Preprints.org logo
Preprint
Article

This version is not peer-reviewed.

Phloroglucinol α-Pyrones of Helichrysum arenarium, Helichrysum stoechas and Achyrocline satureioides: Automated UHPLC–Orbitrap–MS Annotation

Submitted:

21 October 2026

Posted:

22 October 2026

You are already at the latest version

Abstract
Phloroglucinol α-pyrones (PG α-pyr) are acylphloroglucinol–α-pyrone conjugates characteristic of Helichrysum and related genera. Helichrysum arenarium, Helichrysum stoechas and Achyrocline satureioides, three medicinal species not yet examined for the class by high-resolution mass spectrometry, were analysed by UHPLC–Orbitrap–MS. The spectra were annotated by an automated R workflow that applies the fragmentation rules established for H. italicum to every MS2 spectrum and matches it against a dictionary of 463 expected compounds. The workflow annotated 201 PG α-pyr (178 in H. arenarium, 55 in H. stoechas, 32 in A. satureioides); with the 58 compounds published for H. italicum, the four species hold 215 PG α-pyr, five of which, arzanol among them, occur in all four. For the 44 compounds shared with H. italicum the automated annotation reproduced the manual one. H. arenarium is the richest species, with 98 compounds detected only there. H. italicum and H. arenarium contain 153 and 78 µg/g PG α-pyr, dominated by arzanol; H. stoechas (8.4 µg/g) and A. satureioides (3.2 µg/g) are each dominated by one compound, a dehydro-geranyl benzopyran and helipyrone A. Unsupervised clustering of the MS/MS spectra recovered the scaffold and the α-pyrone substituent. The class is common to the four species; the dominant subclass and the preferred substituents distinguish them.
Keywords: 
Helichrysum arenarium
;  
Helichrysum stoechas
;  
Achyrocline satureioides
;  
Helichrysum italicum
;  
phloroglucinol α-pyrones
;  
arzanol
;  
UHPLC–Orbitrap–MS
;  
automated annotation
;  
MS/MS fragmentation rules
Subject: 
Chemistry and Materials Science  -   Analytical Chemistry

1. Introduction

Helichrysum arenarium (L.) Moench, Helichrysum stoechas (L.) Moench and Achyrocline satureioides (Lam.) DC. (Asteraceae, Gnaphalieae) are medicinal plants traditionally used for the treatment of hepatobiliary, digestive and respiratory disorders [1,2,3,4,5]. Previous phytochemical investigations on the three species found the presence of flavonoids, chalcones, phenolic acids, terpenoids and pyrone derivatives [1,5,6,7,8,9]. Among these, phloroglucinol–α-pyrone conjugates constitute one of the most structurally distinctive and pharmacologically relevant metabolite classes [10].
Phloroglucinol α-pyrones (PG α-pyr) are derived from the coupling of an acylphloroglucinol unit with substituted α-pyrone rings via methylene bridges. Since the first reports in the 1970s [11,12], more than fifty structurally characterized PG α-pyrones have been isolated from eleven Helichrysum species [10] and from related genera such as Achyrocline [9,13]. Based on the structural features of the α-pyrone and phloroglucinol moieties, they are grouped into monopyrones, dipyrones, 3-non-alkylated, 3-prenyl, 3-geranyl, 2-prenyl and O-methylated PG α-pyr and into the cyclized benzofuran, benzopyran and spiroketal derivatives, as recently reviewed by us [10].
Of this class, arzanol has been the most extensively studied representative. It inhibits NF-κB signaling (IC50 ≈ 12 µM) [14], mPGES-1 and 5-LOX [15,16,17,18], as recently reviewed by us [19], and exhibits antioxidant [17] and antibacterial activity against drug-resistant Staphylococcus aureus [20]. Yields of up to 0.48% (w/w) have been reported from the flowers of H. stoechas, where it occurs with helipyrones A and B and 18,18-bis-desmethyl achyroclinopyrone C [3,21,22]. From H. arenarium, helipyrone B, arenol, homoarenol and arzanol have been reported [1,12,23]; in A. satureioides, 23-methyl-6-O-desmethylauricepyrone, the antibacterial principle of the plant, reaches 0.3% of the dried material and is accompanied by cyclized benzopyran and benzofuran derivatives and the dibenzofuran achyrofuran [9,24,25,26,27].
Preparative isolation followed by NMR elucidation [3,9,21,22,27] remains the reference method for complete structure elucidation, but it is material-intensive, and trace constituents are commonly lost during the multi-step chromatography required to reach a pure compound. Chromatographic profiling with DAD detection quantifies individual targets such as arzanol but conveys no structural information, because all members of the class share the same chromophore [23,28]. LC–MS has been applied to infusions, decoctions and solvent fractions of the three species [29,30,31,32,33,34], but these studies were directed at flavonoids and phenolic acids, and the pyrones were registered as a group or annotated without a class-specific rule set.
Despite the considerable structural diversity and pharmacological interest, to date, no systematic high-resolution mass spectrometry (HRMS) study has characterized the qualitative and quantitative distribution of PG α-pyrone derivatives in any of the three species.
In our previous study, 58 PG α-pyr and the monopyrone micropyrone were annotated in H. italicum by ultra-high-performance liquid chromatography coupled to Orbitrap high-resolution mass spectrometry (UHPLC–HRMS), and diagnostic fragment ions and neutral losses were established that assign the α-pyrone C6 substituent, the C1 acyl chain and the C3 side chain from a single MS2 spectrum and distinguish the cyclized scaffolds from their open-chain counterparts [35]. H. arenarium and H. stoechas were selected for their early and rich records of the class, and A. satureioides as a representative of the class outside Helichrysum.
Therefore, the purpose of the present study is to characterize the phloroglucinol α-pyrone derivatives of the three species using UHPLC–HRMS. By applying the annotation rules established for H. italicum [35] together with semi-quantitative assessment and multivariate evaluation of fragment-ion relationships, we sought to compare the composition of this metabolite class across H. arenarium, H. stoechas and A. satureioides and to relate it to that of H. italicum. The questions were (i) whether the fragmentation rules, derived from H. italicum, are applicable to other species; (ii) which compounds the four species share and which are confined to one of them; (iii) which of the compounds known from the genus occur in each species, and in what amount; and (iv) whether the unsupervised clustering of the MS/MS spectra reproduces the structural grouping that the rules impose.

2. Results and Discussion

The acetone extracts of A. satureioides, H. arenarium and H. stoechas were analysed by UHPLC–HRMS (Section 3.1, Section 3.2, Section 3.3 and Section 3.4), and the spectra were processed by an automated workflow written in the R programming language that applies the fragmentation rules established in our previous study of H. italicum [35] to every MS2 spectrum. We then pooled the compounds from the four species into one list (Section 2.5) and compared the species. The compounds, the features behind them with their MS2 spectra, the published H. italicum data, the structures, the features not retained and the dictionary of the class are listed in full in the supplementary material (Table S1–S7), and the MS2 spectrum and the structure of every compound are drawn in Supplementary Figure S1.
PG α-pyr consist of an acylphloroglucinol unit linked through a methylene bridge to a 4-hydroxy-5-methyl-2H-pyran-2-one that carries an alkyl substituent at C6 [10,35]; arzanol, the best-known member of the class and the main PG α-pyr compound of H. italicum, has a 6-ethyl-α-pyrone, an acetyl group and a prenyl side chain at C1 and C3, respectively, in the PG unit (Scheme 1). Within this common frame the compounds of the class vary generally at three positions: the alkyl group at C6 of the α-pyrone (methyl, ethyl, propyl); the acyl group at C1 of the PG unit (acetyl to octanoyl, or 3-phenylpropanoyl); and the side chain at C3 of the PG unit, which is absent in the 3-non-alkylated (3-NA) derivatives and is generally a prenyl or a geranyl chain, with or without a hydroxyl group. Cyclization of this side chain onto a neighboring phenolic hydroxyl converts the open-chain compound into a benzofuran (BF) or a benzopyran (BP) derivative. In spiroketals (SKs), the pyrone ring is replaced with a furan-3-one spiro-fused to a dihydrobenzofuran, and in the dipyrones (DPs), two α-pyrones are joined by the methylene bridge without a phloroglucinol unit. The five scaffolds — open-chain, benzofuran, benzopyran, spiroketal and dipyrone — are referred to as subclasses below, the open-chain PG α-pyr is further divided by its C3 side chain; Figure 1 shows representatives of each subclass. We recently reviewed the class’s structural diversity and its distribution in the genus [10], and we developed a comprehensive UHPLC–Orbitrap–MS analysis of H. italicum [35]. The combination of scaffold, C6 substituent, C3 side chain and C1 acyl group is called the substitution pattern of a compound (scaffold | C6 | C3 | C1).
Under negative-ion HCD PG α-pyr initially cleave at the methylene bridge (Scheme 1) [35]. Cleavage on the α-pyrone side retains the bridging carbon on the phloroglucinol unit (Fragment A) and releases the α-pyrone, as a neutral or as the anion [α-pyr−H]−, which loses CO2 to give [α-pyr−CO2−H]−; the masses of this α-pyrone pair (m/z 139.0401 and 95.0502 for a 6-methyl, 153.0557 and 109.0659 for a 6-ethyl α-pyrone, etc.) identify the C6 substituent. Cleavage on the phloroglucinol side yields the acylphloroglucinol anion without the bridging carbon (Fragment B), one 12C atom (12.000 Da) lighter than Fragment A and the base peak of most spectra; its mass is the sum of the phloroglucinol core, the C3 side chain and the C1 acyl group, and the side chain and the acyl group are assigned from it. The relative intensities of these ions distinguish the scaffolds [35]: Fragment A reaches 57–95% of the base peak in the open-chain compounds and at most 15% in the cyclised ones; the α-pyrone pair is weak in the open-chain compounds (6–16% and 6–20%), moderate in the BFs (16–51% and 15–33%) and dominant in the BPs (70–100% and 50–72%); the SKs give Fragment B without an α-pyrone pair; the DPs give little fragmentation information besides the α-pyrone pair, which dominates their spectra while [M−H]− is a minor ion. Each subclass produces in addition a few minor ions, typically 1–10% of the base peak, that confirm the C3 side chain — m/z 205.0870 and 193.0870 for the 3-prenyl derivatives, for instance — referred to as class ions below and listed per subclass in Section 2.2; the acyl group gives a low-intensity anion of its own (m/z 83.0138 for acetyl, 97.0295 for propanoyl, 111.0452 for butanoyl, etc.).

2.1. The Dictionary of Expected PG α-Pyrones

In our previous study of H. italicum [35] we established the principles of fragmentation for PG α-pyrone derivatives: which fragment ions identify the scaffold, the alkyl group at C6 of the α-pyrone, the acyl group at C1 and the side chain or ring at C3. In the present study, we wanted to apply these principles automatically, by a workflow built in the R programming language. For this purpose, we first generated all possible combinations of the structural units known from H. italicum — the five scaffolds, the C6 alkyl groups, the C1 acyl groups and the C3 side chains and rings — extended by the homologues with one more carbon atom, so that compounds not yet reported could also be recognised. The 463 combinations obtained in this way form the dictionary (Section 3.9; Table S6).
For each dictionary compound, the name, molecular formula, [M−H]−, and the masses of the ions its MS2 spectrum should contain are stored: Fragments A and B, the α-pyrone pair, the core ions, the acyl anion and the minor class ions. Forty-nine of the 463 compounds were found in H. italicum [35]. For every feature the script selects the dictionary compounds with the same [M−H]− (within 10 ppm) and checks which of their expected ions are present in the spectrum (within 15 ppm); the compound whose ions are found is assigned to the feature and is given with it in Table S2.

2.2. Features, Confidence Grades and the Compound Pool

The fragmentation rules of the previous study [35] were written in the R programming language and applied to every MS2 scan of the three samples (Section 3.6). The workflow is outlined in Scheme 2. The MS2 scans recorded across one chromatographic peak were merged into one consensus spectrum (Section 3.7), from which the scaffold, the C6 substituent, the C3 side chain and the C1 acyl group were determined. Such a peak, with its molecular ion, retention time, peak area, consensus spectrum and substitution pattern, is called a feature below. The assignment of a feature was graded high when, besides Fragments A and B and the α-pyrone pair, at least two class ions of the assigned subclass were present at ≥ 0.5% of the base peak, and medium otherwise. The principal class ions are m/z 137.0244 and 125.0244 for the 3-NA compounds, 205.0870 and 193.0870 for the 3-prenyl, 221.0819, 209.0819 and 191.0714 for the 3-(hydroxyprenyl), 223.0976 and 211.0976 for the 3-(hydroxydihydroprenyl), 261.1496 and 203.0714 for the 3-geranyl and 289.1445 and 277.1445 for the 3-(hydroxygeranyl) derivatives, 191.0714 and 147.0815 for the BFs, 193.0506 and 149.0608 for the BPs and 243.0663, 215.0350 and 229.0506 for the SKs (complete list in Table S6). Because the class ions are minor ions, a weak spectrum may not show them, and the grade is then medium (Section 2.10). A feature was retained when it met the five conditions of Section 3.10; the features that failed one of them are listed in Table S5.
The retained features of one species with the same substitution pattern and formula that elute within 0.12 min of each other were merged into one compound, and the compounds of the three species and the 58 PG α-pyr published for H. italicum [35] were matched across species by substitution pattern, formula and retention time (Section 3.13). The 215 compounds of the four species together are referred to below as the compound pool (Table 4). Each compound of the pool received an identifier made of the abbreviation of its subclass and a number in order of retention time: PG, open-chain phloroglucinol α-pyrones; BF, benzofurans; BP, benzopyrans; SK, spiroketals; DP, dipyrones. Arzanol, for instance, is PG18. All counts, tables and figures below refer to these compounds; their consensus MS2 spectra are listed in Table S1 and Table S2 and drawn in Supplementary Figure S1.

2.3. PG α-Pyrones Detected in the Three Species

The rules accepted MS2 scans for 742 features in the three samples, of which 323 met the five conditions: 206 in H. arenarium, 78 in H. stoechas and 39 in A. satureioides, 137 with high and 186 with medium confidence (Table 1). These features represent 201 distinct compounds: 178 detected in H. arenarium, 55 in H. stoechas and 32 in A. satureioides, 58 of them in more than one of the three. Together with the 58 compounds published for H. italicum [35], 44 of which were detected in at least one of the three species, the pool of this study holds 215 PG α-pyr (Figure 2). H. arenarium is thus by far the richest of the three samples, with three times the compounds of H. stoechas and more than five times those of A. satureioides, and richer than the H. italicum extract of the previous study, in which 58 PG α-pyr had been annotated manually under the same conditions [35].

2.4. Subclasses and Substitution Patterns

Table 2 gives the compounds of each species by subclass and Table 3 the substituents they carry. All five scaffolds occur in all four species, but in very different proportions. H. arenarium is the richest in every respect: all ten subclasses are represented, and its 3-(hydroxygeranyl) PG α-pyr (45 compounds) and BPs (44) are the two largest groups of the whole study. In H. italicum [35] the 3-prenyl PG α-pyr, the BPs, the 3-NA derivatives and the BFs are the main groups, and the hydroxygeranyl series that dominates H. arenarium is represented by three compounds only. H. stoechas is a BP species — 19 of its 55 compounds — with the 3-geranyl derivatives as its second group and no hydroxylated prenyl derivative at all. A. satureioides has 32 compounds spread over all subclasses except the hydroxygeranyl derivatives; the BFs, the BPs and the 3-prenyl PG α-pyr are its largest groups.
The substituents separate the species more sharply than the subclasses (Table 3). The 6-ethyl-α-pyrone prevails in H. arenarium (92% of the compounds with a determined C6 substituent) and in A. satureioides (27 of 28), whereas H. italicum carries the 6-methyl-α-pyrone almost as often as the 6-ethyl one (22 against 27 compounds) and H. stoechas lies in between (69% ethyl); the 6-propyl-α-pyrone is rare in all four. The C3 side chain is dominated by the hydroxylated and the cyclised geranyl chains in H. arenarium, by the prenyl chain and the cyclised prenyl ring in H. italicum and A. satureioides, and by the cyclised dehydro-geranyl ring and the geranyl chain in H. stoechas. The C1 acyl groups show the clearest differences between the species. In H. stoechas and H. italicum the acetyl group predominates. H. arenarium carries the whole homologous series from acetyl to octanoyl — acetyl, propanoyl, butanoyl and pentanoyl (37–50 compounds each), hexanoyl to octanoyl in a few — and 3-phenylpropanoyl in addition. In A. satureioides the C4 and C5 acyl groups, butanoyl and pentanoyl, predominate (25 of its 30 compounds with a determined acyl group); the PG α-pyr isolated from the species carry the branched 2-methylbutanoyl group, which the rules read as pentanoyl (Section 2.9). H. italicum thus varies the α-pyrone where H. arenarium varies the acyl group and hydroxylates the geranyl chain, and the two species that are poor in the class, H. stoechas and A. satureioides, each vary little at any position.

