Plant materials and seed collection

The seeds of seven forest tree species—A. japonica, C. obtusa, C. kousa, P. amurense, P. densiflora, P. sargentii, and Q. glauca—were obtained from the National Forest Seed and Variety Center (NFSV; Chungju, Korea). All seeds were harvested between 2017 and 2023 from state-managed seed orchards located in various regions of South Korea. Until extraction, the seeds were stored under low-temperature, dry conditions (4℃). Table 1 provides the scientific names of these species, their family names, production sites, and quality metrics. Here, only seed lots classified as preliminary grade under the NFSV quality control standards were used. According to this classification, seeds are excluded from national afforestation and propagation programs if their germination rate or efficiency values fall below the standardized thresholds for forest restoration. All seed materials used in this study were obtained through the official seed-distribution procedure of the NFSV under formal seed-distribution authorization, and originated exclusively from state-managed seed orchards rather than from natural or protected populations; accordingly, no field collection permit under the relevant Korean forest management regulations was required.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. List of forest trees species used.

https://doi.org/10.1371/journal.pone.0345845.t001

Extraction process

The seeds of the seven species were freeze-dried and ground using a milling machine. To minimize processing variability across species, all extractions were performed in a single batch under identical protocols. For extraction, 1 g of seed powder was mixed with 50 mL of 70% ethanol, followed by ultrasonic extraction for 30 min. The extracts were used to analyze the total polyphenol and flavonoid content (TPC and TFC, respectively) and antioxidant activity. For the evaluation of anti-inflammatory, anticancer, and tyrosinase-inhibitory activities, the extracts were prepared as follows: after solvent removal using a rotary vacuum evaporator (N-1000-SW, EYELA, Tokyo, Japan), the residues were re-dissolved in dimethyl sulfoxide (DMSO) at a concentration of 100 mg·mL ⁻ ¹ and passed through a 0.45-μm syringe filter.

Determination of TPC and TFC

TFC was determined using the diethylene glycol (DEG)-sodium hydroxide (NaOH) colorimetric method described by Davis [20]. A reaction mixture consisting of 200 μL of the extract, 2 mL of DEG, and 200 μL of 1 N NaOH was prepared and incubated in a water bath at 37°C for 1 h. Thereafter, the absorbance was measured at 420 nm using a spectrophotometer. A standard calibration curve was constructed using quercetin, and the TFC is expressed as quercetin equivalents (QE) per gram of dry weight (QE·mg g ⁻ ¹ DW).

TPC was determined using a modified Folin–Denis method [21]. A 100-μL aliquot of the extract was mixed with 2 mL of 2% sodium carbonate solution and allowed to react for 3 min, and then 100 μL of 50% Folin–Ciocalteu reagent was added. The mixture was incubated in the dark for 30 min, and the absorbance was measured at 750 nm using a spectrophotometer. A standard curve was prepared using gallic acid, and TPC is expressed as milligrams of gallic acid equivalents (GAE) per gram of dry weight. It should be noted that TPC (expressed as mg GAE·g ⁻ ¹) and TFC (expressed as mg QE·g ⁻ ¹) are reported relative to different reference standards (gallic acid, MW 170.12; quercetin, MW 302.24) and reflect different reaction chemistries (Folin–Ciocalteu vs DEG–NaOH); consequently, the two values are not directly comparable on a mass basis, and a higher numerical TFC value does not imply a higher absolute molar content of flavonoids than total polyphenols.

Determination of antioxidant activities

2,2′-Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical scavenging activity was determined using a modified version of the decolorization method described by Re et al. [22]. To generate ABTS⁺ radicals, 50 mL of ABTS solution was mixed with 12.31 mL of 2.6 mM potassium persulfate and stirred in the dark. The resulting solution was then diluted with phosphate-buffered saline to an absorbance of 0.70 ± 0.03 at 732 nm. For the assay, 950 μL of the diluted ABTS solution was mixed with 50 μL of the extract and incubated in the dark for 10 min. The ABTS radical scavenging activity is expressed as electron donating ability (EDA, %), and the concentration of the extract required to reduce 50% of ABTS radicals (RC₅₀, mg·mL ⁻ ¹) was calculated as follows:

.

