cells) (Fig. 5A). These findings suggest that MeOH-E does not affect the maturation of immature DCs but can inhibit the maturation process of DCs during LPS stimulation. Based on this observation, all subsequent experiments focused on the effects of MeOH-E on DC maturation during the LPS-induced maturation process. Next, we assessed the antigen-uptake ability, which typically decreases during DC maturation. When LPS-treated DCs were stimulated with MeOH-E, their antigen-uptake ability, measured using dextran as a model antigen, was significantly enhanced compared to LPS-treated DCs alone (Fig. 5B). This phenomenon was observed only under physiological conditions (37 °C) and was absent at non-permissive temperatures (4 °C), which indicated that the effects were dependent on active cellular processes. Finally, we evaluated the antigen-presenting ability, which increases during DC maturation. When DCs were treated with the Eα₄₄₋₇₆ antigen, the presentation of the Eα₅₂₋₆₈ peptide via MHC-II (measured by Y-Ae⁺CD11c⁺ cells) was suppressed by MeOH-E treatment (Fig. 5C). These findings indicate that CS-AR MeOH-E can broadly regulate various functions that increase or decrease during DC maturation, thereby promoting the maintenance of dendritic cells in an immature state.

Fig. 5.

Surface molecule expression, antigen-uptake ability, and antigen-presenting ability in CS-AR MeOH-E-treated DCs. (A) BMDCs were stimulated with CS-AR extracts (10, 50, or 100 µg/mL) in the absence or presence of lipopolysaccharide (LPS) for 18 h. After stimulation, the surface molecules CD80, CD86, and MHC-II were stained, and the expression levels of these surface markers on DCs (CD11c⁺) were measured. (B) BMDCs (untreated, LPS-treated, or LPS + MeOH-E-treated for 18 h) were cultured with dextran for 30 min, and the dextran-uptake ability (CD11c⁺dextran⁺ cells) was evaluated. (C) The antigen-presenting ability of MHC-II was assessed by measuring the frequency of CD11c⁺Y-Ae⁺ cells in BMDCs treated with Eα₅₂₋₆₈ (positive control), Eα₄₄₋₇₆, or Eα₄₄₋₇₆ + MeOH-E. All experiments were performed in triplicate, yielding consistent results. Data shown represent the most typical findings. Data are presented as mean ± SD (n = 3). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001. MeOH-E: methanol extract.

Inhibitory effects of CS-AR MeOH-E on mitogen-activated protein kinase (MAPK) and nuclear factor kappa-B (NF-κB) signaling

The activation of the MAPK and NF-κB signaling pathways is critical for DC maturation. Therefore, we hypothesized that MeOH-E from CS-AR mediated its inhibitory effect on DC maturation by suppressing MAPK and NF-κB signaling. LPS alone-treated DCs exhibited time-dependent phosphorylation of ERK, JNK, and p38, thereby indicating the activation of MAPK signaling. However, co-treatment with LPS and MeOH-E reduced the LPS-induced phosphorylation of JNK and p38. Interestingly, ERK phosphorylation decreased at early time points (10 min post-treatment) with MeOH-E but showed an increase at later time points (30 and 60 min). Next, we assessed NF-κB activation by evaluating the nuclear translocation of p65. MeOH-E treatment demonstrated a significant inhibitory effect on NF-κB activation both at early and late time points (Fig. 6). These findings suggest that MeOH-E inhibits DC maturation by suppressing MAPKs, particularly JNK and p38, as well as the NF-κB signaling pathway.

Fig. 6.

Inhibition of MAPK and NF-kB signaling in DCs co-treated with LPS and CS-AR MeOH-E. BMDCs were treated with LPS alone or co-treated with LPS and MeOH-E at various time points. Phosphorylation levels of ERK, JNK, and p38, as well as translocation of NF-κB p65 to the nucleus, were analyzed using western blot. Experiments were performed in triplicate, yielding consistent results. Data shown represent the most typical findings. Data are presented as mean ± SD (n = 3). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001. MeOH-E: methanol extract.