2.5. The PG α-Pyrones of the Four Species

Table 4 lists the 215 compounds of the pool with their substitution pattern, formula, mean [M−H]− and retention time, and for each species the retention time at which the compound was found, its content and the confidence of its assignment; the trivial names and the numbers of the previous paper [35] identify the compounds known from H. italicum. Of the 58 compounds published for H. italicum [35], 44 were detected in at least one of the three species — 40 in H. arenarium, 16 in H. stoechas and 10 in A. satureioides — at the published retention time and with the same substitution pattern.
We annotated the three species automatically using this study’s workflow, and we annotated H. italicum manually in our previous study [35], using the same fragmentation rules and UHPLC–HRMS method. For the 44 compounds detected both in H. italicum and in at least one of the three species, the automated and the manual annotation gave the same substitution pattern, and the retention times agree within 0.01 min; this agreement confirms the automated annotation. The contents of the three species were calculated with the calibration curve of this study and those of H. italicum with the calibration curve of the previous study [35]; both were obtained with the same external standard (THAP) and the same sample preparation (Section 3.12).
Table 4. The 215 PG α-pyr of the four species.
Table 4. The 215 PG α-pyr of the four species.
ID C6; C3; C1 Formula m/z RT A. satureioides H. arenarium H. italicum H. stoechas Name (no.)
PG1 ethyl; 3-(hydroxyprenyl); acetyl C22H25O8− 417.1560 4.50 – 4.50; <LOQM – – –
PG2 methyl; 3-NA; acetyl C16H15O7− 319.0824 5.13 – – 5.12; 0.59 5.14; 0.055H 3
PG3 methyl; 3-(hydroxydihydroprenyl); acetyl C21H25O8− 405.1555 5.44 – – 5.44; 0.096 – 25
PG4 ethyl; 3-NA; acetyl C17H17O7− 333.0981 6.78 – 6.77; 0.90H 6.77; 1.84 6.80; 0.074H 5
PG5 ethyl; 3-(hydroxydihydroprenyl); acetyl C22H27O8− 419.1712 7.21 – 7.20; 0.043H 7.22; 0.20 – 32
PG6 methyl; 3-(hydroxygeranyl); acetyl C26H31O8− 471.2024 7.71 – – 7.71; 0.080 – 48
PG7 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2187 8.35 – 8.35; <LOQM – – –
PG8 methyl; 3-prenyl; acetyl C21H23O7− 387.1450 8.50 – 8.53; 1.74H 8.46; 24.80 – 14
PG9 methyl; 3-(hydroxygeranyl); acetyl C26H31O8− 471.2026 8.62 – 8.63; <LOQM – 8.62; <LOQM –
PG10 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2186 8.99 – 8.99; <LOQM – – –
PG11 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2227 9.30 – 9.30; <LOQM – – –
PG12 ethyl; 3-NA; propanoyl C18H19O7− 347.1138 9.30 – 9.28; 0.67H – 9.31; <LOQH –
PG13 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2184 9.54 – 9.58; <LOQM 9.50; 0.15 – 53
PG14 ethyl; 3-(hydroxyprenyl); propanoyl C23H27O8− 431.1711 9.71 – 9.71; <LOQM – – –
PG15 methyl; 3-(hydroxyprenyl); acetyl C21H23O8− 403.1401 9.98 – 9.99; 0.077H 9.97; 0.79 – 24
PG16 ethyl; 3-(hydroxydihydroprenyl); propanoyl C23H29O8− 433.1870 10.05 – 10.05; <LOQM – – –
PG17 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2184 10.14 – 10.15; 0.041H – 10.14; <LOQH –
PG18 ethyl; 3-prenyl; acetyl C22H25O7− 401.1607 10.43 10.52; <LOQH 10.31; 15.38H 10.37; 43.20 10.53; <LOQM (2) arzanol (20)
PG19 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2186 10.53 – 10.54; 0.19H – 10.52; <LOQM –
PG20 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2308 10.88 – 10.88; <LOQM – – –
PG21 methyl; 3-NA; pentanoyl C19H21O7− 361.1295 10.93 – 10.94; <LOQH 10.93; 0.39 – 6
PG22 methyl; 3-prenyl; propanoyl C22H25O7− 401.1606 11.21 – 11.19; 0.42H 11.23; 1.20 – 21
PG23 ethyl; 3-NA; butanoyl C19H21O7− 361.1294 11.35 – 11.33; 0.85H 11.35; 0.19 11.36; <LOQM 7
PG24 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2346 11.53 – 11.53; <LOQM – – –
PG25 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2187 11.99 – 11.99; 0.059H – – –
PG26 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2343 12.11 – 12.11; <LOQM – – –
PG27 ethyl; 3-(hydroxyprenyl); acetyl C22H25O8− 417.1557 12.41 – 12.41; 1.01H 12.42; 1.76 – 29
PG28 propyl; 3-prenyl; acetyl C23H27O7− 415.1763 12.47 – 12.50; 0.20H 12.45; 0.64 – helitalone B (26)
PG29 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2339 12.69 – 12.69; <LOQM – – –
PG30 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2495 12.78 – 12.78; <LOQH – – –
PG31 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2213 12.90 – 12.90; 0.24H (2) – – –
PG32 methyl; 3-prenyl; butanoyl C23H27O7− 415.1762 13.13 – – 13.13; 1.52 – arenol B (28)
PG33 ethyl; 3-NA; pentanoyl C20H23O7− 375.1451 13.16 13.19; <LOQH 13.17; 0.75H 13.14; 1.12 13.15; <LOQM 10
PG34 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2338 13.23 – 13.23; 0.17H – – –
PG35 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2495 13.37 – 13.37; <LOQM – – –
PG36 ethyl; 3-prenyl; propanoyl C23H27O7− 415.1764 13.37 – 13.31; 6.01H 13.45; 3.84 13.35; <LOQM 27
PG37 ethyl; 3-NA; pentanoyl C20H23O7− 375.1452 13.55 – 13.55; 0.038H – – –
PG38 methyl; 3-NA; hexanoyl C20H23O7− 375.1449 13.60 – – 13.60; 0.46 – 9
PG39 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2465 13.68 – 13.68; 0.045M – – –
PG40 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2495 14.02 – 14.02; <LOQH – – –
PG41 ethyl; 3-(hydroxydihydroprenyl); pentanoyl C25H33O8− 461.2182 14.14 14.14; <LOQM – – – –
PG42 ethyl; 3-NA; 3-phenylpropanoyl C24H23O7− 423.1453 14.28 – 14.28; <LOQH – – –
PG43 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2496 14.51 – 14.51; 0.048M – – –
PG44 methyl; 3-NA; acetyl C16H15O7− 319.0825 14.94 – – – 14.94; <LOQH –
PG45 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2343 14.95 – 14.95; <LOQH – – –
PG46 methyl; 3-prenyl; pentanoyl C24H29O7− 429.1920 15.00 14.99; <LOQM – 15.00; 5.04 – arenol C (35)
PG47 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2651 15.06 – 15.06; <LOQM – – –
PG48 methyl; 3-NA; heptanoyl C21H25O7− 389.1606 15.10 – – 15.10; 0.20 – 17
PG49 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2493 15.15 – 15.15; 0.56H – – –
PG50 ethyl; 3-NA; hexanoyl C21H25O7− 389.1608 15.36 15.35; <LOQH 15.36; 0.047H 15.37; 0.80 – 15
PG51 ethyl; 3-prenyl; butanoyl C24H29O7− 429.1919 15.36 15.35; 0.11H (3) 15.30; 5.75H 15.40; 3.92 15.39; <LOQM 6-O-desmethylauricepyrone/3-prenyl norauricepyrone (36)
PG52 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2341 15.61 – 15.61; <LOQM – – –
PG53 ethyl; 3-(hydroxyprenyl); propanoyl C23H27O8− 431.1714 15.61 – 15.61; 0.15H – – –
PG54 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2342 15.78 – 15.78; 0.095H – – –
PG55 ethyl; 3-NA; hexanoyl C21H25O7− 389.1608 15.86 – 15.86; 0.13H 15.86; 0.32 – 16
PG56 methyl; 3-geranyl; acetyl C26H31O7− 455.2076 16.02 – – – 16.02; 0.036H –
PG57 methyl; 3-geranyl; acetyl C26H31O7− 455.2077 16.18 – 16.18; 0.91H 16.20; 10.40 16.17; 0.22H (2) 18,18-bis-desmethyl achyroclinopyrone C (42)
PG58 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2655 16.25 – 16.25; <LOQM – – –
PG59 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2497 16.99 – 16.99; 0.043H – – –
PG60 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2655 17.04 – 17.04; 0.094H (2) – – –
PG61 ethyl; 3-prenyl; pentanoyl C25H31O7− 443.2078 17.08 17.08; <LOQH – – – –
PG62 ethyl; 3-prenyl; pentanoyl C25H31O7− 443.2076 17.23 17.21; 0.37H (2) 17.17; 4.90H (2) 17.26; 5.92 17.27; <LOQH 23-methyl-6-O-desmethylauricepyrone (40)
PG63 methyl; 3-prenyl; hexanoyl C25H31O7− 443.2075 17.45 – – 17.45; 5.12 – 61
PG64 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2493 17.76 – 17.76; 0.42H (2) – – –
PG65 ethyl; 3-prenyl; pentanoyl C25H31O7− 443.2077 17.77 – 17.77; 0.11H – – –
PG66 ethyl; 3-(hydroxyprenyl); butanoyl C24H29O8− 445.1872 17.80 17.81; <LOQH 17.80; 0.23H – – –
PG67 methyl; 3-prenyl; hexanoyl C25H31O7− 443.2075 17.95 – – 17.95; 1.52 – 41
PG68 ethyl; 3-geranyl; acetyl C27H33O7− 469.2232 18.11 – 18.11; 6.79H – 18.11; 0.12H (2) –
PG69 ethyl; 3-geranyl; acetyl C27H33O7− 469.2232 18.33 18.34; <LOQM – 18.34; 14.40 18.31; 0.54H (2) 18,18-bis-desmethyl achyroclinopyrone A (47)
PG70 ethyl; 3-prenyl; 3-phenylpropanoyl C29H31O7− 491.2076 18.44 – 18.43; <LOQH 18.46; 0.14 – 55
PG71 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2654 18.88 – 18.88; <LOQM – – –
PG72 methyl; 3-geranyl; propanoyl C27H33O7− 469.2218 19.09 – 19.08; 0.32H – 19.09; <LOQM –
PG73 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2184 19.20 – 19.20; <LOQM – – –
PG74 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2184 19.38 – 19.38; <LOQM – – –
PG75 ethyl; 3-prenyl; hexanoyl C26H33O7− 457.2234 19.40 19.38; <LOQM 19.41; 0.13H – – –
PG76 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2653 19.51 – 19.51; <LOQM – – –
PG77 ethyl; 3-(hydroxyprenyl); pentanoyl C25H31O8− 459.2028 19.62 19.62; 0.078H 19.62; 0.12H – – –
PG78 ethyl; 3-prenyl; hexanoyl C26H33O7− 457.2233 19.70 – 19.69; 0.036H 19.71; 3.12 – 23-ethyl-6-O-desmethyl-auricepyrone (43)
PG79 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2652 19.72 – 19.72; 0.11M – – –
PG80 ethyl; 3-NA; acetyl C17H17O7− 333.0983 19.91 – 19.95; <LOQH – 19.88; <LOQH –
PG81 ethyl; 3-NA; acetyl C17H17O7− 333.0982 20.08 – 20.11; <LOQH – 20.04; <LOQH (2) –
PG82 ethyl; 3-prenyl; hexanoyl C26H33O7− 457.2233 20.13 – 20.11; 0.18H 20.15; 0.96 – 23-ethyl-6-O-desmethyl-auricepyrone (44)
PG83 propyl; 3-geranyl; acetyl C28H35O7− 483.2390 20.38 – 20.36; <LOQH 20.40; 0.11 – 8′-methyl-18,18-bis-desmethyl achyroclinopyrone A (50)
PG84 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2336 20.50 – 20.50; <LOQM – – –
PG85 ethyl; 3-(hydroxygeranyl); acetyl C27H33O8− 485.2182 20.57 – 20.55; 0.054H 20.59; 0.12 – 54
PG86 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2340 20.77 – 20.77; <LOQM – – –
PG87 ethyl; 3-geranyl; propanoyl C28H35O7− 483.2390 20.99 – 21.00; 4.96H – 20.98; <LOQM –
PG88 ethyl; 3-geranyl; propanoyl C28H35O7− 483.2392 21.25 – 21.24; 0.17H – 21.26; <LOQH (2) –
PG89 ethyl; 3-geranyl; propanoyl C28H35O7− 483.2387 21.40 – 21.40; 0.093H (2) – – –
PG90 ethyl; 3-geranyl; propanoyl C28H35O7− 483.2373 21.76 – 21.76; <LOQM – – –
PG91 ethyl; 3-prenyl; heptanoyl C27H35O7− 471.2346 21.83 – 21.83; <LOQM – – –
PG92 ethyl; 3-(hydroxygeranyl); propanoyl C28H35O8− 499.2340 22.28 – 22.28; <LOQM – – –
PG93 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2492 22.46 – 22.46; <LOQM – – –
PG94 ethyl; 3-geranyl; butanoyl C29H37O7− 497.2543 22.62 – 22.62; 2.39M – – –
PG95 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2484 22.64 – 22.64; <LOQM – – –
PG96 ethyl; 3-geranyl; butanoyl C29H37O7− 497.2544 23.01 – 23.01; 1.01M – – –
PG97 ethyl; 3-geranyl; butanoyl C29H37O7− 497.2549 23.25 23.29; <LOQM 23.25; 0.28H (2) – 23.21; <LOQM (2) –
PG98 ethyl; 3-geranyl; butanoyl C29H37O7− 497.2542 23.54 – 23.54; <LOQM – – –
PG99 ethyl; 3-geranyl; butanoyl C29H37O7− 497.2535 23.76 – 23.76; <LOQM – – –
PG100 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2497 23.87 – 23.87; 0.038M – – –
PG101 ethyl; 3-(hydroxygeranyl); butanoyl C29H37O8− 513.2497 24.17 – 24.17; 0.040M (2) – – –
PG102 ethyl; 3-geranyl; butanoyl C29H37O7− 497.2551 24.17 – 24.17; <LOQM – – –
PG103 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2654 24.33 – 24.33; <LOQM – – –
PG104 ethyl; 3-geranyl; pentanoyl C30H39O7− 511.2697 24.62 – 24.62; 1.80H (4) – – –
PG105 ethyl; 3-geranyl; pentanoyl C30H39O7− 511.2699 24.83 – 24.83; 0.71H (3) – – –
PG106 ethyl; 3-geranyl; pentanoyl C30H39O7− 511.2704 25.13 25.12; <LOQM – – 25.15; <LOQM –
PG107 ethyl; 3-NA; butanoyl C19H21O7− 361.1294 25.19 – 25.19; <LOQH – – –
PG108 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2656 25.49 – 25.49; <LOQM – – –
PG109 ethyl; 3-geranyl; pentanoyl C30H39O7− 511.2705 25.65 – 25.65; <LOQM – – –
PG110 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2657 25.69 – 25.69; <LOQM – – –
PG111 ethyl; 3-geranyl; 3-phenylpropanoyl C34H39O7− 559.2709 25.78 – 25.78; <LOQM – – –
PG112 ethyl; 3-(hydroxygeranyl); pentanoyl C30H39O8− 527.2654 26.37 – 26.37; <LOQM – – –
PG113 ethyl; 3-geranyl; hexanoyl C31H41O7− 525.2861 26.96 – 26.96; <LOQM – – –
BF1 methyl; cyclised prenyl; acetyl C21H21O7− 385.1293 13.90 – – 13.90; 0.53 – 11
BF2 ethyl; cyclised prenyl; acetyl C22H23O7− 399.1448 16.42 – 16.41; 0.53H 16.43; 0.88 – italipyrone (18)
BF3 methyl; cyclised prenyl; pentanoyl C24H27O7− 427.1762 16.76 – – 16.76; 0.035 – 33
BF4 methyl; cyclised prenyl; acetyl C21H21O7− 385.1293 17.00 – – 17.00; 0.62 – 12
BF5 ethyl; cyclised prenyl; butanoyl C24H27O7− 427.1764 17.58 17.58; <LOQM 17.58; 0.046M – – –
BF6 methyl; cyclised prenyl; pentanoyl C24H27O7− 427.1762 17.86 – – 17.86; 0.11 – 34
BF7 ethyl; cyclised prenyl; pentanoyl C25H29O7− 441.1920 18.98 – 18.97; 0.039M 18.98; 0.088 – 22-methyl-22-ethyl-italipyrone (38)
BF8 ethyl; cyclised prenyl; pentanoyl C25H29O7− 441.1919 18.98 – – 18.98; 0.17 – 22-methyl-22-ethyl-italipyrone (39)
BF9 ethyl; cyclised prenyl; propanoyl C23H25O7− 413.1608 19.41 – 19.41; 0.044H (2) – – –
BF10 ethyl; cyclised prenyl; pentanoyl C25H29O7− 441.1921 19.43 19.43; <LOQM 19.43; <LOQM – – –
BF11 ethyl; cyclised prenyl; acetyl C22H23O7− 399.1450 19.63 – 19.62; 0.63H 19.64; 1.12 – italipyrone (19)
BF12 ethyl; cyclised prenyl; pentanoyl C25H29O7− 441.1920 20.02 – 20.02; 0.053M – – –
BF13 ethyl; cyclised geranyl; acetyl C27H31O7− 467.2077 20.22 – – – 20.22; 0.067M –
BF14 ethyl; cyclised prenyl; 3-phenylpropanoyl C29H29O7− 489.1923 20.29 – 20.29; <LOQM – – –
BF15 ethyl; cyclised prenyl; 3-phenylpropanoyl C29H29O7− 489.1921 20.44 – 20.44; 0.23M – – –
BF16 ethyl; cyclised prenyl; butanoyl C24H27O7− 427.1764 21.19 21.19; <LOQM 21.19; <LOQM – – –
BF17 propyl; cyclised geranyl; acetyl C28H33O7− 481.2238 22.51 – – – 22.51; <LOQM –
BF18 ethyl; cyclised prenyl; propanoyl C23H25O7− 413.1610 22.57 22.62; <LOQM 22.51; <LOQH (3) – – –
BF19 ethyl; cyclised prenyl; pentanoyl C25H29O7− 441.1918 22.98 22.98; 0.087H (2) 22.98; <LOQM – – –
BF20 ethyl; cyclised geranyl; propanoyl C28H33O7− 481.2236 23.22 – 23.22; <LOQM – – –
BF21 ethyl; cyclised geranyl; acetyl C27H31O7− 467.2076 23.80 – 23.80; 0.052M – 23.79; <LOQM –
BF22 ethyl; cyclised prenyl; butanoyl C24H27O7− 427.1763 24.56 24.57; 0.075H 24.54; <LOQH (3) – – –
BF23 methyl; cyclised geranyl; acetyl C26H29O7− 453.1923 25.02 – – – 25.02; <LOQM –
BF24 ethyl; cyclised geranyl; pentanoyl C30H37O7− 509.2548 25.58 – 25.58; <LOQM – – –
BF25 ethyl; cyclised geranyl; propanoyl C28H33O7− 481.2236 26.14 – 26.14; <LOQM – – –
BF26 ethyl; cyclised prenyl; pentanoyl C25H29O7− 441.1920 26.22 26.22; <LOQH 26.21; <LOQH – – –
BF27 ethyl; cyclised geranyl; pentanoyl C30H37O7− 509.2548 26.23 – 26.27; <LOQM (2) – 26.19; <LOQM –
BF28 ethyl; cyclised geranyl; acetyl C27H31O7− 467.2077 26.81 – 26.81; <LOQM – – –
BF29 ethyl; cyclised geranyl; butanoyl C29H35O7− 495.2393 27.10 – 27.10; <LOQM – – –
BP1 methyl; cyclised prenyl; acetyl C21H23O8− 403.1400 4.84 – 4.83; 0.13H 4.85; 0.64 – 22
BP2 ethyl; cyclised prenyl; acetyl C22H25O8− 417.1556 6.31 – 6.32; 1.62H 6.30; 1.28 – plicatipyrone (30)
BP3 methyl; cyclised prenyl; acetyl C21H23O8− 403.1400 6.90 – 6.90; 0.053H 6.90; 0.42 – 23
BP4 ethyl; cyclised prenyl; propanoyl C23H27O8− 431.1713 8.69 – 8.69; 0.58M – – –
BP5 ethyl; cyclised prenyl; acetyl C22H25O8− 417.1557 8.86 – 8.86; 0.93H 8.87; 0.88 – plicatipyrone (31)
BP6 ethyl; cyclised prenyl; butanoyl C24H29O8− 445.1870 10.54 10.54; 0.037M 10.55; 0.43M – – –
BP7 ethyl; cyclised prenyl; propanoyl C23H27O8− 431.1714 11.46 – 11.46; 0.25M – – –
BP8 methyl; cyclised (hydroxy)geranyl; acetyl C26H31O8− 471.2024 11.56 – – 11.56; 0.10 – 49
BP9 methyl; cyclised (hydroxy)geranyl; acetyl C26H31O8− 471.2026 12.02 – – – 12.02; <LOQM –
BP10 ethyl; cyclised prenyl; pentanoyl C25H31O8− 459.2027 12.35 12.34; 0.13M 12.35; 0.35M 12.37; 0.10 – 45
BP11 ethyl; cyclised prenyl; pentanoyl C25H31O8− 459.2028 12.81 – 12.81; <LOQM – – –
BP12 methyl; cyclised dehydro-geranyl; acetyl C26H29O7− 453.1918 13.20 – 13.21; <LOQM – 13.19; <LOQM –
BP13 ethyl; cyclised prenyl; butanoyl C24H29O8− 445.1870 13.32 13.32; 0.034M 13.31; 0.24M – – –
BP14 ethyl; cyclised (hydroxy)geranyl; acetyl C27H33O8− 485.2174 13.64 – 13.58; <LOQH 13.72; 0.18 13.61; <LOQH 51
BP15 ethyl; cyclised (hydroxy)geranyl; acetyl C27H33O8− 485.2168 14.13 – 14.13; 0.80H – 14.13; <LOQH –
BP16 methyl; cyclised dehydro-geranyl; acetyl C26H29O7− 453.1916 14.15 – – – 14.15; <LOQM –
BP17 methyl; cyclised (hydroxy)geranyl; acetyl C26H31O8− 471.2028 14.36 – 14.41; <LOQH 14.30; 0.037 – 56
BP18 ethyl; cyclised prenyl; hexanoyl C26H33O8− 473.2183 14.64 – 14.64; <LOQM – – –
BP19 ethyl; cyclised prenyl; pentanoyl C25H31O8− 459.2030 15.11 15.12; 0.096M (2) 15.11; 0.16M – – –
BP20 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2076 15.19 – 15.19; 0.14M 15.18; 0.11 15.20; <LOQM (2) 46
BP21 ethyl; cyclised prenyl; pentanoyl C25H31O8− 459.2032 15.61 – 15.61; 0.038M – – –
BP22 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2075 16.20 – 16.21; <LOQM – 16.20; <LOQM –
BP23 ethyl; cyclised (hydroxy)geranyl; propanoyl C28H35O8− 499.2344 16.33 – 16.33; 0.052M – – –
BP24 ethyl; cyclised (hydroxy)geranyl; acetyl C27H33O8− 485.2184 16.42 – 16.51; 0.32H 16.38; 0.074 16.37; <LOQM 52
BP25 ethyl; cyclised (hydroxy)geranyl; propanoyl C28H35O8− 499.2341 16.84 – 16.84; 0.48M – – –
BP26 ethyl; cyclised (hydroxy)geranyl; butanoyl C29H37O8− 513.2493 18.24 – 18.24; <LOQM – – –
BP27 ethyl; cyclised dehydro-geranyl; propanoyl C28H33O7− 481.2232 18.31 – 18.31; 0.077M – 18.32; <LOQM –
BP28 methyl; cyclised dehydro-geranyl; acetyl C26H29O7− 453.1920 18.37 – – – 18.37; 1.31M (5) –
BP29 ethyl; cyclised (hydroxy)geranyl; butanoyl C29H37O8− 513.2495 18.61 – 18.61; 0.92M – – –
BP30 ethyl; cyclised (hydroxy)geranyl; propanoyl C28H35O8− 499.2343 18.84 – 18.84; 0.047M – – –
BP31 ethyl; cyclised (hydroxy)geranyl; propanoyl C28H35O8− 499.2340 19.05 – 19.05; 0.14M – – –
BP32 methyl; cyclised dehydro-geranyl; acetyl C26H29O7− 453.1954 19.89 – – – 19.89; <LOQM –
BP33 ethyl; cyclised (hydroxy)geranyl; pentanoyl C30H39O8− 527.2654 19.92 19.93; <LOQM 19.92; <LOQM – – –
BP34 ethyl; cyclised dehydro-geranyl; butanoyl C29H35O7− 495.2393 20.30 20.30; <LOQM 20.29; 0.16M – – –
BP35 ethyl; cyclised (hydroxy)geranyl; pentanoyl C30H39O8− 527.2651 20.31 – 20.31; 0.45M – – –
BP36 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2073 20.36 – – – 20.36; 0.20M (2) –
BP37 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2074 20.56 – 20.56; 0.30M – 20.57; 4.59M (4) –
BP38 ethyl; cyclised (hydroxy)geranyl; butanoyl C29H37O8− 513.2495 20.67 20.67; <LOQM 20.67; 0.32M (2) – – –
BP39 ethyl; cyclised dehydro-geranyl; butanoyl C29H35O7− 495.2391 20.76 – 20.76; <LOQM – – –
BP40 ethyl; cyclised dehydro-geranyl; butanoyl C29H35O7− 495.2391 20.94 – 20.94; <LOQM – – –
BP41 methyl; cyclised dehydro-geranyl; propanoyl C27H31O7− 467.2076 21.11 – – – 21.11; 0.093M –
BP42 methyl; cyclised dehydro-geranyl; acetyl C26H29O7− 453.1925 21.61 – – – 21.61; <LOQM –
BP43 ethyl; cyclised dehydro-geranyl; pentanoyl C30H37O7− 509.2547 22.10 – 22.10; <LOQM – – –
BP44 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2079 22.13 – – – 22.13; <LOQM –
BP45 ethyl; cyclised (hydroxy)geranyl; pentanoyl C30H39O8− 527.2653 22.29 – 22.29; 0.12M (3) – – –
BP46 ethyl; cyclised dehydro-geranyl; pentanoyl C30H37O7− 509.2545 22.33 – 22.33; <LOQM – – –
BP47 ethyl; cyclised dehydro-geranyl; propanoyl C28H33O7− 481.2233 23.20 – 23.20; <LOQM (2) – 23.21; 0.31M (5) –
BP48 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2079 23.62 – 23.62; <LOQM – – –
BP49 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2078 23.78 – 23.78; <LOQM – – –
BP50 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2079 24.37 – 24.37; <LOQM – – –
BP51 ethyl; cyclised dehydro-geranyl; butanoyl C29H35O7− 495.2393 24.80 – 24.79; <LOQM (2) – 24.80; 0.072M (3) –
BP52 ethyl; cyclised dehydro-geranyl; butanoyl C29H35O7− 495.2393 25.01 – – – 25.01; <LOQM –
BP53 ethyl; cyclised dehydro-geranyl; acetyl C27H31O7− 467.2076 25.81 – 25.81; <LOQM – – –
BP54 ethyl; cyclised dehydro-geranyl; butanoyl C29H35O7− 495.2395 25.94 – 25.94; <LOQM – – –
SK1 –; prenyl-derived; acetyl C20H21O6− 357.1347 9.21 – 9.21; 0.099H 9.21; 3.20 – helispiroketal A (59)
SK2 –; prenyl-derived; acetyl C20H21O6− 357.1344 9.57 – – 9.57; 0.61 – helispiroketal A (60)
SK3 –; prenyl-derived; acetyl C21H23O6− 371.1500 11.25 – 11.25; 1.13H 11.24; 5.44 – helispiroketal F (8)
SK4 –; prenyl-derived; acetyl C21H23O6− 371.1502 11.45 – 11.45; 0.20H – – –
SK5 –; prenyl-derived; propanoyl C22H25O6− 385.1659 13.61 – 13.62; 0.39H 13.61; 0.12 – 13
SK6 –; geranyl-derived; acetyl C26H31O6− 439.2129 13.75 – – – 13.75; <LOQM –
SK7 –; prenyl-derived; butanoyl C23H27O6− 399.1816 15.50 15.49; 0.039H (2) 15.50; 0.33H – – –
SK8 –; geranyl-derived; acetyl C25H29O6− 425.1974 16.37 – 16.36; 0.036H – 16.38; <LOQH –
SK9 –; prenyl-derived; heptanoyl C26H33O6− 441.2284 16.60 – 16.60; 0.58M – – –
SK10 –; prenyl-derived; pentanoyl C24H29O6− 413.1956 17.05 17.06; 0.078H 17.03; 0.077H – – –
SK11 –; geranyl-derived; acetyl C26H31O6− 439.2127 18.36 – 18.36; 0.61H 18.35; 0.58 18.38; 0.051H (2) 37
SK12 –; prenyl-derived; octanoyl C27H35O6− 455.2443 19.14 – 19.14; 0.37M – – –
SK13 –; geranyl-derived; propanoyl C27H33O6− 453.2285 20.43 – 20.43; 0.14H – – –
SK14 –; geranyl-derived; propanoyl C27H33O6− 453.2286 20.61 – 20.60; 0.23H – 20.62; <LOQH –
SK15 –; geranyl-derived; butanoyl C28H35O6− 467.2439 22.30 – 22.30; 0.27H (6) – – –
SK16 –; geranyl-derived; pentanoyl C29H37O6− 481.2596 23.66 – 23.66; 0.063H – – –
DP1 methyl + methyl; –; – C15H15O6− 291.0877 6.55 – – – 6.55; 0.066H –
DP2 methyl + ethyl; –; – C16H17O6− 305.1032 8.74 8.74; <LOQH – 8.74; 0.10 8.73; 0.12H helipyrone B (2)
DP3 ethyl + ethyl; –; – C17H19O6− 319.1186 11.23 11.22; 2.04H (2) 11.23; 0.84H 11.25; 0.45 11.24; 0.45H helipyrone A (4)
Identifier (subclass abbreviation and number in order of retention time; PG, open-chain phloroglucinol α-pyrones; BF, benzofurans; BP, benzopyrans; SK, spiroketals; DP, dipyrones), substitution pattern as α-pyrone C6 substituent; C3 side chain or ring (the SKs: prenyl- or geranyl-derived core); C1 acyl group (–, not applicable), formula of the deprotonated molecule, mean measured [M−H]− and mean retention time; for each species the retention time (min) and the content (µg/g dry weight (DW), 2′,4′,6′-trihydroxyacetophenone (THAP) equivalents, uncorrected for recovery; <LOQ, below the limit of quantification (LOQ), 0.033 µg/g) with the confidence grade of the assignment (H high, M medium) and, in parentheses, the number of features merged into the compound where it is more than one; –, not detected. The column of H. italicum gives the published retention time and content [35]; last column, trivial name and compound number in that paper (bold).

