In vitro digestion modifies Lepidium meyenii extract, altering its major glucosinolates, lepidilines, β-carbolines, and gut bacterial growth Authors: Dominik Tarabasz, Ewa Baranowska-Wójcik, Paweł Szczeblewski, Tomasz Laskowski, Klaudia Gustaw, Marzena Furtak-Niczyporuk, Dominik Szwajgier, Rita Cristina Orihuela-Campos, Henry O Meissner, Julia Borzyszkowska-Bukowska, Wirginia Kukula-Koch Published: 2026 ABSTRACT AbstractThe tubers of Lepidium meyenii (commonly known as maca) have been used in the Andean cuisine for ages. The study aims to determine the behavior of the extract from the red phenotype of maca in the in vitro gastrointestinal tract model and also to define the overall impact of the extract on the microbiota. Glucosinolates, lepidilines, and a derivative of β-carboline alkaloid, i.e. 1-methyltetrahydro-β-5,6-hydride-carboline-3-carboxylic acid, as representatives of one of the main groups of metabolites were selected for the quantitative analysis after in vitro digestion at every stage of the digestion model. The results showed that lepidilines remained highly stable throughout the digestive simulation, whereas glucosinolates and the β-carboline derivative were substantially degraded, under the influence of the enzymes and bacteria applied in the “large intestine”. In addition, the red maca extract exhibited prebiotic activity by promoting the growth of selected lactic acid bacteria, with Lactococcus lactis showing the most pronounced response, suggesting that maca-derived preparations may hold promising potential as ingredients in functional foods and microbiota-targeted nutritional formulations.Keywords: Glucosinolates, HPLC-MS, In vitro gastro-intestinal digestion model, Lepidium meyenii, Prebiotics 1. Introduction The ancient plant grown in the Andean highlands known as Peruvian maca (Lepidium meyenii synonym L. peruvianum) has been known and cultivated exclusively in the mountainous areas for most likely for ca. 6000 years. Growing at altitudes between 2800 and 5000 m above sea level, maca is adapted to harsh climate conditions, i.e. low oxygen levels, low temperatures, strong winds, and higher UV radiation [1]. Native Americans of Peru and Bolivia have consumed maca hypocotyls as food for ages because of its high nutritional content and ethnomedical benefits related to reproductive health, vitality and antidepressant effects [2]. However, a special pretreatment of the root is necessary prior to consumption, including either heat treatment or drying in the sun. Maca has a distinctive but rather mild aroma and taste with a slight hint of caramel. The tuber comes in several colors (yellow, red, black, grey, purple and others), out of which yellow is the most abundant and constitutes ca. 60% of the entire annual harvest, followed by the red maca which accounts for ca. 25%, while the least popular black maca represents less than 15% of the harvest [3]. The distribution of the above colors (also referred to in research terminology as phenotypes) may be determined by geographic location where Maca is cultivated [4]. The composition of bioactive ingredients is influenced by cultivation conditions, color, genetics, plant organ, and extraction conditions [4–7]. The metabolites of maca include dietary fibers, minerals, non-starch polysaccharides, glucosinolates, polyphenols, alkaloids, macaenes, and macamides [1]. Maca, as a representative of the Cruciferous vegetable family including well-known vegetables such as broccoli, mustard, cabbage, or bok choy, is capable of synthesizing glucosinolates. It became known that the hydrolysis products of glucosinolates (GSLs), such as isothiocyanates (ITCs), nitriles (NITs), thiocyanates (TCs), and epithionitriles, are primarily responsible for the biological properties of GSLs [8]. Notwithstanding, GSLs also play a significant role as flavoring agents in vegetables [9]. The amount of glucosinolates in cruciferous vegetables varies greatly depending on the age of the plant and environmental factors, which can explain a wide range of values reported for GSLs and β-carboline alkaloids in the same maca variety [10]. It is worth noting that maca is one of the richest sources of GSLs among cruciferous plants [11]. Numerous articles, describe the medicinal properties of GSLs against various diseases. Among several confirmed biological effects of GSLs that include antimicrobial, anticancer and digestive action, these compounds were proved to be metabolized by the gut microflora into active compounds [12]. So far the effects of the digestion process (mainly in in vitro models) on single glucosinolates had been studied. Former studies indicate that Bifidobacterium strains convert sinigrin (SNG) and glucotropaeolin to NITs derivatives [13]; similarly, lactic acid species metabolize glucoraphanin and glucoiberin to NITs [14], while Bacteroides thetaiotaomicron is able to convert glucosinolates to ITCs derivatives [15]. Luang-In et al. proved that SNG and GLT are metabolized to ITCs and NITs derivatives in the presence of Enterococcus casseliflavus, Ligilactobacillus agilis and Escherichia coli bacteria [8]. Although their effects on the intestinal microbiota