Effects of CS-AR MeOH-E treatment on T cell proliferation and activation

To further validate the immunosuppressive properties of CS-AR MeOH-E, we extended our investigation to T cells. Stimulation with anti-CD3/CD28 induces T cell proliferation and activation. Interestingly, when CFSE-labeled T cells were co-treated with anti-CD3/CD28 and MeOH-E, T cell proliferation was significantly reduced in a concentration-dependent manner (Fig. 7A). Moreover, the production of IFN-γ, a key T cell activation marker induced by anti-CD3/CD28 stimulation, was markedly suppressed by MeOH-E treatment (Fig. 7B). These results demonstrate that CS-AR MeOH-E not only inhibited DC function, but also suppressed T cell proliferation and activation, thereby highlighting its potential anti-inflammatory and immunosuppressive effects.

Fig. 7.

Inhibition of T cell proliferation and activation induced by the CS-AR MeOH-E. Spleen cells isolated from mouse spleens were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE) and co-stimulated with anti-CD3/CD28 and MeOH-E. T cell proliferation and activation levels were measured as follows: (A) T cell proliferation levels (CFSE division levels) were analyzed via flow cytometry. (B) T cell activation levels were assessed by measuring IFN-γ levels in the culture supernatant. Experiments were performed in triplicate, yielding consistent results. Data are presented as mean ± SD (n = 3). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001. MeOH-E: methanol extract.

Effects of CS-leaf MeOH-E and CS-AR MeOH-E on cytotoxicity and anti-inflammatory activity

We next compared the anti-inflammatory activities of MeOH-E derived from CS-AR (CS-AR MeOH-E) and CS-leaf (CS-leaf MeOH-E). Prior to evaluating their anti-inflammatory effects, we first evaluated their cytotoxic effects. DCs were treated with 100 µg/mL of either extract for 18 h, and cell viability was assessed via Annexin V and PI staining. As shown in Fig. 8A, neither extract induced cytotoxicity under the tested conditions. Next, to assess their anti-inflammatory potential, DCs were stimulated with LPS and subsequently treated with 100 µg/mL of either CS-leaf MeOH-E or CS-AR MeOH-E. CS-AR MeOH-E treatment resulted in a markedly greater suppression of pro-inflammatory cytokine production compared to CS-leaf MeOH-E. Specifically, CS-AR MeOH-E significantly reduced levels of TNF-α (p < 0.01), IL-6 (p < 0.001), and IL-12p70 (p < 0.01), while no significant difference in IL-1β production was observed between the two groups (Fig. 8B). These results indicate that the MeOH-E from CS-AR exhibits superior anti-inflammatory activity relative to that of CS-leaf MeOH-E.

Fig. 8.

Cytotoxicity and anti-inflammatory activity of CS-leaf MeOH-E and CS-AR MeOH-E. (A) BMDCs were treated with 100 µg/mL CS-leaf MeOH-E, 100 µg/mL CS-AR MeOH-E, or 50 nM staurosporine (STS) for 18 h. Apoptosis and necrosis were assessed using Annexin V and propidium iodide (PI) staining followed by flow cytometry. Bar graphs showing the analysis of necrotic cell death (Annexin V⁻PI⁺), late apoptotic cell death (Annexin V⁺PI⁺), and early apoptotic cell death (Annexin V⁺PI⁻). Data are presented as mean ± SD (n = 3 samples). ***p < 0.001. All experiments were independently repeated three times with consistent results. (B) BMDCs were stimulated with 100 µg/mL CS-leaf MeOH-E or CS-AR MeOH-E in the presence of lipopolysaccharide (LPS) for 18 h. The levels of pro-inflammatory cytokines (TNF-α, IL-6, IL-12p70, and IL-1β) in culture supernatants were measured by ELISA. All experiments were performed in triplicate, yielding consistent results. Data shown represent the most typical findings. Data are presented as mean ± SD (n = 3). Statistical significance: **p < 0.01, ***p < 0.001. Leaf-MeOH-E: methanol extract from CS-leaf; AR-MeOH-E: methanol extract from CS-AR.