2.6. Compounds Shared by the Species and Compounds Confined to One Species

Figure 3 shows which compounds the four species share. Five compounds occur in all four (Table 5): arzanol (PG18), its butanoyl and pentanoyl homologues 6-O-desmethylauricepyrone (PG51) and 23-methyl-6-O-desmethylauricepyrone (PG62), the 3-NA pentanoyl derivative PG33 and the DP helipyrone A (DP3). Thirteen compounds occur in three species, among them the 3-NA acetyl, butanoyl and hexanoyl derivatives PG4, PG23 and PG50, the propanoyl homologue of arzanol PG36, the two 18,18-bis-desmethyl achyroclinopyrones (PG57 and PG69), helipyrone B (DP2) and four BPs; 67 compounds occur in two species and 130 in one. The compounds common to all four are the simplest members of the class: the 6-ethyl-α-pyrone with an acetyl, butanoyl or pentanoyl group on a 3-prenyl or unsubstituted phloroglucinol, and the ethyl dipyrone.

2.6.1. Species-Specific Compounds and Their Chemotaxonomic Reading

The compounds detected in one species only — the species-specific compounds — are those that may carry chemotaxonomic information. H. arenarium has 98, more than half of its 178 compounds, and they are of a few kinds: 40 are 3-(hydroxygeranyl) PG α-pyr (only five of the 45 hydroxygeranyl compounds of the species were detected elsewhere), 22 are BPs and 9 BFs, mostly with a propanoyl, butanoyl or pentanoyl group, 12 are 3-geranyl derivatives and 6 are SKs, among them the heptanoyl and octanoyl members of the series. Forty of the 98 were quantifiable and together make up a sixth of the PG α-pyr content of the species (Section 2.9). H. arenarium is thus the species that hydroxylates the geranyl chain and extends the acyl series, and it does so in every subclass. These compounds were detected only in H. arenarium. H. arenarium gave the richest extract and by far the largest number of MS2 spectra with a PG α-pyr fragmentation pattern: 480 features, against 160 in H. stoechas and 102 in A. satureioides (Table 1). We therefore searched the full-scan (MS1) data of the other two samples for the [M−H]− of each of the 98 compounds at its retention time (Section 3.13; Table S7). Seventy-six of the 98 were found in A. satureioides or H. stoechas, typically at 2% of their intensity in H. arenarium; 21 were found in neither sample.
Of the 58 compounds published for H. italicum [35], 14 were detected in none of the three species, and twelve of them carry a 6-methyl-α-pyrone where the other species carry the 6-ethyl compound: the 6-methyl BFs with a cyclised prenyl ring (BF1, BF3, BF4, BF6), the 6-methyl 3-prenyl butanoyl and hexanoyl derivatives (arenol B, PG32; PG63, PG67), the 6-methyl 3-NA hexanoyl and heptanoyl derivatives (PG38, PG48) and three 6-methyl compounds with a hydroxylated or cyclised side chain (PG3, PG6, BP8). H. stoechas has 16 species-specific compounds: nine BPs with a cyclised dehydro-geranyl or (hydroxy)geranyl ring (six of them with a 6-methyl-α-pyrone), the 6-methyl/6′-methyl DP (DP1) and three BFs with a cyclised geranyl ring. The distinctiveness of H. stoechas lies less in these compounds than in the abundance of the dehydro-geranyl BPs (Section 2.9). A. satureioides has two species-specific compounds, both below the LOQ: a 3-(hydroxydihydroprenyl) pentanoyl derivative (PG41), and a minor 3-prenyl pentanoyl compound (PG61). A. satureioides shares 26 of its 32 compounds with H. arenarium.
The pairwise overlap of the compound lists follows the same pattern: H. arenarium and H. italicum share 40 compounds (Jaccard index 0.20), H. arenarium and H. stoechas 35 (0.18), H. italicum and H. stoechas 16 (0.16), A. satureioides and H. arenarium 26 (0.14), A. satureioides and H. italicum 10 (0.12), A. satureioides and H. stoechas 9 (0.12). H. arenarium is the species that shares most with every other, and A. satureioides, the only species outside Helichrysum, the one that shares least; the indices are low throughout because most compounds of H. arenarium are species-specific. Further accessions are needed to determine whether these differences reflect species characters or the single sample of each species analyzed here.

2.7. Proposed Structures

The 215 compounds carry 91 distinct structures, each drawn once in Supplementary Figure S2 and listed in Table A1: 46 open-chain PG α-pyr, 13 BFs, 17 BPs, 12 SKs and 3 DPs. One structure is drawn per substitution pattern, because the MS2 spectrum does not distinguish the isomers of a pattern: a butanoyl from a 2-methylpropanoyl group, the position of the ring closure of the BFs and BPs, or the position of the hydroxyl group in the hydroxylated side chains (drawn at C8, as reported for the class [35]). The alternatives are listed with each structure in Table S4. The structures correspond to Level 2b of the Schymanski scheme [36] (Section 3.10).

2.8. MS/MS-Based Clustering of the Compounds of Each Species

As in the previous study [35], the consensus MS2 spectra of the retained features were converted into two intensity matrices, one of fragment ions (FI) and one of neutral losses (NL), and clustered without using the assigned substitution patterns. The clusters were then compared with the patterns by the adjusted Rand index (ARI) and by the cluster purity (Section 3.14 and Section 3.15). The analysis was run on each of the three species and on the three together; the published values for H. italicum [35] are given for comparison (Table 6).
For the three species together, the fragment-ion clusters follow the scaffold (ARI 0.45, purity 0.76) and the neutral-loss clusters the C6 substituent of the α-pyrone (ARI 0.87, purity 0.97). Within one species the clusters follow the C3 side chain and the full substitution pattern, with a purity against the scaffold of 0.84–0.89. Neither matrix recovers the acyl group (ARI ≤ 0.12), whose acyl anion is a minor ion and whose main effect is a mass shift of Fragments A and B, and neither recovers the species (ARI 0.06): the spectra of a substitution pattern are the same in the different species. These results agree with those of H. italicum [35]: the neutral losses carry the information on the α-pyrone and the scaffold, the fragment ions that on the phloroglucinol side chain.
The heatmaps of Figure 4 show these effects within one subclass. The 13 3-prenyl PG α-pyr of H. arenarium group by their acyl group, through the Fragment B/A pairs that move by 14 Da from acetyl (m/z 235.0976/247.0976) to hexanoyl (291.1602/303.1602), and the 6-methyl (PG8, PG22) and 6-propyl (PG28) compounds stand apart from the ten 6-ethyl compounds by their α-pyrone pair. The 19 BPs of H. stoechas group by the α-pyrone pair of the 6-methyl and 6-ethyl compounds and by the Fragment B of the acetyl, propanoyl and butanoyl derivatives of the cyclised dehydro-geranyl ring (m/z 301.1445, 315.1602 and 329.1758).

