Denatonium in Camellia japonica seeds: A synthetic bitterant found in nature

Abstract

Denatonium is widely recognized as an intensely bitter synthetic compound used to prevent accidental ingestion of toxic or undesirable products. Here, we report the natural occurrence of denatonium in seed-derived tissues of Camellia japonica. Direct chewing of the seeds produced exceptional bitterness, consistent with LC-MS/MS detection. To confirm reproducibility, samples from different sources or growth conditions, including germinated sprouts and oil-removed seed cake, were analyzed. Denatonium was detected in the radicle, plumule, cotyledons, and seed cake. Its identity was confirmed by comparison with an authentic denatonium benzoate standard, based on retention time at 5.63 min, precursor ion, and diagnostic MS/MS fragment ions. Quantitative analysis revealed high levels of denatonium, reaching approximately 3 g/kg dry weight in cotyledons and 2 g/kg dry weight in seed cake. These findings demonstrate that C. japonica seeds naturally accumulate denatonium, suggesting a role in seed-associated chemical defense.

Graphical abstract

Highlights

• •Denatonium was identified in Camellia japonica seed-derived tissues.
• •Authentic standard matching confirmed the denatonium cation by LC-MS/MS.
• •Denatonium reached 3 g/kg dry weight in cotyledons.
• •Defatted seed cake retained denatonium at 2 g/kg dry weight.
• •Denatonium may explain the extreme bitterness of C. japonica seeds.

1. Introduction

Bitterness is an important sensory cue that allows animals to avoid potentially toxic or harmful substances. Bitter taste perception is closely associated with the detection of noxious or defensive chemicals in food sources and can strongly influence feeding behavior and survival (Wooding, Ramirez, & Behrens, 2021). Many plants produce bitter secondary metabolites, including alkaloids, saponins, phenolic compounds, terpenoids, and glycosides, as chemical defenses against herbivores, insects, and microbial attacks. Seeds are particularly important targets for chemical protection because they contain embryonic tissues and nutrient reserves required for the next generation. This view is consistent with general concepts of plant defense and seed/fruit toxicity, in which specialized metabolites can protect valuable reproductive tissues from herbivores and predators (Beckman, 2013; War et al., 2012).

In particular, C. japonica seeds have been reported to contain triterpenoid saponins and antioxidant phytochemicals, including camelliasaponins and phenolic antioxidants (Saenjum, Pattananandecha, & Nakagawa, 2020; Yoshikawa et al., 1994). These compounds, together with saponins and polyphenols, have generally been considered possible contributors to the bitter or astringent taste of Camellia seeds and seed cake. However, the unusually intense bitterness perceived upon chewing C. japonica seeds appears stronger and more persistent than ordinary plant bitterness, suggesting the presence of an additional potent bitter principle.

Denatonium is widely recognized as one of the most intensely bitter compounds and is commonly used as a synthetic bittering agent to prevent accidental ingestion of toxic or undesirable products. Denatonium salts are used as aversive agents in products such as denatured alcohol, household chemicals, and other materials that should not be ingested (Lege, Guillet, Merel, Yanez Heras, & Zwiener, 2017). Human and animal studies have also shown that denatonium benzoate produces strong taste aversion and can reduce ingestion of treated liquids (Berning, Griffith, & Wild, 1982; Gibbs et al., 2022). Because denatonium has been regarded primarily as a synthetic bitterant rather than a naturally occurring plant metabolite, its detection in plant tissues would be unexpected and scientifically significant.

During our analysis of C. japonica seeds, we observed that direct chewing of the seeds produced an exceptionally intense bitterness distinct from typical bitterness caused by common plant metabolites. This observation prompted us to investigate whether a highly potent bitter compound might be present in the seeds. In the present study, mature C. japonica seeds were germinated in coarse decomposed granite or New Zealand sphagnum moss, and the resulting sprouts were dissected into the radicle, plumule, and cotyledons. These tissues, together with seed cake obtained after oil removal, were analyzed by LC-MS/MS. Using authentic denatonium benzoate as a standard, we identified denatonium in multiple seed-derived tissues and quantified its abundance in cotyledons and seed cake.