Discussion

This study aimed to evaluate the potential of the CS-AR system as an alternative and sustainable source of bioactive compounds, with particular emphasis on its applicability in regulating inflammation and immune-mediated diseases. To assess the biological relevance of CS-AR-derived metabolites, we first compared the cytotoxic and anti-inflammatory effects of extracts obtained using three different solvents—methanol (MeOH-E), chloroform (CHCl₃-E), and hexane (Hexane-E). Functional assays using DCs revealed distinct extract-dependent effects, with MeOH-E exhibiting the strongest anti-inflammatory activity while maintaining minimal cytotoxicity compared to the other two extracts. Metabolomic profiling further identified the presence of various cannabinoids and their derivatives across all extracts, including cannabinoid glycoside derivatives, CBGA-O-acetate, cannabidiol diacetate derivatives, and cannabidiol mono-acetate mono-benzoate. Notably, cannabinoid glycoside derivatives and CBGA-O-acetate were more abundant in MeOH-E than in CHCl₃-E or Hexane-E. Given the superior anti-inflammatory activity of MeOH-E, we further investigated its immunomodulatory effects on LPS-stimulated, functionally mature DCs. MeOH-E effectively maintained DCs in an immature state by downregulating the expression of surface maturation markers and antigen-presenting molecules, while simultaneously enhancing antigen uptake and attenuating activation of the MAPK and NF-κB signaling pathways. Collectively, these findings support the feasibility of the CS-AR system as a viable platform for the production of immunoregulatory metabolites from C. sativa. Furthermore, MeOH-E demonstrates strong potential as a functionally safe and effective extract for modulating DC maturation and suppressing pathological immune responses.

One of the most noteworthy findings of this study is the demonstration that major bioactive constituents of C. sativa can be alternatively produced through the AR system. Moreover, the MeOH-E obtained from CS-AR exhibited higher biological activity compared to CS-leaf MeOH-E, thereby providing critical foundational evidence supporting the utility and superiority of the CS-AR system. In fact, the AR system has been widely recognized as a valuable biotechnological platform for the stable and large-scale production of pharmacologically active compounds in various medicinal plant species^25,31. Although a complete annotation of all metabolites was not performed in our analysis, the presence of cannabinoid derivatives and analogs—key functional constituents of C. sativa—was clearly confirmed in the CS-AR extract^1,32. This observation underscores the biosynthetic capacity of the CS-AR system and highlights its potential as an alternative source for cannabinoid-based therapeutics, similar to how AR systems are applied in other medicinal plants. Furthermore, the superior anti-inflammatory activity observed in CS-AR MeOH-E, compared to the CS-leaf MeOH-E, suggests that the AR system may offer a distinct advantage by enabling the production of compounds with enhanced immunoregulatory potential. These findings collectively point to the CS-AR system as a promising and scalable strategy for generating bioactive metabolites with therapeutic relevance.

Next, it is worth noting that the ability of MeOH-E to modulate DC maturation suggests promising therapeutic potential for the management of inflammatory diseases. DCs are pivotal mediators in the pathogenesis of inflammatory conditions owing to their role in initiating and amplifying immune responses through antigen presentation, co-stimulatory signaling, and cytokine production^33,34. These three signals are critical for T cell activation, which is a key driver in numerous inflammatory and autoimmune diseases^35,36. Dysregulation of one or more of these signals can inhibit T cell activation or induce T cell anergy, thereby offering a strategic target for therapeutic intervention^37,38. For instance, RNA interference approaches and active compounds such as 1,25-dihydroxyvitamin D3 have shown efficacy in suppressing co-stimulatory molecules and pro-inflammatory cytokine production, validating their potential in immune regulation^39,40. Furthermore, therapeutic strategies targeting antigen-presenting pathways, particularly in conditions such as rheumatoid arthritis, have demonstrated significant potential to disrupt chronic T cell activation and inflammation⁴¹. Aligned with these insights, the ability of MeOH-E to inhibit key aspects of DC maturation, including cytokine production, surface molecule expression, and antigen presentation, positions it as a strong candidate for therapeutic development. Its pronounced inhibition of MAPK and NF-κB signaling further underscores its mechanism of action, as these pathways are crucial in the activation of innate immune cells and are well-established therapeutic targets in inflammatory disease management^42–44.

