Introduction

Alpinia is one of the foremost genera in the ginger family (Zingiberaceae) that has been used as food spice, flavoring agent, and in ethnomedicine in several countries as China, Japan, and India1,2. The genus comprises ca. 250 species that are widely distributed throughout tropical and subtropical regions of the globe2,3. Volatile oil, terpenes, phenylpropanoids, diarylheptanoids, and flavonoids are the major classes of chemical constituents commonly found in Alpinia species, as reported in several research studies2,4,5. Essential oil is a principal component of this genus with a complex chemical profile rich in monoterpenes and sesquiterpenes. It is not only responsible for the characteristic aroma of Alpinia but also contributes to a wide range of potential bioactivities, including anti-inflammatory, antimicrobial, cytotoxic, anti-hypertensive, and antioxidant activities, supporting their potential therapeutic applications1,2,4,5,6.

Alpinia katsumadai Hayata is an herbaceous species originating from India and is widely cultivated in Southeast Asia, including China7,8. Phytochemical analyses of A. katsumadai have uncovered a remarkable and impressive spectrum of metabolites, including stilbenes, chalcones, diarylheptanoids, kavalactones, monoterpenes, sesquiterpenes, and flavonoids9,10,11. This distinctive blend of phytochemicals underpins the special therapeutic potential of A. katsumadai in both traditional and modern medicine5,8. The seeds of A. katsumadai (AKS) are documented in the Korean pharmacopeia5,8 and are currently regarded as a valuable and important Traditional Chinese Medicine (known as Cao Dou Kou) used in the treatment of numerous ailments, such as emesis and digestive disorders9,10. Seeds are characterized by a distinctive warm aroma and a pungent, faintly bitter taste that could be derived from their constituents, especially essential oil content, imparting a distinctive flavor profile12.

Culinary and medicinal uses of AKS is attributed for its complex chemical composition, which is marked by the prominence of stilbenes, chalcones, diarylheptanoids, terpenes, and flavonoids, contributing to its distinctive medicinal and culinary properties5,13. Recent research highlights that AKS phenolic-rich extract could serve as effective natural preservatives and functional additives in food products through their inhibitory activity against foodborne pathogens such as Campylobacter jejuni and Staphylococcus aureus, thereby extending product shelf life and enhancing food safety13. A wide range of pharmacological properties have been reported for various AKS extracts and their isolated compounds, including anti-emetic, gastroprotective, anti-inflammatory, antioxidant, neuroprotective, antiviral, and anticancer attributes, underscoring their therapeutic potential7,14,15,16. Few studies highlighted the phytoconstituents of AKS that were reviewed in4,5. The essential oil composition of AKS is dominated by methyl cinnamate (64.2%) alongside alcohols (7.3%), sesquiterpenes (6.8%), and monoterpenes (5.9%)10. In comparison, other studies reported farnesol (14.59 − 21.56%), and 1,8-cineole (18.32–28.20%), as principal constituents in seeds from Liaoning and Yunnan5,10. Shell-derived oil showed 1,8-cineole (19.18%),β-pinene (11.76%),terpinen-4-ol (10.42%),α-thujone (10.01%),and p-cymene (9.28%) s the major components17. Furthermore, AKS is considered a rich source of diarylheptanoids and flavonoids11,18. Diarylheptanoids constitute a broad class of secondary metabolites, dominating in Alpinia drugs, especially AKS, known for their diverse biological activities12,19. A recent study have reported the isolation of a suite of bioactive compounds including acyclic triterpenoids, an acyclic sesquiterpenoid, an arylheptanoid, and two diarylheptanoids, offering promising leads for the development of novel therapeutics targeting hypercholesterolemia20. One of the major bioactive flavonoids in AKS is cardamonin, a chalcone compound that is also reported in cardamom spice and contributes to seeds’ characteristic cardamom-like aroma21. Cardamonin is known for its diverse health benefits, including antitumor, antioxidant, and various other medicinal properties22. This multifaceted compound not only enhances AKS culinary attributes, but also significantly contributes to its overall medicinal value.

