Investigation of Secondary Metabolites and Their Bioactive Potential in Various Iris Species and Cultivars Grown under Different Cultivation Conditions

Investigation of Secondary Metabolites and Their Bioactive Potential in Various Iris Species and Cultivars Grown under Different Cultivation Conditions Tereza Jaegerova, Jitka Viktorova, Marie Zlechovcova, Miroslav Vosatka, Petr Kastanek, and Jana Hajslova* Cite This: ACS Omega 2025, 10, 51256−51271 Read Online ACCESS Metrics & More Article Recommendations * sı Supporting Information ABSTRACT: This study analyzed the bioactive secondary metabolites in leaves and roots with rhizomes of various Iris species and cultivars (subgenus Iris and Limniris) grown under different conditions. Plants were cultivated in aeroponics and hydroponics and treated with a feather hydrolysate as an alternative nutrient source or arbuscular mycorrhizal fungi (AMF) to promote metabolite production. Phytochemical profiling was performed by using UHPLC-HRMS/MS, aided by in silico fragmentation for compound identification. Chemometric analyses revealed differences in the metabolic profiles between plant parts, including upregulation of phenolic acids, steroids, and C-glycosyl xanthones in the leaves and of stilbenes, triterpenoids, and benzophenones in the roots with rhizomes. Variations between the subgenera Iris and Limniris revealed possible chemotaxonomic markers, while cultivars of the same species showed a close similarity. Hydroponics resulted in upregulation of 232 metabolites, including flavone and isoflavonoid glycosides, compared to 42 upregulated metabolites in aeroponics, like xanthones and benzofurans. Neither feather hydrolysate nor AMF significantly altered the profile of detected metabolites. The correlation of the phytochemical profiles with biochemical assays of the Iris extracts indicated metabolites with cytotoxic, antimicrobial, and anti-inflammatory potential, including both known bioactive compounds (e.g., iridin, phalerin, or pcoumaric acid) and compounds with no published bioactivity data (e.g., irisjaponin A). 1. INTRODUCTION The Iris plants are cultivated both for ornamental purposes and industrial applications, with one of the most renowned uses being the production of orris butter, an esteemed substance highly sought after in the fragrance and cosmetics industries. 1,2 In recent years, irises have received considerable attention for their array of biological and therapeutic properties, including antimicrobial, antiviral, antioxidant, anti-inflammatory, antidiabetic, antiosteoporotic, hepatoprotective, phytoestrogenic, anticholinesterase, antihyperglycemic, antihyperlipidemic, anticarcinogenic, antiplasmodial, and molluscicidal activities. 2−6 These diverse attributes render them invaluable resources for medicine as well as various industrial sectors, focusing on diet, health, and beauty. Additionally, Iris extracts find application in the food sector as additives, contributing color and flavor. 1 At present, the species of industrial significance, include Iris pallida,Iris germanica, and Iris Florentina. Nevertheless, the Iris genus, the largest within the Iridaceae family, comprises over 300 species, 7 each with a potentially distinct metabolic profile, bioactive properties, and industrial potential. Yet, the phytochemical profile and bioactivity of many of these species and cultivars remain underexplored or unexplored. A review of the existing literature revealed nearly 300 reported secondary metabolites across various Iris species, many of which are responsible for the biological activities observed in the extracts. These metabolites encompass a wide range of compound groups, including flavonoids, isoflavonoids, phenolic acids, phenols, stilbenes, xanthones, quinones, sterols, steroids, and terpenoids, some of which have been attributed a chemotaxonomic value. 4,8−17 Environmental factors, such as soil composition, nutrient availability, pH, temperature, or humidity, have been shown to influence the production, accumulation, and distribution of secondary metabolites in plants. 18−20 For instance, phosphorus and potassium levels in the soil, along with sunlight exposure, significantly affect the accumulation of phenolic compounds in various Iris species. 19 Received: July 2, 2025 Revised: September 29, 2025 Accepted: October 1, 2025 Published: October 20, 2025 Article http://pubs.acs.org/journal/acsodf © 2025 The Authors. Published by American Chemical Society 51256 https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 This article is licensed under CC-BY 4.0 Downloaded via 212.111.30.228 on November 6, 2025 at 16:02:22 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. These external factors are difficult to control in conventional field cultivation, resulting in variable concentrations of bioactive metabolites in the same plants cultivated across different regions and times. 20 Alternative cultivation methods, such as hydroponics and aeroponics, offer more precise control over environmental variables, allowing for the potential standardization of the bioactive metabolite content. In hydroponics, nutrients are delivered in liquid nutrient-rich media rather than through traditional soil, 20 while aeroponics involves the aerosolized application of nutrients to plant roots and stems suspended in the air. 21 Recently, a trend has emerged to use industrial byproducts as a source of these essential nutrients. An example is chicken feathers, which pose environmentally problematic waste that is challenging to dispose of, but, in the hydrolyzed form, could serve as a low-cost growth medium. 22 While these alternative cultivation approaches undisputably benefit from the sustainable water usage without the need for soil and space saving through vertical cultivation, contradictory results have been reported for different plants on the ability to increase the production of biomass and bioactive metabolites. 