Finally, a notable observation is that the differences in cytotoxicity and anti-inflammatory activity among the extracts may be attributed to compositional and quantitative variations in their metabolite profiles. Although our metabolomic analysis showed that MeOH-E, CHCl₃-E, and Hexane-E share a substantial portion of common metabolites, several compounds—including cannabinoid glycoside derivatives, CBGA-O-acetate, cannabidiol diacetate derivatives, and cannabidiol mono-acetate mono-benzoate—exhibited marked differences in abundance across the extracts. These discrepancies are likely responsible for the distinct bioactivities observed among the solvent fractions. More specifically, CHCl₃-E and Hexane-E demonstrated anti-inflammatory effects on TNF-α, IL-6, and IL-12p70 but failed to suppress IL-1β production. In contrast, MeOH-E exhibited broad-spectrum anti-inflammatory activity by effectively reducing the levels of all tested inflammatory cytokines, including IL-1β. This is of particular interest, as sustained IL-1β production has been implicated in promoting cell death⁴⁵suggesting that the elevated IL-1β levels observed in response to CHCl₃-E and Hexane-E may be due to the higher abundance of IL-1β-inducing metabolites in these extracts compared to MeOH-E. Furthermore, unlike CHCl₃-E and Hexane-E, MeOH-E was enriched in modified or derivative forms of cannabinoids—such as cannabinoid glycoside derivatives and CBGA-O-acetate—that are known to possess anti-inflammatory properties^46–49. The absence or lower levels of these metabolites in CHCl₃-E and Hexane-E may account for their weaker anti-inflammatory activity and limited ability to regulate IL-1β production. Taken together, although the solvent-dependent differences in CS-AR-derived metabolite profiles are expected, our findings suggest that solvent selection can serve as a simple yet effective strategy to enrich beneficial metabolites while excluding potentially cytotoxic components. This highlights the utility of MeOH extraction-based approaches for selectively isolating bioactive compounds from CS-AR for therapeutic applications.

Despite these promising findings, several limitations should be acknowledged. This study was conducted exclusively in vitro, and the in vivo safety and efficacy of MeOH-E remain to be determined. Therefore, comprehensive preclinical evaluation is required to assess its therapeutic potential for inflammatory diseases. Future studies should also focus on evaluating the individual bioactivities of key metabolites that are likely to contribute most significantly to the observed immunomodulatory and cytotoxic effects. In addition, expanded metabolomic profiling is warranted to identify currently uncharacterized compounds, and structural elucidation of non-annotated metabolites should be pursued using complementary analytical techniques such as nuclear magnetic resonance (NMR) spectroscopy. These efforts will enable a more accurate interpretation of our current findings and provide deeper insights into the additional advantages and potential applications of the AR system proposed in this study.

In conclusion, our findings provide strong evidence that the CS-AR system can serve as a plant tissue culture-based platform for the production of major bioactive constituents of C. sativa. Moreover, this system offers a sustainable and effective source of bioactive metabolites with therapeutic potential for the treatment of inflammatory diseases. These discoveries offer valuable insights into the development of CS-AR-based natural therapeutics targeting inflammatory and immune-related disorders, and propose an innovative strategy for pharmaceutical development using C. sativa.

Methods

Animals

All experiments involving BMDCs and T cells derived from C57BL/6 mice were approved by the Animal Ethics Committee of the Korea Research Institute of Bioscience and Biotechnology (KRIBB; approval number KRIBB-AEC-23070), and conducted in accordance with institutional guidelines. The study is reported in accordance with the ARRIVE guidelines (https://arriveguidelines.org). Female C57BL/6 mice (7 weeks old, ~ 18 g) were purchased from Orient Bio (Seongnam, Republic of Korea) and acclimated for one week before use. At 8 weeks of age, mice were anesthetized with 2–3% isoflurane (Ifran; Hana Parm., Seoul, Republic of Korea) via inhalation and euthanized by cervical dislocation. All animals were housed under barrier conditions in a BL-2 biohazard animal facility at KRIBB, maintained at a constant temperature (24 ± 1 °C) and relative humidity (50 ± 5%). Sterilized commercial mouse chow and autoclaved water were provided ad libitum under a 12 h light/dark cycle.