Despite the growing interest in the AKS for its diverse bioactive compounds, a notable gap remains in their thorough chemical characterization. Few studies have conducted detailed metabolomic profiling of this species, and no comprehensive analysis employing multiple complementary analytical techniques has yet been reported.

Driven by the remarkable and distinctive chemical profile previously reported for AKS, alongside the lack of comprehensive chemical exploration of its phytochemical potential, this study presents an in-depth profiling of its phytoconstituents targeting aroma, primary and secondary metabolites. Employing advanced analytical platforms exemplified by UHPLC-MS/MS-based molecular networking, SPME-GC–MS, GC–MS post-silylation, to provide a comprehensive profile of bioactive compounds, and adding to its seed rich chemical makeup for nutraceutical and pharmaceutical development.

UHPLC-high resolution MS/MS led to the detection of a broader range of phytochemicals, while molecular networking maps complex phytochemical relationships, identifying novel compounds and searching bioactive clusters against known libraries. SPME-GC–MS provided the true aroma composition in AKS as cold collection method, whereas, post-silylation GC–MS provided insights into nutritive and low molecular weight secondary metabolites post derivatization. Such a multiplex analytical approach provides unparalleled resolution of AKS’s metabolome, and aid to validate traditional uses of AKS based on such detailed analysis, leading to a paradigm shift from descriptive phytochemistry to functional metabolomics.

Results and discussion

Metabolites profiling of AKS via UHPLC-ESI-QTOF-MS/MS-based molecular networking

Metabolites profiling of AKS was carried out in positive and negative modes via UHPLC-ESI-MS/MS-Based Molecular Networking to provide comprehensive metabolome coverage. Representative chromatograms are depicted in (Fig. S1). Metabolites were tentatively identified based on their molecular formulas, retention times and their fragmentation patterns, compared to previous reported data aided with GNPS spectral library search. As listed in Table 1 and 82 metabolites were annotated, belonging to different classes including sugars (L1–L3), amino acids and nitrogenous compounds, (L4–L7), organic acids (L8–L15), phenolic acids (L16–L22), flavonoids (L23–L50), chalcones and dihydrochalcones (L51–L65), calyxins (L66–L68), linear diarylheptanoids (L69–L73), kavalactones (L74–L75), and miscellaneous metabolites (L76–L82). UHPLC-ESI- MS/MS peak numbers are preceded by “L” to be discriminated from GC/MS peaks. This study is the first to assess the phytochemical makeup of AKS holistically using UHPLC-ESI-MS/MS and visualized using molecular networking. Visual analysis of MS/MS data via molecular networking aiding in identification of new metabolites that were tentatively identified for the first time in AKS. Two molecular networks were separately displayed in positive (Fig. 1) and negative (Fig. 2) ionization modes for assignment of compounds as explained in the next subsection for each class in details. All tentatively identified compounds chemical structures are presented in Table S1.

Table 1 Identified metabolites detected in AKS using UHPLC-ESI-MS/MS analysis in positive and negative ionization modes.
Full size table
Fig. 1
figure 1

GNPS molecular network of AKS in positive ion mode. The node label represents precursor mass (m/z). Cluster A, B, and C represents (diarylheptanoids), (flavanones & flavonochalcones & chalcones) and (calyxins), respectively.

Full size image
Fig. 2
figure 2

GNPS molecular network of AKS in negative ion mode. The node label represents precursor mass (m/z). Cluster A, B, C and D represents (proanthocyanidins), (dihydrochalcone mono-C-hexoside), (dihydrochalcone di-C-glycosides) and (flavonol glycosides), respectively.

Full size image

Identification of flavonoids

The tentatively identified flavonoids belonged to 4 main subclasses namely, flavonols, flavanone, flavanonols, proanthocyanidins and their corresponding glycosides. Flavonoids generally have the framework C6-C3-C6 that mostly undergo Retro Diels-Alder (RDA) cleavage23. Flavonol glycosides represented by quercetin-O-rhamninoside (L23), rutin (L25), quercetin-O-hexoside (L26), isorhamnetin-O-hexoside (L28) were clustered in cluster (D) in negative mode (Fig. 2), exhibiting distinctive losses of 454, 308 and 162 Da corresponding to the cleavage of rhamninoside rutinoside and hexoside moieties, respectively.