This study reports the natural occurrence of denatonium in C. japonica seeds and germinated seed tissues. The finding suggests that denatonium may represent a major bitter principle contributing to the extraordinary bitterness of C. japonica seeds and may function as a chemical defense factor during seed development and germination.

2. Materials and methods

2.1. Chemicals

Denatonium benzoate was purchased from Sigma-Aldrich (St. Louis, MO, USA) and used as an authentic reference standard. All solvents used in this study were of analytical grade or higher. LC-MS-grade solvents were used for LC-MS/MS analysis.

2.2. Plant materials and germination conditions

Mature seeds of Camellia japonica were harvested in late October, washed thoroughly with water to remove surface contaminants, and used for germination experiments. For germination, the seeds were placed in either coarse decomposed granite substrate or New Zealand sphagnum moss and maintained on a balcony under Korean winter conditions until late March. During the germination period, the minimum ambient temperature was approximately 10 °C. Germinated sprouts were harvested in late March.

After collection, the germinated sprouts were washed thoroughly with water to remove residual substrate particles. The samples were then dried and dissected into three anatomical parts: radicle, plumule, and cotyledons. Each dissected tissue was separately prepared for LC-MS/MS analysis.

2.3. Preparation of seed cake

Defatted Camellia japonica seed cake, obtained after removal of seed oil from mature seeds, was provided by the Camellia Research Institute (Tongyeong, Gyeongsangnam-do, Republic of Korea). The sample was dried, ground into a fine powder, and used for denatonium analysis. Denatonium concentration was expressed on a dry weight basis.

2.4. Sample extraction and preparation of analytical solutions

Dried germinated seed tissues, including the radicle, plumule, and cotyledons, as well as defatted Camellia japonica seed cake, were used for LC-MS/MS analysis. Approximately 1 g of each cotyledon and seed cake sample was placed in a 100 mL Erlenmeyer flask, and 100 mL of 80% ethanol was added. The samples were extracted for 2 h using an ultrasonicator (Branson Ultrasonics Co., Danbury, CT, USA), and the extraction was repeated three times under identical conditions. The combined ethanol extracts were transferred to a 500 mL round-bottom flask and evaporated under reduced pressure at 40 °C using a rotary evaporator. The resulting concentrate was dissolved in methanol, transferred into a 20 mL glass vial using a Pasteur pipette, and dried again overnight at 40 °C.

For LC-MS/MS analysis, 0.5 mg of each dried extract was dissolved in 1 mL of LC-grade methanol (Honeywell, Charlotte, NC, USA) to prepare a 500 ppm analytical solution. The solution was collected using a 1 mL Covex syringe and filtered through a 0.22 μm regenerated cellulose syringe filter (GVS S.p.A., Bologna, Italy) before analysis. A denatonium standard solution was prepared by dissolving 0.1 mg of authentic denatonium benzoate in 1 mL of LC-grade methanol to obtain a 100 ppm solution, followed by filtration through a 0.22 μm regenerated cellulose syringe filter.

The extraction yields were calculated only for the cotyledon and seed cake samples. The yields were 8.4% for the cotyledon extract and 27.1% for the seed cake extract, corresponding to 84 mg/g and 271.3 mg/g of dried sample, respectively. The radicle and plumule samples were analyzed by LC-MS/MS without extraction yield determination.

2.5. LC-MS/MS analysis

LC-MS/MS analysis was performed using a Thermo Fisher Scientific Orbitrap Exploris 120 mass spectrometer coupled to an LC system. Chromatographic separation was achieved using a Waters ACQUITY UPLC® BEH C18 column (50 mm × 2.1 mm, 1.7 μm). The column temperature was maintained at 30 °C, and the injection volume was 5 μL. The mobile phases consisted of 0.1% formic acid in water as solvent A and 0.1% formic acid in acetonitrile as solvent B. The flow rate was 0.3 mL/min.

Gradient elution was performed as follows: 10% B was held from 0 to 1.0 min, increased to 90% B from 1.0 to 15.0 min, maintained at 90% B from 15.0 to 17.5 min, and then returned to 10% B at 20.0 min for column re-equilibration. The total run time was 20 min.