2.9. Semi-Quantitative Distribution of the PG α-Pyrones in the Four Species and Comparison with Literature Data

The contents of the compounds were obtained from the areas of their MS1 peaks against an external calibration with 2′,4′,6′-trihydroxyacetophenone (THAP) and are therefore THAP equivalents (Section 3.12); those of H. italicum are the published values of our previous study [35], obtained in the same way. Table 7 and Figure 5 summarise them by species and subclass; the content of every compound is in Table 4. The values are relative, since all compounds were quantified against one standard; the contents in % dry weight, corrected for recovery, are given for the comparison with the isolation yields of the literature. The total PG α-pyr content of a species is the sum of the contents of its compounds above the LOQ, and the percentages given below are shares of this total. H. italicum holds 152.9 µg/g (0.0191% DW; all 58 compounds quantified in the previous study [35]), H. arenarium 77.6 µg/g (0.0097% DW; 99 of its 178 compounds above the LOQ), H. stoechas 8.4 µg/g (0.0010% DW; 17 of 55) and A. satureioides 3.2 µg/g (0.0004% DW; 12 of 32).
In H. italicum, as reported in our previous study [35], the total PG α-pyr content is dominated by the 3-prenyl (66%) and the 3-geranyl (16%) PG α-pyr, with the SKs at 7%, the 3-NA derivatives at 4% and the BPs and BFs at 2–3% each. Arzanol (PG18; 43.2 µg/g, 28% of the total) is the main compound, followed by its 6-methyl-α-pyrone analogue PG8 (16%); with the two 18,18-bis-desmethyl achyroclinopyrones (PG69, PG57), 23-methyl-6-O-desmethylauricepyrone (PG62), helispiroketal F (SK3), the 6-methyl 3-prenyl hexanoyl derivative PG63, arenol C (PG46) and the butanoyl and propanoyl homologues of arzanol (PG51, PG36) these make up the ten most abundant compounds, 80% of the total (Table 4 and Table 7).
In H. arenarium the total PG α-pyr content is likewise dominated by the 3-prenyl (45%) and the 3-geranyl (25%) PG α-pyr, with the BPs at 12% and the SKs at 6%, whereas the 3-(hydroxygeranyl) derivatives, the most numerous group of the species, contribute only 3%. Arzanol (15.4 µg/g, 20% of the total) is the main compound here too, followed by a 3-geranyl acetyl isomer of 18,18-bis-desmethyl achyroclinopyrone A eluting 0.2 min before it (PG68; 6.8 µg/g), the propanoyl and butanoyl homologues of arzanol (PG36, PG51; 6.0 and 5.8 µg/g) and the 3-geranyl propanoyl derivative PG87 (5.0 µg/g); the ten most abundant compounds make up 66% of the total. The arzanol content of H. arenarium is a third of that of H. italicum and its total PG α-pyr content half of it. The two species differ in the compounds that follow arzanol: the 6-methyl analogue of arzanol carries a sixth of the total in H. italicum [35] but is a minor compound in H. arenarium (2%). The arzanol content of both species is of the order of the 0.002% (w/w) isolated from H. italicum aerial parts [37] and one to two orders of magnitude below the yields reported for H. italicum ssp. microphyllum (0.05–0.5%) [14,17] and for the flowers of H. stoechas (0.48%) [3].
In H. stoechas 79% of the total PG α-pyr content falls on the BPs, and more than half of it (55%) on a single compound, BP37, a BP with a cyclised dehydro-geranyl ring, a 6-ethyl-α-pyrone and an acetyl group (4.6 µg/g). This compound is minor in H. arenarium (0.3 µg/g) and was not detected in A. satureioides or in H. italicum, whose dehydro-geranyl BP BP20 [35] is an isomer eluting at 15.2 min that H. arenarium and H. stoechas also hold. BP37 is followed by its 6-methyl-α-pyrone analogue BP28 (16%), 18,18-bis-desmethyl achyroclinopyrone A (PG69), helipyrone A (DP3) and its own propanoyl homologue BP47 (Table 4). The literature on H. stoechas describes an arzanol-rich chemistry — 0.48% (w/w) in the flowers, with helipyrone A (0.041%) and 18,18-bis-desmethyl achyroclinopyrone C (0.086%) [3] — and the isolation of helipyrones and of geranyl derivatives [21,22]. In the sample analysed here arzanol is below the LOQ and the whole class amounts to 8.4 µg/g, two to three orders of magnitude below these yields. The sample is a commercial dried herb (Section 3.1) whose plant part and age are not documented.
In A. satureioides helipyrone A (DP3) alone makes up 64% of the total PG α-pyr content (2.0 µg/g), and the rest is spread over the compounds that the isolation work reports for the species: 23-methyl-6-O-desmethylauricepyrone (PG62; 12%), two pentanoyl BPs with a cyclised prenyl ring (BP10, BP19), the butanoyl homologue of arzanol (PG51) and the pentanoyl and butanoyl BFs with a cyclised prenyl ring (BF19, BF22), each between 0.07 and 0.4 µg/g (Table 4); the ten most abundant compounds make up 98% of the total. The isolation literature on the species is short and fits this profile. Casero et al. [9] obtained from the plant four PG α-pyr, all carrying the (S)-2-methylbutanoyl group that the rules read as pentanoyl: 23-methyl-6-O-desmethylauricepyrone (described from H. odoratissimum [38] and A. satureioides [13,24]) and three compounds in which its prenyl chain is cyclised: to a 2-isopropenyl-2,3-dihydrobenzofuran (the pattern of BF19 and of 22-methyl-22-ethyl-italipyrone), to a 3-hydroxy-2,2-dimethylchroman (the pattern of BP10 and BP19) and to a 2,2-dimethyl-2H-chromene.
The four samples fall into two groups by their PG α-pyr content. H. italicum (153 µg/g) and H. arenarium (78 µg/g) are rich in the class; H. stoechas (8.4 µg/g) and A. satureioides (3.2 µg/g) contain nine to eighteen times less than the former and more than twenty times less than the latter. In the two rich species the main compounds are the same: the open-chain 3-prenyl and 3-geranyl derivatives, with arzanol as the most abundant compound in both. The two species differ in the compounds that accompany arzanol. In H. italicum these are the 6-methyl-α-pyrone analogues, the 6-methyl analogue of arzanol (PG8) being its second most abundant compound. In H. arenarium they are the propanoyl, butanoyl and pentanoyl homologues of the acetyl compounds and the 3-(hydroxygeranyl) derivatives. H. stoechas and A. satureioides, the two species poor in the class, are each dominated by a single compound that is minor or absent in H. italicum and H. arenarium: BP37, a dehydro-geranyl BP, makes up 55% of the content of H. stoechas, and helipyrone A (DP3) 64% of the content of A. satureioides. For the chemotaxonomy of the group this means two things. The class itself is common to all four species: five compounds, arzanol among them, were detected in each of the four species (Table 5). The species differ in the subclass that dominates and in the substituent preferred at the α-pyrone or at the acyl position [10].

2.10. Limitations

The substitution patterns are assignments made by a rule set, not identifications. The rules were derived from the manual annotation of one extract of H. italicum [35]; applied to the other species they recognise the same building units in new combinations, and a combination that was not observed in H. italicum (46 of the 91 patterns) rests on the same diagnostic ions but has not been checked against an isolated compound. Isomers that differ in the branching of the acyl group, the position of a ring closure or of a hydroxyl group, or the geometry of a double bond, receive the same pattern; chromatographically resolved compounds of one pattern are kept apart in Table 4, but which of them corresponds to a reported structure can be said only where the retention time of the published annotation decides.
The confidence grade depends on the intensity of the spectrum. The four parts of a pattern rest on the major ions — the α-pyrone pair, Fragments A and B — and are the same in a strong and in a weak spectrum, but the class ions that separate a high from a medium grade are minor ions, the first to fall below the 0.5% threshold when the precursor is weak. A medium grade of a minor compound therefore means in most cases that the spectrum was too weak to show the class ions, not that they were absent, and the medium-graded compounds are included in all counts and in the quantitation in the same way as the high-graded ones.
The features were built on the MS1 peak of the precursor (Section 3.7). The merging of features within 0.12 min, and the matching of compounds across species by retention time, can in principle join two isomers that elute close together or separate one compound whose weak peak has an imprecise apex; the feature numbers in Table 4 and the feature table S2 make every such decision traceable.
The quantitative comparison between species is relative. The totals and the ranking of the species are therefore indicative, and the comparison with isolation yields from the literature can establish the order of magnitude only; the contents of H. italicum come from the previous study [35] and its own calibration of the same standard. Each species is represented by one sample of one commercial batch; the differences between the species described here are differences between these samples, and their generality has to be tested with further accessions.

3. Materials and Methods

3.1. Plant Material and Sample Preparation

Dried inflorescences of Helichrysum arenarium (L.) Moench were obtained from Greekherbay (Greece), dried plant material of Helichrysum stoechas (L.) Moench, certified as an organic product, from Theodoros Pervoli Ltd (Maroni, Cyprus), and dried Achyrocline satureioides (Lam.) DC. from Diochi spol. s r.o. (Prague, Czech Republic). The data of Helichrysum italicum (Roth) G. Don are those published in our previous study [35]: dried inflorescences from Greekherbay (Patras, Greece), extracted and analysed as described there under the conditions of Section 3.2, Section 3.3 and Section 3.4.
Each plant material was mechanically ground to approximately 0.3 mm. An accurately weighed portion of the powder (H. arenarium, 1.0009 g; H. stoechas, 1.0004 g; A. satureioides, 1.0002 g) was placed in a 100 mL glass flask and extracted with 20 mL of acetone (100%) by ultrasonication (15 min) followed by orbital stirring (600 rpm, 15 min); the extraction was repeated three times (3 × 20 mL). The combined extracts were filtered through qualitative filter paper and dried under vacuum, yielding 90.7 mg (H. arenarium), 31.1 mg (H. stoechas), and 33.4 mg (A. satureioides) of dry residue (9.1, 3.1, and 3.3% of the dried material). Each residue was quantitatively transferred with methanol (MeOH, 100%) into a 10 mL volumetric flask and made up to the mark with the same solvent (Solution I). For LC–HRMS analysis, 100 µL of Solution I was diluted to 1.0 mL with MeOH and filtered through a 0.45 µm syringe filter, and 5 µL was injected into the UHPLC–HRMS system. The three runs of the study are listed in Table 8; each was processed independently through Section 3.5, Section 3.6, Section 3.7, Section 3.8, Section 3.9, Section 3.10, Section 3.11 and Section 3.12, and the species were compared on the resulting compound lists (Section 3.13, Section 3.14 and Section 3.15).

3.2. UHPLC–HRMS Instrument

UHPLC–HRMS analyses were performed on a Q Exactive Plus Orbitrap mass spectrometer (Thermo Fisher Scientific, Karlsruhe, Germany) equipped with a heated electrospray ionisation (HESI-II) source. The chromatographic system was a Dionex UltiMate 3000 RSLC comprising an SRD-3600 degasser, an HPG-3400RS binary pump with solvent selection valve, a thermostatted autosampler (WPS-3000TRS) and a column compartment (TCC-3000RS) (Thermo Fisher Scientific, Germany). Separation was achieved on a Kromasil 100-1.8-C18 column (2.1 × 100 mm, 1.8 µm; part number MF1CLD10; Nouryon, Bohus, Sweden).

3.3. Chromatographic Parameters

The mobile phase consisted of water containing 0.1% formic acid (A) and acetonitrile/water (95:5, v/v) containing 0.1% formic acid (B) at a flow rate of 0.3 mL/min. The gradient was 30% B at 0.5 min, 40% B at 2.0 min, 80% B at 24.0 min and 95% B at 27.0 min, held until 29.5 min, followed by re-equilibration to 50% B at 30.0 min. The method follows the previous study [35], and the retention times of the compounds common to the species agree with the published values within 0.01 min (Section 3.13).

3.4. Mass Spectrometric Parameters

Detection was performed in negative ion mode using full-scan (FS) and data-dependent MS/MS (DD-MS2) acquisition over 0.9–29.06 min, with in-source CID disabled and a default charge state of 1. Full-scan spectra were acquired at a resolution of 35,000 (FWHM at m/z 200) with one microscan, an AGC target of 1 × 106, a maximum injection time of 128 ms and a scan range of m/z 150–1000, in profile mode.
Data-dependent MS/MS spectra were acquired in Top5 mode at a resolution of 17,500 with one microscan, an AGC target of 1 × 105, a maximum injection time of 64 ms, a loop count of 5, an MSX count of 1 and an isolation window of 0.4 m/z with zero offset; no fixed first mass was set. Fragmentation used stepped normalised collision energies (NCE, %) of 20 and 50, as in the previous study, from which the fragmentation rules in Section 3.6 were derived. Precursor selection used a minimum AGC target of 2.5 × 103 (intensity threshold 3.9 × 104), apex triggering between 2 and 4 s and dynamic exclusion of 5.0 s; charge-state and isotope exclusion were not applied, no inclusion or exclusion list was used, and the instrument did not pick additional precursors when idle.

3.5. Data Conversion and Import

Vendor *.raw files (Thermo Fisher Scientific) were converted to *.ms1 (full-scan data) and *.mgf (MS2 data) files with msConvertGUI 3.0 (ProteoWizard) [39] and imported into R 4.6.1 [40] operated from RStudio (Posit, Boston, MA, USA). Full-scan signals below 1 × 104 counts and MS2 signals below 8 × 103 counts were excluded, and fragment intensities were expressed as a percentage of the base peak of each scan. The import was parallelised over runs (packages parallel and future.apply [41]), and all tables were held as data.table objects [42].
Each MS2 scan was linked to the full-scan spectrum that preceded it. The precursor m/z written in the *.mgf file was replaced by the accurate mass of the matching ion in that full scan (15 ppm, 0.05 min), and the full-scan intensity of the ion was stored with the scan. MS2 scans whose precursor was the 13C isotope peak of a more intense full-scan ion were discarded. The twenty most intense ions persisting through each run above 1 × 105 counts were recorded as background ions. The import thus yielded, per run, a list of MS2 scan tables (corrected molecular ion, fragment m/z and intensity, retention time, full-scan intensity, MS1 and MS2 scan numbers) and the complete MS1 ion table from which the extracted-ion chromatograms were built.

3.6. The Fragmentation Rules

Every MS2 scan with at least three fragment ions was tested against the fragmentation rules established for phloroglucinol α-pyrones in H. italicum [35], written as an R function and applied to every scan independently (Scheme 2). All masses are exact monoisotopic masses, matched at 15 ppm with a minimum window of 0.003 Da; intensities are relative to the base peak of the scan.
The α-pyrone unit was sought as the ion pair [α-pyr−H]− (m/z 139.0401 + n × 14.0157) and [α-pyr−CO2−H]− (m/z 95.0502 + n × 14.0157) for the C6 homologues n = 0 (methyl) to 5 (hexyl), both at ≥ 1%, together with Fragment A at [M−H]− − (140.0473 + n × 14.0157) Da and Fragment B at 12.0000 Da below Fragment A. A homologue was accepted when the ion pair was present, or when Fragment B reached ≥ 20% together with Fragment A ≥ 20% or [α-pyr−H]− ≥ 1%; of several qualifying homologues the one with the more complete evidence and the larger summed intensity was retained, and n defined the C6 substituent.
With Fragment B present at ≥ 20%, the scaffold was assigned from the intensity pattern of the previous study: open-chain phloroglucinol α-pyrone when Fragment A was ≥ 30% and A/B ≥ 0.4; BP when Fragment A was ≤ 15% and [α-pyr−H]−/B ≥ 0.6; BF when Fragment A was ≤ 15% and [α-pyr−H]−/B ≥ 0.1; otherwise the scaffold was left unresolved. Scans showing the α-pyrone pair but no Fragment B were labelled as α-pyrone-containing with an unrecognised phloroglucinol unit, or as monopyrone-like when [M−H−CO2]− was the base peak; both received low confidence.
Scans without an α-pyrone pair, and scans with the pair but without Fragment B, were tested for the SK pattern: a base peak (≥ 50%) at [M−H]− minus a furan-3-one unit (C5H6O2, 98.0368 Da, or its CH2 homologues), supported by at least one of the class ions m/z 243.0663, 215.0350, 229.0506, 161.0244, 162.0322, 204.0428 and 217.0870 (≥ 0.5%) or by an admissible resolution of the base peak into an SK core and an acyl ketene. Scans in which [M−H]− was ≤ 5%, [α-pyr−H]− was ≥ 50% and [M−H]− equalled the sum of an α-pyrone anion, a neutral α-pyrone and 12.0000 Da were assigned to the DPs, the two homologues giving the C6 substituents.
The C3 side chain and the C1 acyl group were read from Fragment B, which was resolved into a phloroglucinol core ion and an acyl ketene. The cores admissible in the assigned scaffold (3-NA m/z 125.0244; 3-prenyl 193.0870; 3-(hydroxyprenyl) 209.0819; 3-(hydroxydihydroprenyl) 211.0976; 3-geranyl 261.1496; 3-(hydroxygeranyl) 277.1445; BF cores 191.0714 and 259.1340; BP cores 209.0819, 277.1445 and 259.1340; SK cores 217.0870 and 285.1496) were combined with the ketenes C2H2O + k × CH2 (k = 0–6, acetyl to octanoyl) and C9H8O (3-phenylpropanoyl), and the combination whose sum matched Fragment B within tolerance defined C3 and C1. When more than one combination matched, the one supported by observed core ions (the core, the core + 12.0000 Da, and their CO2-loss ions) and by the acyl anion was taken and the alternative was recorded in the notes. The acyl-derived anion series (m/z 83.0138 + k × 14.0157; 173.0608 for the phenylpropanoyl member) was read as an independent indication of C1, and a disagreement with the ketene-derived value was recorded. A BF or BP assignment resting on intensities alone was overruled when Fragment B resolved only under the other scaffold, and an unresolved scaffold was assigned from the core when Fragment B resolved under exactly one admissible scaffold; both cases were flagged. A second α-pyrone homologue with its own ion pair and its own Fragment B flagged the scan as a mixture of co-eluting isobaric isomers.
The low-intensity diagnostic ions of the assigned subclass (≥ 0.5%; Supplementary Table S1 of [35]) were counted, and the anion formula was computed from the assigned substituents. Each assignment was graded: high when Fragment B resolved, at least two class ions were present, the scaffold was resolved and no reassignment or mixture had been flagged; medium when Fragment B resolved but one of these conditions failed; low otherwise. Spiroketals were graded high with a resolution of the base peak and at least two class ions, medium with one class ion; DPs satisfying the mass relationship were graded high. Scans graded low were kept in the scan tables for inspection but did not enter the feature building.

