Comparative Phytochemical Analysis of Five South African Bulbine Species Using Chromatography‐Based Methods

ABSTRACT

Bulbine Wolf (Asphodelaceae) comprises over 80 species, with five holding significant medicinal value in South Africa for treating skin and gastrointestinal ailments. While roots are traditionally prioritised, the chemical profiles of leaves and rhizomes remain underexplored although they are also medicinally important. This study used chromatographic and mass spectrometry‐based analytical techniques (HPTLC, LC‐QToF‐MS, and UPLC‐MS) combined with chemometric analysis to chemically profile the metabolites of five Bulbine species from the Eastern Cape. Chemical profiling tentatively identified 73 compounds, including flavonoids, phenolic acids, and anthraquinones. Phenylanthraquinones, specifically knipholone and knipholone 6′‐methyl ether, were ubiquitous across all taxa. Species‐specific markers were identified, such as bulbine‐knipholone in B. abyssinica, gaboroquinones in B. frutescens, and bulbnatalonosides in B. latifolia. Chemometric modelling revealed that while rhizome chemistry is relatively conserved, root profiles provide the most robust inter‐species differentiation. Bulbine frutescens and B. latifolia displayed the highest chemical diversity, whereas B. asphodeloides was notably rich in anthocyanins. The significant overlap in bioactive constituents between leaves, roots and rhizomes suggests that aerial parts could serve as viable substitutes in traditional remedies. These findings establish a chemical basis for species authentication and support sustainable harvesting practices by reducing the destructive exploitation of rhizomes and roots.

Comprehensive chromatographic and chemometric profiling of five medicinal South African Bulbine species across leaves, rhizomes, and roots reveals 73 metabolites, including species‐specific phenylanthraquinones, flavonoids, and anthocyanins, that serve as reliable chemical markers for species authentication. Crucially, the significant overlap in bioactive compounds between leaves and underground parts opens a sustainable harvesting pathway, reducing the destructive exploitation of roots and rhizomes.

1. Introduction

Bulbine Wolf of the Asphodelaceae family comprises over 80 documented species, and 74 of these are endemic to Southern Africa. Five species hold ethnobotanical relevance in South Africa, and these are Bulbine abyssinica, Bulbine asphodeloides, Bulbine frutescens, Bulbine latifolia, and Bulbine narcissifolia [1, 2, 3, 4, 5]. Collectively, these five Bulbine species occur across all nine provinces of South Africa, although their individual distributions differ. These Bulbine sp. are also widely cultivated as ornamental garden plants. While Bulbines are perennial plants with adaptations for survival in arid and semi‐arid regions, they differ mainly in their root, leaf, and rhizome structures. Bulbine abyssinica has spindle‐shaped roots and sheathing leaves from a solitary rhizome [4, 6]. Bulbine asphodeloides has a small subcylindrical rhizome with compact or arching leaves [7]. Bulbine frutescens bears thin, wiry roots with cylindrical and grooved leaves [3, 4]. Bulbine latifolia has a bulbous rhizome with fleshy, red‐tinged roots and firm, rosette‐arranged Aloe‐like leaves [3, 4, 8, 9], and B. narcissifolia bears spindle‐shaped roots with strap‐shaped leaves that emerge from a vertical and woody rhizome [4, 10].

Ethnobotanical records highlight the extensive medicinal applications of Bulbine species, where B. abyssinica is traditionally used to manage diabetes, alleviate pain and inflammation, treat gastrointestinal disorders and infections, and soothe cracked lips [11, 12]. Bulbine asphodeloides is primarily used to treat dermatological conditions, including wounds, burns, inflammation, and insect bites [5, 13], similar to B. frutescens, which is widely used for burns, insect bites, sores, and skin infections [14, 15]. Bulbine latifolia is used for gastrointestinal ailments, arthritis, and general maintenance of skin health [16, 17], while B. narcissifolia is used for both skin and gastrointestinal disorders [3, 18]. In all cases, the leaf gel is freshly applied topically to manage skin conditions, while decoctions prepared from the roots or rhizomes are administered orally. There is a clear overlap in the traditional uses of Bulbine species, necessitating the investigation of their phytochemistry to determine if these applications are supported by their chemical composition.

Earlier phytochemical investigations of Bulbine species have predominantly focused on the roots, resulting in the isolation of key compounds such as anthraquinones, isofuranonaphthoquinones, phenolics, flavonoids, and triterpenes [2]. Notably, the identification of Aloe‐emodin in the roots of B. frutescens and B. abyssinica established anthraquinones as a major class of compounds in the genus, consistent with observations in related members of the Asphodelaceae family, including Aloe and Kniphofia species [19, 20]. Within this class, phenylanthraquinones in Bulbine species exhibit greater structural diversity than in other genera, contributing to the phytochemical uniqueness of the genus [21, 22].

Although roots have been extensively characterised, research on the leaves and rhizomes, which are also considered medicinally important and display good pharmacological potential, remains scarce. Some progress has, however, been made toward closing this gap. A study by [23] identified ten antidiabetic compounds in a polyphenol‐rich fraction of B. abyssinica leaves, including astilbin, galangin, petunidin 3‐rutinoside, ψ‐tectorigenin, rutin, and variabiloside A, among others. Another comparative study by [24] revealed the presence of higher levels of phenolics, flavonoids, tannins, and proanthocyanidins in the leaves compared to the roots and rhizomes of B. abyssinica. In a subsequent study, the leaves of B. frutescens were found to exhibit greater chemical diversity, whereas the roots contained higher levels of total phenolics, flavonoids, and tannins [25]. Similarly, in B. natalensis, the rhizomes were reported to possess higher levels of total phenolics, flavonoids, and proanthocyanidins compared to the leaves and roots [26]. The study by [27] isolated the dimeric anthracene derivatives, abyquinones A, B, and C from B. abyssinica fruits. In one of the few phytochemical studies on B. narcissifolia roots, two new anthraquinone glycosides, namely knipholone‐8‐β‐D‐gentiobioside and chrysalodin‐10‐β‐D‐gentiobioside, were identified [18]. Other compounds identified include chrysophanol, knipholone, isoknipholone, and chrysalodin, further confirming anthraquinones as one of the major compound classes of Bulbines.

The background outlined above highlights the need for a systematic investigation of the chemical profiles of the leaves and rhizomes, in addition to the roots, of these five ethnobotanically important Bulbine species. Such an investigation will support the standardisation of these herbal remedies through the identification and quantification of marker compounds, which are also essential for species differentiation. Accordingly, in this study, five Bulbine sp. from the Eastern Cape Province were analysed using chromatography‐based techniques combined with chemometrics to elucidate interspecies variation.

2. Results and Discussion

2.1. The HPTLC Profiles of Bulbine Species

2.1.1. Leaf Profiles of the Five Bulbine Species

The methanol leaf profiles of the five Bulbine species are displayed in Figure 1, where three representative samples were analysed for each species. A total of 15 bands were observed across the five species, and 11 of these were tentatively identified, while four remain unknown (Table 1). Bulbine abyssinica displayed the most comprehensive profile with nine clear bands of high intensity. The green intense bands present in B. abyssinica at Rf 0.3, 0.35, 0.62, and 0.9 were identified as knipholone, bulbine‐knipholone, knipholone‐glucopyranoside, and knipholone 6′‐methyl ether, respectively. While bulbine‐knipholone and knipholone‐glucopyranoside were unique to B. abyssinica, knipholone was also identified in B. latifolia and knipholone 6′‐methyl ether in both B. latifolia and B. frutescens. A brown band at Rf 0.8, uniquely present in B. asphodeloides, was identified as isochlorogenic acid C 4‐O‐glucoside. Other compounds present in B. asphodeloides at high intensity are pelargonidin 3‐O‐sambubioside (R f 0.18) and dicaffeoylquinic acid (R f 0.55). Bulbine frutescens largely displayed a similar profile to B. asphodeloides; however, the presence of a pink band at Rf 0.49 identified as gaboroquinone A/B, was unique to B. frutescens. Although the fingerprint profiles for B. latifolia were not as highly resolved as the other four species, the prominent yellow band at Rf 0.89, identified as kaempferol glucoside, was unique to B. latifolia. The most prominent and unique compound to B. narcissifolia was identified as apigenin‐galacturonide, which presented as a maroon band at Rf 0.48. These results indicate clear differences in the leaf chemistry of the five Bulbine species, based on HPTLC analysis.

FIGURE 1.

The HPTLC leaf profiles of the five Bulbine species observed under white light.

TABLE 1.

Compounds identified in the leaf profiles of the five Bulbine sp using HPTLC analysis.

Rf value a	Band colour	m/z [M+H]+ b	Identification	Species
0.05	Orange‐dark brown	—	—	All species
0.17–1.90	Black	565.1564	Pelargonidin 3‐O‐sambubioside	B. asphodeloides
		595.1638	Peonidin 3‐O‐sambubioside /Cyanidin‐3‐O‐rutinoside
0.30	Dark green	435.1071	Knipholone	B. abyssinica, B. latifolia
0.35	Dark Green	435.1071	Bulbine‐knipholone	B. abyssinica
0.41	Light orange	—	—	B. abyssinica
0.48	Maroon	447.0944	Apigenin‐galacturonide	B. narcissifolia
0.49	Pink	451.1135	Gaboroquinone A/B	B. frutescens
0.49	Brown	—	—	B. abyssinica
0.55	Orange	517.1564	Dicaffeoylquinic acid (diCQA)	B. asphodeloides
0.58	Yellow	531.0613	—	B. frutescens
0.62	Green	597.1627	Knipholone‐glucopyranoside	B. abyssinica
0.73	Violet	—	—	All species
0.80	Brown	679.1701	Isochlorogenic acid C 4‐O‐glucoside	B. asphodeloides
0.89	Yellow	287.0551	Kaempferol glucoside	B. latifolia
0.90	Dark green	449.1066	Knipholone 6′‐methyl ether	B. abyssinica, B. frutescens, B. latifolia

^a Rf value = retardation or retention factor.