2,2-Diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity was evaluated following the method described by Brand-Williams et al. [23]. A 200-μL aliquot of the extract was mixed with 800 μL of 0.2 mM DPPH solution, and the solution was incubated in the dark for 30 min. The absorbance was then measured at 517 nm. The radical scavenging activity was calculated using the same formula that was used in the ABTS assay.

Determination of anti-inflammatory activity

RAW264.7 cells were obtained from the Korea Cell Line Bank (Seoul, South Korea). For cell culture, Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 units·mL ⁻ ¹ penicillin, and 100 μg·mL ⁻ ¹ streptomycin was used. The cells were sub-cultured every 2 days in an incubator maintained at 37℃ with 5% CO₂ and 95% humidity.

Cell viability was evaluated using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. RAW264.7 cells were seeded in 96-well plates at a density of 2.5 × 10⁵ cells·well ⁻ ¹ and incubated for 6 h. Thereafter, 200 μL of each treatment agent was added to the wells: lipopolysaccharide (LPS, 1 μg·mL ⁻ ¹), quercetin (positive control, 25 μM), and each extract (100 μg·mL ⁻ ¹). Following 18 h of treatment, 10 μL of MTT solution (1 mg·mL ⁻ ¹) was added to each well, and the samples were incubated for an additional 90 min. The medium was then removed, and the resulting formazan crystals were dissolved in 200 μL of DMSO. Absorbance was measured at 550 nm using a microplate reader (Epoch, BioTek, Vermont, USA). Cell viability is expressed as percentage relative to that of the untreated control. Based on the criteria described in ISO 10993−5 (2009), treatments that resulted in cell viability below 80% were considered cytotoxic and excluded from further analysis.

Nitric oxide (NO) production was evaluated using the Griess assay [24]. Following cell seeding and sample treatment as described in the MTT assay, 100 μL of the cell culture supernatant was mixed with an equal volume of Griess reagent and incubated for 10 min. Thereafter, absorbance was measured at 550 nm. NO concentration was calculated using a standard curve constructed with sodium nitrate.

Determination of anticancer activity

Seven human-derived cancer cell lines were used: A549 (lung cancer), B16F10 (melanoma), Caco-2 and HCT15 (colorectal cancer), LNCaP-LN3 (prostate cancer), MDA-MB-231 (breast cancer), and SK-OV-3 (ovarian cancer). All cell lines were obtained from the Korea Cell Line Bank. The cells were cultured in RPMI 1640 medium supplemented with 10% FBS and 1% penicillin–streptomycin. All cell lines were subcultured every 2 days under standard incubation conditions of 37℃, 5% CO₂, and 95% humidity.

To evaluate the anticancer activity of the seven forest tree seed extracts, MTT assays were performed. The seven cancer cell lines were seeded in 96-well plates at a density of 1 × 10⁴ cells·well ⁻ ¹ and incubated for 24 h. The cells were exposed to one of the following three treatments and incubated for an additional 48 h: control (no treatment), mock control (treated with DMSO at the same concentration as that in the sample), and extract treatment (100 µg·mL ⁻ ¹). Thereafter, 1 mg·mL ⁻ ¹ MTT solution was added to each well, and the samples were incubated for 4 h. The medium was then removed, and the resulting formazan crystals were completely dissolved in DMSO. Absorbance was measured at 550 nm, and the relative cell viability of the extract-treated group was calculated relative to that of the mock control group.