AR induction and growth conditions

To obtain ARs from the seeding of C. sativa, the seeds of the Korean hemp cultivar ‘Cheungsam’ were surface-sterilized by immersion in a 1% hydrogen peroxide solution. The seeds were then incubated in the dark on a rotary shaker, with the 1% H₂O₂ solution renewed daily for 2 days. After removing the seed coat from the germinated seeds, the embryos were surface-sterilized with 1% H₂O₂ and rinsed thrice with sterilized water. They were then dried with 3 M paper and placed on a 1/2 Murashige & Skoog (M-S) medium containing 15 g/L sucrose and 4 g/L gelite (pH 5.8). The embryos were cultivated at 25 °C with 16 h light and 8 h dark photoperiod. To induce CS-ARs, leaf explants were excised from 7-day-old seedlings and cultured on an M-S medium supplemented with 0.1 mg/L α-naphthaleneacetic acid (NAA), 15 g/L sucrose, and 4 g/L gelite (pH 5.8) under dark conditions at 22 °C. The CS-ARs were then sub-cultured every 2 weeks to promote proliferation.

Extraction of metabolites from CS-ARs and CS-leaf using hexane, chloroform, and methanol

CS-ARs and CS-leaf were dried in a dry oven at 65 °C for 4 days yielding 30 g of dry sample. The dried sample was ground into a fine powder using a high-capacity blender (HBL-S650R, Samyang Electronics). The powdered sample was then divided into three equal portions (10 g each) for extraction using either 150 mL of hexane, chloroform, or methanol solvents, respectively. The samples were placed in 1 L bottles and subjected to shaking for 2 h using a funnel shaker (EYELA MMV-1000 W). After shaking, the mixtures were stored at 25 °C for 1 week. Following the storage period, the mixtures were filtered using qualitative filter paper (No.20, ⦰110 nm, HYUNDAI MICRO) and a vacuum pump (GAST) to obtain pure extracts. The collected extracts were then concentrated using a rotary evaporator (EYELA N-1100) with a water bath (EYELA OBS-2100). The final yields of the concentrated extracts from CS-ARs were as follows: hexane extract (Hexane-E: 29.7 mg), chloroform extract (CHCl₃-E: 74.9 mg), and methanol extract (MeOH-E: 2.6042 g). For CS-leaf, the yield of methanol extract (CS-leaf MeOH-E) was 3.6701 g.

UPLC-Q-TOF MS analysis

The metabolite analysis of CS-AR extracts (Hexane-E, CHCl₃-E, and MeOH-E) was performed using an ACQUITY I-Class UPLC system (Waters Corp., Milford, MA, USA). Chromatographic separation was achieved on an ACQUITY BEH C18 column (1.7 μm, 2.1 mm × 100 mm). The column oven was maintained at 35 °C, and the sample tray was kept at 4 °C. The mobile phases consisted of solvent A (0.1% formic acid in water, v/v) and solvent B (0.1% formic acid in acetonitrile, v/v). The gradient elution profile was as follows: 0–1 min, 5% B; 1–20 min, linear increase to 100% B; 20–22.3 min, held at 100% B; 22.3–22.4 min, returned to 5% B; and 22.4–25 min, equilibrated at 5% B. The flow rate was set to 0.4 mL/min, and the injection volume was 2 µL. Mass spectrometric analysis was performed using a VION IMS Q-TOF mass spectrometer (Waters Corp.) operated in both positive and negative electrospray ionization (ESI) modes. Data were acquired in MS^E mode, which alternates between low and high collision energy scans. The parameters were configured as follows: capillary voltage, 3.0 kV; source temperature, 110 °C; cone voltage, 40 V; cone gas flow, 50 L/h; desolvation temperature, 350 °C; and desolvation gas flow, 800 L/h. Accurate mass measurements were achieved using a continuously infused lock-mass reference compound (leucine-enkephalin, m/z 556.2771 for ESI + and m/z 554.2615 for ESI−) via an automated calibration delivery system. Data collection, normalization, and alignment were carried out using UNIFI software (version 1.9.2.045, Waters). Statistical analysis and data visualization were conducted using SIMCA-P + 12.0 (Umetrics, Umeå, Sweden).