Mass spectrometric analysis was performed using a heated electrospray ionization source in positive ion mode. High-purity nitrogen was used as the nebulizing and auxiliary gas. The spray voltage was set to 3500 V in positive ion mode and 2500 V in negative ion mode. The sheath gas and auxiliary gas flow rates were 50 and 10 arbitrary units, respectively. The ion transfer tube temperature was 325 °C, and the vaporizer temperature was 350 °C.

Full-scan MS spectra were acquired over an m/z range of 100–1500 at a resolution of 120,000. Data-dependent MS/MS acquisition was performed with four dependent scans. The isolation window was set to 2 m/z, and higher-energy collisional dissociation was performed using normalized collision energies of 15, 30, and 60. The Orbitrap resolution for MS/MS scans was 30,000. Denatonium was identified by comparison with an authentic denatonium benzoate standard based on retention time, precursor ion, and diagnostic MS/MS fragment ions.

2.6. Quantitative UPLC-QTRAP/MS analysis of denatonium

Denatonium content was quantitatively determined using a Nexera X2 UPLC system (Shimadzu Corporation, Japan) coupled to a QTRAP 4500 mass spectrometer (NFEC-2019-03-255,243; SCIEX, Framingham, MA, USA) at the High-Tech Materials Analysis Core Facility, Gyeongsang National University, Jinju, Korea. Before analysis, sample extracts and standard solutions were diluted with methanol to appropriate concentrations and filtered through a syringe filter. Denatonium benzoate was used as an authentic standard, and denatonium was detected in positive ion mode as the denatonium cation.

Chromatographic separation was performed on a Waters ACQUITY BEH C18 UHPLC column (1.7 μm, 2.1 × 100 mm) maintained at 40 °C. The mobile phases consisted of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B). The gradient elution program was as follows: 10% B at 0.1 min, increased to 95% B at 8 min, maintained at 95% B until 10 min, and returned to 10% B at 12 min. The flow rate was 0.3 mL/min.

Mass spectrometric detection was performed in positive ion mode using multiple reaction monitoring (MRM). The ion spray voltage was set at 5500 V, the ion source temperature was maintained at 500 °C, and Gas 1 and Gas 2 were both set at 50 psi. Data acquisition and processing were performed using Analyst software version 1.7.0 (SCIEX).

Denatonium was detected at a retention time of 4.43 min. The precursor ion was monitored at m/z 325.27, and the optimized MRM transitions were m/z 325.27 → 90.9, 325.27 → 86.1, and 325.27 → 223.2. The transition m/z 325.27 → 90.9 was used for quantification, whereas the other transitions were used for confirmation. Quantification was performed by comparing the chromatographic peak area of denatonium in each sample extract with the peak areas of calibration standards analyzed under identical UPLC-QTRAP/MS conditions.

2.7. Confirmation of denatonium identity

To confirm the identity of denatonium in C. japonica seed-derived samples, the retention time, precursor ion, and MS/MS fragment ions of the detected peak were compared with those of authentic denatonium benzoate standard under identical LC-MS/MS conditions. The sample-derived denatonium peak showed the same retention time and diagnostic fragmentation pattern as the authentic standard, confirming the presence of denatonium in C. japonica seeds and germinated seed tissues.

2.8. Experimental reproducibility

Two independent experiments were performed using germinated C. japonica seed samples. In both experiments, denatonium was consistently detected in the radicle, plumule, and cotyledons. Seed cake was analyzed separately to determine denatonium abundance in the oil-removed seed fraction.

2.9. Statistical analysis

All quantitative data are presented as mean ± SD. Statistical significance between cotyledon and seed cake samples was determined using an unpaired two-tailed Student's t-test. A p-value of less than 0.05 was considered statistically significant. Statistical significance is indicated as follows: ***p < 0.0001.