Determination of tyrosinase-inhibitory activity

Tyrosinase-inhibitory activity of the extracts was evaluated using an L-DOPA oxidation-based tyrosinase inhibition assay. Each extract (100 mg·mL ⁻ ¹ in DMSO) was diluted with 0.175 M sodium phosphate buffer (pH 6.8) to a final concentration of 1 mg·mL ⁻ ¹. To assess the effect of DMSO contained in the extracts, an equivalent volume of DMSO diluted in buffer was used as the control. Ascorbic acid (AsA, 1 mg·mL ⁻ ¹) was used as a positive control. For the assay, 20 μL of sample, 40 μL of 10 mM L-DOPA, 100 μL of buffer, and 40 μL of tyrosinase (110 units·mL ⁻ ¹) were added sequentially. The enzymatic reaction was monitored kinetically at 475 nm using a microplate reader, with absorbance measured every minute for 20 min. Tyrosinase activity (%) was calculated using the inhibition formula described by Tsai et al. [25]:

Data collection and statistical analysis

All experimental measurements were conducted using technical replicates obtained from independent wells of the same extract preparation and the same cell passage. The extract yield, TPC, TFC, antioxidant (ABTS and DPPH), and tyrosinase-inhibition assays were performed in triplicate (n = 3), whereas the anti-inflammatory and anticancer assays were conducted with six technical replicates (n = 6). Data are presented throughout the manuscript as mean ± standard deviation (SD). Statistical analyses were conducted using the SAS software (version 9.4; SAS Institute, Cary, NC, USA). Multi-group comparisons among species were evaluated by one-way ANOVA followed by Tukey’s honest significant difference (HSD) test, with statistical significance set at p < 0.05. For comparisons of LPS-stimulated NO production (Fig 2B), the LPS-treated group was compared against the unstimulated control by Welch’s t-test, and each extract-treated group was compared against the LPS-treated reference group using Dunnett’s multiple-comparison test. For comparisons of cell viability against the corresponding control group (Figs 2A and 3), Dunnett’s test was also used. Significance levels in Figs 2 and 3 are denoted as * p < 0.05, ** p < 0.01, *** p < 0.001 (and ### p < 0.001 for the LPS-vs-control comparison in Fig 2B). Pearson correlation coefficients (r) among species-level mean values of TPC, TFC, ABTS RC₅₀, DPPH RC₅₀, NO concentration, tyrosinase-inhibition, and mean cancer cell viability across the seven cancer cell lines were calculated to examine relationships among the measured variables, with two-tailed significance evaluated at p < 0.05.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Antioxidant compound content and activities of forest seed extracts.

(A) Total polyphenol content (TPC) in mg GAE·g ⁻ ¹ dry weight. (Bactivity. Species abbreviations (e.g., AJ, CO, and CK) are listed in Table 1. Data are presented as mean ± standard deviation (SD) of three technical replicates (n = 3). Different lowercase letters indicate significant differences among species based on Tukey’s HSD test (p < 0.05).

https://doi.org/10.1371/journal.pone.0345845.g001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Anti-inflammatory activity of forest seed extracts in RAW 264.7 macrophages.

(A) Cell viability is expressed as percentage relative to that of the control group, and significant differences among seed extract treatments were determined using Tukey’s HSD test (p < 0.05). (B) Nitric oxide (NO) production was measured in LPS-stimulated cells. ### and * indicate significant differences between the control and LPS (p < 0.001, Welch’s t-test) and between LPS and extract-treated (*p < 0.05, **p < 0.01, ***p < 0.001, Dunnett’s test) groups, respectively. Data are presented as mean ± standard deviation (SD) of six technical replicates (n = 6). Species abbreviations (e.g., AJ, CO, and CK) are defined in Table 1.

https://doi.org/10.1371/journal.pone.0345845.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Anticancer activity of forest seed extracts on various cancer cell lines.