Noticeably, cluster (B) in positive mode (Fig. 1) characterized by presence of 5 flavanones; naringenin (L34), pinocembrin (L36), alpinetin (L38), obovatin methyl ether (L39), obovatin (L40). Obovatin and its methyl ether derivative were previously isolated in genera Tephrosia and Dalea (family Fabaceae)24,25. Mass fragmentation of flavanones typically yields1,3A+ and1,3B+ fragment ions through RDA fragmentation as base peak. Besides, flavanone itself undergoes cleavage to yield1,4B+ fragment ions (Fig. S2)26,27. Naringenin and pinocembrin showed the same1,3A+ fragment ion at m/z 153.018, but differing in1,3B+ and 1,4B+ fragment ions due to the presence of hydroxyl group on ring B in naringenin, yielding fragment ions that differ by 16 Da (Fig. S2A and B). Likewise, alpinetin showed similar fragment ions1,3B+ and 1,4B+ at m/z 103.054 and 131.049, respectively, with pinocembrin, but differ in 1,3A+ fragment ion at m/z167.033 due to the presence of 5-methoxy group on ring A in alpinetin (Fig. S2C). Besides 1,3,A+ fragment ions at m/z 233.080 and 219.065 for obovatin methyl ether and obovatin indicated the presence of the methyl group (Fig. S2D and E).

Proanthocyanidins (PACs), condensed tannins, are oligomers and polymers consisting of flavan-3-ol monomeric units28. Herein, identified proanthocyanidins have the C4→C8 linkages. Catechin represents the basic common building block of proanthocyanidin. Catechin (L46) displayed fragment ion at m/z 245.082 from the loss of 44 Da (CH2 = CH-OH). Fragment ions at m/z 179.035 and 109.028 were detected due to the loss of dihydroxybenzene moiety, loss of ring B. Heterocyclic ring fission (HRF) led to the formation of fragment ions at m/z 165.018 and 125.023. retro-Diels–Alder (RDA) cleavage of ring C was characterized by presence of fragment ions at m/z 151.039 and 137.023 (Table 1)29. Two PACs dimer type B were detected represented by catechin-(4→8)-catechin (procyanidin B) (L45) and catechin-(4→8)-guibourtinidol (L50), Two PACs trimer type B namely, catechin-(4→8)-catechin-(4→8)-catechin (procyanidin C) (L47) and catechin-(4→8)-catechin − (4→8)-guibourtinidol (L49) were identified and clustered in cluster (A) in negative mode (Table 1; Fig. 2). The main fragmentation mechanisms of PACs can be explained by RDA, quinine methide (QM), and HRF fragmentation pathways28,30,31. Their MS/MS chromatograms were depicted in (Fig. S3). In general, MS/MS of PACs dimer and trimer type B showed the characteristic fragment ions differing by 32 Da as they distinguished in terminal monomer, guibourtinidol differs from catechin by 32 Da, this was further confirmed by QM cleavage fragment ions at m/z 287.056 and 257.082 indicating the presence of catechin and guibourtinidol moieties, respectively (Fig. S4). Interestingly, two new PACs (L49 and L50) were tentatively identified for the first time in Alpinia genus based on their first postulated fragmentation patterns (Fig. S3 and S4). The pronounced biological activities of AKS are strongly attributable to its richness in phenolic metabolites which are well recognized for their anti-inflammatory, antimicrobial, antioxidant and anticancer, etc. activities32.

Identification of chalcones and dihydrochalcones

UHPLC-ESI-MS/MS based molecular networking in positive and negative modes revealed the abundance of chalcone and dihydrochalcones, being detected in clusters (B) in positive mode (Fig. 1) and clusters (B&C) in negative mode (Fig. 2). Among the tentatively identified chalcones and dihydrochalcones in the present study, (L59) flavokawain B and (L63) cardamonin have been previously reported in Alpinia genus4,5,33, Chalcones and dihydrochalcones have inhibitory activities on enzymes, anti-inflammatory, antioxidant, antibacterial, anticancer, antifungal, antimalarial, anti-filarial and antiprotozoal activity34,35. Main fragmentation of chalcones is derived from α-cleavage mechanism leading to the loss of phenyl or styryl radical. In cluster (B) in positive mode (Fig. 1), flavokawain b (L59) and 2’,6’-dihydroxy-4,4’-dimethoxychalcone (L61) showed characteristic fragment ions at m/z 181.049 and 167.033, respectively due to the loss of substituted styryl radical and fragment ions at m/z 131.049 and 161.059, due to loss of substituted phenyl radical36,37 (Fig. S5).