3. Results and discussion

3.1. Germination and dissection of Camellia japonica seeds

Mature Camellia japonica seeds were germinated to examine the distribution of denatonium in seed-derived tissues. During germination, the radicle elongated from the seed body, while the plumule emerged from the upper region of the germinating seed (Fig. 1A and B). Germinated seed tissues were collected, dried, and prepared for LC-MS/MS analysis (Fig. 1C). Based on seedling morphology, the germinated seeds were divided into major anatomical regions, including the plumule, cotyledon, and radicle (Fig. 1D). This separation allowed us to determine whether denatonium was restricted to a specific tissue or distributed throughout seed-derived tissues during early germination.

Fig. 1.

Germination and tissue dissection of Camellia japonica seeds. (A) Germinated C. japonica seeds showing emerging radicles. (B) Freshly germinated seed sprouts grown in the germination substrate. (C) Dried germinated sprouts prepared for LC-MS/MS analysis. (D) Schematic representation of a germinated C. japonica seed showing the major seed-derived tissues analyzed in this study, including the radicle, plumule, and cotyledons.

3.2. Detection of denatonium in plumule and radicle tissues

To determine whether denatonium was present in germinated seed tissues, LC-MS/MS analysis was performed using extracts prepared from the plumule and radicle. In both tissues, total ion chromatograms (TICs) showed a prominent peak at approximately 5.73–5.74 min (Fig. 2A). Extracted ion chromatograms (XICs) corresponding to the denatonium cation also showed a clear peak at the same retention time, supporting the selective detection of denatonium-related ions in both plumule and radicle samples.

Fig. 2.

LC-MS/MS detection and confirmation of denatonium in germinated Camellia japonica seed tissues. (A) Representative TICs and XICs of germinated C. japonica seed tissues, including the radicle and plumule. A prominent denatonium peak was detected at approximately 5.73–5.74 min in both seed-derived tissues. The corresponding XIC peak appeared at the same retention time, supporting the selective detection of the denatonium cation. (B) Full-scan MS¹ and MS² spectra of the denatonium peak detected in germinated seed tissues. The MS¹ spectrum showed the denatonium cation at m/z 325.2272, and MS² fragmentation of the precursor ion generated diagnostic product ions, including m/z 86.0965, 91.0543, 148.1121, and 233.1650. The matching retention time, precursor ion, and MS/MS fragmentation pattern confirmed the presence of denatonium in the radicle and plumule of germinated C. japonica seeds. MS and MS/MS spectra were acquired in positive ion mode.

The identity of the detected peak was further examined by full-scan MS and MS/MS fragmentation analysis. The MS¹ spectrum showed a dominant ion at m/z 325.2272, corresponding to the denatonium cation (Fig. 2B). MS² fragmentation of this ion generated diagnostic product ions, including m/z 86.0965, 91.0543, 148.1121, and 233.1650. These results indicate that denatonium is present in germinated seed-derived tissues, including both plumule and radicle.

As an independent untargeted MS/MS-based analysis, GNPS molecular networking also revealed a recurrent feature at m/z 325.2272 with a retention time of 5.73 min. This feature was annotated as denatonium by GNPS library matching and was detected in the combined radicle–plumule sample, plumule, and radicle tissues. Extracted ion chromatogram analysis confirmed a reproducible peak at RT 5.73 min, and MS/MS spectra showed comparable fragmentation patterns among these tissues. Because this GNPS-based annotation was initially based on library matching, the feature was conservatively described as a putative denatonium-like feature in the supplementary analysis (Supplementary Figs. S1 and S2).

3.3. Comparison of denatonium standard, cotyledon extract, and seed cake extract

To further confirm the presence of denatonium in C. japonica seed-derived samples, an authentic denatonium benzoate standard, cotyledon extract, and seed cake extract were analyzed under the same LC-MS/MS conditions. TIC analysis showed a common prominent chromatographic peak at approximately 5.50–5.63 min in the denatonium standard, cotyledon extract, and seed cake extract (Fig. 3A). In parallel, extracted ion chromatogram (XIC) analysis for the denatonium cation further revealed a corresponding peak at the same retention time in all three samples (Fig. 4A). The denatonium benzoate standard was analyzed at 100 ppm, corresponding to 72.88 ppm denatonium cation, whereas cotyledon and seed cake extracts were analyzed at 500 ppm. The radicle and plumule analyses were performed primarily to confirm the presence and tissue distribution of denatonium rather than for quantitative comparison among germination replicates.