^b m/z [M+H] = molecular ion in positive‐ion mode mass spectrometry

2.1.2. Rhizome Profiles of the Five Bulbine Species

The HPTLC rhizome profiles of the five Bulbine species are displayed in Figure 2. A total of 16 bands of different colour and intensities were observed across the species, and 9 of these were tentatively identified (Table 2). The species with the most comprehensive rhizome profile was B. latifolia, where nine bands were visibly resolved, and five of these distinctively occurred in this species. The dark brown band at R f 0.18–0.21, corresponding to dehydrobulbnatalonoside A isomer I; the brown band at R f 0.25, identified as bulbnatalonoside C; the yellow‐green and light‐green unidentified bands at R f 0.35 and 0.50, respectively; along with bulbnatalone (yellow‐green, Rf 0.60) and dehydrobulbnatalonoside A isomer II (light green, R f 0.80), were distinctive compounds found exclusively in B. latifolia. The B. abyssinica profile had a distinct orange band at Rf 0.42 (unidentified) and a green band at R f 0.68 identified as demethylknipholone. B. asphodeloides and B. narcissifolia fingerprints were visually distinct; however, the two species shared a compound at R f 0.28 (yellow band), which was identified as Aloe‐emodin. A prominent unidentified orange band at R f 0.55 was observed in B. frutescens. Although the green band at R f 0.3 was observed in both B. abyssinica and B. latifolia, the compound was identified as knipholone in B. abyssinica and bulbnatalonoside D in B. latifolia. One compound that remains unidentified at R f 0.48 (pink) was the only common compound to B. abyssinica, B. asphodeloides, B. frutescens, and B. narcissifolia. The dark green band at R f 0.9, representing knipholone 6′‐methyl ether, was observed in both B. abyssinica and B. latifolia.

FIGURE 2.

The HPTLC rhizome profiles of the five Bulbine species observed under white light.

TABLE 2.

Compounds identified in the rhizome profiles of five Bulbine sp using HPTLC analysis.

R f value a	Band colour	m/z [M+H]+ b	Identification	Species
0.05–0.08	Dark green	—	—	All species
0.18–0.21	Dark green	655.1829	Dehydrobulbnatalonoside A isomer I	B. latifolia
0.25	Brown	839.2169	Bulbnatalonoside C	B. latifolia
0.28	Yellow	271.0603	Aloe‐emodin	B. asphodeloides, B. narcissifolia
0.30	Green	435.1071	Knipholone	B. abyssinica
		795.2107	Bulbnatalonoside D	B. latifolia
0.35	Yellow‐green	—	—	B. latifolia
0.42	Orange	—	—	B. abyssinica
0.48	Pink	—	—	B. abyssinica, B. asphodeloides, B. frutescens, B. narcissifolia
0.50	Light green	—	—	B. latifolia
0.55	Orange	491.2146	—	B. frutescens
0.60	Yellow‐green	449.1258	Bulbnatalone	B. latifolia
0.68	Green	421.0912	Demethylknipholone	B. abyssinica
0.70	Violet	—	—	B. narcissifolia
0.73–0.75	Violet	—	—	All species
0.80	Light green	655.1829	Dehydrobulbnatalonoside A isomer II	B. latifolia
0.90	Dark green	449.1098	Knipholone 6’‐methyl ether	B. latifolia, B. abyssinica

^a Rf value = retardation or retention factor.

^b m/z [M+H] = molecular ion in positive‐ion mode mass spectrometry

2.1.3. Root Profiles of the Five Bulbine Species

The representative root profiles of the five Bulbine species are shown in Figure 3. Nineteen distinct bands were observed across the five species, and only seven were tentatively identified (Table 3). The green band at R f 0.3 was identified as knipholone, which was present in B. abyssinica, B. asphodeloides, B. frutescens, and B. latifolia. However, at the same R f 0.3, the purple band was identified as chrysalodin in B. narcissifolia. Bulbnatalonoside D was a distinct marker of B. latifolia identified as a brown band at R f 0.32, while a purple band at R f 0.35, identified as bichrysophanol, was unique to B. narcissifolia, and an orange band at R f 0.41, identified as pelargonidin 3‐O‐sambubioside, was unique to B. asphodeloides. The green band (R f 0.9) was identified as knipholone 6′‐methyl ether in B. abyssinica, B. asphodeloides, and B. latifolia, while for B. frutescens, an additional compound was identified as gaboroquinone A/B. The identities of the other 12 compounds could not be verified; the profiles of the three species are highly diverse, with few compounds shared among the five species.

FIGURE 3.

The HPTLC root profiles of the five Bulbine species observed under white light.

TABLE 3.

Compounds identified in the root profiles of five Bulbine sp using HPTLC analysis.

R f value a	Band colour	m/z [M+H]+ b	Identification	Species
0.06–0.10	Dark brown‐black	609.1755	—	B. latifolia
		689.1329	—	B. narcissifolia
0.12–0.19	Dark brown‐black	—	—	All species
0.30	Green	435.1071	Knipholone	B. abyssinica, B. asphodeloides, B. frutescens, B. latifolia
	Purple	525.1196	Chrysalodin	B. narcissifolia
0.32	Brown	795.2107	Bulbnatalonoside D	B. latifolia
0.35	Green	—	—	B. abyssinica, B. asphodeloides, B. frutescens
	Light purple	509.1206	Bichrysophanol	B. narcissifolia
0.41	Orange	—	—	B. frutescens, B. latifolia
		546.1564	Pelargonidin 3‐O‐sambubioside	B. asphodeloides
0.46	Light green	625.1741	—	B. narcissifolia
0.52	Brown	359.0969	—	B. latifolia
0.55	Orange	449.1066	—	B. frutescens
0.57	Brown	—	—	B. latifolia
0.64	Brown	689.1329	—	B. latifolia
0.65	Green, brown‐yellow	— 233.1191	—	B. frutescens, B. narcissifolia
0.67	Brown	—	—	B. asphodeloides
0.70	Green	—	—	B. frutescens
0.74	Violet	669.1755	—	B. narcissifolia
0.80	Brown	—	—	B. latifolia
0.89‐0.9	Dark green	451.1039	Gaboroquinone A/B	B. frutescens
		449.1210	Knipholone 6′‐methyl ether	B. abyssinica, B. asphodeloides, B. frutescens, B. latifolia

^a Rf value = retardation or retention factor.

^b m/z [M+H] = molecular ion in positive‐ion mode mass spectrometry.

Overall, B. abyssinica displays an anthraquinone‐ and phenolic‐rich chemical profile across its leaves, rhizomes, and roots, dominated by structurally related compounds such as knipholone, knipholone 6′‐methyl ether, and demethylknipholone. The recurrence of knipholone and its derivatives across all three plant parts suggests a conserved biosynthetic pathway for anthraquinone production within the species. However, plant part‐specific metabolites such as bulbine‐knipholone and knipholone‐glucopyranoside in the leaves, and demethylknipholone in the rhizomes, indicate differential metabolic modifications depending on the plant part. The leaves displayed the highest chemical diversity and the most intense HPTLC profiles, while the rhizomes contained unique anthracene‐type derivatives, and the roots showed a simpler but related chemical profile enriched in knipholone‐type compounds.

Across the plant parts, B. asphodeloides displays a phenolic‐ and anthraquinone‐dominated profile, with consistent detection of pelargonidin 3‐O‐sambubioside and knipholone derivatives. The leaves and roots share a common anthocyanin profile, while the rhizomes show a shift toward Aloe‐emodin–type quinones. The occurrence of both phenolic acids (e.g., dicaffeoylquinic acid) and anthraquinones (e.g., knipholone, Aloe‐emodin) highlights a dual biosynthetic tendency toward both flavonoid and polyketide pathways.

Bulbine frutescens displays a chemically conserved profile across the plant parts, characterised by the presence of gaboroquinone A/B and knipholone derivatives. The leaves and roots show similar composition, while the rhizomes displayed distinct orange‐pigmented compounds, possibly representing oxidised quinone derivatives.

Bulbine latifolia demonstrates clear differences across the plant parts, with kaempferol glucoside dominating in the leaves, multiple bulbnatalonosides and bulbnatalone confined to the rhizomes, and bulbnatalonoside D characterising the roots. The presence of both anthracene glycosides and flavonoid derivatives reflects a highly diversified secondary metabolism. The overlap of knipholone‐type compounds across the plant parts suggests a shared biosynthetic core, while the bulbnatalonosides show a distinct pattern in the underground parts.

Bulbine narcissifolia presents a chemically distinctive profile dominated by anthraquinone and flavone derivatives. The leaves are rich in apigenin‐galacturonide, while the roots contain unique dimers such as bichrysophanol and chrysalodin, suggesting specialised oxidative coupling mechanisms in the roots. The rhizomes serve as an intermediate, featuring Aloe‐emodin typical of the genus but lacking the strongly pigmented flavones of the leaves.