Clusters (B&C) in negative mode (Fig. 2) mainly represented dihydrochalcones-C-glycosides. In C-glycosides, cleavage of sugar moiety, hexose, showed the loss of neutral fragments 120 Da, representing cleavage of 1’→2’ linkage versus 90 Da representing the cleavage of 1’→3’ linkage38,39. Cluster (B, Fig. 2) represented dihydrochalcone mono-C-hexoside; aspalathin (L51), nothofagin (L53), and deoxy phloretin-3’-C-hexoside (L55) that yielded fragment ions at m/z 361.093, 345.098, 329.103 ([M-H-90]−) and m/z 331.082, 315.088, 299.092 ([M-H-120]−), respectively. Interestingly, fragment ions at m/z 289.072, 273.077, 257.082 ([M-H-162]−), loss of six-member ring sugar moiety, were also observed38 (Fig. S6). Additionally, nothofagin (L53), and deoxy phloretin-3’-C-hexoside (L55) yielded fragment ions at m/z 167.034 and 125.023, resulting from β- and α- cleavages of the carbonyl group, respectively (Fig. S6B and C). It is noteworthy that deoxy phloretin-3’-C-hexoside (L55) is a putative compound and the first time to be identified in Alpinia genus based on its mass fragmentation.

On the other hand, cluster (C, Fig. 2) exhibited two dihydrochalcone di-C-glycosides represented by phloretin 3’,5’-di-C-hexoside (L52), deoxy phloretin 3’,5’-di-C-hexoside (L54). They were clustered with 2ʹ,4ʹ,6ʹ-trihydroxy-acetophenone-3ʹ,5ʹ-di-C-hexoside (L78) to produce characteristic fragments ions at m/z 477.138, 461.146, 371.097 ([M-H-120]−), m/z 387.108, 371.113, 281.067 ([M-H-210]−) and m/z 357.099, 341.103, 251.056 ([M-H-240]−), respectively, and inferring that peaks L52, 54 and L78 belonged to di-C-glycosides (Fig. S7).

Other dihydrochalcones appearing in positive ion mode, derived network included de-O-methyl rotundaflavanochalcone (L62) and rotundaflavanochalcone (L65), flavonochalcones, to encompass a flavanone connected to a dihydrochalcone moiety via a C–C bond, previously isolated from Boesenbergia rotunda genus (family Zingiberaceae)40. They appeared in cluster (B) in positive mode (Fig. 1). The MS/MS spectrum of both compounds showed the same fragment ion at m/z 387.122 due to α cleavage of carbonyl group in the dihydrochalcone moiety leading to loss of phenyl radical. Whilst, they differ in phenyl group substituents leading to the formation of fragment ions at m/z 153.018 and 167.033, respectively (Fig. S8). Moreover, C-C glycosidic bond cleavage led to fragment ions at m/z 257.080 and 271.096, respectively (Fig. S8) corresponding to dihydrochalcone and flavanone moieties. This is the first time that a fragmentation pattern mechanism of those compounds has been proposed in the literature.

Identification of calyxins (chalcone/flavanone-diarylheptanoids)