3.7. Feature Construction on MS1 Peaks and Consensus Spectra

The scans that passed the rules were grouped into features on the MS1 chromatographic peak of their precursor with the functions of the targeted workflow of the pipeline. Scans were processed in order of decreasing full-scan intensity of the precursor. The most intense unassigned scan served as seed; candidate scans were those within 10 ppm of its molecular ion and ±0.30 min of its retention time. An extracted-ion chromatogram (EIC; 10 ppm, ±0.60 min around the seed) was built from the MS1 data, and a peak was picked on it with the MS1 scans of the candidates nearest the seed as mandatory anchors. Peak detection worked on a 9-point smoothed trace; the boundaries were placed on the raw trace at the valleys at which the signal fell below 10% of the apex, or at which the rise after a valley exceeded 3% of the valley, three times the noise and 2% of the apex; a peak required at least five points. The peak was integrated by the trapezoid rule after subtraction of a baseline set to the median absolute deviation of the noise of the trace, and the same baseline was subtracted from the apex height. MS2 scans were assigned to the feature when their MS1 scan lay within the picked peak. A seed falling on a peak already taken by a feature of the same precursor (its apex inside the new segment, the maximum of the new segment at the edge of the existing one, or a tail scan within 1.0 min and below 2% of the existing apex) was merged into that feature. Seeds without a pickable peak were grouped with the scans within ±0.05 min and discarded afterwards, as no MS1 quantity could be assigned to them. A broad or shouldered peak with several seeds can nevertheless be built as several features, which the merging of Section 3.13 joins again.
The MS2 scans of each feature were combined into one consensus spectrum: the fragment ions of all member scans were clustered at 3 ppm, each cluster represented by its intensity-weighted mean m/z and summed intensity and rescaled to percent of the base peak, and ions below 0.1% were removed. The consensus spectrum was classified with the rules of Section 3.6, which gave the substitution pattern of the feature, and the number of member scans whose own assignment agreed with it was recorded. Features whose molecular ion lay 1.0034 Da above that of a feature at the same retention time (±0.08 min) with at least twice the intensity were flagged as 13C-isotope features. Each feature carried the molecular ion, the apex retention time, the peak boundaries, width, area, apex intensity (FSint) and baseline, the number and identity of its MS2 scans, the EIC, and the substitution pattern (scaffold, C6, C3, C1, predicted formula, confidence, notes).
The features were then written into the compound and fragment tables of the targeted pipeline, so that formula assignment, matrix construction, clustering and quantitation ran unchanged. The consensus fragments were filtered as in the reference workflow: ions with a fractional mass between 0.40 and 0.97 (incompatible with CHO ions), ions below 60 Da and ions beyond the 30 most intense were removed unless they carried a role label from the rules (Fragment A, Fragment B, core ions, α-pyrone pair, acyl anion, class ions); duplicate ions within 3 ppm were collapsed with scan-weighted intensities. Features without an integrable MS1 peak were dropped. After formula assignment (Section 3.8), fragments with a ring-and-double-bond equivalent below 1, without carbon or below 0.5% (role-labelled ions excepted) were removed, and the measured fragment masses were replaced by the exact masses of their formulae. The per-run feature tables, the scan-level assignments and the summaries by subclass and substituent were written to xlsx, and the EIC of every feature with its MS2 scans was plotted to PDF for inspection.

3.8. Elemental Composition Assignment

The molecular formula was taken from the assigned substituents whenever the assignment was complete. Otherwise CHO formulae were solved for the molecular ion by integer linear programming (package lpSolve [43]) under the constraints of a ring-and-double-bond equivalent between 0 and 10, at least one carbon, at most 49 C, 62 H and 14 O, and a mass tolerance widened in ten steps from 0.6 ppm to 15 ppm, the solution found at the tightest step being retained. Fragment formulae were solved in the same way within the elemental composition of the parent ion (15 ppm), and the exact mass and mass error of every fragment were recorded with it.

3.9. Predicted Compound Dictionary and Library Search

A dictionary of all phloroglucinol α-pyrone derivatives that can be assembled from the substituents observed in H. italicum was generated. The building units were the α-pyrone C6 alkyl (methyl to pentyl, n = 0–4), the C1 acyl group (acetyl, propanoyl, butanoyl, pentanoyl, hexanoyl, heptanoyl, 3-phenylpropanoyl), the phloroglucinol core (3-NA, 3-prenyl, 3-(hydroxyprenyl), 3-(hydroxydihydroprenyl), 3-geranyl, 3-(hydroxygeranyl) and 3-(hydroxydihydrogeranyl); BF with cyclised prenyl or geranyl; BP with cyclised prenyl, (hydroxy)geranyl or dehydrogeranyl; SK with a prenyl- or geranyl-derived core) and, for the SKs, the furan-3-one unit (C5H6O2 or C6H8O2); DPs were enumerated as all pairs of C6 homologues. Combination gave 463 entries (420 open-chain, BF and BP derivatives, 28 SKs and 15 DPs; Table S6), each with its anion formula and exact [M−H]−, the exact masses of Fragments A and B, of the core ions (A and B after loss of the acyl ketene), of the α-pyrone pair and of the acyl anion, the corresponding neutral losses, the expected base peak and intensity pattern of the scaffold, the class ions, the isobaric entries of the same formula and the measurement that separates them. Entries were marked as observed when they corresponded to one of the 58 PG α-pyr annotated in H. italicum, as predicted from observed substituents, or as extended homologues when they used a unit not observed so far (butyl and pentyl α-pyrones, hydroxydihydrogeranyl, geranyl benzofuran).
Every feature of every run was matched to the dictionary by its molecular ion (10 ppm). For each candidate entry the expected ions were looked up in the consensus spectrum (15 ppm, ≥ 0.5%) and a score was computed as the fraction of the expected-ion weights that were found (Fragment B 3, Fragment A 2, [α-pyr−H]− 2, [α-pyr−CO2−H]− 1, core ions 1 each, acyl anion 1, B−CO2 0.5, each class ion 0.5; for SKs base peak 3, core 1, acyl anion 0.5; for DPs the two α-pyrone anions 3 each and their CO2-loss ions 1 each). Candidates were ranked by agreement of their subclass with the pattern assigned by the rules and then by score; the best entry with its score, the ions found and missing and the ranked alternatives were reported per feature, and the matched entries were summarised by species (78 entries were matched by a retained feature with a score ≥ 0.5 — 37 observed, 35 predicted from observed substituents and 6 extended homologues; 70 in H. arenarium, 33 in H. stoechas and 20 in A. satureioides; Table S2).

3.10. The Five Conditions for Retaining a Feature

Each feature carried the grade of the assignment on its consensus spectrum (Section 3.6). A feature was retained when (i) its scaffold was resolved, (ii) the assignment was graded high or medium, (iii) it was built from at least two MS2 scans, (iv) it was not a 13C-isotope feature (Section 3.7) and (v) its notes did not flag a mixture of co-eluting isomers, a scaffold reassignment, an alternative resolution of Fragment B or a disagreement between the acyl anion and the ketene-derived acyl group. Features that matched the SK pattern but whose base peak could not be resolved into an admissible core and acyl group were removed as well, since no substituents and no structure can be given for them. The features that failed one of these conditions were kept in the feature tables for inspection (Table S5) but do not enter any count, table or figure of Section 2.
In the scheme of Schymanski et al. [36], a high-confidence assignment — an accurate-mass formula consistent with the assigned substituents and the complete set of diagnostic fragments — corresponds to Level 2b (probable structure by diagnostic evidence); a medium-confidence assignment, in which the scaffold, C6, C3 and C1 were read from the fragments but fewer than two class ions were observed, is a tentative candidate (Level 3). Because the class ions are minor ions, the grade depends on the intensity of the spectrum (Section 2.10). No reference standard of the class was available. Stereochemistry and the position of side-chain hydroxylation are not accessible by MS2 and, where stated, were inferred by analogy with the reported structures.

3.11. Structure Generation

For every substitution pattern and formula of the pool (Section 3.13), a structure was assembled from the assigned substituents with RDKit 2026.03.6 [44] under Python 3.13 [45]. Templates for the five scaffolds (open-chain phloroglucinol α-pyrone, benzofuran, benzopyran, spiroketal, dipyrone) were combined with the C6 alkyl, the C1 acyl group and the C3 side chain or ring; for the SKs the furan-3-one unit (C5H6O2 or C6H8O2) was the one that reproduced the formula of the pattern. The acyl groups were drawn as the isomers reported for the class (2-methylpropanoyl and 3-methylbutanoyl for the C4 and C5 members), the hydroxylated side chains with the hydroxyl on C8, the second carbon of the chain, and the double bond on C9–C10 (2-hydroxy-3-methylbut-3-enyl, 2-hydroxy-3-methylbutyl and 2-hydroxy-3,7-dimethylocta-3,6-dienyl), the position reported for the hydroxylated side chains of the class [35], and the fused rings closed onto the hydroxyl adjacent to the methylene bridge; the alternative isomers were listed with every structure. Canonical SMILES [46], standard InChI and InChIKey [47], the neutral formula, the exact mass and the calculated [M−H]− were derived and checked against the formula of the pattern, and depictions were written as PNG and SVG files (Figure 1, Supplementary Figures S1 and S2). All structures were drawn in one orientation, that of the drawing of the previous study [35]: the α-pyrone ring on the left and the phloroglucinol ring on the right, both with a vertex at the top, the C5–CH2–C3′ bridge above the two rings, the 4-hydroxyl upwards, the C3 side chain or the fused ring of the BFs and BPs to the right, the C1 acyl group below the ring and the 2- and 6-hydroxyls downwards; the side chains were laid out as zig-zags to the right. The SKs keep the phloroglucinol ring in the same place with the spiro carbon in the place of C3′, and the DPs are drawn as two pyrones bridged by the methylene. The two-dimensional coordinates of these core atoms were fixed and the remaining atoms were placed by the depiction routine of RDKit (GenerateDepictionMatching2DStructure). The templates reproduced the formulae and [M−H]− of all 463 dictionary entries and the InChIKeys of the compounds of the class available in public databases (arzanol, italipyrone, plicatipyrone, helipyrones A and B, helispiroketals A and F). SMILES and InChIKey were appended to the compound and feature tables (Table S1, Table S2 and Table S4).

3.12. Semi-Quantitative Assessment

Phloroglucinol α-pyrones were assessed semi-quantitatively against 2′,4′,6′-trihydroxyacetophenone monohydrate (THAP; BLDPharm, Kaiserslautern, Germany; Cat. No. BD33753-5g) as external calibration standard, using the baseline-corrected areas of the picked MS1 peaks of the molecular ions (Section 3.7). External calibration was performed over 0.000333–1.000 µg/mL at six levels (0.000333, 0.001, 0.01, 0.05, 0.1 and 1 µg/mL) injected at the same volume as the samples (5 µL), and fitted by 1/x2 weighted linear regression, A = b0 + b1C (b0 = 4.66 × 104, b1 = 4.371 × 108; weighted R2 = 0.9973; Figure 6), where A is the peak area and C the concentration in the injected vial. The weighting was chosen on the back-calculated concentrations of the calibration levels: under 1/x2 all six levels fall within −6 to +9% of nominal (97.1, 108.5, 101.2, 100.0, 98.9 and 94.2% from the lowest to the highest level), whereas the lowest level deviates by −21% under 1/x weighting and by several hundred per cent without weighting. Concentrations were calculated as Cvial = (A − b0)/b1 and converted to µg/g dry weight through the preparation scheme (Section 3.1): residue in 10 mL and 100 µL diluted to 1.0 mL, i.e. a factor of 100 from the injected vial to the extract and 99.91, 99.96 and 99.98 µg/g per µg/mL for the H. arenarium, H. stoechas and A. satureioides samples. The limit of quantification (LOQ) was set at the lowest calibration level whose back-calculated concentration was within ±20% of nominal, 0.000333 µg/mL, which corresponds to 0.033 µg/g dry weight; the limit of detection (LOD) was set at one third of it (0.011 µg/g). Features below the LOQ were flagged and left out of the totals; no feature exceeded the upper calibration level (1 µg/mL, about 100 µg/g). Matrix effect and recovery were those determined in the previous study [35] with an extract of A. satureioides spiked with THAP at 0.001 and 0.1 µg/mL (signal in the matrix 95–105% of that in solvent; recoveries 85% and 75%); the mean recovery of 80% was applied as a correction factor to the contents expressed in % dry weight, whereas the contents in µg/g are reported uncorrected. The content of a compound is the sum of the contents of the features merged into it (Section 3.13). The contents of H. italicum are the published ones [35], obtained with the same standard and preparation scheme against the calibration of that study (1/x weighted) and converted from the published recovery-corrected % dry weight to uncorrected µg/g (% DW × 104 × 0.8).

3.13. From Features to Compounds: Merging, Matching Across Species and Identifiers

Within a run, retained features that were assigned the same scaffold, the same C6, C3 and C1 substituents and the same molecular formula, and whose apex retention times lay within 0.12 min of each other, were merged into one compound (split peaks); the compound takes the retention time, molecular ion and consensus spectrum of its most intense feature, the sum of the peak areas and contents of its features, and the better of their grades. The 58 PG α-pyr of the published annotation of H. italicum [35] (micropyrone, a monopyrone outside the rules, excepted; Table S3) entered the comparison as compounds with their published substitution pattern — subclass, α-pyrone C6 substituent, C3 side chain and C1 acyl group as assigned there — formula, retention time and content.
Compounds of different species were taken as the same compound when they had the same substitution pattern and formula and their retention times formed a chain in which consecutive values differ by at most 0.20 min; a chain holding two compounds of one species was split at the largest step between the two. Since the runs used the chromatographic method of the previous study, no alignment was needed: for the compounds common to H. arenarium and the published annotation the retention times differ by −0.01 min on median (median absolute deviation 0.02 min, 36 of the 40 within 0.1 min and none beyond 0.15 min). The compounds of the four species thus form one pool (Table 4, Table S1). Each compound of the pool received an identifier made of the abbreviation of its subclass (PG, BF, BP, SK, DP; Section 2.2) and a number in order of the mean retention time of its records. The compound numbers of the previous study are given with the identifiers where a compound of the pool is one of the 58 (Table 4). The sets of compounds of the species were compared as the compounds shared by all four, by three, by two and confined to one species, and pairwise by the Jaccard index (shared compounds divided by the compounds of either species; Section 2.6). For every substitution pattern and formula of the pool one structure was drawn (Section 3.11; Supplementary Figure S2); the structure is labelled with the identifiers of all compounds that carry the pattern. For every compound of the pool, the consensus MS2 spectrum of its most intense feature (Section 3.7), the structure and the species in which it was detected, with retention time, content and grade, were plotted on one page (Supplementary Figure S1; matplotlib 3 [48] and RDKit); the 14 compounds known from the published annotation only are shown with their structure and published data, and the consensus spectrum of one feature (H. arenarium, feature 448), which the feature table lacked, was rebuilt from its seven MS2 scans.
Detection in one sample only was distinguished from absence by a full-scan check: for every compound of the pool, the MS1 data of each of the three samples (both acquisitions) were searched for its [M−H]− within 10 ppm and ±0.30 min of its retention time, and the compound was recorded as present in full scan when the ion reached ≥ 1 × 104 counts in at least three scans with its apex within 0.15 min of the retention time; a signal was disregarded when a detected compound of the same formula eluted within 0.30 min in that sample (an isomer or a split peak). The intensity was compared with the MS2 trigger threshold (3.9 × 104 counts; Section 3.4), the MS2 scans recorded within 10 ppm and 0.15 min were inspected for Fragments A and B of the compound, and the features of the same substitution pattern that had not been retained (Section 3.10) were listed. The result per compound and sample is Table S7.

3.14. Fragment-Ion and Neutral-Loss Matrices

The compounds entering the multivariate analysis were the retained features (Section 3.10) of each species, each species analysed separately with its consensus spectra, and the three species together. Fragment ions were collected across compounds on their formula-derived exact masses and retained when present in at least two compounds at ≥ 0.5%; ions carrying a role label from the rules were retained regardless, ions above the smallest molecular ion of the set were removed, and non-detected ions were encoded as zero. The number of retained features was selected as described previously [35]: reduced matrices of increasing size were compared with the full candidate matrix by the Spearman correlation of their pairwise compound dissimilarities, the first size reaching r ≥ 0.98 being taken, with a cap of 65 features. Neutral losses were computed for every fragment with an assigned formula as the exact-mass difference between the parent and the fragment formula (losses below 17 Da discarded), and the neutral-loss matrix was built and reduced in the same way.

3.15. Dissimilarity Analysis, Clustering and Validation

Bray–Curtis dissimilarities between compounds and between features were calculated with vegdist (package vegan) [49]. The number of compound clusters was determined by partitioning around medoids (PAM) [50] on the dissimilarity matrix (package cluster) [51] as the smallest k within 0.02 of the maximum average silhouette width [52] for 2 ≤ k ≤ min(20, n − 1), and the PAM membership at that k was the primary partition. Hierarchical clustering with average linkage (UPGMA) on the same dissimilarities gave the dendrograms and the heatmaps (package pheatmap) [53], drawn with factoextra [54], dendextend [55], ggdendro [56] and ggplot2 [57]; the cophenetic correlation was recorded as a measure of the fit of each dendrogram, and k-means (Euclidean, 25 starts, k from PAM) was computed as a secondary comparison. Principal component analysis was performed on the Hellinger-transformed matrix [58], and the compound dissimilarities were additionally displayed as ordered distance heatmaps.
Cluster memberships from PAM, hierarchical clustering and k-means, for the fragment-ion and the neutral-loss matrix, were compared with the assigned substitution patterns at five levels (scaffold, C6, C1, C3 and their combination), and, for the three species together, with the species, by the adjusted Rand index [59] and by cluster purity, the fraction of compounds belonging to the majority group of their cluster; the agreement between the fragment-ion and the neutral-loss partitions was quantified by the adjusted Rand index (Table 6); the composition of every cluster by scaffold was tabulated in the output workbook of the workflow and is not reproduced here.
Because one heatmap of all compounds is unreadable above about fifty rows, the same matrix was additionally drawn per subclass (the ten subclasses of Table 2): for each subclass with at least eight compounds in a species, the compounds were plotted with the fragment ions present in at least two of them at ≥ 1% (at most 40, by summed intensity), Bray–Curtis dissimilarity and average linkage, rows labelled by compound identifier, molecular ion and retention time and annotated by the substituents that vary within the subclass, and the rows were cut at the number of clusters with the largest average silhouette width (2 ≤ k ≤ 8) (Figure 4). Cluster memberships, validation metrics and cluster compositions were written to xlsx.

3.16. Software

All processing was performed in R 4.6.1 [40] with the packages data. Table 1.18.6 [42], lpSolve 5.6.23 [43], vegan 2.7-6 [49], cluster 2.1.8 [51], pheatmap 1.0.13 [53], factoextra 2.2.0 [54], dendextend 1.19.1 [55], ggdendro 0.2.0 [56], ggplot2 4.0.3 [57], openxlsx 4.2.9 [60], future.apply 1.20.2 [41] and stringr 1.6.0 [61]; the annotation and the multivariate analysis were run on up to ten worker processes (package parallel), one run or one set of compounds per worker. Structures were generated with RDKit 2026.03.6 [44] under Python 3.13 [45] with openpyxl 3.1.5 [62], and the pages of Supplementary Figure S1 were drawn with matplotlib 3.11 [48]. Raw data were converted with ProteoWizard msConvert [39]. The scripts of the workflow (data import, fragmentation rules, feature construction, dictionary and library search, structure generation, quantitation, merging and matching of the compounds, multivariate analysis and report) are available from the corresponding author on request.