Cell viability was measured using the MTT assay in seven human cancer cell lines: lung cancer (A549), melanoma (B16F10), ovarian cancer (SK-OV-3), colon cancer (Caco-2, HCT15), prostate cancer (LNCaP-LN3), and breast cancer (MDA-MB-231). Data are presented as mean ± standard deviation (SD) of six technical replicates (n = 6). Significant differences between each species and the corresponding control group within each cell line were determined using Dunnett’s test (*p < 0.05, **p < 0.01, ***p < 0.001). Species abbreviations (e.g., AJ, CO, and CK) are defined in Table 1.

https://doi.org/10.1371/journal.pone.0345845.g003

Extract yield and antioxidant activities of forest tree seeds

Q. glauca had the highest polyphenol content (30.7 ± 0.28 mg GAE·g ⁻ ¹), followed by A. japonica (25.2 ± 0.1 mg GAE·g ⁻ ¹) and C. obtusa (21.5 ± 0.1 mg GAE·g ⁻ ¹; Fig 1A). Similar trends were observed for TFC (Fig 1B), with Q. glauca also showing the highest flavonoid content (46.0 ± 0.25 mg QE·g ⁻ ¹). The ABTS and DPPH assays, performed to determine the radical-scavenging capacity, demonstrated that Q. glauca had the highest TPC and TFC and the lowest RC₅₀ values, indicating the strongest antioxidant activity among all tested species (Fig 1C, 1D). These results are consistent with those of Kumar et al. [26], who reported strong positive correlations between TPC and ABTS (r = 0.998, R² = 0.997) and DPPH (r = 0.994, R² = 0.988) radical scavenging activities of various plant extracts. The high correlation coefficients suggest that the polyphenol content is a critical contributor to antioxidant potential. The correlations observed between compound content (TPC and TFC) and radical-scavenging activities support the notion that Q. glauca contains abundant phenolic compounds and highly efficient radical-scavenging compounds. These results emphasize the importance of both the quantity and quality of the antioxidant constituents. The combination of high antioxidant content and radical-scavenging activity of Q. glauca highlights its potential as a valuable source of natural antioxidants. Furthermore, the strong correlation observed between TPC/TFC and radical-scavenging activities in Q. glauca suggests its potential as a premium functional forest bioresource.

Anti-inflammatory activity of the seed extracts on RAW 264.7 macrophages

To examine the anti-inflammatory properties of the seed extracts, we evaluated their cytotoxicity against RAW 264.7 macrophage cells. As shown in Fig 2a, most extracts maintained cell viability above 80% at 100 µg·mL ⁻ ¹, indicating low cytotoxicity at the tested dose. However, the viability of cells treated with the seed extracts of A. japonica and C. obtusa was 77.5% and 7.0%, respectively; therefore, these extracts were classified as cytotoxic based on the ISO 10993−5 standard. Accordingly, these extracts were excluded from subsequent NO inhibition assays. The exclusion of A. japonica and C. obtusa extracts from the subsequent NO-inhibition analysis reflects a methodological limitation of the present screening setup: at 100 µg·mL ⁻ ¹, the cytotoxic effects of these extracts prevent evaluation of NO production independently of cell viability. The absence of NO-inhibition data for these two species should therefore not be interpreted as absence of anti-inflammatory activity, but rather as an inability to evaluate this activity at the tested concentration. Lower-concentration or sub-cytotoxic exposure conditions would be required to evaluate the anti-inflammatory activity of A. japonica and C. obtusa seed extracts in future studies.

Following LPS stimulation, NO production increased markedly to 46.51 ± 1.10 µM, compared with that in the control group (9.81 ± 0.10 µM), confirming successful induction of the inflammatory response (Fig 2b). The positive control quercetin significantly reduced the NO level to 14.50 ± 0.84 µM. Among the seed extracts tested, Q. glauca, C. kousa, and P. sargentii extracts considerably suppressed NO production, lowering levels to 33.76 ± 1.88, 31.96 ± 2.68, and 36.16 ± 1.58 µM, respectively. These reductions were statistically significant compared with those in the LPS-treated group (p < 0.01). In contrast, treatment with the extracts of P. amurense (51.27 ± 1.18 µM) and P. densiflora (49.42 ± 0.33 µM) did not result in a meaningful decrease in NO production.