Fig. 3.

TIC-based comparison of denatonium standard, cotyledon, and seed cake extracts. (A) TICs of authentic denatonium standard, C. japonica cotyledon extract, and seed cake extract. A common prominent peak corresponding to denatonium was observed at approximately 5.50–5.63 min in all samples. The denatonium benzoate standard was analyzed at 100 ppm, corresponding to 72.88 ppm denatonium cation, whereas cotyledon and seed cake extracts were analyzed at 500 ppm. MS and MS/MS spectra were acquired in positive ion mode. (B) Full-scan MS¹ and MS² spectra of the denatonium peak in the authentic standard, cotyledon extract, and seed cake extract. The MS¹ spectra showed the denatonium cation at m/z 325.227, and MS² fragmentation produced diagnostic product ions at m/z 86.096, 91.054, 148.112, and 233.165. The matching retention time, precursor ion, and fragmentation patterns confirmed the presence of denatonium in C. japonica cotyledon and seed cake samples.

Fig. 4.

XIC-based confirmation of denatonium in cotyledon and seed cake extracts. (A) XICs of authentic denatonium standard, C. japonica cotyledon extract, and seed cake extract. Denatonium was detected at a similar retention time in all three samples, with peaks observed at approximately 5.50–5.63 min. The denatonium standard was analyzed at 100 ppm, corresponding to 72.88 ppm denatonium cation, whereas cotyledon and seed cake extracts were analyzed at 500 ppm. (B) Full-scan MS¹ and MS² spectra of the denatonium peak in the authentic standard, cotyledon extract, and seed cake extract. The MS¹ spectra showed the denatonium cation at m/z 325.227, and MS² fragmentation generated diagnostic product ions, including m/z 86.096, 91.054, 148.112, and 233.165. The matching retention time, precursor ion, and MS/MS fragmentation patterns confirmed the presence of denatonium in C. japonica cotyledon and seed cake samples. MS and MS/MS spectra were acquired in positive ion mode.

Full-scan MS analysis of the corresponding chromatographic peak showed the denatonium cation at m/z 325.227 in all three samples (Fig. 3B). Consistently, XIC-based MS analysis also confirmed the presence of the denatonium-related ion signal in the standard, cotyledon extract, and seed cake extract (Fig. 4B). In addition, MS² spectra of the denatonium standard, cotyledon extract, and seed cake extract showed highly similar fragmentation patterns. Diagnostic product ions, including m/z 86.096, 91.054, 148.112, and 233.165, were consistently observed. The agreement in TIC peak position, XIC peak position, precursor ion, and MS/MS fragmentation pattern confirmed that the compound detected in cotyledon and seed cake samples was denatonium.

3.4. Quantitative determination of denatonium in cotyledon and seed cake

Denatonium in cotyledon and seed cake was quantified by LC-MS/MS using an authentic denatonium benzoate standard. Denatonium was detected in positive ion mode as the denatonium cation. Quantitative LC-MS/MS analysis showed that the denatonium content was approximately 3 g/kg dry weight in cotyledons and 2 g/kg dry weight in seed cake (Fig. 5). The denatonium level was significantly higher in cotyledons than in seed cake. These findings demonstrate that denatonium is present not only in germinated seed tissues such as plumule and radicle, but also in cotyledons and oil-removed seed cake. The relatively high level of denatonium in cotyledons suggests that denatonium is intrinsically accumulated in seed tissue, whereas its detection in seed cake indicates that this bitter compound is retained in the non-oil fraction after oil extraction.

Fig. 5.

Quantitative determination of denatonium in cotyledon and seed cake. Denatonium content in cotyledon and seed cake was quantified by LC-MS/MS using an authentic denatonium benzoate standard. Denatonium was detected in positive ion mode as the denatonium cation. The denatonium content was approximately 3 g/kg dry weight in cotyledons and 2 g/kg dry weight in seed cake. Data are presented as mean ± SD. ***p < 0.0001.