Various calyxins were previously reported from Alpinia genus4,5, to exert various biological activities relevant to antidiabetic, antibiotic, antiproliferative, and vasodilative activities5,41. Herein, the identified calyxins were clustered in cluster (C) in positive mode (Fig. 1). They consist of pinocembrin (calyxin N/O isomers (L66), and calyxin P (L68) or de-O-methyl flavokawain B (calyxin Q, (L67)) attached with diarylheptanoid, mostly identified based on their RDA fragmentation pathway in ring C followed by the elimination of side chain. The MS/MS spectrums of calyxins are similar to each other due to similarity in structures, calyxin P is a demethylated derivative of calyxin N/O, yielding fragment ions that differ from each other by 14 Da, indicating the presence of a methyl group in calyxin N and O (Fig. S9A andC). Whilst, MS/MS spectrum of calyxin Q was identical to that of calyxin N/O, likely due to isomerization of chalcone into their corresponding flavanones42,43 (Fig. S9B). It should be noted that this is the first suggested fragmentation pattern of calyxins (Fig. S10). Additionally, fragment ions at m/z 147.044, 131.049, 117.070 and 91.055, resulting from mass fragmentation of diarylheptanoid moiety, aided in identification.

Identification of diarylheptanoids

Diarylheptanoids are mostly reported from Alpinia genus4,5, and unique constituents of the Zingiberaceae family to impart several health effects, including hepatoprotective, anticancer, antioxidant, and melanogenesis activities, besides their nutraceutical applications as organoleptic additives in foods44. They were identified in cluster (A) in positive mode (Fig. 1). Five diarylheptanoids were detected, represented by 1,7-diphenyl-5-hydroxy-1-heptene (L69), 5-hydroxy-1-(4-hydroxyphenyl)−7-phenylhepta-1,6-dien-3-one (L70), 1,7-diphenyl-4,6-heptadien-3-one (L71), 1,7-diphenyl-5-hydroxy-4,6-heptadien-3-one (L72), 1,7-diphenyl-1,6-heptadiene-3,5-dione (73). (L70) and (L73) were previously detected in turmeric, Curcuma longa, (family Zingiberaceae)45 and in Eugenia jambos (family Myrtaceae)46, respectively. Diarylheptanoids encompass two phenyl groups joined by a 7-carbon chain (heptane) and have various substituents and undergo the intermolecular cleavage of the tautomeric isomers, followed by the cleavage of the heptanoid chain between the carbonyl and methylene groups47,48. The mass fragmentation spectrum of identified diarylheptanoids is presented in Fig. S11.

SPME/GC–MS analysis of AKS

Essential oil is regarded as a key component of Alpinia taxa, primarily consisting of various terpenoids4, to contribute to its culinary uses and further health benefits. This study explored the aroma profile of AKS using SPME/GC–MS analysis. GC–MS peak numbers are preceded by “G”. A total of 30 different volatile constituents were identified and categorized into various classes, with monoterpene and sesquiterpene hydrocarbons most abundant detected at 36.8% and 55.1% of the total volatiles, respectively. A detailed summary of the identified compounds and their absolute quantities (µg/g) is presented in Table 2, while Fig. 3 illustrates GC–MS chromatogram.

Table 2 Absolute quantification (µg/g) of volatiles of AKS identified using GC–MS analysis. The values are expressed as average ± st. dev. (n = 3).
Full size table
Fig. 3
figure 3

Representative GC–MS chromatogram of AKS for its volatiles analyzed by SPME coupled to GC–MS. The corresponding volatile names for each major peak follow that listed in Table 2. G4: α-Phellandrene, G6: p-Cymene, G8: 1,8 Cineole, G18: Daucene, G19: β-Caryophyllene, and G21: α-Humulene.

Full size image

Sesquiterpene hydrocarbons

Sesquiterpene hydrocarbons constituted the major class in AKS, with 7 peaks. Daucene (G18) represented the major form at highest level of 274.41 µg/g, followed by α-humulene (G21) and β-caryophyllene (G19) at 197.56 µg/g and 62.05 µg/g, respectively. α-Humulene and β-caryophyllene are key odor-active compounds in A. katsumadai seed oil (AKSO), and their high abundance may contribute significantly to its woody and spicy aroma49. Other sesquiterpene hydrocarbons detected at much lower levels included bicyclosesquiphellandrene (G22), α-muurolene (G23) and germacrene D (G28).