The above results suggest that the seed extracts of Q. glauca, C. kousa, and P. sargentii exhibit anti-inflammatory potential by effectively suppressing NO production. Numerous studies have demonstrated a strong and complex interrelationship between antioxidant activity and inflammation [27–29]. Consistent with the findings of these studies, the seed extracts of Q. glauca, C. kousa, and P. sargentii displayed pronounced anti-inflammatory activity and were also among the species with the lower ABTS and DPPH RC₅₀ values in the present dataset. However, Pearson correlation analysis across the seven species (S1 Fig) showed that NO concentration was not significantly correlated with TPC, TFC, ABTS RC₅₀, or DPPH RC₅₀ (|r| ≤ 0.55, p > 0.19 in all cases), indicating that the co-occurrence of strong anti-inflammatory and antioxidant activity in these three species cannot be fully attributed to phenolic content alone and may reflect additional, compound-specific contributors [27–29].

Additionally, the strong anti-inflammatory activities of Q. glauca, C. kousa, and P. sargentii are consistent with those previously reported for each species. Kim et al. [30] reported that Q. glauca acorn shell extracts exert anti-inflammatory activity by alleviating oxidative stress and promoting recovery from hydrogen peroxide-induced cellular damage. Similarly, leaf extracts of C. kousa with high polyphenol content have potential benefits in the prevention and treatment of inflammatory diseases [31]. Furthermore, hexane and chloroform fractions of P. sargentii leaves reportedly inhibited NO production in LPS-stimulated RAW264.7 macrophages [32]. While previous studies primarily focused on leaves or bark, here we demonstrated comparable anti-inflammatory activities of seed extracts of these species. This finding suggests that multiple parts, including the seeds, of these species have anti-inflammatory potential.

Anticancer activity against various cancer cell lines

The seed extracts exhibited varying degrees of anticancer activity against seven human cancer cell lines: lung (A549), skin (B16F10), colorectal (Caco-2 and HCT15), prostate (LNCaP-LN3), breast (MDA-MB-231), and ovarian (SK-OV-3) (Fig 3). Among these extracts, C. obtusa extract showed the most pronounced effect on cell viability, significantly reducing the viability of multiple cancer cell lines under the present screening conditions. Cell viability was reduced to less than 10% in A549, B16F10, Caco-2, and HCT15 cells, indicating marked growth-inhibitory effects across the tested cancer cell lines at the screening concentration. LNCaP-LN3 and MDA-MB-231 cells also exhibited significant reductions in viability, further supporting the notable cell-line–dependent activity at the tested concentration of C. obtusa extract. In contrast, the effect on SK-OV-3 cells was relatively moderate, suggesting cell-line–specific sensitivity of the extract. In addition to C. obtusa seed extract, other seed extracts exhibited distinct cell-line–specific anticancer activities. Alnus japonica significantly inhibited the proliferation of A549, LNCaP-LN3, and HCT15 cells, reducing cell viability to 76.6% (p < 0.01), 35.1% (p < 0.001), and 78.0% (p < 0.001), respectively, compared to the vehicle control. However, no significant effects were observed in B16F10, Caco-2, or MDA-MB-231 cells, while a stimulatory effect on cell proliferation was detected in SK-OV-3 cells (p < 0.01). C. kousa significantly suppressed cell proliferation in six of the seven cancer cell lines tested, with the exception of SK-OV-3. P. amurense exhibited anticancer activity against B16F10 (p < 0.05) and LNCaP-LN3 cells (p < 0.001). Pinus densiflora exhibited significant inhibitory effects on A549 (p < 0.05), LNCaP-LN3 (p < 0.01), and HCT15 cells (p < 0.001). P. sargentii demonstrated antiproliferative activity in Caco-2 and HCT15 cells (both p < 0.05), while Q. glauca significantly reduced the viability of A549 (p < 0.01), Caco-2 (p < 0.05), and LNCaP-LN3 cells (p < 0.001).