2.2. The LC‐QToF‐MS Profiles of Bulbine Species

The LC‐QToF‐MS profiles of the five Bulbine species obtained in the positive and negative ionisation modes are represented in Figures 4 and 5, respectively. The three different plant parts are indicated for each species, allowing for a clear comparison of the profiles. The generated molecular formulas were searched in a specific library generated in‐house, the Bulbine library, using 98 compounds. The qualitative data related to the Bulbine species provided mass fragmentation information necessary for the putative identification of 73 compounds (Table 4). The compounds were classified into five different classes, namely: amino acids, anthocyanins, anthraquinones, along with phenolic compounds, and flavonoids. The occurrence of the compounds in the various species and the plant parts varied significantly, and the variation is described under each compound class.

FIGURE 4.

The LC‐QToF‐MS chromatograms (CPS intensity vs. retention time) of the five Bulbine species (B1‐B15) obtained in the positive ionisation mode.

FIGURE 5.

The LC‐QToF‐MS chromatograms (CPS intensity vs. retention time) of the five Bulbine species (B1‐B15) obtained in the negative ionisation mode.

TABLE 4.

Tentatively identified untargeted metabolomics of methanolic extracts of various Bulbine species using LC‐QToF‐MS.

#	t R (min) a	Molecular formula	m/z [M‐H]− b	Negative mode, fragment ions	m/z [M+H]+ c	Positive mode, fragment ions	Identified compound	Source
Amino acid
1	1.4	C9H11NO2	—	—	166.0865 (166.0863)*	120.0768 [M+H‐H2O‐CO]+	Phenylalanine	B1 to B15 (ALL)
Phenolic acids and Flavonoids
2	1.5	C16H18O9	353.0863, (353.0878)*	191.0556 [Quinic acid‐H]−, 179.0339 [Caffeic acid‐H]−, 173.0445 [Quinic acid‐H‐H2O]−, 135.0450 [Caffeic acid‐H‐CO2]−	355.1017, (355.1024)	163.0395 [Caffeic acid+H‐H2O]+	Chlorogenic acid [28]	B4, B5, B6
3	1.7							B4, B5, B6
4	1.8							B4, B5, B6
5	3.6	C31H34O17	677.1699, (677.1723)	515.1378 [M‐H‐Glc]−, 431.0962, 353.0860 [Chlorogenic acid‐H]−, 191.0548 [Quinic acid‐H]−, 179.0341 [Caffeic acid‐H]−, 135.0444 [Caffeic acid‐H‐CO2]−	679.1868 (679.1865)	517.1393 [M+H‐Glc]+, 433.1152, 355.0893 [Chlorogenic acid+H]+, 163.0393 [Caffeic acid+H‐H2O]+	Isochlorogenic acid C 4′‐O‐glucoside [29]	B4, B5(tr), B6(tr)
6	4.08	C27H30O15	593.1509, (593.1512)	447.0889 [M‐H‐Rha]−, 285.0399 [M‐H‐Rha‐Glc]−	595.1666 (595.1657)	—	Kaempferol rutinoside [28]	B8(tr), B9, B14, B15
7	4.1	C21H20O11	447.0925, (447.0933)	285.0396 [M‐H‐Glc]−	449.1098 (449.1078)	287.0580 [M+H‐Glc]+	Kaempferol glucoside [28]	B4(tr), B6(tr), B7, B9(tr), B10(+++), B13, B14
8	4.9	C25H24O12	515.1186, (515.1194)	353.0868 [Chlorogenic acid‐H]−, 191.0557 [Quinic acid‐H]−, 179.0345 [Caffeic acid‐H]−, 135.0450 [Caffeic acid‐H‐CO2]−	517.1356 (517.1341)	355.1038 [M+H‐caffeoyl]+, 337.0996 [M+H‐caffeoyl‐H2O]+, 181.0593 [Caffeic acid+H]+, 163.0399 [Caffeic acid+H‐H2O]+	Dicaffeoylquinic acids (diCQA) [28]	B4, B5(tr), B6(tr)
9	5.8							B4, B5(tr), B6(tr)
10	5.6	C21H18O11	445.0770, (445.0776)	269.0452 [Apigenin‐H]−	447.0932 (447.0922)	271.0619 [Apigenin+H]+	Apigenin glucuronide [30]	B13
11	6.0	C29H30O17	649.1399, (649.1410)	563.1406 [M‐H‐C3H2O3]−, 605.1483 [M‐H‐CO2]−, 269.0442 [Apigenin‐H]−	651.1567 (651.1556)	565.1572[M+H‐C3H2O3]+, 519.1143 [M+H‐132]+, 433.1188 [M+H‐C3H2O3‐132]+, 271.0621 [Apigenin+H]+, 153.0217 (RDA)	Malonylapiin [31]	B4, B5(tr), B6(tr)
12	6.8							B4, B5(tr), B6(tr)
13	6.1	C21H18O12	461.0704, (461.0725)	285.0391 [Luteolin‐H]−	463.0888 (463.0871)	287.0574 [Luteolin+H]+	Luteolin glucuronide [28]	B10
14	6.9	C30H26O14	609.1241, (609.1250)	285.0394 [Kaempferol‐H]−	611.1405 (611.1395)	287.0572 [Kaempferol+H]+, 163.0393	Kaempferol‐caffeoylglucoside [28]	B4, B7, B10
15	7.2	C24H22O13	517.0963, (517.0988)	473.1059 [M‐H‐CO2]−, 431.0946 [M‐H‐C3H2O3]−, 269.0443 [Apigenin‐H]−	519.1144 (519.1133)	475.1252 [M+H‐CO2]+, 433.1126 [M+H‐C3H2O3]+, 271.0626 [Apigenin+H]+	Apigenin malonylglucoside [32]	B4
16	8.3	C15H10O6	285.0402, (285.0405)	151.0032 [C7H4O4‐H]−, 133.0289 [C8H6O2‐H]−	287.0570 (287.0550)	241.0514, 153.0193 [C7H4O4+H]+	Luteolin [28]	B4, B5(tr), B6, B7, B9, B10, B11(tr), B12(tr), B13, B14, B15(tr)
Anthocyanins
17	2.9	C27H31O15 +	—	—	595.1669 (595.1663)	449.1111 [M‐Rha]+, 287.0582 [M‐Rha‐Glc]+	Cyanidin‐3‐O‐rutinoside [28]	B1, B4
18	3.9	C26H29O15 +	—	—	581.1511 (581.1506)	449.1085 [M‐Xyl]+, 287.0578 [M‐Xyl‐Glc]+	Cyanidin‐3‐O‐sambubioside [28]	B4, B5(tr), B6(tr)
19	5.1	C26H29O14 +	—	—	565.1555 (565.1557)	433.1123 [M‐Xyl]+, 271.0622 [M‐Xyl‐Glc]+	Pelargonidin 3‐O‐sambubioside [28]	B4, B5, B6
20	5.4	C27H31O15 +	—	—	595.1667 (595.1663)	463.1232 [M‐Xyl]+, 433.1163 [M‐Glc]+, 301.0692 [M‐Xyl‐Glc]+, 273.0717[Peonidin‐CO]+, 153.0202 (RDA)	Peonidin 3‐O‐sambubioside [28]	B4, B5, B6
Anthraquinones
21	1.5	C19H24O9	395.1336, (395.1348)	351.1081 [M‐H‐C2H4O]−, 275.0923 [M‐H‐120]−, 247.0967 [M‐H‐120‐CO]−, 231.0670 [M‐H‐C2H4O‐120]−, 213.0562 [M‐H‐C2H4O‐H2O‐120]−, 203.0718 [M‐H‐C2H4O‐120‐CO]−	397.1518 (397.1493)	361.1305 [M+H‐2H2O]+, 299.0921[M+H‐2H2O‐C2H4O‐H2O]+, 233.0827 [M+H‐C2H4O‐120]−, 217.0883 [M+H‐Glc‐ H2O]−	8‐C‐glycosyl‐aloesol [33]	B3, B11, B12
22	2.0	C21H22O11	449.1104, (449.1089)	269.0454 [M‐H‐Glc‐H2O]−, 241.0495[M‐H‐Glc‐H2O‐CO]−, 253.0500[Chrysophanol‐H]−, 225.0549 [Chrysophanol‐H‐CO]−	—	—	Rumejaposide D, (6,10‐Dihydroxyaloin A) [33]	B5(tr), B6, B8(tr), B9(tr), B10(tr), B11, B12
23	5.38	C10H12O4	195.0661, (195.0663)	—	197.0805, (197.0808)	179.0697 [M+H‐H2O]+, 167.0325 [M+H‐30]+, 165.0445 [M+H‐33]+, 151.0749 [M+H‐33‐14]+, 149.0231 [M+H‐33‐16]+	Hydroxy‐dimethoxy‐acetophenone [34]	B1(tr), B2, B3, B4(tr), B5, B6, B7(tr), B8, B9
24	6.6	C30H32O18	679.1493, (679.1516)	299.0546 [M‐H‐C3H2O3‐132‐162]−	681.1659, (681.1661)	595.1695[M+H‐C3H2O3]+, 549.1278 [M+H‐132]+, 301.0724 [M+H‐C3H2O3‐132‐162]+	Chrysoeriol 7‐O‐(6''‐malonyl‐apiosyl‐glucoside) [35]	B4, B5(tr), B6(tr)
25	7.0							B4, B5(tr), B6(tr)
26	6.7	C16H10O7	313.0343, (313.0354)	269.0437 [M‐H‐CO2]−	—	—	Laccaic acid D isomers [34], (trihydroxy anthraquinone)	B1(tr), B5, B6, B7(tr), B8, B9(tr), B12, B14, B15
27	9.4							B1, B2, B3, B4, B5, B6, B7(tr), B8, B9, B10, B11, B12, B13, B14, B15
28	7.16	C22H20O11	459.0920, (459.0933)	253.0495 [Chrysophanol‐H]− or [M‐H‐Glc‐CO2]−	—	—	Rhamnalpinoside [33], (3‐Methyl‐3,6,8‐trihydroxy‐1‐methylanthraquinone‐2‐carboxylic acid‐6‐O‐ β‐D‐apiofuranoside)	B1(tr), B2, B3, B5, B6, B7(tr), B8, B9, B11(+++), B12(+++), B14(tr), B15(tr)
29	8.7	C23H22O12	489.1029, (489.1038)	325.0709 [M‐H‐CO2(44)‐120]−, 283.0603 [M‐H‐Glc‐CO2]−, 268.0375 [M‐H‐Glc‐CO2‐CH3(15)]−, 240.0444 [M‐H‐Glc‐CO2‐CH3‐CO]−, 212.0488 [M‐H‐Glc‐CO2‐CH3‐2CO]−	491.1179 (491.1184)	329.0654 [M+H‐Glc]+, 311.0538 [M+H‐Glc‐H2O]+	Bulbnatalonoside E [33]	B11, B12