have been partially studied, in the scientific literature there is scarce information on the effects of digestion on maca extracts as well as on the viability of lactic acid bacteria (LAB) upon the administration of maca. The latter action is of particular importance as maca is a source of valuable polysaccharides that can certainly have an impact on the microbiota and, possibly, play an important prebiotic role [16–18]. The gut microbiota, mainly bacteria, significantly affects host metabolism and immunity [19]. Antibiotic disruption of this microbiota can lead to dysregulated immune responses, increasing the risk of TNF-associated autoimmune or chronic inflammatory diseases [19,20], like inflammatory bowel diseases (IBDs), that have been strongly linked to antibiotic use, especially when exposure occurs early in life [21,22]. Also, little is known about the actual behavior of L. meyenii metabolites in the digestive tract. As the tubers are widely consumed around the world in the form of food products, dietary supplements and plant drugs, studies on digestion of L. meyenii tubers, also colloquially called “Peruvian ginseng” [10], appear to be of high importance. Therefore, this study aimed to evaluate the impact of in vitro gastrointestinal digestion on the stability of major metabolites of red maca (L. meyenii), including glucosinolates, lepidilines, and a β-carboline derivative, and to investigate the potential prebiotic effects of maca extract on selected lactic acid bacteria. 2. Materials and methods 2.1. Plant material and extraction Dried powdered red maca (Lepidium meyenii Walp.) was used as the plant material in this study. The material originated from Peru and was obtained as a commercial product labeled as 100% organic (BIO) red maca. The product was distributed by BRAINMARKET s. r.o. (Hladnovská 83/93, 710 00 Ostrava, Czech Republic), batch number 2528 5, with an expiration date of 22 June 2026 (see Figures S3 and S4 (https://doi.org/10.38212/2224-6614.3602) in the Supplementary File). According to the manufacturer’s information, the material was stored in a dry place in sealed packaging prior to use. The powdered tubers (20 g) were extracted with 50% ethanol for 30 min in an ultrasonic bath, three times, using fresh portions of solvent every time, in a solvent to solute ratio 1:5 (m/v). The selection of 50% ethanol was dictated by the phytochemical profile of Lepidium meyenii, which contains both highly polar compounds (e.g., glucosinolates) and moderately lipophilic constituents (lepidilines, macamides, and macaenes). A 1:1 ethanol:water mixture enables efficient simultaneous extraction of these different metabolite groups. Ethanol solutions of 40–60% are widely used for the extraction of glucosinolate-rich plants, which further supports the choice of this concentration. Consequently, 50% ethanol provides an optimal compromise ensuring high extraction efficiency. The obtained extracts were filtered through a hard filter paper, collected together, and centrifuged. The supernatant was transferred to a round-bottomed flask and evaporated to dryness on a rotary evaporator at 45 °C. The dried residue weighed 7.6 g, resulting in an extraction yield of 38%. The extract was stored in a refrigerator before testing but for no longer than a week. 2.2. Qualitative and quantitative analysis of the extracts by high performance liquid chromatography coupled with mass spectrometry (HPLC-ESI-QTOF-MS/MS) The extract from L. meyenii was analyzed using the HPLC-ESI-QTOF-MS/MS platform produced by Agilent Technologies, located in Santa Clara, California, USA. The composition of the extracts was determined by utilizing various components of the mentioned equipment, such as a photodiode array detector (G1315D), a binary pump (G1312C), an autosampler (G1329B), and a degasser (G1322A) had belonged to the HPLC chromatograph, and an ESI-QTOF-MS/MS mass spectrometer (G6530B) with electrospray ionization that was attached to the HPLC chromatograph (1200 series). For the analysis of the sample’s composition, a Zorbax Eclipse Plus RP-18 chromatographic column (3.5 μm; 150 mm × 2.1 mm) was used. To perform the analysis, Agilent MassHunter Workstation Software (version B.10.00) was utilized to obtain the MS spectra and handle the recorded data. For the analysis, the HPLC thermostat was set at 25 °C, while the selected UV detection wavelengths were 254, 280, 320, and 365 nm. The UV–Vis/DAD detector’s operating wavelength range was 190–600 nm. Using a 5 μL injection volume and 0.2 mL/min flow rate applied during a 36 min gradient elution program the chromatographic separation was achieved. The mobile phase consisted of an eluent A (0.1% formic acid in water, v/v) and eluent B (acetonitrile with 0.1% formic acid added). Gradient elution was as follows: 0 min - 1% B, 2 min - 20% B, 3 min - 30% B, 10 min - 40% B, 17 min - 95% B, 19 min - 95% B, 19.5–36 min - 1% B. Mass spectrometer measurements were performed under the following conditions: gas temperature and sheath gas temperature were 275 °C and 325 °C, respectively; gas flow rate was set at 12 L/min; capillary voltage was 3000 V; fragmentation voltage was 110 