Our results indicate that, under the present screening conditions, the cell viability effects of the forest tree seed extracts range from broad effects across multiple cancer cell lines to more cell-line–specific responses. The seed extract of C. obtusa exhibited the most pronounced reduction in cell viability across cancer cell lines of diverse tissue origins under the tested screening concentration. Similarly, previous studies have reported that extracts of the leaves and heartwood of C. obtusa significantly inhibit the proliferation of various cancer cell lines, including PANC-1, HCT116, KB, HONE-1, and TSCH cells [33–35]. In the literature, such activity has been attributed to the combined effects of diverse phytochemicals abundant in coniferous species, including terpenoids, flavonoids, and phenolic acids [36]. Accordingly, the effects observed here may also be driven by similar phytochemical compositions in the seed extracts, although this hypothesis remains to be tested through compound-level characterization. Additionally, the remaining seed extracts exhibited selective inhibitory effects across cancer cell lines, consistent with the notion that activity in such screening assays varies with both plant species and cancer cell line type. Cell-line–specific sensitivity of plant extracts has been reported [37,38], and such selective responses are presumed to arise from variations in the bioactive compounds in each plant species and their distinct

Tyrosinase-inhibitory activity

The tyrosinase-inhibitory activity of the seed extracts was evaluated as an in vitro indicator of potential relevance to skin-pigmentation–related applications, while recognizing that the present assay does not directly assess cellular melanogenesis. As shown in Fig 4 and the accompanying quantitative data, all extracts exhibited varying degrees of tyrosinase activity inhibition, ranging from 28.48% to 58.90%. Ascorbic acid, used as the positive control, demonstrated the highest inhibition at 98.31% ± 0.15%, validating the assay.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Tyrosinase-inhibitory activity of forest tree seed extracts.

The inhibitory effect of seed extracts on tyrosinase activity is expressed as percentage. Different lowercase letters indicate significant differences among seed extracts based on Tukey’s HSD test (p < 0.05, n = 3). AsA (ascorbic acid, 1 mg·mL ⁻ ¹) was used as a positive control. Species abbreviations (e.g., AJ, CO, and CK) are defined in Table 1.

https://doi.org/10.1371/journal.pone.0345845.g004

Among the extracts, A. japonica seed extract exhibited the strongest tyrosinase-inhibitory effect (63.82% ± 3.45%), followed by Q. glauca (58.40% ± 3.24%) and C. obtusa seed extracts (53.42% ± 0.60%). These three species exhibited significantly higher activity than the other species (p < 0.05). In contrast, P. amurense, P. densiflora, and P. sargentii exhibited relatively low inhibition rates (37.04–39.60%), indicating limited effectiveness in tyrosinase activity inhibition.

The aforementioned species-dependent patterns suggest that qualitative and quantitative differences in phenolic compounds and other secondary metabolites may underlie the observed variations in tyrosinase-inhibitory activity. Various phenolic compounds, including flavonoids, phenolic acids, and tannins, are known to inhibit tyrosinase activity through multiple mechanisms: (i) direct interaction with the catalytic pocket of the enzyme, (ii) chelation of the two copper ions at the active site, and (iii) acting as alternative substrates or redox competitors altering L-DOPA oxidation [39–41]. Here, the seed extracts of A. japonica, C. obtusa, and Q. glauca exhibited strong antioxidant activities along with significant tyrosinase-inhibitory effects in the present screening. Consistent with previous reports of positive correlations among TPC, TFC, tyrosinase-inhibitory activity, and antioxidant capacity in other plant systems [42,43], Pearson correlation analysis across the seven species in the present study (S1 Fig) revealed strong positive correlations of tyrosinase-inhibition with both TPC (r = +0.95, p < 0.001) and TFC (r = +0.88, p < 0.01), and a strong negative correlation with DPPH RC₅₀ (r = −0.86, p < 0.05). This pattern provides direct, dataset-level support for the proposition that the antioxidant-associated phenolic constituents in these seed extracts are major contributors to tyrosinase-inhibitory activity [44]. Certain limitations of the present study should nonetheless be acknowledged. Individual phenolic compounds were neither identified nor quantified here, and the specific molecules responsible for tyrosinase activity inhibition, as well as their direct mechanisms of interaction with the enzyme, were not elucidated. In addition, all assays were performed at a single fixed concentration of crude extracts, without dose–response or selectivity characterization. Future studies employing chromatographic and spectrometric approaches, such as HPLC and LC–MS, combined with bioactivity-guided fractionation and dose–response analysis, will be necessary to identify the key active constituents and to clarify their roles in tyrosinase activity inhibition.