30	9.5	C36H38O18	757.1984, (757.1985)	433.0912 [M‐H‐2Glc]−	759.2137 (759.2131)	435.1123 [M+H‐2Glc]+	Knipholone‐O‐β‐gentiobioside [33]	B1(tr), B2(tr), B6(tr), B7(tr), B8, B9, B10, B11, B12
31	10.3							B1, B2, B3, B6(tr), B7, B8, B9, B10, B11, B12
32	10.7							B1, B2, B3, B5(tr), B6, B7(tr), B8, B9, B10, B11(+++), B12, B13(tr), B14(tr)
33	11.1							B1, B2, B3, B5(tr), B6, B7(tr), B8, B9, B10, B11(+++), B12, B13(tr), B14(tr), B15(tr)
34	11.3							B1(tr), B2(tr), B3(tr), B11
35	9.7	C42H40O17	815.2193 (815.2193)	668.1522 [M‐2H‐Rha]−, 653.1633 [M‐H‐Glc]−, 506.0998 [M‐2H‐Rha‐Glc]−, 253.0480 [Chrysophanol‐H]−	839.2166, (839.2158), [M+Na]+	799.2245 [M+H‐H2O]+, 655.1810 [M+H‐Glc]+, 509.1243 [M+H‐Glc‐Rha]+	Bulbnatalonoside C [33]	B11(tr), B12(tr)
36	10.4	C15H10O5	269.0458, (269.0455)	—	271.0614, (271.0601)	153.0190, 145.0285, 119.0491	Aloe emodin [19]	B4, B5, B6, B7(tr), B10, B13
37	11.2	C42H40O17	815.2204, (815.2193)	668.1536 [M‐2H‐Rha]−, 653.1641 [M‐H‐Glc]−, 506.0993 [M‐2H‐Rha‐Glc]−, 253.0483 [Chrysophanol‐H]−	839.2148, (839.2158), [M+Na]+	799.2240 [M+H‐H2O]+, 655.1814 [M+H‐Glc]+, 509.1245 [M+H‐Glc‐Rha]+	Bulbnatalonoside C, atropisomer [33]	B11(tr), B12
38	12.1	C10H12O4	195.0661, (195.0663)	—	197.0815 (197.0808)	167.0723 [M+H‐CH2O]+	Acetosyringone	B1(tr), B2(tr), B14, B15
39	12.2	C42H40O19	847.2077, (847.2091)	523.1033 (aglycone)	—	—	Chrysalodin 10‐β‐D‐gentiobioside [33]	B6(tr), B13, B14, B15
40	12.4	C30H28O13	595.1447, (595.1447)	433.0924 [M‐H‐Glc]−, 253.0501 [Chrysophanol‐H]− or [M‐H‐Glc‐180]−	597.1618, (597.1603)	435.1073 [M+H‐Glc]+, 273.1125	Bulbineloneside A, (Knipholone ‐glucopyranoside), and its atropisomer [33]	B1, B9(tr), B10, B11, B12
41	12.8							B1(tr), B9(tr), B10, B11, B12
42	12.7	C29H26O16S	661.0859, (661.0869)	581.144 [M‐H‐SO3]−, 433.0921, 419.0749 [M‐H‐Glc]−, 299.0544	—	—	Demethylknipholone‐glucopyranoside‐O‐sulfate [33]	B1(tr), B2, B3, B8(tr), B9
43	12.8	C16H10O6	297.0405, (297.0405)	253.0505 [Chrysophanol‐H]− or [M‐H‐CO2]−	—	—	3,8‐dihydroxy‐1‐methylanthraquinone‐2‐carboxylic acid	B1, B2, B3, B4, B5, B6, B7, B8, B9, B10, B11, B12, B13(tr), B14, B15
44	13.2	C37H40O18	771.2129 (771.2142)	447.1070 [M‐H‐2Glc(gentiobioside)]−, 401.0751 [M‐H‐2Glc‐H2O‐CO]−	795.2099 (795.2107), [M+Na]+	755.2178 [M+H‐H2O]+, 449.1249 [M+H‐2Glc (gentiobioside)]+, 431.1134 [M+H‐2Glc‐H2O]+	Bulbnatalonoside D [33]	B3(tr), B6(tr), B7(tr), B8, B9, B11, B12, B13(tr), B14, B15
45	13.3							B11, B12, B14(tr), B15(tr)
46	13.8	C24H18O8	433.0924, (433.0929)	253.0485 [Chrysophanol‐H]−	—	—	Isoknipholone [33]	B14, B15
47	14.8	C29H26O13	581.1295, (581.1301)	419.0752 [M‐H‐Glc]‐, 291.0649	—	—	Demethylknipholone glucopyranoside [33]	B1, B2, B3, B8, B9
48	15.5	C15H14O4	—	—	259.0962, (259.0965)	244.0745 [M+H‐CH3]+, 241.0872 [M+H‐H2O]+, 229.0520 [M+H‐2CH3]+, 200.0848 [M+H‐OCH3‐CO]+, 173.0601 (lactone fragment), 169.0656 [M+H‐2OCH3‐CO]	4‐O‐Methyl eleutherol [29]	B7, B8, B9
49	16.7	C23H16O11S	499.0337, (499.0341)	419.0771 [M‐H‐SO3]−, 291.0648	—	—	Demethylknipholone 6’‐O‐sulfate [36]	B1, B2, B3, B8, B9(+++),
50	17.0	C24H18O9	449.0869, (449.0878)	—	451.1032, (451.1024)	405.0978 [M+H‐CH2O2]+, 283.0610 [C16H10O5+H]+	Gaboroquinone A/B [33]	B1(tr), B2(tr), B3(tr), B5(tr), B6(tr), B7, B8, B9, B10(tr), B11, B12
51	17.3							B1(tr), B2(tr), B3(tr), B7, B8, B9
52	17.9							B1, B2, B3(tr), B7(tr), B8(tr), B9(tr),
53	18.5							B1(tr), B2(tr), B3, B5(tr), B6, B7(tr), B8(tr), B9, B11, B12
54	18.2	C24H18O11S	513.0487, (513.0497)	433.0909 [M‐H‐SO3]−	—	—	Knipholone sulfate [36]	B1, B2, B3, B6(tr), B8, B9, B15(tr)
55	19.5	C23H16O8	419.0780, (419.0772)	—	421.0929, (421.0918)	403.0845 [M+H‐H2O]+, 385.0727 [M+H‐2H2O]+, 357.0779 [M+H‐3H2O]+, 267.0672	Demethylknipholone [22]	B1, B2, B3, B9,
56	20.2	C41H40O15	771.2282 (771.2294)	639.1906 [M‐H‐Xyl]−, 625.1694 [M‐H‐Rha]−, 475.1253 [M‐H‐Xyl‐Rha‐H2O]−, 457.1234 [M‐H‐Xyl‐Rha‐2H2O]−	773.2431 (773.2440)	641.2016 [M+H‐Xyl]+, 627.1789 [M+H‐Rha]+, 495.1468 [M+H‐Xyl‐Rha]+, 255.0610 [Chrysophanol+H]+, 240.0801 [Chrysophanol+H‐CH3]+	Bulbnatalonoside B [33]	B10(tr), B11(+++), B12(+++)
57	20.5	C36H30O12	653.1658 (653.1664)	506.0999 [M‐H‐C6H11O4]−, 489.0989 [M‐H‐Rha‐H2O]−, 253.0499 [Chrysophanol‐H]−	655.1804 (655.1810)	637.1672 [M+H‐H2O]+, 619.1590 [M+H‐2H2O]+, 509.1214 [M+H‐Rha]+, 491.1131 [M+H‐Rha‐H2O]+, 255.0652 [Chrysophanol+H]+	Dehydrobulbnatalonoside A isomers	B10(tr), B11, B12
58	21.4							B10(tr), B11, B12
59	21.4	C24H20O8	435.1079, (435.1085)	253.0492, 181.0494, 166.0281	437.1241, (437.1231)	419.1132 [M+H‐H2O]+, 163.0400	Isofoliosone/Foliosone [33]	B9, B10, B11, B12
60	24.8							B1, B2, B3, B8(tr), B9, B11(tr), B12(tr), B14(tr), B15
61	20.9	C36H32O12	655.1819, (655.1821)	509.1194 [M‐H‐Rha]−	657.1975, (657.1967)	233.1839 [M+H‐H2O]+, 511.1404 [M+H‐Rha]+, 255.0601 [Chrysophanol+H]+	Bulbnatalonoside A isomers [33]	B10(tr), B11, B12
62	21.5							B10(tr), B11, B12
63	22.7	C24H18O8	433.0926 (433.0929)	418.0701 [M‐H‐CH3]−, 402.0735 [M‐H‐CH2OH]−, 390.0762 [M‐H‐CH3CO]−, 265.0504 [M‐H‐acetylphloroglucinol]−, 253.0505 [Chrysophanol‐H]−	435.1077, (435.1074)	417.0995 [M+H‐H2O]+, 393.0980 [M+H‐42]+, 375.0880 [M+H‐42‐H2O]+, 359.0578 [M+H‐42‐H2O‐CH3]+, 267.0601 [M+H‐42‐2H2O‐CH3]+	Knipholone [33]/Bulbine‐knipholone [33]	B1, B2, B3(tr), B4(tr), B5, B6, B7(tr), B8, B9, B10, B11, B12, B13(tr), B14, B15
64	23.2							B1, B2(tr), B3, B7, B8(tr), B9(tr), B10(tr), B11(tr),
65	24.5	C30H20O9	523.1023 (523.1035)	505.0899 [M‐H‐H2O]−, 477.0946 [M‐H‐H2O‐CO]−, 253.0490 [C15H10O4‐H]−	525.1191, (525.1180)	507.1081 [M+H‐H2O]+, 489.0955 [M+H‐2H2O]+, 461.1047 [M+H‐2H2O‐CO]+	Chrysalodin [37]	B5(tr), B6, B13, B14, B15
66	25.5	C25H20O8	447.1075, (447.1085)	401.0646 [M‐H‐C2H6O]−	449.1242, (449.1231)	431.1150 [M+H‐H2O]+	Bulbnatalone [33]	B8(tr), B11(tr), B12(tr)
67	27.2	C25H20O8	447.1086 (447.1085)	253.0524 [Chrysophanol‐H]− or [M‐H‐(hydroxy‐dimethoxyacetophenone)]−	449.1226, (449.1231)	431.1150 [M+H‐H2O]+, 399.0883 [M+H‐H2O‐CH3OH]+, 373.0721 [M+H‐H2O‐CH3OH‐C2H2]+	Knipholone 6′‐methyl ether [33]	B1(tr), B2(tr), B3, B4(tr), B5, B6, B7(tr), B8, B9, B10(tr), B11(+++), B12(+++), B13(tr), B14, B15
68	28.8	C30H20O8	507.1072, (507.1085)	489.0983 [M‐H‐H2O]−, 418.1048, 253.0498 [Chrysophanol‐H]−	509.1220, (509.1231)	491.1073 [M+H‐H2O]+	Bichrysophanol [38]	B1(tr), B2(tr), B5(tr), B6, B9(tr), B11, B12, B13, B14, B15
69	29.4	C30H18O8	505.0923, (505.0929)	487.0790 [M‐H‐H2O]−	507.1082, (507.1074)	489.0989 [M+H‐H2O]+	Abyquinone A [39]	B8(tr), B15(tr)
70	29.44	C48H36O15	851.1949, (851.1946)	773.1795, 431.0748, 239.0693	853.2126, (853.2127)	295.0637	Joziknipholone A/B [40]	B1, B2, B6(tr), B8(tr), B9(tr), B10(tr), B11(tr), B12(tr)
71	29.6							B1(tr), B2(tr), B6, B8, B9, B10, B11, B12
72	29.9							B1(tr), B2, B6(tr), B8(tr), B9(tr), B10(tr), B11(tr), B12(tr)
73	30.1							B1(tr), B2(tr), B9(tr), B10(tr), B11(tr), B12(tr)