23,24 To date, to the best of our knowledge, no studies have compared the metabolic profiles of Iris species cultivated in aeroponic and hydroponic systems. Another factor influencing plant metabolism is the treatment with arbuscular mycorrhizal fungi (AMF), which has been shown to enhance the nutrient uptake, reduce phytotoxicity from heavy metals, increase stress tolerance, and protect against pathogens. 25−29 In I. germanica, AMF inoculation has been linked to the increased accumulation of primary metabolites, such as starch and myristic acid. 30 However, to the best of our knowledge, the influence of AMF on secondary metabolite production in Iris species remains unexplored. Despite extensive research on the chemical composition and biological activity of extracts prepared from various Iris species (e.g., ref 2,4,5,8,10,15 and 31−42), many species and cultivars have until now received little to no attention. Furthermore, many analytical studies tend to focus on a limited range of phytochemicals, failing to capture the full metabolic diversity of these plants and possibly overlooking metabolites of significant biological and industrial value. Reliable compound identification is crucial for enabling further biochemical investigation and ensuring safe and proper industrial applications. Conventional LC-UV/DAD methods generally provide a low identification confidence for compounds in crude plant extracts, unless analytical reference standards are used. Tandem high-resolution mass spectrometry (HRMS/MS) is a technique that produces characteristic fragmentation (MS/MS) spectra of compounds used to deduce the molecular structure. Particularly when coupled to separation techniques, such as liquid chromatography (LC), HRMS/MS offers a more reliable compound identification. However, the fragmentation spectra generated by LC-HRMS/ MS can be rather difficult to compare when obtained under different instrumental conditions. Moreover, the laborintensive process of manually reviewing and comparing MS/ MS spectra to external spectral libraries, together with the lack of MS/MS spectra and the absence of reference standards for many plant metabolites, makes identification challenging. Under these circumstances, in silico fragmentation is a tool that utilizes advanced computational algorithms to predict molecular structures from fragmentation data, enabling the high-throughput screening of hundreds of metabolites with an improved identification confidence. 43 The presented bioprospective study explores the phytochemical profiles of a unique collection of various Iris species and cultivars and provides insights into the effect of different cultivation conditions to foster industrial potential. The implemented innovative UHPLC-HRMS/MS aided by in silico fragmentation allowed the screening of 288 Iris secondary metabolites from an in-house database, representing various chemical classes, including flavonoids and isoflavonoids, phenolic acids, xanthones, terpenoids, steroids, and others. In addition, bioactivity tests were performed on the Iris extracts, and the results were correlated to the metabolic profiles to identify potential bioactive compounds. Key aims were to (i) compare the phytochemical profiles of leaves and roots with rhizomes of various Iris species and cultivars from Figure 1. Phytochemical profile of Iris extracts with the lowest and highest number of detected metabolites. Overlaid extracted ion chromatogram (EIC) of compounds detected in positive and negative ionization modes in the extracts of (a) leaves of the hydroponically grown AMF-treated Iris lactea (sample 114, 103 metabolites) and (b) roots with rhizomes of the hydroponically grown, AMF-treated I. squalens (sample 133, 282 metabolites) based on the in-house metabolite database. ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51257 subgenus Iris and Limniris and identify chemotaxonomic markers; (ii) investigate the effect of aeroponic and hydroponic cultivation systems on the detected metabolites; (iii) assess the effect of AMF inoculation and treatment with a feather hydrolysate during growth on the phytochemical profile; and (iv) identify compounds correlated with antimicrobial, anti-inflammatory, antioxidant, and cytotoxic activity. 2. RESULTS AND DISCUSSION This study employed UHPLC-HRMS/MS to analyze 149 samples of 37 different Iris plants cultivated under four distinct conditions (aeroponics, hydroponics, treatment with a feather hydrolysate, and treatment with AMF). A detailed overview of the samples is provided in Supporting Information, S1. The analysis provided a broad data matrix of a total of 747 detected compounds, matching the exact mass and isotopic profile of known Iris secondary metabolites from our in-house database (Supporting Information, S2) and passing through the MSC filter where applicable, as described in Section 4.6. UHPLCHRMS/MS data processing and compound identification. The number of detected compounds surpassed the 288 metabolites included in the in-house database, primarily owing to the recurrent occurrence of isomeric peaks, which inherently complicates unambiguous compound identification, as explained in Section 4.6. The detected compounds represented predominantly (iso)flavonoids (460), terpenoids (78), and phenols (74), with smaller numbers of phenolic acids (41), xanthones (29), quinones (25), steroids (15), stilbenes (8), and other unclassified compounds (17). As expected and consistent with previous reports, (iso)flavonoids were the most abundant secondary metabolites in Iris. 40,44 The lowest number of targeted metabolites (103) was found in the leaves of the hydroponically grown, AMF-treated Iris lactea (sample 114), while the highest number (282) was detected in the roots with rhizomes of the hydroponically grown, AMF-treated Figure 2. PCA graph of the data set for the analyzed Iris extracts. Plots colored by the (a) cultivation method and (b) subgenus, with highlighted plant part separation. ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51258 Table 1. Chemotaxonomic Markers of Subgenus Iris and Limniris marker compound VIP tentative identification a compound group b chemical formula RT (min) plant part c ionization polarity dominant ion/searched m/ztop fragment ions reference MS/MS d Limniris 8 1.92 2,6,4′-trihydroxy-4-methoxybenzophenone benzophenone C14H12O53.93 L, R NEG [M + HCOO]−/305.0667 125.0258, 137.0243, 167.0338, 221.0457, 121.0672 n.a. 109 1.75 alpinone flavanonol C16H14O511.68 L, R NEG [M −H]−/285.0768 124.0168, 139.0408, 117.0350, 179.0325, 97.0309 n.a. 233 1.74 irilone/irisone B isoflavonoid C16H10O612.72 L, R NEG [M −H]−/297.0405 282.0157, 253.0138, 226.0273, 181.0373, 198.0309 10 irisone B n.a. 108 1.62 alpinone flavanonol C16H14O510.87 L, R NEG [M −H]−/285.0768 119.0500, 116.0194, 93.0337, 65.0013, 97.0290 n.a. 297 1.61 isoorientin/orientin/kaempferol 3-O-galactoside/kaempferol 3-O-glucoside flavone glycoside C21H20O11 7.56 L, R NEG [M −H]−/447.0988 n.a. n.a. Iris 607 1.41 irisjaponin B/irigenin S isoflavonoid C19H18O810.25 L, R POS [M + H]+/375.1074 345.0591, 342.0724, 360.0821, 329.0637, 327.0482 n.a. 98 1.37 9-methoxyirispurinol/eupatorin 12a-hydroxyrotenoid/flavone C18H16O710.16 L, R NEG [M −H]−/343.0823 328.0587, 313.0361, 298.0101, 270.0124, 285.0413 9-methoxyirispurinol n.a., 47 591 1.37 irilone 4′-O-beta-D-glucopyranoside isoflavonoid glycoside C22H20O11 8.48 L, R POS [M + H]+/461.1078 299.0549, 85.0278, 314.0643, 69.0325, 97.0275 10 14 1.19 3,5-dihydroxy-4-(4-hydroxybenzoyl)phenyl hexopyranoside benzophenone glycoside C19H20O10 4.49 L, R NEG [M −H]−/407.0984 245.0435, 287.0517, 151.0004, 125.0199, 193.0116 n.a. 21 1.14 3-O-methylgalangin/acacetin/genkwanin/ izalpinin flavone C16H12O511.54 L, R NEG [M −H]−/293.0612 268.088, 240.0428, 117.0338, 195.0474, 163.0374 MoNA, massbank, izalpinin n.a. a Tentative identification based on the in-house metabolite database and further checked against online spectral data (column Reference MS/MS); isomeric compounds may occur and are separated by “/”. b Proposed isomers can belong to different compound groups separated by “/”. c Occurrence of a compound in the leaves (L) and/or roots with rhizomes (R). d Reference MS/MS spectrum was not available (n.a.) or not searched for (−) due to missing MS2 information for the compound in the experiment. ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51259 Iris squalens (sample 133). The phytochemical profiles of these two samples are illustrated in Figure 1. The signals of the detected metabolites across the different samples and their tentative identifications based on the in-house database are provided in Supporting Information, S3. Given the high complexity of the data due to the multiple experimental variables, namely, Iris subgenus, species, cultivar, plant part, cultivation method, and plant treatment during growth, principal component analysis (PCA) was performed on the complete data set to identify general trends (Figure 2). The first two principal components of the PCA explained 61.6% of the variance, and one of the key findings was the clear separation of samples based on plant part, leaves versus roots with rhizomes. Further distinctions were observed among the leaf samples based on the cultivation method, aeroponic versus hydroponic. Among the roots with rhizomes, Iris versicolor grown in aeroponics did not cluster with the other aeroponic samples, suggesting that other factors may be driving its separation. This led to the identification of an additional grouping, expressed especially within the roots with rhizomes, based on subgenus variation, Iris versus Limniris. The separation of I. versicolor roots with rhizomes from other aeroponic samples likely reflects its classification under the subgenus Limniris, which was the only such representative among the aeroponic samples. This observation hints at the presence of chemotaxonomically significant metabolites in I. versicolor. With respect to plant treatments during growth - feather hydrolysate in aeroponics and AMF in hydroponics - the PCA revealed no substantial alterations in metabolic profiles, as treated and untreated samples clustered closely together. The implications of PCA were further investigated and are discussed within the following sections of this chapter. 2.1. Iris Subgenera, Species, and Cultivars. The Iris species and cultivars analyzed in this study were representatives of the subgenera Iris and Limniris, the former characterized by the presence of a beard on the flower petals, and the latter by a crest. 45 As demonstrated earlier, clustering of samples based on the subgenus was observed, justifying further investigation. A Volcano plot was used to reduce the normalized data matrix from all samples, filtering 504 significant metabolites, 395 downregulated and 109 upregulated in the Iris/Limniris direction (Supporting Information, S4). The supervised Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) applied to the reduced data set achieved a clear separation, with the model achieving an excellent predictive value (Q2= 0.90). Table 1 details 10 compounds with the highest VIP scores, 5 for each subgenus, that were identified as possible chemotaxonomic markers. Of these, 8 compounds were (iso)flavonoids and 2 were phenols, all detected in both leaves and roots with rhizomes. Some of the proposed compounds were previously reported to be bioactive, such as the antioxidant irilone 4′-O-beta-D-glucopyranoside 46 or the anticarcinogenic and anti-inflammatory eupatorin. 