Differentiation of BMDCs

BMDCs were differentiated from bone marrow cells isolated from the femur and tibia of 8-week-old female C57BL/6 mice. The BM cells were cultured in RPMI-1640 medium (Gibco BRL, Grand Island, NY, USA) supplemented with 10% heat-inactivated fetal bovine serum (Gibco BRL), 1% penicillin/streptomycin (Gibco BRL), 20 ng/mL Granulocyte-macrophage colony-stimulating factor (GM-CSF; JW Creagene, Gyeonggi, Republic of Korea), and 0.5 ng/mL interleukin-4 (IL-4; JW Creagene) for 8 days to induce differentiation into BMDCs. All immunological evaluations were performed using BMDCs on day 8 of differentiation.

Cellular toxicity analysis

BMDCs (1 × 10⁶ cells per 48-well) were treated with various concentrations (10, 50, 100 µg/mL) of CS-AR extracts (Hexane-E, CHCl₃-E, and MeOH-E) in a 37 °C CO₂ incubator for 18 h. Control and staurosporine (STS; Sigma-Aldrich, St. Louis, MO, USA)-treated cells were exposed to equivalent volumes of DMSO, and corresponded to the amount used in the extract treatments. Following treatment, the cells were stained with Annexin V and propidium iodide (PI) according to the manufacturer’s protocol (Thermo Fisher Scientific, Waltham, MA, USA). The stained cells were then analyzed using a Life Launch Attune Nxt Flow Cytometer (Thermo Fisher Scientific).

Analysis of specific cytokines in BMDCs using specific enzyme-linked immunosorbent assay (ELISA)

BMDCs (1 × 10⁶ cells per 48-well) were stimulated with various concentrations (10, 50, 100 µg/mL) of CS-AR extracts (Hexane-E, CHCl₃-E, and MeOH-E) in the presence or absence of 100 ng/mL lipopolysaccharide (LPS; InvivoGen, San Diego, Calif) in a 37 °C CO₂ incubator for 18 h. Control and LPS-treated cells were exposed to equivalent volumes of DMSO that corresponded to the amount used in the extract treatments. After stimulation, culture supernatants were collected, and cytokine levels (TNF-α, IL-6, IL-1β, IL-12p70 from Thermo Fisher Scientific) were measured using cytokine-specific ELISA kits according to the manufacturer’s protocol.

Analysis of surface molecule expression, antigen uptake, and antigen-presenting ability in BMDCs

Surface molecule expression in BMDCs was assessed following the protocol used for cytokine analysis. Briefly, stimulated cells (non-treated, MeOH-E-treated, LPS-treated, and MeOH-E + LPS-treated) were harvested after 18 h. These cells were then stained with antibodies specific to DC surface molecules (CD11c-PE-Cy7, CD80-APC, CD86-FITC, MHC-II-PerCp-Cy5.5; Thermo Fisher Scientific) at 4 °C for 30 min. Flow cytometry was performed to evaluate the expression levels of these surface markers. For the analysis of antigen uptake, BMDCs were stimulated with either LPS (100 ng/mL) alone or in combination with MeOH-E (100 µg/mL) for 18 h. After stimulation, the cells were incubated with antigen mimic FITC-Dextran (Sigma-Aldrich) for 30 min at either 37 °C–4 °C. Following incubation, the cells were washed thrice with PBS, stained with CD11c-PE-Cy7, and analyzed for Dextran⁺CD11c⁺ cells using flow cytometry. The antigen-presenting ability of BMDCs was assessed by treating the cells with the Eα₅₂₋₆₈ peptide (directly binding to MHC-II; sequence: ASFEAQGALANIAVDKA; AnaSpec, San Jose, CA, USA) for 2 h, or antigen mimic peptide (Eα₄₄₋₇₆; RLEEFAKFASFEAQGALANIAVDKANLDVMKKR) and MeOH-E (100 µg/mL) for 18 h, either individually or in combination. Post-incubation, the cells were collected and stained with CD11c-PE-Cy7 and Y-Ae-FITC antibodies (specific for MHC-II/OVA complex detection; Thermo Fisher Scientific). The antigen-presenting capability was then analyzed using flow cytometry.