Note: ‘*’ indicates exact mass theoretical value; ‘tr’ indicates chromatographic intensity < 103 cps; ‘+’ indicates presence of compounds (intensity > 103 cps). B1 (B. abyssinica‐Leaf); B2 (B. abyssinica‐Rhizome); B3 (B. abyssinica‐Root); B4 (B. asphodeloides‐Leaf); B5 (B. asphodeloides‐Rhizome); B6 (B. asphodeloides‐Root); B7 (B. frutescens‐Leaf); B8 (B. frutescens‐Rhizome); B9 (B. frutescens‐Root); B10 (B. latifolia‐Leaf); B11 (B. latifolia‐Rhizome); B12 (B. latifolia‐Root); B13 (B. narcissifolia‐Leaf); B14 (B. narcissifolia‐Rhizome); B15 (B. narcissifolia‐Root).

^{a t} R (min) = retention time in UPLC.

^b m/z [M‐H]⁻ = molecular ion in negative‐ion mode mass spectrometry.

^c m/z [M+H]⁺ = molecular ion in positive‐ion mode mass spectrometry.

2.2.1. Amino acid (1)

One amino acid was identified in all the species and plant parts, with a molecular ion of m/z 166.0865 [M+H]⁺ along with the characteristic fragment ions such as m/z 120.0768 [M+H‐H₂O‐CO]⁻ and 103.0536 [M+H‐H₂O‐CO‐NH₃]⁺. Based on the relative fragment ion information, the amino acid was characterised as phenylalanine (1), and the corresponding retention time was observed at 1.4 min.

2.2.2. Phenolic Acids and Flavonoids (2–16)

Six phenolic acids and eight flavonoids were identified from various samples of Bulbine species. Chlorogenic acid isomers (2–4) were observed at retention times of 1.5, 1.7, and 1.8 min with molecular ion at m/z 353.0863 [M‐H]⁻ along with fragment ions at m/z 191.0556 [Quinic acid‐H]⁻, 179.0339 [Caffeic acid‐H]⁻, and 135.0450 [Caffeic acid‐H‐CO₂]⁻. In addition, at 3.6 min, isochlorogenic acid C 4‐O‐glucoside (5) was tentatively identified based on the fragment ions such as at m/z 515.1378, 431.0962, 353.0860, 191.0548, 179.0341, and 135.0444. Further, dicaffeoylquinic acids (8–9) were observed at 4.9 and 5.8 min, whose molecular ion and fragment ions are listed in Table 4. The tentatively identified phenolic acids are the characteristic compounds for B. asphodeloides (leaf, rhizome, and root).

Eight flavonoid compounds, the polyphenolic class, were identified based on their molecular features. Luteolin (16), a flavonoid aglycone, was identified in all the species and plant parts at 8.3 min based on the molecular ion at m/z 285.0402 [M‐H]⁻ and corresponding fragment ions at m/z 151.0032, and 133.0289, which were formed by retro‐Diels‐Alder cleavage at 1, 3 positions. Similarly, a few other flavonoids were tentatively identified, such as kaempferol rutinoside (6), kaempferol glucoside (7), apigenin glucuronide (10), malonylapiin (11‐12), luteolin glucuronide (13), kaempferol caffeoylglucoside (14), and apigenin malonylglucoside (15), which are shown in Table 4. In addition to phenolic acid compounds, malonylapiin (11–12) was also identified as a characteristic compound in B. asphodeloides. Among the identified flavonoids, compound 10 was identified only in the B. narcissifolia leaf extract. Similarly, compound 13 was identified only in B. latifolia leaf extract, and compound 15 was identified in B. asphodeloides leaf extract. In addition, kaempferol caffeoyl glucoside (14) was tentatively identified based on the molecular features from leaf extracts of B. asphodeloides, B. frutescens, and B. latifolia.

2.2.3. Anthocyanins (17–20)

Anthocyanins are the coloured pigment containing flavonoid compounds synthesised via the phenylpropanoid pathway. Here, we have identified and characterised four different anthocyanins containing cyanidin, pelargonidin, and peonidin aglycone. At 2.9 min, a chromatographic peak was identified as cyandin‐3‐O‐rutinoside (17) based on the molecular ion observed at m/z 595.1669 [M]⁺ along with m/z 449.1111 [M‐Rha]⁺, and 287.0582 [M‐Rha‐Glc]⁺ fragment ions, which are considered to be the characteristic fragment ions to determine the core aglycone moiety present in the anthocyanin (Table 4). Compound 17 was observed in leaf extracts of B. abyssinica and B. asphodeloides. Further, at 3.9 min, cyanidin‐3‐O‐sambubioside (18) was identified based on the similar type of fragment ions observed in compound 17. For compound 18, the fragment ions of m/z 449.1085 [M‐Xyl]⁺, 287.0578 [M‐Xyl‐Glc]⁺ were observed as listed in Table 4. Additionally, pelargonidin 3‐O‐sambubioside and peonidin‐3‐O‐sambubioside were identified in all plants of B. asphodeloides. Based on the LC‐QToF‐MS/MS data observed for various Bulbine species, anthocyanins were predominantly observed in B. asphodeloides species compared to other Bulbine species.