4. Conclusions

The fragmentation rules established for the PG α-pyr of H. italicum [35] were written into an automated workflow in R programming language and applied to H. arenarium, H. stoechas and A. satureioides. The automated annotation gave the same substitution pattern as the manual annotation of the previous study, and the generated dictionary of 463 expected compounds recognised 46 substitution patterns that had not been observed before. The four species share the core of the class — five compounds, arzanol among them, were detected in each of them — and differ in the subclass that dominates and in the substituents preferred at the α-pyrone and at the acyl position: H. italicum varies the α-pyrone, H. arenarium extends the acyl series and hydroxylates the geranyl chain, H. stoechas is dominated by a dehydro-geranyl BP and A. satureioides by helipyrone A. H. arenarium is the richest source of the class in number of compounds (178) and, after H. italicum, in content (78 µg/g). The unsupervised clustering of the MS/MS spectra recovered the scaffold and the α-pyrone substituent in the three species, as it did in H. italicum.

Supplementary Materials

The following supporting information can be downloaded at the website of this paper posted on Preprints.org. Supplementary Figure S1: consensus MS2 spectrum, structure and species of occurrence of every PG α-pyr of the pool (PDF, 218 pages); Supplementary Figure S2: structures of the 91 substitution patterns (PDF); Supplementary Table S1–S7 (one workbook): Table S1, the 215 compounds; Table S2, the 323 features with their MS2 spectra; Table S3, the published H. italicum annotation; Table S4, the structures (SMILES, InChIKey); Table S5, the features not retained; Table S6, the dictionary; Table S7, the full-scan check.

Author Contributions

Conceptualization, Y.V.; methodology, P.N. and T.M.; software, Y.V.; validation, Y.V.; writing—original draft preparation, Y.V.; writing—review and editing, Y.V., P.N. and T.M; visualization, Y.V.; project administration, Y.V. All authors have read and agreed to the published version of the manuscript.

Funding

The research was supported by the Council of Medical Sciences, Medical University of Sofia, Bulgaria. Grant number: 141/29.05.2024.

Data Availability Statement

The data presented in this study are available in the Supplementary Materials; the R scripts of the workflow are available from the corresponding author on request.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
3-NA 3-non-alkylated (phloroglucinol unit without a side chain at C3)
AGC automatic gain control
ARI adjusted Rand index
BF benzofuran (subclass of the PG α-pyr)
BP benzopyran (subclass of the PG α-pyr)
CID collision-induced dissociation
DAD diode-array detection
DD-MS2 data-dependent MS2 acquisition
DP dipyrone (subclass of the PG α-pyr)
DW dry weight
EIC extracted-ion chromatogram
FI fragment-ion (matrix)
FS full scan
FWHM full width at half maximum
HCD higher-energy collisional dissociation
HESI heated electrospray ionisation
HRMS high-resolution mass spectrometry
IC50 half-maximal inhibitory concentration
InChI, InChIKey IUPAC International Chemical Identifier and its hashed key
LC–MS liquid chromatography–mass spectrometry
LOD limit of detection
LOQ limit of quantification
5-LOX 5-lipoxygenase
MeOH methanol
mPGES-1 microsomal prostaglandin E synthase-1
MS1 full-scan mass spectrum (precursor ions)
MS2, MS/MS tandem mass spectrum (fragment ions)
MSX multiplexing (number of precursors per MS2 scan)
NCE normalised collision energy
NF-κB nuclear factor κB
NL neutral-loss
NMR nuclear magnetic resonance
PAM partitioning around medoids
PG phloroglucinol (unit)
PG α-pyr phloroglucinol α-pyrone(s)
RT retention time
SK spiroketal (subclass of the PG α-pyr)
SMILES simplified molecular-input line-entry system
THAP 2′,4′,6′-trihydroxyacetophenone
UHPLC ultra-high-performance liquid chromatography
UPGMA unweighted pair group method with arithmetic mean

Appendix A

Table A1. The 91 structures (drawn in Supplementary Figure S2), one per substitution pattern and formula: the compounds that carry the pattern, scaffold, α-pyrone C6 substituent, C3 side chain or ring, C1 acyl group (and, for the SKs, the furan-3-one unit), neutral formula, calculated [M−H]−, retention-time range of the compounds, species in which they occur, and trivial name. SMILES, InChIKeys and the positions assumed are in Table S4.
Table A1. The 91 structures (drawn in Supplementary Figure S2), one per substitution pattern and formula: the compounds that carry the pattern, scaffold, α-pyrone C6 substituent, C3 side chain or ring, C1 acyl group (and, for the SKs, the furan-3-one unit), neutral formula, calculated [M−H]−, retention-time range of the compounds, species in which they occur, and trivial name. SMILES, InChIKeys and the positions assumed are in Table S4.
Compounds Scaffold C6 C3 C1 Formula [M−H]− calc. RT (min) Species Name
PG1, PG27 open-chain ethyl 3-(hydroxyprenyl) acetyl C22H26O8 417.1555 4.50–12.41 H, I –
PG2, PG44 open-chain methyl 3-NA acetyl C16H16O7 319.0823 5.13–14.94 I, S –
PG3 open-chain methyl 3-(hydroxydihydroprenyl) acetyl C21H26O8 405.1555 5.44 I –
PG4, PG80, PG81 open-chain ethyl 3-NA acetyl C17H18O7 333.0980 6.78–20.08 H, I, S –
PG5 open-chain ethyl 3-(hydroxydihydroprenyl) acetyl C22H28O8 419.1711 7.21 H, I –
PG6, PG9 open-chain methyl 3-(hydroxygeranyl) acetyl C26H32O8 471.2024 7.71–8.62 H, I, S –
PG7, PG10, PG11, PG13, PG17, PG19, PG25, PG31, PG73, PG74, PG85 open-chain ethyl 3-(hydroxygeranyl) acetyl C27H34O8 485.2181 8.35–20.57 H, I, S –
PG8 open-chain methyl 3-prenyl acetyl C21H24O7 387.1449 8.50 H, I –
PG12 open-chain ethyl 3-NA propanoyl C18H20O7 347.1136 9.30 H, S –
PG14, PG53 open-chain ethyl 3-(hydroxyprenyl) propanoyl C23H28O8 431.1711 9.71–15.61 H –
PG15 open-chain methyl 3-(hydroxyprenyl) acetyl C21H24O8 403.1398 9.98 H, I –
PG16 open-chain ethyl 3-(hydroxydihydroprenyl) propanoyl C23H30O8 433.1868 10.05 H –
PG18 open-chain ethyl 3-prenyl acetyl C22H26O7 401.1606 10.43 A, H, I, S arzanol
PG20, PG24, PG26, PG29, PG34, PG45, PG52, PG54, PG84, PG86, PG92 open-chain ethyl 3-(hydroxygeranyl) propanoyl C28H36O8 499.2337 10.88–22.28 H –
PG21 open-chain methyl 3-NA pentanoyl C19H22O7 361.1293 10.93 H, I –
PG22 open-chain methyl 3-prenyl propanoyl C22H26O7 401.1606 11.21 H, I –
PG23, PG107 open-chain ethyl 3-NA butanoyl C19H22O7 361.1293 11.35–25.19 H, I, S –
PG28 open-chain propyl 3-prenyl acetyl C23H28O7 415.1762 12.47 H, I helitalone B
PG30, PG35, PG39, PG40, PG43, PG49, PG59, PG64, PG93, PG95, PG100, PG101 open-chain ethyl 3-(hydroxygeranyl) butanoyl C29H38O8 513.2494 12.78–24.17 H –
PG32 open-chain methyl 3-prenyl butanoyl C23H28O7 415.1762 13.13 I arenol B
PG33, PG37 open-chain ethyl 3-NA pentanoyl C20H24O7 375.1449 13.16–13.55 A, H, I, S –
PG36 open-chain ethyl 3-prenyl propanoyl C23H28O7 415.1762 13.37 H, I, S –
PG38 open-chain methyl 3-NA hexanoyl C20H24O7 375.1449 13.60 I –
PG41 open-chain ethyl 3-(hydroxydihydroprenyl) pentanoyl C25H34O8 461.2181 14.14 A –
PG42 open-chain ethyl 3-NA 3-phenylpropanoyl C24H24O7 423.1449 14.28 H –
PG46 open-chain methyl 3-prenyl pentanoyl C24H30O7 429.1919 15.00 A, I arenol C
PG47, PG58, PG60, PG71, PG76, PG79, PG103, PG108, PG110, PG112 open-chain ethyl 3-(hydroxygeranyl) pentanoyl C30H40O8 527.2650 15.06–26.37 H –
PG48 open-chain methyl 3-NA heptanoyl C21H26O7 389.1606 15.10 I –
PG50, PG55 open-chain ethyl 3-NA hexanoyl C21H26O7 389.1606 15.36–15.86 A, H, I –
PG51 open-chain ethyl 3-prenyl butanoyl C24H30O7 429.1919 15.36 A, H, I, S 6-O-desmethylauricepyrone/3-prenyl norauricepyrone
PG56, PG57 open-chain methyl 3-geranyl acetyl C26H32O7 455.2075 16.02–16.18 H, I, S 18,18-bis-desmethyl achyroclinopyrone C
PG61, PG62, PG65 open-chain ethyl 3-prenyl pentanoyl C25H32O7 443.2075 17.08–17.77 A, H, I, S 23-methyl-6-O-desmethylauricepyrone
PG63, PG67 open-chain methyl 3-prenyl hexanoyl C25H32O7 443.2075 17.45–17.95 I –
PG66 open-chain ethyl 3-(hydroxyprenyl) butanoyl C24H30O8 445.1868 17.80 A, H –
PG68, PG69 open-chain ethyl 3-geranyl acetyl C27H34O7 469.2232 18.11–18.33 A, H, I, S 18,18-bis-desmethyl achyroclinopyrone A
PG70 open-chain ethyl 3-prenyl 3-phenylpropanoyl C29H32O7 491.2075 18.44 H, I –
PG72 open-chain methyl 3-geranyl propanoyl C27H34O7 469.2232 19.09 H, S –
PG75, PG78, PG82 open-chain ethyl 3-prenyl hexanoyl C26H34O7 457.2232 19.40–20.13 A, H, I 23-ethyl-6-O-desmethyl-auricepyrone
PG77 open-chain ethyl 3-(hydroxyprenyl) pentanoyl C25H32O8 459.2024 19.62 A, H –
PG83 open-chain propyl 3-geranyl acetyl C28H36O7 483.2388 20.38 H, I 8′-methyl-18,18-bis-desmethyl achyroclinopyrone A
PG87, PG88, PG89, PG90 open-chain ethyl 3-geranyl propanoyl C28H36O7 483.2388 20.99–21.76 H, S –
PG91 open-chain ethyl 3-prenyl heptanoyl C27H36O7 471.2388 21.83 H –
PG94, PG96, PG97, PG98, PG99, PG102 open-chain ethyl 3-geranyl butanoyl C29H38O7 497.2545 22.62–24.17 A, H, S –
PG104, PG105, PG106, PG109 open-chain ethyl 3-geranyl pentanoyl C30H40O7 511.2701 24.62–25.65 A, H, S –
PG111 open-chain ethyl 3-geranyl 3-phenylpropanoyl C34H40O7 559.2701 25.78 H –
PG113 open-chain ethyl 3-geranyl hexanoyl C31H42O7 525.2858 26.96 H –
BF1, BF4 benzofuran methyl cyclised prenyl acetyl C21H22O7 385.1293 13.90–17.00 I –
BF2, BF11 benzofuran ethyl cyclised prenyl acetyl C22H24O7 399.1449 16.42–19.63 H, I italipyrone
BF3, BF6 benzofuran methyl cyclised prenyl pentanoyl C24H28O7 427.1762 16.76–17.86 I –
BF5, BF16, BF22 benzofuran ethyl cyclised prenyl butanoyl C24H28O7 427.1762 17.58–24.56 A, H –
BF7, BF8, BF10, BF12, BF19, BF26 benzofuran ethyl cyclised prenyl pentanoyl C25H30O7 441.1919 18.98–26.22 A, H, I 22-methyl-22-ethyl-italipyrone
BF9, BF18 benzofuran ethyl cyclised prenyl propanoyl C23H26O7 413.1606 19.41–22.57 A, H –
BF13, BF21, BF28 benzofuran ethyl cyclised geranyl acetyl C27H32O7 467.2075 20.22–26.81 H, S –
BF14, BF15 benzofuran ethyl cyclised prenyl 3-phenylpropanoyl C29H30O7 489.1919 20.29–20.44 H –
BF17 benzofuran propyl cyclised geranyl acetyl C28H34O7 481.2232 22.51 S –
BF20, BF25 benzofuran ethyl cyclised geranyl propanoyl C28H34O7 481.2232 23.22–26.14 H –
BF23 benzofuran methyl cyclised geranyl acetyl C26H30O7 453.1919 25.02 S –
BF24, BF27 benzofuran ethyl cyclised geranyl pentanoyl C30H38O7 509.2545 25.58–26.23 H, S –
BF29 benzofuran ethyl cyclised geranyl butanoyl C29H36O7 495.2388 27.10 H –
BP1, BP3 benzopyran methyl cyclised prenyl acetyl C21H24O8 403.1398 4.84–6.90 H, I –
BP2, BP5 benzopyran ethyl cyclised prenyl acetyl C22H26O8 417.1555 6.31–8.86 H, I plicatipyrone
BP4, BP7 benzopyran ethyl cyclised prenyl propanoyl C23H28O8 431.1711 8.69–11.46 H –
BP6, BP13 benzopyran ethyl cyclised prenyl butanoyl C24H30O8 445.1868 10.54–13.32 A, H –
BP8, BP9, BP17 benzopyran methyl cyclised (hydroxy)geranyl acetyl C26H32O8 471.2024 11.56–14.36 H, I, S –
BP10, BP11, BP19, BP21 benzopyran ethyl cyclised prenyl pentanoyl C25H32O8 459.2024 12.35–15.61 A, H, I –
BP12, BP16, BP28, BP32, BP42 benzopyran methyl cyclised dehydro-geranyl acetyl C26H30O7 453.1919 13.20–21.61 H, S –
BP14, BP15, BP24 benzopyran ethyl cyclised (hydroxy)geranyl acetyl C27H34O8 485.2181 13.64–16.42 H, I, S –
BP18 benzopyran ethyl cyclised prenyl hexanoyl C26H34O8 473.2181 14.64 H –
BP20, BP22, BP36, BP37, BP44, BP48, BP49, BP50, BP53 benzopyran ethyl cyclised dehydro-geranyl acetyl C27H32O7 467.2075 15.19–25.81 H, I, S –
BP23, BP25, BP30, BP31 benzopyran ethyl cyclised (hydroxy)geranyl propanoyl C28H36O8 499.2337 16.33–19.05 H –
BP26, BP29, BP38 benzopyran ethyl cyclised (hydroxy)geranyl butanoyl C29H38O8 513.2494 18.24–20.67 A, H –
BP27, BP47 benzopyran ethyl cyclised dehydro-geranyl propanoyl C28H34O7 481.2232 18.31–23.20 H, S –
BP33, BP35, BP45 benzopyran ethyl cyclised (hydroxy)geranyl pentanoyl C30H40O8 527.2650 19.92–22.29 A, H –
BP34, BP39, BP40, BP51, BP52, BP54 benzopyran ethyl cyclised dehydro-geranyl butanoyl C29H36O7 495.2388 20.30–25.94 A, H, S –
BP41 benzopyran methyl cyclised dehydro-geranyl propanoyl C27H32O7 467.2075 21.11 S –
BP43, BP46 benzopyran ethyl cyclised dehydro-geranyl pentanoyl C30H38O7 509.2545 22.10–22.33 H –
SK1, SK2 spiroketal – prenyl-derived (C5H6O2) acetyl C20H22O6 357.1344 9.21–9.57 H, I helispiroketal A
SK3, SK4 spiroketal – prenyl-derived (C6H8O2) acetyl C21H24O6 371.1500 11.25–11.45 H, I helispiroketal F
SK5 spiroketal – prenyl-derived (C6H8O2) propanoyl C22H26O6 385.1657 13.61 H, I –
SK6, SK11 spiroketal – geranyl-derived (C6H8O2) acetyl C26H32O6 439.2126 13.75–18.36 H, I, S –
SK7 spiroketal – prenyl-derived (C6H8O2) butanoyl C23H28O6 399.1813 15.50 A, H –
SK8 spiroketal – geranyl-derived (C5H6O2) acetyl C25H30O6 425.1970 16.37 H, S –
SK9 spiroketal – prenyl-derived (C6H8O2) heptanoyl C26H34O6 441.2283 16.60 H –
SK10 spiroketal – prenyl-derived (C6H8O2) pentanoyl C24H30O6 413.1970 17.05 A, H –
SK12 spiroketal – prenyl-derived (C6H8O2) octanoyl C27H36O6 455.2439 19.14 H –
SK13, SK14 spiroketal – geranyl-derived (C6H8O2) propanoyl C27H34O6 453.2283 20.43–20.61 H, S –
SK15 spiroketal – geranyl-derived (C6H8O2) butanoyl C28H36O6 467.2439 22.30 H –
SK16 spiroketal – geranyl-derived (C6H8O2) pentanoyl C29H38O6 481.2596 23.66 H –
DP1 dipyrone methyl + methyl – – C15H16O6 291.0874 6.55 S –
DP2 dipyrone methyl + ethyl – – C16H18O6 305.1031 8.74 A, I, S helipyrone B
DP3 dipyrone ethyl + ethyl – – C17H20O6 319.1187 11.23 A, H, I, S helipyrone A
Species: A, A. satureioides; H, H. arenarium; I, H. italicum; S, H. stoechas.