47 To explore tissue-specific chemotaxonomic markers, a Volcano plot analysis was conducted within leaves and roots with rhizomes separately (Supporting Information, S4). Compounds that were not detected in a given plant part were excluded from the analysis for that part, as including them could produce artifacts: missing values are typically imputed during preprocessing, which can artificially inflate fold changes and statistical significance. Research carried out until now has highlighted the taxonomic significance of certain (iso)flavonoids with restricted occurrence in Iris. For example, the presence of isoflavonoid aglycones, rather than glycosides, has been noted as a characteristic feature of Limniris. 46,48 This observation was partly confirmed in our study with many of the significant (iso)flavonoid markers of Limniris, including those listed in Table 1, being isoflavonoid aglycones. Nevertheless, some Limniris markers were also suspected to be glycosides, such as compound 203 (iridin). Moreover, certain isoflavonoid aglycones were more characteristic of the subgenus Iris, such as compound 607 (irisjaponin B/irigenin S). Notably, when examining the sum of normalized signals for compounds within individual compound groups, the Volcano plot, included in Supporting Information, S5, revealed that isoflavonoid aglycones were not significant for distinguishing the subgenera, whereas glycosides were upregulated in Iris. This observation was supported by OPLS-DA, which indicated a similar distribution pattern for isoflavonoid aglycones and glycosides, with each group contributing to the differentiation between the subgenera. This suggests that the mere presence or absence of these compounds is not a reliable taxonomic marker for a subgenus classification. On the other hand, benzoquinones and cinnamaldehydes were found to be upregulated in Limniris, with some compounds, like compound 264 (irisoquin A), compound 272 (irisoquin F), and compound 532 (coniferaldehyde), surpassing a VIP score of 1.2. Another distinguishing feature of Limniris was the upregulation of phenolic acids, with compounds 369 (protocatechuic acid) and 137 (caffeic acid) serving as the strongest markers. Additionally, sesquiterpenoids, triterpenoids, steroids, and sterols were also among the significant compound groups for characterizing the Limniris subgenus. The metabolic profiles of different species varied significantly in some cases, primarily due to belonging to different subgenera, as discussed earlier. For example, a strong difference was observed between I. germanica (Iris) and I. versicolor (Limniris). In the various cultivars of I. germanica, (iso)- flavonoids predominated, with some of the most prominent signals attributed to compound 226 (irigenin, confirmed with an analytical reference standard), compound 429 (5,3′,4′,5′- tetramethoxy-6,7-methylenedioxyisoflavone), and compound 578 (iridin). In contrast, the (iso)flavonoid profile in I. versicolor was relatively poor, with an average of 98 (iso)flavonoids detected compared to 148 in I. germanica. A few (iso)flavonoids seemed to be a characteristic of I. versicolor, including compound 233 (irilone/irisone B), one of the Limniris markers, as well as compound 187 (formononetin) and compound 200 (iridin), both of which were only detected in the leaves of I. versicolor. Other characteristic compounds included several tentatively identified terpenoids, such as compound 727 (spirobicyclic triterpenoid), compound 107 (adehydroirigermanal/missourin), compound 290, and 650 (iritectol A/22,23-epoxy-21-hydroxyiridal/iritectol B). These terpenoids were generally absent in I. germanica or showed weak signals. Notably, compounds 290 and 650 were detected exclusively in the aeroponically grown roots with rhizomes of I. versicolor, implying a significant effect of the cultivation method. In addition, I. versicolor displayed elevated levels of cinnamaldehydes and benzoquinones compared with other species. Interestingly, some benzoquinones extracted from Iris were recently linked to allelopathic effects. 49 While some metabolites appeared to be species-specific, 40 metabolites were found across all studied species and cultivars, ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51260 suggesting that they could be ubiquitous within the Iris genus. Among these were compound 395 (tectorigenin, confirmed by an analytical reference standard), compound 593, and compound 734, possibly irilone and tectoridin, respectively, which are considered some of the most representative compounds of the Iris genus [40]. Of these 40 common metabolites, 6 were detected in all samples (including both plant parts, aeroponic, hydroponic, treated, and control samples), though at varying levels. These included compound 593 (irilone/irisone B), compound 632 (irispurinol/quercetin 3,4′-dimethyl ether/iristectorigenin A), compound 429 (5,3′,4′,5′-tetramethoxy-6,7-methylenedioxyisoflavone), compound 731 (stigmasterol), compound 292 (isomangiferin/ nigricanside), and compound 325 (mangiferin, confirmed by an analytical reference standard). Similar to our findings, Williams et al. (1997) reported the presence of mangiferin and isomangiferin in all 37 Iris species and cultivars they studied, 50 and these C-glycosyl xanthones have been described in over 40 Iris species to date. 17 The presented study expands on this list, confirming mangiferin in previously uninvestigated species such as Iris nyaradyana or Iris neglecta. In general, cultivars of the same species exhibited similar metabolic profiles. This was evident from the PCA described in chapter 2, Results and discussion, where some clustering was observed among different cultivars of I. germanica, particularly in the roots with rhizomes. For example, I. germanica Florentina Coerulea (sample 61, 260 targeted metabolites) and I. germanica Florentina Alba (sample 53, 266 targeted metabolites) showed almost identical profiles of detected compounds, as shown in Figure 3. 