Relationships among the measured variables (TPC, TFC, ABTS RC₅₀, DPPH RC₅₀, NO concentration, tyrosinase-inhibition, and mean cancer cell viability across the seven cancer cell lines) were further examined by Pearson correlation analysis using the species-level mean values (n = 7 species; S1 Fig). TPC and TFC were strongly and positively correlated with each other (r = +0.97, p < 0.001) and were negatively correlated with ABTS and DPPH RC₅₀ values (TPC: r = −0.78 and −0.82, p < 0.05; TFC: r = −0.77, p < 0.05 for DPPH), indicating that species with higher phenolic content also showed stronger radical-scavenging activity, as expected. TPC and TFC also showed strong positive correlations with tyrosinase-inhibition (r = +0.95, p < 0.001 and r = +0.88, p < 0.01, respectively), and DPPH RC₅₀ was negatively correlated with tyrosinase-inhibition (r = −0.86, p < 0.05), suggesting that the same antioxidant-related phenolic constituents may contribute substantially to the differences in tyrosinase-inhibitory activity observed among species. In contrast, neither TPC nor TFC was significantly correlated with NO concentration or with mean cancer cell viability (|r| ≤ 0.34 and ≤ 0.27, p > 0.4 in both cases), indicating that anti-inflammatory and anticancer activities at the species level cannot be fully accounted for by total phenolic or flavonoid content and likely reflect additional, possibly compound-specific, contributors. Notably, NO concentration was strongly positively correlated with mean cancer cell viability (r = +0.87, p < 0.05) — that is, species whose extracts more strongly suppressed NO production also tended to more strongly reduce cancer cell viability — most clearly exemplified by C. obtusa, which produced the lowest NO concentration and the lowest mean cancer cell viability among the seven species. These correlation patterns indicate that, although antioxidant activity and tyrosinase-inhibition appear to be largely co-driven by phenolic content in the present sample set, anti-inflammatory and anticancer activities are not adequately predicted by phenolic content alone and may depend on additional bioactive constituents that warrant compound-level identification in follow-up work.

Several additional limitations should be acknowledged. First, the seed lots used in this study were collected across different years (2017–2023) and locations within the Republic of Korea. Although all lots were stored under uniform low-temperature, dry conditions at the NFSV and processed in a single extraction batch under identical protocols, we cannot exclude the possibility that variation in pre-harvest environment, storage duration, and intra-species genetic background contributed to the observed differences. Accordingly, the bioactivity differences reported here should be interpreted as comparative species-level observations from the present sample set rather than as strictly species-intrinsic effects. Future work using multiple seed lots per species across years should help disentangle species-level effects from year- and site-related variation. Second, all replicates in this study are technical replicates from a single extract preparation and the same cell passage; biological replication using independent extract preparations and independent cell passages would strengthen the generalizability of the present observations. Third, all functional assays in this study were performed at a single fixed concentration of crude extracts and therefore do not include dose–response characterization, IC₅₀ estimation, selectivity profiling against non-cancer cells, or compound-level identification. The present findings should accordingly be interpreted as a comparative bioactivity screening that provides a basis for prioritizing species for follow-up mechanistic and compound-level investigations, rather than as direct evidence supporting application-ready bioactivity.