References

  1. Pljevljakušić, D.; Bigović, D.; Janković, T.; Jelačić, S.; Šavikin, K. Sandy everlasting (Helichrysum arenarium (L.) Moench): Botanical, chemical and biological properties. Front. Plant Sci. 2018, 9, 1123. [Google Scholar] [CrossRef]
  2. Dănăilă-Guidea, S.M.; Eremia, M.C.; Dinu, L.D.; Miu, D.M. Helichrysum arenarium: From Cultivation to Application. Appl. Sci. 2022, 12, 10241. [Google Scholar] [CrossRef]
  3. Les, F.; Venditti, A.; Cásedas, G.; Frezza, C.; Guiso, M.; Sciubba, F.; Serafini, M.; Bianco, A.; Valero, M.S.; López, V. Everlasting flower (Helichrysum stoechas Moench) as a potential source of bioactive molecules with antiproliferative, antioxidant, antidiabetic and neuroprotective properties. Ind. Crops Prod. 2017, 108, 295–302. [Google Scholar] [CrossRef]
  4. Valero, M.S.; Nuñez, S.; Les, F.; Castro, M.; Gómez-Rincón, C.; Arruebo, M.P.; Plaza, M.Á.; Köhler, R.; López, V. The Potential Role of Everlasting Flower (Helichrysum stoechas Moench) as an Antihypertensive Agent: Vasorelaxant Effects in the Rat Aorta. Antioxidants 2022, 11, 1092. [Google Scholar] [CrossRef]
  5. Retta, D.; Dellacassa, E.; Villamil, J.; Suárez, S.A.; Bandoni, A.L. Marcela, a promising medicinal and aromatic plant from Latin America: A review. Ind. Crops Prod. 2012, 38, 27–38. [Google Scholar] [CrossRef]
  6. Badalyan, A.; Abrahamyan, S.; Abovyan, A.; Badalyan, A.; Semerjyan, G.; Hovhannisyan, N. Helichrysum arenarium as a source of flavonoids: Evaluation of antimicrobial activity and flavonoid content of extracts of Helichrysum flowers in vitro. Funct. Foods Health Dis. 2024, 14, 51–61. [Google Scholar] [CrossRef]
  7. Hussain, M.S.; Azam, F.; Eldarrat, H.A.; Alkskas, I.; Mayoof, J.A.; Dammona, J.M.; Ismail, H.; Ali, M.; Arif, M.; Haque, A. Anti-inflammatory, analgesic and molecular docking studies of Lanostanoic acid 3-O-α-D-glycopyranoside isolated from Helichrysum stoechas. Arab. J. Chem. 2020, 13, 9196–9206. [Google Scholar] [CrossRef]
  8. Moresco, K.S.; Silveira, A.K.; Zeidán-Chuliá, F.; Correa, A.P.F.; Oliveria, R.R.; Borges, A.G.; Grun, L.; Barbé-Tuana, F.; Zmozinski, A.; Brandelli, A.; et al. Effects of Achyrocline satureioides Inflorescence Extracts against Pathogenic Intestinal Bacteria: Chemical Characterization, In Vitro Tests, and In Vivo Evaluation. Evid.-Based Complement. Altern. Med. 2017, 2017, 4874865. [Google Scholar] [CrossRef]
  9. Casero, C.; Machín, F.; Méndez-Álvarez, S.; Demo, M.; Ravelo, A.G.; Pérez-Hernández, N.; Joseph-Nathan, P.; Estévez-Braun, A. Structure and Antimicrobial Activity of Phloroglucinol Derivatives from Achyrocline satureioides. J. Nat. Prod. 2015, 78, 93–102. [Google Scholar] [CrossRef]
  10. Voynikov, Y. Phloroglucinol α-Pyrones from Helichrysum: A Review on Structural Diversity, Plant Distribution and Isolation. Plants 2025, 14, 3460. [Google Scholar] [CrossRef]
  11. Opitz, L.; Hänsel, R. Helipyron, ein methylen-bis-triacetsäurelacton aus Helichrysum italicum. Tetrahedron Lett. 1970, 11, 3369–3370. [Google Scholar] [CrossRef]
  12. Vrkoč, J.; Dolejš, L.; Buděšínský, M. Methylene-bis-2H-pyran-2-ones and phenolic constituents from the root of Helichrysum arenarium. Phytochemistry 1975, 14, 1383–1384. [Google Scholar] [CrossRef]
  13. Schmeda-Hirschmann, G. The constituents of Achyrocline satureioides DC. Rev. Latinoam. Quím. 1984, 15, 134–135. [Google Scholar]
  14. Appendino, G.; Ottino, M.; Marquez, N.; Bianchi, F.; Giana, A.; Ballero, M.; Sterner, O.; Fiebich, B.L.; Munoz, E. Arzanol, an anti-inflammatory and anti-HIV-1 phloroglucinol α-pyrone from Helichrysum italicum ssp. microphyllum. J. Nat. Prod. 2007, 70, 608–612. [Google Scholar] [CrossRef]
  15. Bauer, J.; Koeberle, A.; Dehm, F.; Pollastro, F.; Appendino, G.; Northoff, H.; Rossi, A.; Sautebin, L.; Werz, O. Arzanol, a prenylated heterodimeric phloroglucinyl pyrone, inhibits eicosanoid biosynthesis and exhibits anti-inflammatory efficacy in vivo. Biochem. Pharmacol. 2011, 81, 259–268. [Google Scholar] [CrossRef]
  16. Rosa, A.; Atzeri, A.; Nieddu, M.; Appendino, G. New insights into the antioxidant activity and cytotoxicity of arzanol and effect of methylation on its biological properties. Chem. Phys. Lipids 2017, 205, 55–64. [Google Scholar] [CrossRef]
  17. Rosa, A.; Deiana, M.; Atzeri, A.; Corona, G.; Incani, A.; Melis, M.P.; Appendino, G.; Dessì, M.A. Evaluation of the antioxidant and cytotoxic activity of arzanol, a prenylated α-pyrone–phloroglucinol etherodimer from Helichrysum italicum subsp. microphyllum. Chem.-Biol. Interact. 2007, 165, 117–126. [Google Scholar] [CrossRef]
  18. Rosa, A.; Pollastro, F.; Atzeri, A.; Appendino, G.; Melis, M.P.; Deiana, M.; Incani, A.; Loru, D.; Dessì, M.A. Protective role of arzanol against lipid peroxidation in biological systems. Chem. Phys. Lipids 2011, 164, 24–32. [Google Scholar] [CrossRef]
  19. Voynikov, Y. Arzanol: A Review of Chemical Properties and Biological Activities. Plants 2025, 14, 3474. [Google Scholar] [CrossRef]
  20. Taglialatela-Scafati, O.; Pollastro, F.; Chianese, G.; Minassi, A.; Gibbons, S.; Arunotayanun, W.; Mabebie, B.; Ballero, M.; Appendino, G. Antimicrobial phenolics and unusual glycerides from Helichrysum italicum subsp. microphyllum. J. Nat. Prod. 2013, 76, 346–353. [Google Scholar] [CrossRef]
  21. Rios, J.; Recio, M.; Villar, A. Isolation and identification of the antibacterial compounds from Helichrysum stoechas. J. Ethnopharmacol. 1991, 33, 51–55. [Google Scholar] [CrossRef]
  22. Lavault, M.; Richomme, P. Constituents of Helichrysum stoechas variety olonnense. Chem. Nat. Compd. 2004, 40. [Google Scholar] [CrossRef]
  23. Judzentiene, A.; Budiene, J.; Nedveckyte, I.; Garjonyte, R. Antioxidant and toxic activity of Helichrysum arenarium (L.) Moench and Helichrysum italicum (Roth) G. Don essential oils and extracts. Molecules 2022, 27, 1311. [Google Scholar] [CrossRef]
  24. Joray, M.B.; del Rollán, M.R.; Ruiz, G.M.; Palacios, S.M.; Carpinella, M.C. Antibacterial Activity of Extracts from Plants of Central Argentina—Isolation of an Active Principle from Achyrocline satureioides. Planta Med. 2011, 77, 95–100. [Google Scholar] [CrossRef]
  25. Carney, J.R.; Krenisky, J.M.; Williamson, R.T.; Luo, J. Achyrofuran, a New Antihyperglycemic Dibenzofuran from the South American Medicinal Plant Achyrocline satureioides. J. Nat. Prod. 2002, 65, 203–205. [Google Scholar] [CrossRef]
  26. Casero, C.; Estévez-Braun, A.; Ravelo, Á.G.; Demo, M.; Méndez-Álvarez, S.; Machín, F. Achyrofuran is an antibacterial agent capable of killing methicillin-resistant vancomycin-intermediate Staphylococcus aureus in the nanomolar range. Phytomedicine 2013, 20, 133–138. [Google Scholar] [CrossRef]
  27. Wang, S.; Zhang, D.; Zhu, J.; Liu, H.; Li, B.; Huang, L. Achyrophenols A–F: Polycyclic Polyphenol Lactone Skeletons and a Nor-Ursane-Type Triterpenoid from Achyrocline satureioides. J. Org. Chem. 2021, 86, 12813–12820. [Google Scholar] [CrossRef]
  28. Silva, L.; Rodrigues, A.M.; Ciriani, M.; Falé, P.L.V.; Teixeira, V.; Madeira, P.; Machuqueiro, M.; Pacheco, R.; Florêncio, M.H.; Ascensão, L.; Serralheiro, M.L.M. Antiacetylcholinesterase activity and docking studies with chlorogenic acid, cynarin and arzanol from Helichrysum stoechas (Lamiaceae). Med. Chem. Res. 2017, 26, 2942–2950. [Google Scholar] [CrossRef]
  29. Kenig, S.; Kramberger, K.; Šik Novak, K.; Karnjuš, I.; Bandelj, D.; Petelin, A.; Jenko Pražnikar, Z. Helichrysum italicum (Roth) G. Don and Helichrysum arenarium (L.) Moench infusions in reversing the traits of metabolic syndrome: a double-blind randomized comparative trial. Food Funct. 2022, 13, 7697–7706. [Google Scholar] [CrossRef]
  30. Petelin, A.; Šik Novak, K.; Hladnik, M.; Bandelj, D.; Baruca Arbeiter, A.; Kramberger, K.; Kenig, S.; Jenko Pražnikar, Z. Helichrysum italicum (Roth) G. Don and Helichrysum arenarium (L.) Moench Infusion Consumption Affects the Inflammatory Status and the Composition of Human Gut Microbiota in Patients with Traits of Metabolic Syndrome: A Randomized Comparative Study. Foods 2022, 11, 3277. [Google Scholar] [CrossRef]
  31. Nalkiran, I.; Sevim Nalkiran, H. Phytochemical Profile and Anticancer Potential of Helichrysum arenarium Extracts on Glioblastoma, Bladder Cancer, and Breast Cancer Cells. Pharmaceuticals 2025, 18, 144. [Google Scholar] [CrossRef]
  32. Jabłońska, J.; Obrębski, M.; Kiełkiewicz, R.M.; Macias, I.; Bogusz, J.; Skowrońska, W.; Sykłowska-Baranek, K. Antioxidant and Anti-Aging Activity of Helichrysum arenarium (L.) Moench: Comparative Evaluation of Extracts from Shoots, Inflorescences, Plantlets and Callus Tissue. Molecules 2026, 31, 2625. [Google Scholar] [CrossRef]
  33. González, M.J.; Beoletto, V.G.; Agnese, A.M.; Audisio, M.C.; Marioli, J.M. Purification of Substances from Achyrocline satureioides with Inhibitory Activity Against Paenibacillus larvae, the Causal Agent of American Foulbrood in Honeybees’ Larvae. Appl. Biochem. Biotechnol. 2015, 175, 3349–3359. [Google Scholar] [CrossRef]
  34. Lu, A.W.; Peng, F.H.; Liu, Z.Q.; Hu, J.G.; Chen, J.; Min, J.X.; Li, B. UPLC-Q-TOF-MS/MS-based guided isolation of phloroglucinols from Achyrocline satureioides and anti-lung cancer activity. Zhongguo Zhong Yao Za Zhi 2026, 51, 4030–4039. [Google Scholar] [CrossRef]
  35. Voynikov, Y.; Marinov, T.; Nedialkov, P. Comprehensive UHPLC–Orbitrap–MS Profiling of Phloroglucinol α-Pyrones in Helichrysum italicum (Roth) G. Don (Asteraceae). Plants 2026, 15, 2619. [Google Scholar] [CrossRef]
  36. Schymanski, E.L.; Jeon, J.; Gulde, R.; Fenner, K.; Ruff, M.; Singer, H.P.; Hollender, J. Identifying Small Molecules via High Resolution Mass Spectrometry: Communicating Confidence. Environ. Sci. Technol. 2014, 48, 2097–2098. [Google Scholar] [CrossRef]
  37. Werner, J.; Ebrahim, W.; Özkaya, F.C.; Mándi, A.; Kurtán, T.; El-Neketi, M.; Liu, Z.; Proksch, P. Pyrone derivatives from Helichrysum italicum. Fitoterapia 2019, 133, 80–84. [Google Scholar] [CrossRef]
  38. Hänsel, R.; Cybulski, E.-M.; Çubukçu, B.; Meriçli, A.H.; Bohlmann, F.; Zdero, C. Neue pyron-derivate aus Helichrysum-arten. Phytochemistry 1980, 19, 639–644. [Google Scholar] [CrossRef]
  39. Chambers, M.C.; Maclean, B.; Burke, R.; Amodei, D.; Ruderman, D.L.; Neumann, S.; Gatto, L.; Fischer, B.; Pratt, B.; Egertson, J. A cross-platform toolkit for mass spectrometry and proteomics. Nat. Biotechnol. 2012, 30, 918–920. [Google Scholar] [CrossRef]
  40. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2026. [Google Scholar]
  41. Bengtsson, H. A Unifying Framework for Parallel and Distributed Processing in R using Futures. R J. 2021, 13, 208–227. [Google Scholar] [CrossRef]
  42. Barrett, T.; Dowle, M.; Srinivasan, A.; Gorecki, J.; Chirico, M.; Hocking, T.; Schwendinger, B.; Krylov, I. Data.table Ext. ‘data.frame’, R Package Version [1.18.x] 2026. [CrossRef]
  43. Berkelaar, M. lpSolve: Interface to ‘Lp_solve’ v. 5.5 to Solve Linear/Integer Programs, R package version 5.6.23. 2024. [CrossRef]
  44. Landrum, G.; Tosco, P.; Kelley, B.; Rodriguez, R.; Cosgrove, D.; Vianello, R.; Sriniker; Gedeck, P.; Jones, G.; Schneider, N. RDKit Open-Source Cheminformatics, Release 2024.03.4 2024. [CrossRef]
  45. Python Software, F. Python Lang. Ref.>, Version 3.12 2024. [CrossRef]
  46. Weininger, D. SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules. J. Chem. Inf. Comput. Sci. 1988, 28, 31–36. [Google Scholar] [CrossRef]
  47. Heller, S.R.; McNaught, A.; Pletnev, I.; Stein, S.; Tchekhovskoi, D. InChI, the IUPAC International Chemical Identifier. J. Cheminform. 2015, 7, 23. [Google Scholar] [CrossRef]
  48. Hunter, J.D. Matplotlib: A 2D graphics environment. Comput. Sci. Eng. 2007, 9, 90–95. [Google Scholar] [CrossRef]
  49. Oksanen, J.; Simpson, G.L.; Blanchet, F.G.; Kindt, R.; Legendre, P.; Minchin, P.R.; O’Hara, R.B.; Solymos, P.; Stevens, M.H.H.; Szoecs, E. Vegan Community Ecol. Packag.>, R Package Version 2.7-5 2026. [CrossRef]
  50. Kaufman, L.; Rousseeuw, P.J. Finding Groups in Data: An Introduction to Cluster Analysis; Wiley: New York, NY, USA, 1990. [Google Scholar] [CrossRef]
  51. Maechler, M.; Rousseeuw, P.; Struyf, A.; Hubert, M.; Hornik, K. cluster: Cluster Analysis Basics and Extensions, R package version [2.1.8] 2026. [CrossRef]
  52. Rousseeuw, P.J. Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. J. Comput. Appl. Math. 1987, 20, 53–65. [Google Scholar] [CrossRef]
  53. Kolde, R. pheatmap: Pretty Heatmaps, R package version 1.0.13. 2025. [CrossRef]
  54. Kassambara, A.; Mundt, F. factoextra: Extract and Visualize the Results of Multivariate Data Analyses, R package version 1.0.7. 2020. [CrossRef]
  55. Galili, T. dendextend: an R package for visualizing, adjusting and comparing trees of hierarchical clustering. Bioinformatics 2015, 31, 3718–3720. [Google Scholar] [CrossRef]
  56. de Vries, A.; Ripley, B.D. ggdendro: Create Dendrograms and Tree Diagrams Using ‘ggplot2’, R package version 0.2.0 2024. [CrossRef]
  57. Wickham, H. ggplot2: Elegant Graphics for Data Analysis; Springer-Verlag: New York, NY, USA, 2016. [Google Scholar] [CrossRef]
  58. Legendre, P.; Gallagher, E.D. Ecologically meaningful transformations for ordination of species data. Oecologia 2001, 129, 271–280. [Google Scholar] [CrossRef]
  59. Hubert, L.; Arabie, P. Comparing partitions. J. Classif. 1985, 2, 193–218. [Google Scholar] [CrossRef]
  60. Schauberger, P.; Walker, A. openxlsx: Read, Write and Edit xlsx Files, R package version 4.2.9 2026. [CrossRef]
  61. Wickham, H. stringr: Simple, Consistent Wrappers for Common String Operations, R package version 1.5.1 2023. [CrossRef]
  62. Gazoni, E.; Clark, C. openpyxl: A Python library to read/write Excel 2010 xlsx/xlsm files, version 3.1.5 2024. [CrossRef]
Preprints 239626 g001
Figure 1. Representatives of each subclass of the PG α-pyr derivatives described in this paper: (a) open-chain PG α-pyrone, arzanol (PG18; 6-ethyl-α-pyrone, 3-prenyl side chain, acetyl group); (b) benzofuran, italipyrone (BF2; the prenyl chain closed to a 2-isopropenyl-2,3-dihydrobenzofuran); (c) benzopyran, plicatipyrone (BP2; the prenyl chain closed to a 3-hydroxy-2,2-dimethylchroman); (d) spiroketal, helispiroketal A (SK1; a furan-3-one spiro-fused to the dihydrobenzofuran in place of the α-pyrone); (e) dipyrone, helipyrone A (DP3; two 6-ethyl-α-pyrones joined by the methylene bridge). The structures of all 91 substitution patterns of this study are in Supplementary Figure S2.
Figure 1. Representatives of each subclass of the PG α-pyr derivatives described in this paper: (a) open-chain PG α-pyrone, arzanol (PG18; 6-ethyl-α-pyrone, 3-prenyl side chain, acetyl group); (b) benzofuran, italipyrone (BF2; the prenyl chain closed to a 2-isopropenyl-2,3-dihydrobenzofuran); (c) benzopyran, plicatipyrone (BP2; the prenyl chain closed to a 3-hydroxy-2,2-dimethylchroman); (d) spiroketal, helispiroketal A (SK1; a furan-3-one spiro-fused to the dihydrobenzofuran in place of the α-pyrone); (e) dipyrone, helipyrone A (DP3; two 6-ethyl-α-pyrones joined by the methylene bridge). The structures of all 91 substitution patterns of this study are in Supplementary Figure S2.
Preprints 239626 g001
Preprints 239626 sch001
Scheme 1. Fragmentation of arzanol (PG18; [M−H]− at m/z 401.1606, C22H25O7−) under negative-ion HCD, reproduced from [35] (CC BY 4.0). Cleavage of the methylene bridge on the α-pyrone side releases the α-pyrone as the neutral C8H10O3 (154.0630) or as the anion C8H9O3− (m/z 153.0557), which decarboxylates to m/z 109.0659, and leaves Fragment A (PG–CH2, C14H15O4−, m/z 247.0976); cleavage on the phloroglucinol side releases the methylene-α-pyrone C9H10O3 (166.0630) and leaves Fragment B (PG, C13H15O4−, m/z 235.0976). Fragments A and B lose the acetyl ketene (m/z 205.0870 and 193.0870, the class ions of the 3-prenyl compounds), the prenyl chain (m/z 179.0350 from A; 166.0272 from B, as a radical anion), CO, CO2 or C2H4 (m/z 219.1016, 203.1060 and 219.0652 from A; 191.1070 from B, followed by the loss of C2H2O or C5H8 to m/z 149.0956 and 123.0436); the acetyl group gives the ion C4H3O2− (m/z 83.0138).