Monoterpene hydrocarbons

Compared to sesquiterpenes, monoterpene hydrocarbons accounted for the second most prevalent class of volatile compounds in AKS. A total of 9 monoterpene hydrocarbons were detected in AKSO. In contrast to previous reports4, our findings show major forms included α-phellandrene (G4) at 161.25 µg/g of potential aroma and used in pharmaceutical, food, and cosmetic industries due to its pleasant scent50. The second most abundant volatile was p-cymene (G6) at 101.25 µg/g. Other identified monoterpene hydrocarbons included α-Thujene (G1), sabinene (G2), β-myrcene (G3), isoterpinolene (G5), D-limonene (G7), β-cis-ocimene (G9), and α-ocimene (G11) that were detected at lower levels.

Ketones

Ketones accounted for 4.04% of the total volatiles detected in AKS. Among them, L-fenchone (G10), noted for its herbaceous and fragrance-like aroma51,52, was detected at a concentration of 19.2 µg/g. Other identified ketones include camphor (G12) and benzylacetone (G16), which were detected at 1.03 and 23.02 µg/g, respectively.

Alcohols and oxides

Ocimenol (G14) was the only identified alcohol found at a trace amount of 8.66 µg/g in AKS. whereas 1,8 cineole (G8), a monoterpene oxide, was detected at a relatively high level of 86.27 µg/g compared to other volatile components, which is consistent with findings from previous studies5.

GC–MS profiling of AKS metabolites post-silylation

GC–MS post-silylation analysis was employed to characterize A. katsumadai seed’s chemical profile, with emphasis on primary metabolites not detected using LC-MS, resulting in the identification of 48 metabolites that may contribute to its food or health effects. These metabolites were categorized into diarylheptanoids, organic acids, alcohols, amino acids, phenolic acids, flavonoids, sugars, fatty acids, mono-, and sesquiterpenes. In terms of quantity, the most abundant classes belonged to sugars 58.23%, arylheptanoids 10.05%, and flavonoids 8.88%. A list of identified compounds and their relative abundances is listed in Table 3, with the GC chromatogram displayed in Fig. 4.

Table 3 Quantification of silylated primary metabolites of AKS identified using GC–MS analysis. The values are expressed as average ± st. dev. (n = 3).
Full size table
Fig. 4
figure 4

Representative GC–MS chromatogram of AKS for silylated primary metabolites. The corresponding compound numbers for peaks follow those listed in Table 3.

Full size image

Sugars/sugar alcohols/sugar acids

Sugars account for the seed’s nutritional value and taste, impacting its palatability53. In A. katsumadai seed, sugars amounting to the most dominant primary metabolites comprising 9 peaks. Sucrose (G45) was the main di-sugar, detected at 23.81 mg/g, constituting 26.51% of the total identified metabolites. In contrast, mono-sugars were detected but at much lower levels represented by fructose (G17, G19, G22), arabinose (G20), mannose (G23), and glucose (G27). Regarding sugar alcohols, only D-mannitol (G26) was detected at a low level, whereas D-gluconic acid (G28) was the only identified sugar acid, suggesting that di-sugars account the most for A. katsumadai seed calorie and taste.

Fatty acids

Fatty acids constitute an essential group of primary metabolites, represented by 6 distinct peaks observed at comparatively lower levels. The fatty acid composition of AKS was analyzed, revealing the presence of saturated and unsaturated fatty acids. Among the saturated fatty acids, palmitic acid (G29), a long-chain saturated fatty acid, was detected as the most abundant, constituting 0.96 mg/g. Other detected saturated fatty acids included dodecanoic acid (G15), stearic acid (G31), and lignoceric acid (G46). In addition, two monounsaturated fatty acids were identified: oleic acid (G30), which accounted for 0.67 mg/g and cis-13-docosenoic acid (G47), comprising 0.27 mg/g.

Organic acids

Organic acids play a crucial role in food products as natural preservatives, supporting digestion and enhancing protein utilization while also exhibiting antioxidant activity11,18. Acrylic acid (G25) was the major form accounting for 0.45 mg/g representing 1.61% of total identified metabolites, and likely to contribute for seed taste, followed by oxalic acid (G2), which was detected at 0.17 mg/g. Moreover, lactic acid (G1) and malic acid (G11) were found in trace amounts.

Alcohols

Alcohols were found at much lower levels, primarily represented by glycerol (G6), a simple triose compound with diverse pharmaceutical applications54. Additionally, a trace presence of carotol (G14), a sesquiterpene alcohol, was detected.