3.5. Natural occurrence and significance of denatonium in C. japonica seed-derived tissues

The present study identifies denatonium in Camellia japonica seeds and seed-derived tissues, including the plumule, radicle, cotyledon, and seed cake. This finding is notable because denatonium has generally been regarded as an intensely bitter synthetic compound rather than a plant-derived metabolite. Denatonium is commonly used as an aversive bitterant to prevent accidental ingestion of undesirable or toxic products, particularly in household, cosmetic, pharmaceutical, and industrial materials (Hansen, Janssen, & Beasley, 1993; Lege et al., 2017). Its detection in C. japonica seeds therefore provides an unexpected example of a synthetic bitterant occurring in a natural plant source.

The natural occurrence of denatonium has been reported previously in white kidney bean, Phaseolus multiflorus var. albus (Lee et al., 2021). In that study, denatonium was isolated as compound 1 and structurally identified by NMR and HR-ESI-MS; the authors described it as the first identification of denatonium from a natural source. Importantly, they also considered the possibility of artifact or solvent contamination and reported that denatonium was not detected in the methanol solvent used for extraction, supporting the conclusion that the isolated denatonium was a genuine natural compound. This previous report is highly relevant to the present study because it establishes that denatonium is not necessarily restricted to synthetic or industrial sources.

However, the present study differs from the white kidney bean report in an important way. In the previous study, 2.3 mg of denatonium was isolated from the EtOAc-soluble fraction obtained during the fractionation of 1 kg of plant material. Although this finding was significant as evidence for naturally occurring denatonium, denatonium was not emphasized as a major bitter principle of the plant material. In contrast, our LC-MS/MS analysis detected denatonium as a prominent peak in C. japonica seed-derived samples. Quantitative LC-MS/MS analysis determined the denatonium content to be approximately 3 g/kg dry weight in cotyledons and 2 g/kg dry weight in seed cake.

Although the isolation yield from white kidney bean and the quantitatively determined denatonium content in C. japonica seed-derived materials cannot be directly compared because the plant materials, extraction procedures, and analytical methods differ, the contrast is still noteworthy. Our results showed that denatonium was present at approximately 3 g/kg dry weight in cotyledons and 2 g/kg dry weight in seed cake, indicating that C. japonica seeds contain substantial amounts of denatonium.

The higher denatonium content in cotyledons suggests that denatonium is intrinsically accumulated in the seed tissue, rather than being newly produced during early root growth or germination. The relatively lower denatonium level in seed cake may be partly attributable to partial loss of denatonium during the oil extraction process. Nevertheless, the detection of denatonium at 2 g/kg dry weight in seed cake indicates that a considerable amount of this compound is retained in the non-oil fraction after oil extraction. These findings suggest that C. japonica seed cake may represent a naturally denatonium-containing seed material rather than a plant material containing only trace-level denatonium.

The relatively high abundance of denatonium in C. japonica seed-derived samples is also surprising in the context of known Camellia seed chemistry. Camellia seeds are generally known to contain fatty acids, saponins, flavonoids, phenolic compounds, and other functional metabolites (Guo, Tong, Ren, Tu, & Li, 2018; Pereira et al., 2022; Zeng & Endo, 2019). These constituents have been associated with nutritional, cosmetic, antioxidant, anti-inflammatory, and industrial applications. However, denatonium has not been widely recognized as a representative component of Camellia seeds. The present finding therefore expands the phytochemical profile of C. japonica seeds and suggests that an exceptionally bitter quaternary ammonium compound may be one of the characteristic constituents of the seed non-oil fraction.

The tissue distribution of denatonium provides a clue to its possible biological role. Denatonium was detected in germinated tissues, including the plumule and radicle, and was also confirmed in cotyledon and seed cake. By contrast, it was not detected in extracts from mature roots of large C. japonica trees under the same analytical conditions (Supplementary Fig. S3). This tissue-biased distribution suggests that denatonium is not a constitutive metabolite broadly present in all tissues, but is more closely associated with seeds and early germination. Such localization is consistent with optimal defense theory, which predicts preferential allocation of defensive metabolites to tissues with high fitness value (Hunziker et al., 2021). Because large seeds are nutrient-rich reproductive structures and may be attractive to seed-consuming animals, accumulation of an intensely bitter compound in seeds and young seedlings may represent an efficient chemical defense strategy.