2.2.4. Anthraquinones (21–73)

In total, 53 anthraquinone compounds were identified across various Bulbine species based on the molecular ions and their corresponding characteristic fragment ions. Compound 21 shows a molecular ion peak at m/z 397.1515 [M+H]⁺ along with corresponding fragment ions such as m/z 361.1305 [M+H‐2H₂O]⁺, followed by the loss of ‐C₂H₄O group from the terminal methyl group and subsequent loss of ‐H₂O and ‐Glc moiety. The detailed loss of fragment ions observed is listed in Table 4. Based on the observed fragment ions, compound 21 was tentatively identified as 8‐glucosyl aloesol. Based on the qualitative analysis of tentatively identified anthraquinones across various Bulbine species, a few anthraquinone compounds (43, 63, 67, 27) were identified in all five different Bulbine species, including various plant parts (leaf, rhizome, and roots). Compound 43 shows a molecular ion peak at m/z 297.0405 [M‐H]^−, along with m/z 253.0505 [M‐H‐CO₂]⁻ fragment ion, indicating the presence of a ‐COOH group on the core moiety, which is chrysophanol. Based on the observation of characteristic fragment ion compound 43, identified as 3,8‐dihydroxy‐1‐methylanthraquinone‐2‐carboxylic acid. Similarly, compounds 27, 63, and 67 were identified based on their molecular ions, fragment ions, and other molecular features, which were listed in Table 4. Furthermore, compounds 31‐33 ,28 and 50 were observed in the majority of the Bulbine species. Considering their presence in various Bulbine species reported in this study, it is understandable that anthraquinones are the characteristic compounds for Bulbine species. Bichrysophanol (68) is another compound that has a molecular ion observed at m/z 507.1072 [M‐H]⁻, followed by the fragment ions such as m/z 489.0983 [M‐H‐H₂O]⁻, and 253.0498 [Chrysophanol‐H]⁻. Compound 68 was identified in B. abyssinica leaves and rhizomes, B. asphodeloides rhizomes and roots, B. latifolia rhizomes and roots, B. frutescens roots, and B. narcissifolia leaves, rhizomes, and roots. Although the distribution of anthraquinones is diversified among various species, it is observed that some of the compounds were present in more than one species, such as bulbine‐knipholone (64), identified in B. abyssinica (leaves, rhizomes, and roots), B. frutescens (leaves, rhizomes, and roots), and B. latifolia (leaves and rhizomes). Similarly, compound 23 was identified in all plant parts of B. abyssinica, B. asphodeloides, and B. frutescens. Gaboroquinones (50–53) are majorly characteristic compounds for B. abyssinica and B. frutescens among the five different Bulbine species. Similarly, sulfated anthraquinone compounds (42, 49, and 54) are characteristic of B. abyssinica and B. frutescens compared to other species. The similarity in the MS/MS fragments, such as loss of sugar molecules, presence of similar core moiety, and double bond equivalence (DBE), helps in the differentiation of the anthraquinones from other classes of compounds, such as phenolic acids, flavonoids, and anthocyanins. Detailed information about the observed molecular features of individual identified compounds is provided in Table 4 along with their chromatographic retention time. Molecular ion at m/z 679.1493 [M‐H]⁻ and fragment ion at m/z 299.0546 were observed at 6.6 and 7.0 min, identified as chrysoeriol‐malonyl‐apiosyl glucoside (24–25). Furthermore, Aloe‐emodin was tentatively identified at m/z 271. 0614 [M+H]⁺ along with the corresponding fragment ions at m/z 153.0190, 145.0285, and 119.0491 (positive ion mode) in B. asphodeloides (leaves, rhizomes, and roots) and the leaves of B. frutescens, B. latifolia, and B. narcissifolia. Compound 48, identified as 4‐O‐methyl eleutherol, was identified only in B. frutescens. Bulbnatalosides A‐C and E were characteristic compounds observed only in B. latifolia, especially in rhizomes and roots, compared to leaves. Molecular ion at m/z 523.1023 [M‐H]⁻ followed by fragment ions at m/z 505.0899 [M‐H‐H₂O]⁻, 477.0946 [M‐H‐H₂O‐CO]⁻, and 253.0490 [Chrysophanol‐H]⁻ were found to be characteristic ions and tentatively identified as chrysalodin (65).

2.3. Untargeted Chemotypic Variation in Bulbine Species

2.3.1. Variation Based on Leaf Extracts

The UPLC‐MS chromatographic profiles of the five Bulbine species were analysed using multivariate data analysis tools to reveal their chemical relationships. OPLS‐DA was performed to separate the chemical data related to species differentiation (predictive) from unrelated data (orthogonal). The scores scatter plot illustrates the clustering of the five Bulbine species, with B. abyssinica and B. asphodeloides appearing chemically distinct based on their placement in separate quadrants. In contrast, the remaining three species exhibit a closer chemical relationship, as indicated by their close positioning on the plot (Figure 6A). The modelled variation along the first and second predictive components (R2X[1] and R2X[2]) was 19.6% and 14.1%, respectively, which was responsible for the clustering pattern observed. The variables (Rt_m/z) which contribute significantly to this chemical variation combine high covariation (high model influence) with high correlation (high reliability) and appear at the far ends of the loadings plot (Figure 6B). Although three chemical markers were filtered for B. abyssinica (Rt_m/z 5.50_317.06, 6.46_403.08, and 11.39_421.09) (Figure 6B), only demethylknipholone (11.39_421.09) was tentatively identified, while the other two compounds are unknown (Figure S1A). The chemical markers for B. asphodeloides (blue) were identified as chlorogenic acid (Rt_m/z 33.5_355.10), and dicaffeoylquinic acid (Rt_m/z 5.36_517.13), and these can be observed on the chromatogram in Figure S1B. Although the Rt‐mz pair for B. frutescens was identified as 4.99_287.05 on the loadings plot, tentative identification of the compound was not possible (Figure S1C). The marker compound for B. latifolia was identified as luteolin (6.96_287.05) (Figure S1D), and kaempferol glucoside (5.00_499.10) for B. narcissifolia (Figure S1E).

FIGURE 6.

Orthogonal projections to latent structures discriminant analysis (OPLS‐DA) score plot computed from ESI positive data of the leaves, showing separation of the five species (A) and the corresponding loadings plot showing putative biomarkers (Rt_m/z) for each species (B).

2.3.2. Variation Based on Rhizome Extracts

The OPLS‐DA scores plot modelled using chemical data of the rhizomes is presented in Figure 7A. There is a clear separation of B. abyssinica and B. latifolia from the other three species along both R2X[1] and R2X[2], with a total variation of 30.3%. Three putative biomarkers associated with the B. abyssinica are displayed on the loadings plot as 9.42_403.08, 10.70_393.09 and 13.55_403.08 (Figure 7B, green); however, none of the compounds could be tentatively identified (Figure S2A). Bulbine latifolia rhizomes were associated with three chemical markers at 7.67_435.10; 417.09 (unknown), 8.70_431 (bulbnatalone) and 12.79_417.10 (knipholone) (Figure S2B). The B. asphodeloides, B. frutescens and B. narcissifolia cluster could not be separated on the scores plot, and therefore a single variable was observed on the loadings plot (13.28_319,19) (Figure 7B). This unknown marker was prominent in B. asphodeloides and B. frutescens chromatograms (Figure S2B,C).

FIGURE 7.

Orthogonal projections to latent structures discriminant analysis (OPLS‐DA) score plot computed from ESI positive data of the rhizomes, showing separation of the five species (A) and the corresponding loadings plot showing putative biomarkers (Rt_m/z) for each species (B).

2.3.3. Variation Based on Root Extracts

Chemical variation in Bulbine roots was also assessed using OPLS‐DA, and the modelled variation using 6PCs (4 + 2) was 67.9%, which showed clear separation of the five species (Figure 8A). The root chemistry showed clear distinction of the species; however, the low modelled variation showed that there is still significant chemical overlap among the species. The loadings plot revealed various markers that can be used to quantitatively differentiate these species (Figure 8B). Although three variables were identified (Rt_m/z 13.5114_461.12, 13.49_207.06, and 13.05_435.10) for B. abyssinica (green), only one was tentatively identified as bulbine‐knipholone (13.05_435.10), which is also indicated in the chromatogram (Figure S3A). Bulbine asphodeloides was more chemically similar to B. latifolia than the other three species, and both species shared a chemical marker at 14.88_449,12 (knipholone 6′‐methyl ether), and then an unknown compound at 8.03_207.06 was the differentiating chemical marker for B. asphodeloides (Figure S3B). The chemical marker identified for B. frutescens was 4‐O‐methyleleutherol (10.56_259.09; 226.06) and an unknown compound at 11.39_403.08 (Figure S3C). Two chemical markers were identified for B. latifolia, and these were knipholone 6’‐methyl ether (14.88_431.11) and bulbnatalone (8.71_449,12). Both compounds appeared as major peaks in the corresponding chromatogram (Figure S3D). The potential chemical markers for B. narcissifolia were identified as epigenin‐galacturonide (8.26_447.13) and an unknown compound at 12.25_609.17, as shown in Figure S3E.

FIGURE 8.

Orthogonal projections to latent structures discriminant analysis (OPLS‐DA) score plot computed from ESI positive data of the roots, showing separation of the five species (A) and the corresponding loadings plot showing putative biomarkers (Rt_m/z) for each species (B).

The chemometrics results show that the leaf chemistry displays moderate chemical divergence, where B. abyssinica and B. asphodeloides are most distinct, while B. frutescens, B. latifolia, and B. narcissifolia are chemically similar. In contrast, the rhizomes show B. abyssinica and B. latifolia to be chemically distinct, while the other three species show a closer association. The root chemistry enabled the greatest separation among species, but with partial overlap in knipholone‐type compounds, suggesting both species‐specific markers and shared anthraquinone biosynthesis. In summary, the leaf chemistry is moderately discriminative, rhizomes are more conserved among most Bulbine species, and the roots provide the most robust chemical differentiation.

3. Discussion

The combined application of HPTLC‐MS and LC‐QToF‐MS in the current study provides a comprehensive overview of the phytochemical composition of five Bulbine species. The HPTLC enabled visualisation of compound distribution across plant parts, while LC‐QToF‐MS offered high‐resolution tentative identification and structural elucidation of both known and unknown metabolites. Together, these techniques revealed complex phytochemical profiles that varied not only among species but also between aerial and underground parts. Compounds identified were largely consistent between both analytical approaches, thereby validating their presence and enhancing confidence in compound annotation.