2.2. Leaves vs Roots with Rhizomes. The chemical profiles of the leaves and roots with rhizomes of the studied Iris species were distinctly different, as illustrated by the PCA described in chapter 2, Results and discussion. Considering the entire sample set, more targeted metabolites were identified in the roots with rhizomes (503) compared with the leaves (438), this trend being true for most species and cultivars, with 194 metabolites common to both plant parts. According to the Volcano plot analysis included in the Supporting Information, S6, 228 metabolites showed an upregulated expression in the leaves, while 278 metabolites were upregulated in the roots with rhizomes. The upregulated compound groups in the leaves included phenolic acids, steroids, C-glycosyl xanthones, and flavone/flavone glycosides. In contrast, compounds, such as stilbenes, triterpenoids, flavanonol glycosides, isoflavonoids, and their glycosides, cinnamaldehydes, and benzophenones were more abundant in the roots with rhizomes. This finding aligns with the literature, which reports that isoflavonoids are primarily found in the rhizomes, while the leaves are rich in apigeninand luteolin-based Cand O-glycosylflavones, along with their derivatives, such as vitexin, isovitexin, orientin, isoorientin, and swertisin. 46 Subsequently, OPLS-DA was performed on the Volcanoreduced data set, generating a robust model capable of predicting the plant part with a Q2value of 0.99. Table 2 presents the 10 metabolites with the highest VIP score: 5 for each plant part. 2.3. Aeroponics vs Hydroponics. As mentioned in chapter 1, Introduction, the cultivation method influences the availability and uptake of nutrients by plants and can also induce stress, which in turn may alter the profile of secondary metabolites. In this study, we evaluated the differences in the profiles of detected metabolites between all samples of aeroponically and hydroponically grown irises. The Volcano plot analysis conducted on the whole data set revealed that 232 metabolites were upregulated in hydroponically grown plants, while only 42 metabolites showed upregulation in aeroponically grown plants (Supporting Information, S7). This leads to the presumption that hydroponics, under the conditions used in the experiment, is more effective for the enhanced production of a broader range of Iris metabolites. Chromones, anthocyanins, flavone, and isoflavonoid glycosides were among the compound groups most upregulated in hydroponics, whereas in aeroponics, Figure 3. Phytochemical profile of Iris germanica cultivars. Extracted ion chromatograms (EIC) of compounds detected in the negative ionization mode in the extracts of roots with rhizomes of (a) I. germanica Florentina Coerulea and (b) I. germanica Florentina Alba based on the in-house metabolite database. ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51261 Table 2. Differential Metabolites of the Leaves and Roots with Rhizomes of Iris marker compound VIP tentative identification a compound group b chemical formula RT (min) plant part c ionization polarity dominant ion/searched m/ztop fragments reference MS/MS d roots with rhizomes 182 1.79 formononetin isoflavonoid C16H12O48.32 R NEG [H + HCOO]-/313.0718 298.0472, 283.0238, 297.0370, 269.0431, 239.0398 n.a. 220 1.76 iriflophenone benzophenone C13H10O55.87 L, R NEG [M −H]−/245.0455 93.0336, 107.0144, 151.0028, 65.0017, 63.0219 MoNA 13 1.74 3′,4′,6-trimethoxy-5′,7,8-trihydroxyisoflavone isoflavonoid C18H16O89.21 R NEG [M −H]−/359.0772 329.0300, 344.0527, 314.0069, 286.0110, 297.0124 51 599 1.72 irisdichotin A flavonol glycoside C23H24O12 7.18 L, R POS [M + H]+/493.1341 331.081, 316.0562 n.a. 69 1.70 6,4′-dimethoxy-5-hydroxyflavone 7-glucoside/4′-methyltectorigenin-7-glucoside (1. conformer)/ 4′-methyltectorigenin-7-glucoside (2. conformer)/ irisolidone-7-O-a-D-glucoside flavone glycoside/isoflavonoid glycoside C23H24O11 7.22 L, R NEG [H + HCOO]-/521.1301 359.0773, 343.0467, 344.0540, 506.1073, 328.0203 n.a. leaves 44 1.66 5,3′,4′,5′-tetramethoxy-6,7-methylenedioxyisoflavone isoflavonoid C20H18O86.66 L NEG [H + HCOO]-/431.0984 311.0503, 341.0624, 283.0582, 323.0508, 281.0396 n.a. 391 1.62 swertisin/4′-O-methylapigenin 6-C-hexoside/4′-O-methylapigenin 8-C-hexoside flavone glycoside C22H22O10 6.83 L NEG [M −H]−/445.1140 297.0383, 325.0696, 282.0514, 231.0297, 117.0305 MoNA 12 438 1.61 5,7-dihydroxy-6-methoxychromon chromone C10H8O51.79 L POS [M + H]+/209.0444 n.a. 82 1.61 7-O-methylmangiferin/irisxanthone/7-O-methylisomangiferin C-glycosyl xanthone C20H20O11 6.44 L NEG [M −H]−/435.0933 315.0516, 345.0615, 272.0330, 330.0377, 300.0260 51, irishanthone n.a. 295 1.61 isomangiferin/nigricanside C-glycosyl xanthone C19H18O11 7.16 L NEG [M −H]−/421.0776 259.0205, 301.0323, 331.0435, 227.0271, 313.0333 51 a Tentative identification based on the in-house metabolite database and further checked against online spectral data (column Reference MS/MS); isomeric compounds may occur and are separated by “/”. b Proposed isomers can belong to different compound groups separated by “/”. c Occurrence of a compound in the leaves (L) and/or roots with rhizomes (R). d Reference MS/MS spectrum was not available (n.a.) or not searched for (−) due to missing MS2 information for the compound in the experiment. ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51262 Table 3. Differential Metabolites of Iris Cultivated in Aeroponics and Hydroponics marker compound VIP tentative identification a compound group b chemical formula RT (min) plant part c ionization polarity dominant ion/searched m/ztop fragments reference MS/MS d hydroponics 439 1.85 5,7-dihydroxy-6-methoxychromon chromone C10H8O52.03 L, R POS [M + NH4]+/226.0710 97.0264, 84.0440, 180.0626, 106.0635, 121.0729 n.a. 78 1.84 7-O-methylaromadendrin/hesperetin/dihydrokaempferide/5,7,4′-trihydroxy-6-methoxyflavanone flavanone/flavanonol C16H14O69.83 L, R NEG [M-H]-/301.0718 n.a. 276 1.75 irispurinol/quercetin 3,4′-dimethyl ether/iristectorigenin A 12a-hydroxyrotenoid/ flavone/isoflavonoid C17H14O79.60 L, R NEG [M-H]-/329.0667 n.a. 316 1.57 isovitexin/vitexin flavone glycoside C21H20O10 7.41 L NEG [H + HCOO]-/477.1038 n.a. n.a. 529 1.55 cinnamic acid phenolic acid C9H8O25.64 L, R POS [M + NH4]+/166.0863 80.0482, 136.0758, 78.0323, 53.0368, 69.0322 n.a. aeroponics 589 1.97 irilone 4′-O-beta-D-glucopyranoside isoflavonoid glycoside C22H20O11 7.72 L, R POS [M + H]+/461.1078 299.0551, 85.0276, 69.0329, 341.0628, 127.0372 10 524 1.87 betavulgarin/irisolone/irisone A isoflavonoid C17H12O69.61 L, R POS [M + H]+/313.0707 298.0475, 297.0398, 180.0053, 240.0404, 270.0515 n.a. 114 1.82 alpinone flavanonol C16H14O58.71 L, R NEG [H + HCOO]-/331.0823 165.9913, 316.0584, 110.0016, 273.0392, 82.0056 n.a. 684 1.79 kanzakiflavone-1/iriflogenin/soforanarin A/irisoid A flavone/isoflavonoid/ peltogynoid C17H12O78.83 L, R POS [M + H]+/329.0656 314.0408, 180.0054, 297.0395, 269.0440, 301.0681 n.a. 321 1.77 luteolin-4′,7-dimethyl ether/irilin A/irisolidone/5,7dihydroxy-2′,6-dimethoxyisoflavone flavone/isoflavonoid C17H14O67.24 L NEG [H + HCOO]-/359.0772 329.0295, 344.0535, 314.0056, 230.0187, 285.0407 n.a. a Tentative identification based on the in-house metabolite database and further checked against online spectral data (column Reference MS/MS); isomeric compounds may occur and are separated by “/”. b Proposed isomers can belong to different compound groups separated by “/”. c Occurrence of a compound in the leaves (L) and/or roots with rhizomes (R). d Reference MS/MS spectrum was not available (n.a.) or not searched for (−) due to missing MS2information for the compound in the experiment. ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51263 groups such as xanthones and benzofurans showed a stronger response. Subsequently, supervised OPLS-DA was applied to identify the secondary metabolites most significantly influenced by the cultivation method. The model successfully discriminated between the two groups with a predictive ability Q2= 0.90. Table 3 details 10 compounds with the highest VIP score, 5 for aeroponics, and 5 for hydroponics. Although the hydroponic experiment encompassed a substantially larger number of samples and species than the aeroponic experiment, analyses were performed on the combined data set to maximize statistical power and robustness. Both cultivation modes included representatives of the Iris and Limniris subgenera, which partially controlled for potential phylogenetic bias. To validate the results, a paired Volcano plot analysis (Supporting Information, S7) was conducted on the subset of common samples, showing that 90% of the significant metabolites were also significant in the full data set, reinforcing the reliability of the applied approach. The use of the full data set also enabled the tissue-specific evaluation of the cultivation mode, which, while technically feasible with the limited overlap of species and cultivars, would have lacked statistical strength and representativeness. In both leaves and roots with rhizomes, the trend of hydroponics to enhance the production of a larger number of targeted Iris metabolites was confirmed. Upand downregulated metabolites, as determined by Volcano plots, are presented in Supporting Information, S7. As explained earlier, compounds not detected in a given plant part were excluded from the analysis for that part to avoid reporting metabolites with an apparent statistical significance arising from the missing value imputation. 2.4. Supportive Plant Treatment during Growth. The effect of treatment with a chicken feather hydrolysate on the profile of targeted metabolites was examined in aeroponically grown plants. According to the Volcano plot analysis, no significant differences were observed between the profiles of detected metabolites of the treated plants and the control group. While the feather hydrolysate provided additional amino acids and peptides, the plants were simultaneously fertilized. It is possible that the fertilization supplied all of the essential nutrients needed for optimal plant growth, making the feather hydrolysate redundant in this experimental setup and offering no additional benefit over the standard fertilizer. Future studies should explore the use of a feather hydrolysate as the sole nutrient source to determine its effectiveness as a cost-effective growth medium while also addressing feather waste disposal and potentially supporting a circular economy. The effect of inoculation with arbuscular mycorrhizal fungi on the profile of targeted metabolites was investigated in hydroponic samples. Comparing the treated samples and controls, the Volcano plot revealed no significantly upor downregulated metabolites. This finding is in contrast with the general trend reported in the literature, where plant treatment with AMF is commonly associated with the increased production of biomass and secondary metabolites, such as terpenoids, phenolics, alkaloids, and various flavor compounds. 52−54 Only a few studies reported no significant effect or even a decrease in the biosynthesis of certain secondary metabolites. 55,56 For example, Crisan et al. (2019) observed that AMF inoculation enhanced the production of some metabolites in certain I. germanica cultivars while having no significant effect on others. 