Amino acids/Nitrogenous compounds

In AKS, amino acids were detected at trace levels compared to other chemical classes represented by valine (G3) and 2-pyrrolidone carboxylic acid (G12) which were detected at 0.07 and 0.05 mg/g, respectively. Besides, a higher level of unknown nitrogenous compound (G8) was detected at 0.14 mg/g.

Arylheptanoids

A total of 6 arylheptanoids were detected in A. katsumadai seeds including three diarylheptanoids (G4, G32, G39), two phenylheptanoids (G35), and 1-phenyl-1-hepten-3-ol (G38), which exhibited the highest level at 8.51% of the total identified metabolites and 84.13% of the identified arylheptanoids, in addition to a peak of an arylbutanoid compound (G37).

Flavonoids and non-phenolic acid compounds

Flavonoids constituted a major class represented by two catechins peaks (G21, G48) and two peaks (G33, G42) of cardamonin, the predominant flavonoid compound accounting for 6.46% of the total identified metabolites. It plays a key role in imparting the characteristic strong aroma of AKS, giving them a scent like that of cardamom5,21. Additionally, alizarin (G34), a dihydroxyanthraquinone metabolite, was detected at a low abundance percentage of 0.07%. Other phenolic derivatives, such as phloroglucinol (G10) and 3,4-dihydroxybenzophenone (G41), were also detected at levels of 0.14% and 1.87%, respectively.

Phenolic acids

Compared to other metabolite classes, phenolic acids were also identified in AKS at relatively lower levels. Three phenolic acids were detected, including benzoic acid (G5) and cinnamic acid (G13), with trace levels. Notably, methyl mandelic acid (G43) was the predominant phenolic acid, comprising 0.5% of the total identified metabolites.

Conclusion

The current study presented a comprehensive investigation of AKS metabolome, targeting primary and secondary metabolites as well as aroma constituents to decode for seed health-promoting properties and culinary value. UHPLC-ESI-MS- MS/MS-based GNPS networking allowed for the annotation of 82 secondary metabolites across diverse chemical classes. Such extensive profiling led to the detection of two newly annotated catechin-guibourtinidol derivatives and deoxy phloretin-3’-C-hexoside, identified for the first time in the Alpinia genus. As well, SPME headspace aroma profiling identified a total of 30 volatiles belonging to various classes, with the abundance of monoterpenes and sesquiterpenes accounting for its discrete aroma. Results of GC–MS post-silylation analysis resulting in the identification of 48 metabolites, revealing the predominance of sugars, arylheptanoids, and flavonoids. Among these, sucrose emerged as the most abundant sugar, while cardamonin was identified as the predominant flavonoid. These components significantly underpin the seed’s dual role in culinary and medicinal applications: sucrose imparts a delicate sweetness, while cardamonin contributes to its distinctive aromatic flavor profile alongside certain potent pharmacological actions, such as anti-inflammatory and anticancer activities. Future investigations employing other broader analytical techniques targeting minerals, vitamins in AKS are recommended to identify optimal resources and further elucidate their health-promoting effects.

Materials and methods

Plant material

Alpinia katsumadai seeds were purchased from local store in Hangzhou China and authenticated by Dr. Zuying Zhang, Zhejiang A&F University, Hangzhou, Zhejiang, China. A voucher specimen (No. 10-9-25-F) was placed at the herbarium of the Faculty of Pharmacy, Cairo University.

Chemicals

Formic acid and acetonitrile (HPLC grade) were provided by Baker (Deventer, The Netherlands). All other solvents, standards, and chemicals were obtained from Sigma Aldrich (St Louis, MO, USA). For SPME sampling, Special fibers of polydimethylsiloxane or divinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS, 50 μm/30 µm of 1 cm in length) were obtained from Supelco (Bellefonte, PA, USA). All solvents, standards and chemicals were obtained from Sigma Aldrich (St. Louis, MO, USA).