The identification of diverse classes of compounds in the Bulbine species provides valuable insight into their ethnopharmacological relevance and supports the traditional uses of these plants. The occurrence of phenylalanine across all species and plant parts is also noteworthy, as this amino acid serves as an important precursor in the biosynthesis of phenolic compounds and other secondary metabolites with therapeutic significance. Anthraquinones/phenylanthraquinones were among the phytochemicals detected, and this group of compounds has been reported to possess significant antiplasmodial properties in previous studies on Bulbine and Kniphofia species [22, 40, 41, 42]. In the current study, B. abyssinica displayed several knipholone‐type compounds, including knipholone, bulbine‐knipholone, knipholone‐glucopyranoside, and knipholone 6′‐methyl ether, which were more abundant in the leaves compared to the underground parts. The other Bulbine species displayed abundant anthraquinones, largely in the roots and rhizomes. The current study also confirmed the presence of demethylknipholone in the leaves, rhizomes, and roots of B. abyssinica, a compound previously reported in the underground parts of B. frutescens and used as a discriminator between B. frutescens and B. natalensis [22, 33]. Similarly, gaboroquinone A/B, which was identified exclusively in the underground parts of B. frutescens, had also been previously reported to differentiate B. frutescens from B. natalensis based on the root chemistry [33]. The current study also reports the identification of gaboroquinone in the leaves, rhizomes, and roots of B. abyssinica, B. asphodeloides, B. latifolia, and B. frutescens. The rhizomes of B. latifolia contained the highest diversity of phenylanthraquinones, including bulbnatalonoside D, bulbnatalone, and knipholone 6′‐methyl ether. Earlier studies linked bulbnatalonoside D and bulbnatalone to B. natalensis underground parts [33, 43]. The detection of these compounds in the current study confirms the chemical similarity between B. natalensis and B. latifolia. Across all five species, compounds such as knipholone, knipholone 6′‐methyl ether, and 3,8‐dihydroxy‐1‐methylanthraquinone‐2‐carboxylic acid were consistently detected in both aerial and underground parts, which aligns with [38], who reported that Bulbine species produce high levels of phenylanthraquinones. These compounds, particularly knipholone‐type anthraquinones, may thus serve as reliable marker compounds for chemotaxonomic identification within the genus.

Bulbine asphodeloides was rich in phenolic and flavonoid compounds, with chlorogenic acid, luteolin, kaempferol, and apigenin derivatives dominating its profile. Notably, thirteen flavonoids were unique to B. asphodeloides, highlighting its distinct chemistry within the genus. The presence of pelargonidin‐3‐O‐sambubioside in B. asphodeloides further indicated an abundance of phenolic compounds. Phenolic acids such as chlorogenic and dicaffeoylquinic acids enhance fibroblast proliferation and collagen synthesis, promoting tissue regeneration. Chlorogenic acid and related caffeoylquinic acid derivatives are widely recognised for their anti‐oxidant, anti‐inflammatory, antimicrobial, and antidiabetic activities through mechanisms involving free radical scavenging, modulation of oxidative stress pathways, suppression of inflammatory mediators such as NF‐κB, and regulation of glucose metabolism [45, 46, 47]. Kaempferol derivatives exert anti‐inflammatory and antimicrobial effects that protect wounds from secondary infection [48, 49]. The dominance of these flavonoids and phenolic acids in B. asphodeloides aligns with its traditional use for skin disorders and wound care. This is consistent with recent bioassays, where leaf extracts of B. asphodeloides promoted wound closure in fibroblast scratch assays and improved fin regeneration in zebrafish larvae [50].

The detection of anthocyanins, including pelargonidin‐ and cyanidin‐based glycosides predominantly in B. asphodeloides, further extends the pharmacological profile of the genus. Anthocyanins are well known for their antioxidant and anti‐inflammatory activities, protecting tissues from oxidative stress and UV‐induced damage [51]. Their occurrence in B. asphodeloides correlates with its ethnomedicinal use for skin soothing and regeneration and may contribute to its wound‐healing efficacy.

Bulbine narcissifolia displayed a less complex profile overall. However, the anthraquinone isoknipholone was unique to its rhizomes and roots, while abyquinone A and chrysalodin were shared only with a few other species. Aloe‐emodin, a compound largely distributed among Asphodelaceae taxa, including Aloe species, was identified in the rhizomes of both B. asphodeloides and B. narcissifolia. This finding supports the close phylogenetic and chemical relationship among members of the Asphodelaceae family [19, 42].

Multivariate data analysis allowed for the variation between the five species to be investigated using three plant parts. Chemical profiles of the five important Bulbines have not been extensively researched. Additionally, research into chemical variation between the plant parts of the five species is very scarce. The current study identified five known compounds, including demethylknipholone (B. abyssinica), chlorogenic acid and dicaffeoylquinic acid (B. asphodeloides), luteolin (B. latifolia), and kaempferol glucoside (B. narcissifolia), as potential chemical markers for the leaves of four species; however, the only chemical marker confirmed for B. frutescens was unknown. While the highest number of chemical markers were confirmed for rhizomes, the majority of the compounds were unknown. The only known compounds confirmed as chemical markers were for B. asphodeloides (pelargonidin 3‐O‐sambubioside) and B. latifolia (bulbnatalone, knipholone, and knipholone 6’‐methyl ether). The chemometrics results of the rhizomes exposed the need for further research into this understudied plant part to ensure quality control. Although phytochemical research of Bulbine roots commenced decades ago, the various species were investigated in isolation; therefore, the chemical variation between the species has been unknown. The current study revealed bulbine‐knipholone as a chemical marker for B. abyssinica roots, while 4‐O‐methyleleutherol and knipholone were confirmed as chemical markers of B. frutescens roots. The chemical marker knipholone 6’‐methyl ether was common between B. asphodeloides and B. latifolia; the presence of bulbnatalone was noted as a chemical marker for the latter species. The known chemical marker for B. narcissifolia roots was apigenin‐galacturonide. Further investigation of Bulbine phytochemistry using multivariate analysis ensures that variation between the species is known, which enables a deeper understanding of the plants.

Key discriminant metabolites in the leaves included demethylknipholone in B. abyssinica, chlorogenic and dicaffeoylquinic acids in B. asphodeloides, and luteolin and kaempferol glucoside in B. latifolia and B. narcissifolia. Several detected Rt_m/z features in B. frutescens and B. abyssinica remained unidentified, underscoring the need for further compound elucidation. The rhizome dataset revealed stronger differentiation, with B. abyssinica and B. latifolia clearly separated from the remaining species along Pp1 and Pp2, capturing 30.3% of the total variation. Biomarkers such as bulbnatalone, exclusive to B. latifolia, support previous findings that certain phenylanthraquinones may be species‐specific [33, 39, 43]. In contrast, B. asphodeloides, B. frutescens, and B. narcissifolia could not be chemically distinguished and clustered together, showing only one shared variable. Analysis of root extracts showed the highest modelled variation (67.9%) across six components, confirming that the root chemistry provides more comprehensive species differentiation. Unique markers such as bulbine‐knipholone, knipholone 6’‐methyl ether, and 4‐O‐methyleleutherol were linked to B. abyssinica, B. latifolia, and B. frutescens, respectively, further supporting the findings that roots are chemically richer and taxonomically informative. Nevertheless, a number of unidentified peaks across all organs highlight the need for expanded spectral libraries and deeper compound annotation.

Although the reported biological activities of these compounds provide a relevant pharmacological context, any potential pharmacological mechanisms of the studied Bulbine extracts remain tentative and speculative until confirmed using biological evaluation of the studied extracts and compounds. Comparative analysis revealed that bioactive compounds such as knipholone, knipholone 6′‐methyl ether, and 3,8‐dihydroxy‐1‐methylanthraquinone‐2‐carboxylic acid were distributed across both leaves and roots of the studied species. This finding suggests that pharmacologically important metabolites are not restricted to the underground parts but are also abundant in leaves. Such chemical overlap between leaves and roots provides a basis for promoting leaf‐based harvesting, thereby reducing destructive root collection. Harvesting leaves allows plants to grow and regenerate after collection, while the harvesting of roots or rhizomes destroys the entire plant, which threatens wild‐harvested populations. This approach supports conservation, sustainable use of medicinal plants, and helps maintain biodiversity while still providing plant material for traditional use.

4. Conclusions

The comprehensive phytochemical profiling and chemometrics analysis of the five Bulbine species revealed interspecies and inter‐organ chemical diversity, with rhizomes and roots generally showing the highest number of detected compounds. Species‐specific phenylanthraquinones, knipholone‐type compounds, and flavonoids emerged as potential chemotaxonomic markers, underscoring the unique chemical fingerprints of B. abyssinica, B. latifolia, and B. asphodeloides in particular. The identification of demethylknipholone, bulbnatalone, and knipholone 6’‐methyl ether across various organs highlights their potential utility as diagnostic markers for quality control and authentication of bulbine‐based herbal products. Additionally, the multivariate analysis confirmed clear chemical variation, especially in roots, reinforcing the importance of targeted organ‐specific studies in phytochemical investigations. Despite the valuable insights gained, a significant number of compounds remain unidentified, particularly among the chemical markers associated with B. frutescens and within the rhizomes of several species. This indicates a pressing need for further compound elucidation through techniques such as nuclear magnetic resonance spectroscopy and expansion of the spectral libraries. Overall, this study enhances our understanding of the phytochemical landscape of Bulbine species and provides a foundational framework for future research in chemotaxonomy, pharmacological exploration, and the standardisation of bulbine‐derived therapeutic agents.