57 In our study, the absence of a detectable metabolite response may be due to several factors. First, colonization appeared relatively weak in many plants, limiting the potential impact of AMF. Second, compatibility between the plant genotypes and the specific AMF strain used may have been suboptimal, preventing the establishment of a symbiotic relationship. Third, the hydroponic environment may reduce the reliance of plants on AMF-mediated nutrient intake, thereby attenuating the metabolic response; consequently, AMF treatment could exert different effects in conventional cultivation. Additional environmental factors, such as media composition, nutrient balance, and the presence of growth regulators, may also have influenced the outcome. 58,59 Importantly, however, AMFtreated plants still exhibited increased antimicrobial and cytotoxic activity, indicating that the symbiosis may have modulated metabolic pathways, leading to metabolic products beyond the scope of our targeted metabolites. In spite of our extensive in-house database of Iris metabolites, other constituents, including mycorrhizal-derived metabolites, may underlie the observed bioactivity, providing a rationale for follow-up studies, such as nontargeted metabolomic screening to identify these differential compounds. 2.5. Identification of Potential Bioactive Metabolites. In addition to the phytochemical analysis, the Iris extracts were tested for cytotoxic, antimicrobial, anti-inflammatory, and antioxidant activities, evaluated using human dermal fibroblasts and melanoma cells, selected microorganisms (Staphylococcus aureus,Salmonella enterica,Cutibacterium acnes,Candida albicans), RAW 264.7 macrophages, and the ORAC assay, respectively, following established protocols as detailed in Section 4.3, Bioactivity assays. Many of the extracts showed strong bioactive properties. To identify potential bioactive compounds, the bioassay data and normalized metabolite signals obtained through the targeted UHPLC-HRMS/MS screening of secondary metabolites were correlated using the Pearson correlation, as described in Section 4.7 Chemometric analysis. The data matrix used for the correlational analysis is presented in Supporting Information, S8. All samples, including those exhibiting no detectable bioactivity, were incorporated into the correlation analysis to ensure an unbiased evaluation of the relationships between the chemical composition and biological activity. A moderate correlation was found between some individual metabolites and cytotoxic, anti-inflammatory, and antimicrobial activity (inhibition of C. albicans and S. enterica). The details of these metabolites are listed in Table 4.Figure 4 depicts the structures of selected metabolites that to our knowledge are associated for the first time with the observed bioactivity. Altogether, 15 metabolites showed a moderate correlation with the cytotoxic activity. Among them was compound 206, tentatively identified as iriflophenone 2-O-rhamnoside, previously reported to exhibit significant toxicity against human tumor cells, 61 compound 358, tentatively identified as phalerin, which has shown inhibitory effects against myeloma and leukemia cell lines, 62 and compound 578, tentatively identified as iridin, described as possessing promising anticarcinogenic properties. 63 Some of the correlated compounds lack bioactivity data and are thus candidates for further testing. Most of the compounds were detected in the roots with rhizomes only or showed a strong upregulation in this plant part, such as iridin (compound 578), indicating that the roots with rhizomes contain metabolites that the warrant closer examination for potential cytotoxic activity. Nonetheless, it ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51264 (43) Zhu, B.; et al. Knowledge-based in silico fragmentation and annotation of mass spectra for natural products with MassKG. Comput. Struct. Biotechnol. J. 2024,23, 3327−3341. (44) Jaegerova, T.; et al. Investigation of Iris versicolor metabolic profile and optimization of the isolation of bioactive components on a semi-operation scale. Process Biochem. 2024,146, 97−108. (45) Wilson, C. Patterns in Evolution in Characters That Define Iris Subgenera and Sections. Aliso 2006,22, 425−433. (46) Wang, H.; Cui, Y.; Zhao, C. Flavonoids of the Genus Iris (Iridaceae). Mini-Rev. Med. Chem. 2010,10, 643−661. (47) Li, L.; Chen, Y.; Feng, X.; Yin, J.; Li, S.; Sun, Y.; Zhang, L. Identification of Metabolites of Eupatorin in Vivo and in Vitro Based on UHPLC-Q-TOF-MS/MS. Molecules 2019,24, 2658. (48) Qin, M.-J.; Xu, L.-S.; Toshihiro, T.; Wang, Q.; Xu, G.-J. A preliminary study on the distribution pattern of isoflavones in rhizomes of Iris from China and its systematic significance. J. Syst. Evol. 2000,38, 343−349. (49) Yu, J.; Lee, J.-H.; Song, M.-H.; Keum, Y.-S. Metabolomic Responses of Lettuce (Lactuca sativa) to Allelopathic Benzoquinones from Iris sanguinea Seeds. J. Agric. Food Chem. 2023,71, 5143−5153. (50) Williams, C. A.; Harborne, J. B.; Colasante, M. Flavonoid and xanthone patterns in bearded Iris species and the pathway of chemical evolution in the genus. Biochem. Syst. Ecol. 1997,25, 309−325. (51) Zhang, Y.-Y.; Wang, Q.; Qi, L.-W.; Qin, X.-Y.; Qin, M.