Scheme 1. Fragmentation of arzanol (PG18; [M−H]− at m/z 401.1606, C22H25O7−) under negative-ion HCD, reproduced from [35] (CC BY 4.0). Cleavage of the methylene bridge on the α-pyrone side releases the α-pyrone as the neutral C8H10O3 (154.0630) or as the anion C8H9O3− (m/z 153.0557), which decarboxylates to m/z 109.0659, and leaves Fragment A (PG–CH2, C14H15O4−, m/z 247.0976); cleavage on the phloroglucinol side releases the methylene-α-pyrone C9H10O3 (166.0630) and leaves Fragment B (PG, C13H15O4−, m/z 235.0976). Fragments A and B lose the acetyl ketene (m/z 205.0870 and 193.0870, the class ions of the 3-prenyl compounds), the prenyl chain (m/z 179.0350 from A; 166.0272 from B, as a radical anion), CO, CO2 or C2H4 (m/z 219.1016, 203.1060 and 219.0652 from A; 191.1070 from B, followed by the loss of C2H2O or C5H8 to m/z 149.0956 and 123.0436); the acetyl group gives the ion C4H3O2− (m/z 83.0138).
Preprints 239626 sch001
Preprints 239626 sch002
Scheme 2. Workflow of the study. The acetone extracts of the three species were analysed by UHPLC–Orbitrap–MS, and every MS2 scan was read by the fragmentation rules established for H. italicum (Section 3.6). The scans of one chromatographic peak were merged into a feature (Section 3.7); the features retained under the five conditions of Section 3.10 were merged into compounds and matched across the species (Section 3.13), and the compounds of the three species were pooled with the 58 PG α-pyr published for H. italicum. Dashed boxes, the inputs taken from the previous study and the dictionary of expected compounds (Section 3.9); the frame marks the automated part of the workflow.
Scheme 2. Workflow of the study. The acetone extracts of the three species were analysed by UHPLC–Orbitrap–MS, and every MS2 scan was read by the fragmentation rules established for H. italicum (Section 3.6). The scans of one chromatographic peak were merged into a feature (Section 3.7); the features retained under the five conditions of Section 3.10 were merged into compounds and matched across the species (Section 3.13), and the compounds of the three species were pooled with the 58 PG α-pyr published for H. italicum. Dashed boxes, the inputs taken from the previous study and the dictionary of expected compounds (Section 3.9); the frame marks the automated part of the workflow.
Preprints 239626 sch002
Preprints 239626 g002
Figure 2. The PG α-pyr of the four species by subclass: (a) number of compounds, with the total of each species at the end of the bar; (b) the same as the share of the compounds of the species. The compounds of H. italicum are those published in [35]; the compounds of the other three species are the retained features of this study after the merging of split peaks (Table 1).
Figure 2. The PG α-pyr of the four species by subclass: (a) number of compounds, with the total of each species at the end of the bar; (b) the same as the share of the compounds of the species. The compounds of H. italicum are those published in [35]; the compounds of the other three species are the retained features of this study after the merging of split peaks (Table 1).
Preprints 239626 g002
Preprints 239626 g003
Figure 3. The 215 PG α-pyrones detected in the four species, by the species in which each was detected. Each compound is counted once, in the bar of the exact combination of species in which it was detected (black dots; grey, not detected), so that the bars add up to 215; a compound detected in all four species is counted in the first bar only. The columns are grouped by the number of species and ordered by size within each group; the horizontal bars on the left give the number of compounds detected in each species. The compounds of H. italicum are those published in [35].
Figure 3. The 215 PG α-pyrones detected in the four species, by the species in which each was detected. Each compound is counted once, in the bar of the exact combination of species in which it was detected (black dots; grey, not detected), so that the bars add up to 215; a compound detected in all four species is counted in the first bar only. The columns are grouped by the number of species and ordered by size within each group; the horizontal bars on the left give the number of compounds detected in each species. The compounds of H. italicum are those published in [35].
Preprints 239626 g003
Preprints 239626 g004
Figure 4. Fragment-ion heatmaps per subclass (Bray–Curtis dissimilarity, average linkage; rows cut at the number of clusters with the largest average silhouette width; rows labelled by compound identifier, [M−H]− and retention time; columns, the 40 fragment ions present in at least two compounds at ≥ 1%, ordered by their own clustering; colour bars, the substituents that vary within the subclass): (a) the 13 3-prenyl PG α-pyr of H. arenarium (six clusters), which group by their acyl group through the shifted Fragment B/A pairs; (b) the 19 BPs of H. stoechas (seven clusters), which group by the α-pyrone C6 substituent and the acyl group.
Figure 4. Fragment-ion heatmaps per subclass (Bray–Curtis dissimilarity, average linkage; rows cut at the number of clusters with the largest average silhouette width; rows labelled by compound identifier, [M−H]− and retention time; columns, the 40 fragment ions present in at least two compounds at ≥ 1%, ordered by their own clustering; colour bars, the substituents that vary within the subclass): (a) the 13 3-prenyl PG α-pyr of H. arenarium (six clusters), which group by their acyl group through the shifted Fragment B/A pairs; (b) the 19 BPs of H. stoechas (seven clusters), which group by the α-pyrone C6 substituent and the acyl group.
Preprints 239626 g004
Preprints 239626 g005
Figure 5. Semi-quantitative PG α-pyr content of the four species by subclass: (a) sum of the compounds above the LOQ, µg/g dry weight as THAP equivalents (uncorrected for recovery), with the total of each species; (b) the same as the share of the total PG α-pyr content of the species. The values of H. italicum are the published ones [35].
Figure 5. Semi-quantitative PG α-pyr content of the four species by subclass: (a) sum of the compounds above the LOQ, µg/g dry weight as THAP equivalents (uncorrected for recovery), with the total of each species; (b) the same as the share of the total PG α-pyr content of the species. The values of H. italicum are the published ones [35].
Preprints 239626 g005
Preprints 239626 g006
Figure 6. Calibration of 2′,4′,6′-trihydroxyacetophenone: MS1 peak area against the concentration in the injected vial, six levels from 0.000333 to 1 µg/mL (logarithmic axis), 1/x2 weighted linear regression (A = 4.66 × 104 + 4.37 × 108·C, weighted R2 = 0.9973). Back-calculated concentrations of the six levels: 97.1, 108.5, 101.2, 100.0, 98.9 and 94.2% of nominal.
Figure 6. Calibration of 2′,4′,6′-trihydroxyacetophenone: MS1 peak area against the concentration in the injected vial, six levels from 0.000333 to 1 µg/mL (logarithmic axis), 1/x2 weighted linear regression (A = 4.66 × 104 + 4.37 × 108·C, weighted R2 = 0.9973). Back-calculated concentrations of the six levels: 97.1, 108.5, 101.2, 100.0, 98.9 and 94.2% of nominal.
Preprints 239626 g006
Table 1. Outcome of the automated annotation of the three species analysed.
Table 1. Outcome of the automated annotation of the three species analysed.
Species Features with a PG α-pyr pattern Retained features High Medium Compounds PG BF BP SK DP
A. satureioides 102 39 19 20 32 14 7 7 2 2
H. arenarium 480 206 93 113 178 98 21 44 14 1
H. stoechas 160 78 25 53 55 24 5 19 4 3
H. italicum (published) – – – – 58 33 8 10 5 2
Pool of the four species – – – – 215 113 29 54 16 3
The high-confidence count of a species is the number of retained features graded high; a compound built from several features carries the better grade. The 58 compounds of H. italicum are the manual annotation of the previous study, matched to the pool by substitution pattern, formula and retention time (Section 3.13).
Table 2. PG α-pyr per structural subclass in the four species: number of compounds (in parentheses, for the three species analysed here, the number of retained features behind them).
Table 2. PG α-pyr per structural subclass in the four species: number of compounds (in parentheses, for the three species analysed here, the number of retained features behind them).
Subclass A. satureioides H. arenarium H. italicum H. stoechas Four species
3-NA PG α-pyr 2 (2) 12 (12) 9 8 (9) 16
3-prenyl PG α-pyr 6 (9) 13 (14) 14 4 (5) 18
3-(hydroxyprenyl) PG α-pyr 2 (2) 7 (7) 2 0 (0) 7
3-(hydroxydihydroprenyl) PG α-pyr 1 (1) 2 (2) 2 0 (0) 4
3-geranyl PG α-pyr 3 (3) 19 (26) 3 9 (14) 22
3-(hydroxygeranyl) PG α-pyr 0 (0) 45 (49) 3 3 (3) 46
benzofurans 7 (8) 21 (27) 8 5 (5) 29
benzopyrans 7 (8) 44 (49) 10 19 (34) 54
spiroketals 2 (3) 14 (19) 5 4 (5) 16
dipyrones 2 (3) 1 (1) 2 3 (3) 3
All 32 (39) 178 (206) 58 55 (78) 215
Table 3. Frequency of the individual substituents among the PG α-pyr of the four species (number of compounds in which the substituent was assigned).
Table 3. Frequency of the individual substituents among the PG α-pyr of the four species (number of compounds in which the substituent was assigned).
Position Substituent A. satureioides H. arenarium H. italicum H. stoechas Four species
C6 (α-pyrone) ethyl 27 150 27 33 159
C6 (α-pyrone) methyl 1 11 22 14 34
C6 (α-pyrone) propyl 0 2 2 1 3
C6 (α-pyrone) methyl + methyl 0 0 0 1 1
C6 (α-pyrone) methyl + ethyl 1 0 1 1 1
C6 (α-pyrone) ethyl + ethyl 1 1 1 1 1
C3 (side chain) 3-(hydroxygeranyl) 0 45 3 3 46
C3 (side chain) cyclised prenyl 11 27 13 0 32
C3 (side chain) cyclised dehydro-geranyl 1 17 1 15 25
C3 (side chain) 3-geranyl 3 19 3 9 22
C3 (side chain) 3-prenyl 6 13 14 4 18
C3 (side chain) 3-NA 2 12 9 8 16
C3 (side chain) cyclised (hydroxy)geranyl 2 14 4 4 16
C3 (side chain) cyclised geranyl 0 7 0 5 10
C3 (side chain) prenyl-derived 2 8 4 0 9
C3 (side chain) 3-(hydroxyprenyl) 2 7 2 0 7
C3 (side chain) geranyl-derived 0 6 1 4 7
C3 (side chain) 3-(hydroxydihydroprenyl) 1 2 2 0 4
C1 (acyl) acetyl 2 50 32 34 71
C1 (acyl) pentanoyl 14 37 9 4 44
C1 (acyl) butanoyl 11 38 3 5 40
C1 (acyl) propanoyl 1 37 3 9 38
C1 (acyl) hexanoyl 2 7 7 0 10
C1 (acyl) 3-phenylpropanoyl 0 5 1 0 5
C1 (acyl) heptanoyl 0 2 1 0 3
C1 (acyl) octanoyl 0 1 0 0 1
Spiroketals carry no α-pyrone and DPs no acylphloroglucinol unit; substituents that could not be determined are not counted.
Table 5. The compounds detected in all four species and in three of them.
Table 5. The compounds detected in all four species and in three of them.
ID Subclass C6 C3 C1 RT Species Name
PG18 open-chain ethyl 3-prenyl acetyl 10.43 A, H, I, S arzanol
DP3 dipyrone ethyl + ethyl – – 11.23 A, H, I, S helipyrone A
PG33 open-chain ethyl 3-NA pentanoyl 13.16 A, H, I, S –
PG51 open-chain ethyl 3-prenyl butanoyl 15.36 A, H, I, S 6-O-desmethylauricepyrone/3-prenyl norauricepyrone
PG62 open-chain ethyl 3-prenyl pentanoyl 17.23 A, H, I, S 23-methyl-6-O-desmethylauricepyrone
PG4 open-chain ethyl 3-NA acetyl 6.78 H, I, S –
DP2 dipyrone methyl + ethyl – – 8.74 A, I, S helipyrone B
PG23 open-chain ethyl 3-NA butanoyl 11.35 H, I, S –
BP10 benzopyran ethyl cyclised prenyl pentanoyl 12.35 A, H, I –
PG36 open-chain ethyl 3-prenyl propanoyl 13.37 H, I, S –
BP14 benzopyran ethyl cyclised (hydroxy)geranyl acetyl 13.64 H, I, S –
BP20 benzopyran ethyl cyclised dehydro-geranyl acetyl 15.19 H, I, S –
PG50 open-chain ethyl 3-NA hexanoyl 15.36 A, H, I –
PG57 open-chain methyl 3-geranyl acetyl 16.18 H, I, S 18,18-bis-desmethyl achyroclinopyrone C
BP24 benzopyran ethyl cyclised (hydroxy)geranyl acetyl 16.42 H, I, S –
PG69 open-chain ethyl 3-geranyl acetyl 18.33 A, I, S 18,18-bis-desmethyl achyroclinopyrone A
SK11 spiroketal – geranyl-derived acetyl 18.36 H, I, S –
PG97 open-chain ethyl 3-geranyl butanoyl 23.25 A, H, S –
Species: A, A. satureioides; H, H. arenarium; I, H. italicum; S, H. stoechas. RT, mean retention time (min); C6, C3, C1, substitution pattern as in Table 4.
Table 6. Validation of the MS/MS-based clustering against the assigned substitution patterns.
Table 6. Validation of the MS/MS-based clustering against the assigned substitution patterns.
Set FI n FI k ARI scaffold Purity scaffold ARI C6 ARI C3 ARI pattern NL n NL k ARI scaffold Purity scaffold ARI C6 ARI C3 ARI FI vs NL
A. satureioides 38 12 0.22 0.84 −0.01 0.23 0.56 39 2 0.12 0.51 n/a 0.09 0.03
H. arenarium 210 18 0.23 0.84 0.03 0.29 0.20 203 2 0.18 0.60 0.44 0.06 0.05
H. stoechas 79 10 0.36 0.89 0.25 0.42 0.31 75 10 0.43 0.88 0.28 0.21 0.30
three species together 327 5 0.45 0.76 0.20 0.19 0.07 319 3 0.15 0.56 0.87 0.05 0.25
H. italicum (published) 58 – 0.14 0.95 0.04 0.20 – 58 – 0.42 0.95 0.26 0.03 –
n, number of compounds entering the analysis; k, number of clusters of the partitioning around medoids (PAM); ARI, adjusted Rand index, and purity against the scaffold; ARI against the α-pyrone C6 substituent, the C3 side chain and the full substitution pattern, for the fragment-ion (FI) and the neutral-loss (NL) matrix; last column, ARI between the FI and NL partitions. The row of H. italicum gives the published values of the previous study (best-performing method per level; ARI against the subclass, C6 and C3; no value for the full pattern or for FI vs NL). Three species together: A. satureioides, H. arenarium and H. stoechas. The analysis was run on the retained features before the merging of split peaks and before the removal of the eight SK features without a resolved core (Section 2.2), so that n differs from Table 1. Not shown: the purity of the neutral-loss clusters of the three species together against C6 is 0.97; the ARI against the C1 acyl group is between −0.02 and 0.12 in every set; the ARI against the species is 0.06 for both matrices. n/a: the 36 compounds of A. satureioides with a C6 substituent fall into a single neutral-loss cluster, for which the ARI is undefined.
Table 7. Semi-quantitative content of the PG α-pyr of each species, expressed as 2′,4′,6′-trihydroxyacetophenone (THAP) equivalents: number of compounds, number above the LOQ (0.033 µg/g DW); total content (µg/g dry weight, uncorrected; % dry weight after correction for the 80% recovery); the most abundant compound with its identifier, content and share of the total PG α-pyr content of the species, and the share of the ten most abundant compounds in that total. The row of H. italicum is calculated from the published contents [35].
Table 7. Semi-quantitative content of the PG α-pyr of each species, expressed as 2′,4′,6′-trihydroxyacetophenone (THAP) equivalents: number of compounds, number above the LOQ (0.033 µg/g DW); total content (µg/g dry weight, uncorrected; % dry weight after correction for the 80% recovery); the most abundant compound with its identifier, content and share of the total PG α-pyr content of the species, and the share of the ten most abundant compounds in that total. The row of H. italicum is calculated from the published contents [35].
Species Compounds > LOQ Total, µg/g Total, % DW Most abundant compound µg/g % of the total Ten most abundant, % of the total
A. satureioides 32 12 3.2 0.0004 DP3 — helipyrone A 2.04 64 98
H. arenarium 178 99 77.6 0.0097 PG18 — arzanol 15.38 20 66
H. italicum (published) 58 58 152.9 0.0191 PG18 — arzanol 43.20 28 80
H. stoechas 55 17 8.4 0.0010 BP37 — benzopyran | ethyl | cyclised dehydro-geranyl | acetyl 4.59 55 95
Table 8. The three LC–HRMS runs of the study: species, run identifier, supplier of the plant material, mass of dried material extracted, dry residue and extraction yield, and preparation scheme (volume of methanol in which the residue was taken up; dilution before injection; injection volume). The H. italicum extract of the previous study [35] was prepared by the same scheme from 1.001 g of dried inflorescences (85.4 mg of residue, 8.5%) and injected at 1 µL.
Table 8. The three LC–HRMS runs of the study: species, run identifier, supplier of the plant material, mass of dried material extracted, dry residue and extraction yield, and preparation scheme (volume of methanol in which the residue was taken up; dilution before injection; injection volume). The H. italicum extract of the previous study [35] was prepared by the same scheme from 1.001 g of dried inflorescences (85.4 mg of residue, 8.5%) and injected at 1 µL.
Species Run Material Mass (g) Residue (mg) Yield (%) Scheme
A. satureioides Asat Diochi, Prague 1.0002 33.4 3.3 10 mL; 1:10; 5 µL
H. arenarium Hare Greekherbay 1.0009 90.7 9.1 10 mL; 1:10; 5 µL
H. stoechas Hsto Theodoros Pervoli 1.0004 31.1 3.1 10 mL; 1:10; 5 µL
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.