UHPLC-ESI-QTOF-MS/MS analysis and GNPS feature-based molecular MS/MS network

Finely ground AKS (10 mg) was subjected to extraction with 2 mL of 70% MeOH, containing 10 µg mL−1 umbelliferone (internal standard) and sonication for 20 min with intermittent shaking, then centrifugation at 12 000 × g for 10 min. Then the extract was filtered through a 0.22-µm filter and subjected to solid-phase extraction using a C18 cartridge (SepPack, Waters, Milford, MA, USA). 2 µL of the extract were injected on an HSS T3 column (100 × 1.0 mm, particle size 1.8 μm; Waters, Milford, MA, USA) installed on an ACQUITY UPLC system (Waters) equipped with a 6540 Agilent Ultra-High-Definition (UHD) Accurate-Mass Q-TOF-LC–MS (Palo Alto, CA, USA) coupled to an ESI interface, operated in positive or negative ion mode under the exact conditions as previously described55. Duplicate runs per sample were performed.Metabolites were characterized by their exact masses, MS/MS fragmentation in both positive and negative ionization modes, retention time, and comparisons to the Natural Products database and reference literature. Using the FBMN workflow on GNPS2, a molecular networking was created. The resulting aligned list of features was exported in an mgf file besides their feature quantification table in csv format. The values of feature quantification table were uploaded onto the FBMN page of GNPS2. Edges of the MN were filtered to have a cosine score above 0.65 and more than 4 matched peaks between the connected nodes. The edges between two nodes were kept in the network. The MNs were visualized using Cytoscape 3.9.1. For molecular networking of AKS in positive mode (https://gnps2.org/status? task=ae9207b6aa644d48ac4e7f9515cb6f90 ), and in negative mode (https://gnps2.org/status? task=bc6a6741185343c0a06afb68aaf4828c ), the precursor ion mass tolerance was set to 0.02 Da and the MS/MS fragment ion tolerance to 0.02 Da.

SPME GC–MS volatiles analysis

100 mg of freeze-dried seed powder was loaded into 1.5 mL SPME vials spiked with 10 µg (Z)−3-hexenyl acetate. After fiber insertion above, samples were equilibrated at 50 °C for 30 min. Volatile analysis was performed by HS-SPME/GC–MS using an Agilent 5977B GC/MSD with a DB-5 column (30 m × 0.25 mm × 0.25 μm film thickness; Supelco) and a quadrupole mass spectrometer. The injector and interface temperatures were maintained at 220 °C. The oven temperature program was initiated at 40 °C for 3 min, ramped to 180 °C at 12 °C/min (held 5 min), then to 240 °C at 40 °C/min (held 5 min). Helium carrier gas flowed at 0.9 mL/min. Post-analysis, fibers were reconditioned at 220 °C for 2 min. Triplicate runs per sample were performed, included interspersed blanks, with EI-MS operating at 70 eV (scan range: m/z 40–500)56. Volatiles were quantified relative to the amount of recovered hexenyl acetate as an internal standard.

GC–MS analysis of silylated primary metabolites

Freeze-dried seed powder (100 mg) was extracted with 5 mL 100% methanol with sonication (30 min, frequent vortex shaking). Triplicate extracts per accession were processed. Aliquots (100 µL) were evaporated under a nitrogen gas stream, then derivatized with 150 µL N-methyl- N-(trimethylsilyl)-trifluoroacetamide (MSTFA, Sigma, St. Louis, MO, USA)/anhydrous pyridine (1:1) at 60 °C for 45 min prior to GC–MS analysis. Silylated derivatives were separated on a Rtx-5MS column (30 m × 0.25 mm, 0.25 μm). Quantitative analysis of primary metabolites followed54. Soluble sugars, amino acids, organic acids and fatty acids were quantified using standard curves of glucose, glycine, citric and stearic acid standards (Four serial dilutions of each, from 10 to 600 µg/mL; R² ≈ 0.99). Results were expressed as mg/g dry weight56. GC–MS data were processed using AMDIS software (www.amdis.net) for peak deconvolution before mass spectral matching. Identification of both volatile and silylated components was based on calculated KI using alkane standard C8-C40, mass matching to NIST database and standards whenever available as the exact protocol detailed in57. Peak abundance was obtained using MS-DIAL software following default parameters for GC/MS53.