This interpretation is consistent with the sensory observation that direct chewing of C. japonica seeds produced an exceptionally intense bitterness distinct from ordinary plant bitterness. Saponins and phenolic compounds may contribute to bitterness and astringency, but the extreme bitterness of the seeds may be more directly explained by the presence of denatonium. Because denatonium is one of the most intensely bitter compounds known, its accumulation in seed-derived tissues could serve as a strong feeding deterrent. Therefore, denatonium may function as a seed-specific bitter principle that reduces the likelihood of seed predation and helps protect the next generation of the plant. The biosynthetic origin of denatonium in higher plants remains unknown, and the present data do not allow specific precursors or enzymatic steps to be proposed without speculation. Interestingly, immature embryonic tissues showed little bitterness, whereas mature seeds exhibited extremely intense bitterness, suggesting that denatonium accumulation may increase during seed maturation. Because cotyledons are nutrient-rich storage tissues that support early seedling development, the accumulation of such an intensely bitter compound may contribute to the chemical protection of mature seeds against predation. Further studies combining isotope tracing, transcriptomic analysis, and metabolomic profiling will be required to elucidate the biosynthetic pathway of denatonium and its developmental regulation in C. japonica.

The detection of denatonium in seed cake has practical implications. C. japonica seeds are used for oil production, and the remaining seed cake is often treated as a by-product. Since denatonium is a cationic and non-lipid compound, it is likely to remain in the non-oil fraction after oil extraction. This may explain why denatonium was still detected at a substantial level in the seed cake. Therefore, C. japonica seed cake may have potential value as a natural source of intensely bitter compounds. In addition to denatonium, camellia seed materials contain saponins, flavonoids, phenolics, and fatty acids, suggesting that seed cake could be developed as a multifunctional natural material for cosmetic, industrial, deterrent, or functional applications (Ma, Shen, Fang, Chen, & Wang, 2022; Teixeira & Sousa, 2021).

Several potential applications can be considered. Denatonium-containing C. japonica seed cake may be useful as a natural bittering material, animal feeding deterrent, cosmetic ingredient, industrial bitterant, or functional plant-derived material. The fact that denatonium was detected in a plant material associated with traditional oil production may broaden the perception of this compound from a purely synthetic bitterant to a naturally occurring bitter principle. However, practical application will require careful standardization and safety evaluation. Although camellia seed oil and related materials have been used in food and cosmetic contexts, concentrated denatonium-rich fractions should not be assumed safe without dose-dependent toxicological assessment and regulatory review.

4. Conclusions

Overall, this study demonstrates that C. japonica seeds contain denatonium at substantial levels, despite denatonium having been regarded mainly as a synthetic bitterant. Its detection in multiple seed-derived tissues, including plumule, radicle, cotyledon, and defatted seed cake, and its confirmation using an authentic standard support its presence as a reproducible seed-associated constituent rather than an analytical artifact. Importantly, the g/kg-level abundance of denatonium in cotyledons and seed cake indicates that it is a major bitter principle, not a trace-level contaminant. These findings provide a chemical basis for the exceptionally intense bitterness of C. japonica seeds and suggest that denatonium may contribute to a previously unrecognized seed-associated defense system. The retention of denatonium in defatted seed cake further identifies camellia seed cake as a naturally denatonium-containing non-oil fraction. With appropriate standardization, purification, and safety evaluation, naturally derived denatonium from camellia seed cake may provide an alternative source of bittering material for safer and more sustainable industrial applications.

CRediT authorship contribution statement

Jung-Hwan Kim: Writing – review & editing, Writing – original draft, Visualization, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Chaeyeon Cho: Software, Data curation. Heejung Yang: Validation.

Funding

This work was supported by the National Research Foundation of Korea (NRF) (NRF-2022R1I1A3070407 and RS-2026-25470399) and the Korea Basic Science Institute (National Research Facilities and Equipment Center) grant funded by the Ministry of Education, Republic of Korea (Grant No. 2022R1A6C101B724).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.