-J. Characterization and determination of the major constituents in Belamcandae Rhizoma by HPLC−DAD−ESI-MSn. J. Pharm. Biomed. Anal. 2011,56, 304−314. (52) Amani Machiani, M.; Javanmard, A.; Habibi Machiani, R.; Sadeghpour, A. Arbuscular mycorrhizal Fungi and Changes in Primary and Secondary Metabolites. Plants 2022,11, 2183. (53) Schmid, C.; et al. Quantitative Mapping of Flavor and Pharmacologically Active Compounds in European Licorice Roots (Glycyrrhiza glabra L.) in Response to Growth Conditions and Arbuscular Mycorrhiza Symbiosis. J. Agric. Food Chem. 2021,69, 13173−13189. (54) Araim, G.; Saleem, A.; Arnason, J. T.; Charest, C. Root Colonization by an Arbuscular Mycorrhizal (AM) Fungus Increases Growth and Secondary Metabolism of Purple Coneflower, Echinacea purpurea (L.) Moench. J. Agric. Food Chem. 2009,57, 2255−2258. (55) Zhao, Y.; Cartabia, A.; Lalaymia, I.; Declerck, S. Arbuscular mycorrhizal fungi and production of secondary metabolites in medicinal plants. Mycorrhiza 2022,32, 221−256. (56) Walid, N.; Kassam, R.; Ashfaq, M. Interaction between mycorrhizal and medicinal plants towards enhancement of secondary metabolites. Int. J. Chem. Stud. 2021,9, 140−152. (57) Crisan, I.; Vidican, R.; Olar, L.; Stoian, V.; Morea, A.; Stefan, R. Screening for Changes on Iris germanica L. Rhizomes Following Inoculation with Arbuscular Mycorrhiza Using Fourier Transform Infrared Spectroscopy. Agronomy 2019,9, 815. (58) Ran, Z.; et al. Arbuscular mycorrhizal fungi: Effects on secondary metabolite accumulation of traditional Chinese medicines. Plant Biol. 2022,24, 932−938. (59) Selwal, N.; et al. Enhancing secondary metabolite production in plants: Exploring traditional and modern strategies. J. Agric Food Res. 2023,14, 100702. (60) Hu, T.; et al. Metabolite identification of iridin in rats by using UHPLC-MS/MS and pharmacokinetic study of its metabolite irigenin. J. Chromatogr. B 2021,1181, 122914. (61) BioCat. Iriflophenone 2-O-Rhamnoside. https://www.biocat. com/products/TN3870-5mg-TM (accessed 4.2.2025). (62) Wahyuningsih, M.; et al. Phalerin, A New Benzophenoic glucoside isolated from the methanol extract of Mahkota Dewa (Phaleria macrocarpa (Scheff). Boerl. Leaves. Indones. J. Pharm. 2005, 16, 51−57. (63) Bhosale, P.-B.; Vetrivel, P.; Ha, S. E.; Kim, H. H.; Heo, J. D.; Won, C. K.; Kim, S. M.; Kim, G. S. Iridin Induces G2/M Phase Cell Cycle Arrest and Extrinsic Apoptotic Cell Death through PI3K/AKT Signaling Pathway in AGS Gastric Cancer Cells. Molecules 2021,26, 2802. (64) Pei, K.; Ou, J.; Huang, J.; Ou, S. Coumaric acid and its conjugates: dietary sources, pharmacokinetic properties and biological activities. J. Sci. Food Agric. 2016,96, 2952−2962. (65) Hu, J.; et al. p-coumaric acid prevents Colletotrichum gloeosporioides by inhibiting membrane targeting and organic acid metabolism. Postharvest Biol. Technol. 2023,204, 112447. (66) Xu, J.-G.; Hu, H. X.; Chen, J. Y.; Xue, Y. S.; Kodirkhonov, B.; Han, B. Z. Comparative study on inhibitory effects of ferulic acid and p-coumaric acid on Salmonella Enteritidis biofilm formation. World J. Microbiol. Biotechnol. 2022,38, 136. (67) Jaudszus, A.; et al. Vaccenic acid-mediated reduction in cytokine production is independent of c9,t11-CLA in human peripheral blood mononuclear cells. Biochim. Biophys. Acta, Mol. Cell Biol. Lipids 2012,1821, 1316−1322. (68) Jacome-Sosa, M.; et al. Vaccenic acid suppresses intestinal inflammation by increasing anandamide and related N-acylethanolamines in the JCR:LA-cp rat[S]. J. Lipid Res. 2016,57, 638−649. (69) Mostafaie, A.; Mansouri, K.; Abbasi, A.; Bahrami, G.; Sisakhtnejad, S. The effect of Cis and trans vaccenic acids on expression of ICAM-1 and VCAM-1 in human microvascular endothelial cells (HMEC). J. Rep. Pharm. Sci. 2015,4, 65. (70) Vaou, N.; Stavropoulou, E.; Voidarou, C.; Tsigalou, C.; Bezirtzoglou, E. Towards Advances in Medicinal Plant Antimicrobial Activity: A Review Study on Challenges and Future Perspectives. Microorganisms 2021,9, 2041. (71) Wagner, H.; Ulrich-Merzenich, G. Synergy research: Approaching a new generation of phytopharmaceuticals. Phytomedicine 2009, 16, 97−110. (72) Deprez, M.; Robinson, E. C. Chapter 8 - Feature extraction and selection. in Machine Learning for Biomedical Applications, Deprez, M., Robinson, E. C., Eds., pp 175−192, Academic Press, 2024. doi:DOI: . (73) Turk, A.; Addam, K.; Hwang, B. Y.; Lee, M. K. Metabolomic Profiling of Iris palaestina via Molecular Networking and Its AntiDiabetic Potential. Molecules 2025,30, 2509. (74) Zuzana, C.; et al. Botanical Gardens As A Part Of European Cultural Heritage Iris, 2020. (75) Szemerédi, N.; Dobiasová, S.; Salardón-Jiménez, N.; Kincses, A.; Nové, M.; Habibullah, G.; Sevilla-Hernández, C.; Benito-Lama, M.; Alonso-Martínez, F. J.; Viktorová, J.; et al. Cyanoand KetoneContaining Selenoesters as Multi-Target Compounds against Resistant Cancers. Cancers 2021,13, 4563. (76) Viktorová, J.; Dobiasová, S.; Rehorová, K.; Biedermann, D.; Kánová, K.; Seborová, K.; Václavíková, R.; Valentová, K.; Ruml, T.; Kren, V.; et al. Antioxidant, Anti-Inflammatory, and Multidrug Resistance Modulation Activity of Silychristin Derivatives. Antioxidants 2019,8, 303. (77) Tran, V. N.; et al. Cannabidiol nanoemulsion for eye treatment −Anti-inflammatory, wound healing activity and its bioavailability using in vitro human corneal substitute. Int. J. Pharm. 2023,643, 123202. (78) Schymanski, E. L.; et al. Identifying Small Molecules via High Resolution Mass Spectrometry: Communicating Confidence. Environ. Sci. Technol. 2014,48, 2097−2098. (79) Ghosh, P., Das, C., Biswas, S., Nag, S. K., Dutta, A., Biswas, M., Sil, S., Hazra, L., Ghosh, C., Das, S., et al., Phytochemical composition analysis and evaluation of in vitro medicinal properties and cytotoxicity of five wild weeds: A comparative study. (2020).9, 493, , (80) Zhang, P.; Wang, H.; Xu, X.; Ye, Y.; Zhang, Y. Correlation analysis between phytochemical composition and biological activities of Artemisia scoparia. Food Biosci. 2024,62, 105342. (81) Koss-Mikołajczyk, I.; Kusznierewicz, B.; Bartoszek, A. The Relationship between Phytochemical Composition and Biological Activities of Differently Pigmented Varieties of Berry Fruits; Comparison between Embedded in Food Matrix and Isolated Anthocyanins. Foods 2019,8, 646. ACS Omega http://pubs.acs.org/journal/acsodf Article https://doi.org/10.1021/acsomega.5c06354 ACS Omega 2025, 10, 51256−51271 51271