Western blotting

BMDCs (2 × 10⁶ cells per 6-well plate) were treated with either LPS (100 ng/mL) alone or with a combination of LPS and MeOH-E (100 µg/mL). To obtain cytosolic fractions, cells were harvested at various time points (0, 10, 30, and 60 min), washed twice with PBS, and lysed using 200 µL of M-PER Mammalian Protein Extraction Reagent containing Halt Protease and Phosphatase Inhibitor Cocktail (Thermo Fisher Scientific). Nuclear extracts were obtained by lysing cells using a ClLytic NuCLEAR Extraction Kit (Sigma-Aldrich) according to the manufacturer’s protocol. The cells were lysed by incubating them at 4 °C for 10 min. Following lysis, the cell extracts were centrifuged at 12,000 rpm for 10 min to obtain clear protein lysates. Protein concentrations were quantified using the BCA protein assay. The quantified proteins were subjected to western blot analysis according to previously established protocols⁵⁰. The antibodies used for the Western blot analysis-including anti-phosphorylated extracellular signal-regulated kinase (anti-p-ERK1/2), anti-phosphorylated c-Jun N-terminal kinase (anti-p-JNK), anti-phosphorylated p38 (anti-p-p38), and anti-nuclear factor kappa B (NF-κB, p65)-were purchased from Cell Signaling Technology (Beverly, MA, USA) and used at a dilution of 1:1,000. Anti-β-actin and anti-Lamin B antibodies were purchased from Sigma-Aldrich and used at a dilution of 1:2,000.

Inhibition assay for T cell activation and proliferation

Splenocytes were isolated from 8-week-old female C57BL/6 mice, followed by red blood cell (RBC) lysis using RBS lysis buffer (Thermo Fisher Scientific). The resulting single-cell suspensions were labeled with CellTrace™ carboxyfluorescein diacetate succinimidyl ester (CFSE) dye (1 µM; Thermo Fisher Scientific) by incubating at 37 °C in the dark for 15 min. CFSE-labeled cells (0.5 × 10⁶ cells) were seeded into 96-well U-bottom plates and subsequently stimulated with Dynabeads™ Mouse T-Activator CD3/CD28 (1 × 10⁵ beads; Thermo Fisher Scientific) in the presence of MeOH-E (50 and 100 µg/mL). After 3 days of culture, the cells were harvested, washed once with PBS, and analyzed for CFSE dilution using flow cytometry to assess T cell proliferation. Culture supernatants were collected and used to measure IFN-γ levels using ELISA.

Statistical analysis

All experiments were performed in triplicate under identical conditions. Statistical analysis was conducted using GraphPad Prism Software (version 9; GraphPad Software, San Diego, CA, USA). The significance of differences between samples was determined using one-way ANOVA followed by Tukey’s multiple comparison test and unpaired t-test.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Su Hyun Park: Writing – original draft, Methodology, Formal analysis, Resources, Data curation. Jeong Moo Han: Methodology, Formal analysis, Data curation, Software, Validation. Yun Hye Kim: Formal analysis, Data curation. Hyeon Jin Lee: Formal analysis, Data curation. Young Bae Ryu: Investigation, Funding acquisition. Hyung Won Ryu: Methodology, Formal analysis, Data curation. Jae Cheol Jeong: Methodology, Resources, Project administration, Funding acquisition. Seon Min Oh: Methodology, Formal analysis, Data curation, Software, Writing – review & editing. Woo Sik Kim: Conceptualization, Writing – original draft, Methodology, Data curation, Validation, Supervision, Writing – review & editing. All authors reviewed the manuscript.

Funding

This work was supported by the Korea Research Institute of Bioscience and Biotechnology Research Initiative Programs [grant numbers KGM5382521 and KGM1052511].

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.

Ethics statement

All experiments involving bone marrow-derived dendritic cells (BMDCs) and T cells derived from C57BL/6 mice were approved by the Animal Ethics Committee of the Korea Research Institute of Bioscience and Biotechnology (KRIBB; approval number KRIBB-AEC-23070), and conducted in accordance with institutional guidelines. The study is reported in accordance with the ARRIVE guidelines (https://arriveguidelines.org).