5. Experimental Section

5.1. Chemicals and Reagents

Analytical grade dichloromethane, formic acid and methanol were purchased from Thembane Chemicals (Johannesburg, South Africa). Ethyl acetate was purchased from Merck Millipore (Johannesburg, South Africa), methanol and acetonitrile (HPLC grade, Romil), polyethylene glycol, and diphenylboric acid‐β‐ethylamino ester were purchased from Sigma–Aldrich (Johannesburg, South Africa).

5.2. Sourcing of Plant Material and Extraction

A total of 76 whole plant samples, representing the five Bulbine species, were collected from 10 natural populations in the Eastern Cape Province of South Africa in January 2021. Samples were collected under plant collection permit number HO/RSH/13/2022, issued by the Eastern Cape Department of Economic Development, Environmental Affairs & Tourism (ECDE‐ DEAT). Species identification was confirmed by a botanist (R.R. Klopper), and the voucher specimens were prepared and stored in the Department of Pharmaceutical Sciences at the Tshwane University of Technology under reference numbers TUT Bulbine 1–5. The whole plants were separated into three parts, namely leaves, rhizomes and roots, then oven dried at 50°C (Labcon laboratory equipment (Pty) Ltd, South Africa). Subsequently, the leaves, rhizomes, and roots were finely ground into powder in a blade coffee grinder (Russell Hobbs, South Africa). In total, 228 powder samples comprising leaves, rhizomes and roots were placed in 125 mL plastic jars and stored. Approximately 1 g of finely ground powder from each plant part was extracted with methanol (10 mL), using ultra‐sonication‐assisted extraction at 30°C for 30 min. Following each extraction, the supernatant was filtered through Munktell filter paper (Ahlstrom, Finland). The extraction process was repeated twice on the residue to ensure maximum extraction of phytochemicals. The filtrates from the three extractions were pooled into a conical flask, and the solvent was evaporated in a fume hood over a period of two days to yield the dry crude extracts. The crude extracts were then stored in the fridge to maintain their integrity for subsequent analysis.

5.3. Chromatographic Analysis

5.3.1. High‐Performance Thin‐Layer Chromatography Analysis

High‐performance thin‐layer chromatography (HPTLC) was performed on a semi‐automated HPTLC system (CAMAG, Muttenz, Switzerland) using the method of [52]. The crude extracts were dissolved in methanol to obtain a concentration of 100 mg/mL and applied to HPTLC silica gel 60 F254 plates (20 × 10 cm; Merck Millipore, Johannesburg, South Africa) as 10 mm bands at the bottom of the plate. Separation was achieved using ethyl acetate: water: acetic acid: formic acid (8:1:1:1 v/v/v/v) as the mobile phase. The plates were developed in a pre‐saturated developing chamber for 20 min at 23 ± 2°C and 33% relative humidity, at a distance of 70 mm from the lower edge. The plates were derivatised using p‐anisaldehyde reagent (methanol, acetic acid, sulphuric acid, anisaldehyde (8.5: 1: 0.5: 0.05 v/v/v/v) and band intensities were increased by heating at 100°C for 3 min. The HPTLC plates were visualised under 254 nm and 366 nm radiation, and images captured using winCATS software.

Compound identification was achieved using UPLC‐MS analysis. The HPTLC bands were scraped off the silica gel plate with a spatula and extracted with 1 mL HPLC‐grade methanol prior to filtering through 0.22 µm syringe filters and subjected to UPLC‐MS analysis to obtain the chemical profile of each band. Mass spectrometry (Xevo G3QToF mass spectrometer, Waters, USA) analysis was performed in the positive ion electron spray mode (ESI+). Nitrogen was used as the desolvation gas at a flow rate of 500 L/h, while maintaining a desolvation temperature of 350°C. The source temperature was set to 100°C, and the capillary and cone voltages were adjusted to 3500 and 10 V, respectively. Data collected over the range m/z 100 to 1000 were centroided during acquisition using independent reference lock‐mass ions via the LockSpray interface to ensure mass accuracy and reproducibility. Tentative identification of the compounds was carried out using spectral fingerprint match to literature and databases.

5.3.2. Liquid Chromatography Quadrupole Time‐of‐Flight Mass Spectrometry for Phytochemical Profiling

Fifteen samples representing the three plant parts (leaves, rhizomes, and roots) of the five Bulbine species (B. abyssinica, B. asphodeloides, B. frutescens, B. latifolia, and B. narcissifolia) were randomly selected. About 5–10 mg of each sample (dry crude) was sonicated in 1 mL of methanol for 5 min followed by centrifugation for 10 min at 5000 rpm. The supernatant was filtered and analysed on a liquid chromatography quadrupole time‐of‐flight mass spectrometry (LC‐QToF‐MS) using the method by [33]. Briefly, the liquid chromatographic system (Agilent 1290 series) coupled to 6530 Agilent Accurate‐Mass Q‐ToF LC/MS with an ESI interface was used. Separation was achieved on an Agilent Poroshell 120 EC‐C18 column, and the mobile phase consisted of water with 0.1% formic acid (solvent A) and acetonitrile with 0.1% formic acid (solvent B) at a flow rate of 0.25 mL/min, with gradient elution of 15% B to 55% B in 25 min and in 5 min to 100% B. All data analysis was processed with MassHunter Qualitative Analysis software Ver. B.07.00 in the mass range of 100–1700 Da for both the raw MS and MS/MS data analysed in positive and negative ionisation modes. Analysis included molecular feature extraction, background subtraction, data filtering, and molecular formula estimation using the accurate mass of all isotopes, their relative abundances, and all detected adducts with their measured isotopes. To perform subtraction of molecular features (MFs) originating from the background, analysis of a blank sample (methanol) was carried out under identical instrument settings and background MFs were removed.

5.3.3. Ultra‐Performance Liquid Chromatography System Coupled With Mass Spectrometry Analysis for Chemometrics

A Waters Acquity Ultra Performance Liquid Chromatographic system with photodiode array (PDA) detector combined with Xevo G3‐QToF mass spectrometer (Waters, USA) was used to perform UPLC‐MS analysis using the method by [52]. Crude extracts were reconstituted in HPLC‐grade methanol to achieve a concentration of 5 mg/mL. Each extract was injected three times into the UPLC‐MS system. The UPLC conditions were optimised to provide superior separation on a Cortecs UPLC C18 column (150 mm × 2.1 mm, i.d., 1.6 µm particle size, Waters). The mobile phase consisted of 0.1% formic acid (Solvent A) and acetonitrile (Solvent B) at a flow rate of 0.3 mL/min. The gradient elution started at 95.0% A: 5.0% B and changed to 15.0% A: 85.0% B over 16 min, followed by a quick transition to 5.0% A: 95.0% B in 0.50 min, and finally returning to the initial condition of 95.0% A: 5.0% B in 0.10 min. The total runtime for each analysis was 16.60 min. The extracts were injected into the mobile phase using a 2 µL full loop injection. Data collection and processing were carried out using the chromatographic software MassLynx V4.1, with the PDA detector spectral range set to 200–400 nm. Mass spectrometry was performed in the ESI+ mode. Nitrogen was used as the desolvation gas at a flow rate of 500 L/h, while maintaining a desolvation temperature of 350°C. The source temperature was set to 100°C, and the capillary, sampling cone, and extraction cone voltages were adjusted to 3.5, 10, and 4 kV, respectively. Data were collected over the mass range of m/z 80–1000 [29].

5.4. Chemometric Data Analysis

The UPLC‐MS data were pre‐processed in MarkerLynx V4.1 (Waters Corporation, USA) before conducting chemometrics analyses. The preprocessing parameters were optimised to correct for baseline, minimise noise, and select peaks to ensure accurate spectral alignment of mass fragments across all samples. The resulting dataset, containing retention time and mass‐to‐charge (Rt_m/z) pairs with their corresponding peak intensities, was exported as a CSV file for further analysis in SIMCA‐P + 14.0 (Umetrics AB, Umeå, Sweden). To identify natural groupings and variability within the dataset, unsupervised principal component analysis (PCA) was performed. Score plots were examined to observe outliers, clusters, and groups, while loading plots revealed compounds contributing to the observed clustering patterns. Subsequently, OPLS‐DA was applied to assess the chemical differences among the five Bulbine species. The resulting score plots displayed class clustering based on similarities in chemical composition. Potential marker compounds responsible for the observed classes were identified from the corresponding loading plots.

Author Contributions

Ramoagi T. Segone: investigation, writing – original draft, writing – review and editing, data curation, formal analysis, methodology, conceptualisation. Maxleene Sandasi: writing – review and editing, methodology, conceptualisation, supervision, administration. Efficient Ncube: writing – review and editing, formal analysis, methodology. Weiyang Chen: writing – review and editing, conceptualisation. Bharathi Avula: LC‐QToF‐MS analysis and data generation. Kumar Katragunta: LC‐QToF‐MS analysis and data generation. Ikhlas A. Khan: LC‐QToF‐MS analysis and data generation. Alvaro M. Viljoen: writing – review and editing, supervision, project administration, resources, funding acquisition. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting Information

Supporting information for this article is available on the WWW under https://doi.org/10.1002/cbdv.71385.

Supporting information

Supporting File: cbdv71385‐sup‐0001‐SuppMat.pdf

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.