V; collision energy was 10 and 20 V; skimmer voltage 65 V; atomizer pressure 35 psig. Collected spectra were scanned in the m/z range 40–1200 Da in the negative ionization mode (to visualize glucosinolates) and in the positive ionization mode (to analyze the behavior of alkaloids). Two of the most intense signals in the TIC spectrum were automatically fragmented to obtain their MS/MS spectrum in every scan. After acquiring two spectra for a given m/z value, an exclusion of the selected signal from further fragmentation was used for the following 0.2 min. The tentative identification of metabolites in the tested extract was performed based on the retention time readings, MS/MS fragmentation spectra, and in accordance with the scientific literature and open mass databases (Metlin, HMDB, Mass Bank). The quantitative analysis of the main components of red maca extract was performed on the total extract used for digestion for glucotropaeolin, glucolimnanthin, lepidilines A, B, C, D, and E, and 1-methyltetrahydro-β-5,6-hydride-carboline-3-carboxylic acid (CC). For the sake of quantitative analysis the standards (all in the purity of 95%) of glucotropaeolin (Extrasynthese, Cas 5115-71-9), glucolimnanthin (Extrasynthese, Cas 111810-95-8), lepidiline A (MedchemExpress; HY-129583), and a 1-Methyl-2,3,4,9-tetrahydro-1H-β-carboline-1-carboxylic acid (Sigma CDS002614) as an equivalent for 1-methyltetrahydro-β-5,6-hydride-carboline-3-carboxylic acid were used. The quantification of all lepidilines was performed using the standard of lepidiline A only – as it was the only commercially available standard – and expressed as calculated to lepidiline A. The quantification was achieved after triple injection of the total extract and standard solutions. For this purpose, the solutions of standards were first prepared at the concentration of 0.5 mg/mL and then diluted to cover the range of concentrations from 0.005 to 0.5 mg/mL. Also, the listed standard solutions were diluted until the detection was not possible – to calculate the LOD and LOQ values. LOD values of glucotropaeolin, lepidiline A, and CC were 2.56, 2.26, 1.22 and 1.02 μg/mL, whereas the corresponding LOQ values were equal to 8.45, 6.79, 4.03, and 3.06 μg/mL. 2.3. Growth curves of lactic acid bacteria obtained with bioscreen C/examining the effects of maca extracts on LAB The activity was tested using thermostated microplate reader Bioscreen C (Labsystem, Helsinki, Finland). The aim was to examine the impact of the extracts on microflora, therefore the following species available in the collection were selected: Lactiplantibacillus plantarum B 4496 (recently Lactobacillus plantarum), Lactococcus lactis DSM 20729, Lactobacillus intermedius B 3693, Lacticaseibacillus rhamnosus B-1445, Lacticaseibacillus casei Lby. The bacterial cultures were centrifuged and removed from the media after 24 h of incubation. The bacterial cells were suspended in physiological saline, and the optical density at 600 nm was adjusted to 1 OD (optical density). The studies were carried out on log phase cells. MRS broth was used to culture the bacterium, and as a control medium in the actual experiment. The media used in the Bioscreen experiment were made on the basis of MRS along with the percentage addition of the maca. The following concentration was used 5% and if no limit was set for bacterial growth 20% were applied. The applied concentration range (3–25%, v/v) was selected to enable assessment of dose-dependent effects of maca extract on lactic acid bacteria, including potential growth inhibition at higher concentrations. A 2% concentration of DMSO was used to dissolve the extracts used in the article, and was then added to the media. In triplicate, 350 μL of the media were put onto 100-well honeycomb plates, and 10 μL of the bacterial suspension was put into each well. The experiments were carried out in conditions under oxygen-limited conditions achieved by overlaying the culture medium with sterile liquid paraffin, with the OD at 600 nm being measured every 2 h for 72 h. The growth curves were analyzed by the PYTON script. For the in vitro digestions, Lactiplantibacillus plantarum B 4496 was chosen as described below. The concentrations of maca extract were chosen to reflect physiologically relevant scenarios. The lower doses (e.g., 3–10%) represent conditions where maca is consumed as a minor dietary supplement (side dish beside the main meal). The upper limit of 25% was selected to simulate a situation where maca constitutes a significant portion of a meal, which aligns with traditional dietary practices where it can comprise up to one-quarter of food intake. This allows us to model the potential impact on gut bacteria that might be encountered during digestion across a spectrum of realistic consumption patterns. 2.4. In vitro digestion conditions and sample preparation for qualitative analysis 2.4.1. Preparation of the inoculum For the in vitro digestions, the inoculum (Lactiplantibacillus plantarum B 4496 strain stored at −80 °C) was restored to normal metabolism through triplicate inoculation onto a new MRS medium and culturing at 30 °C for 24 h. Post-culture fluids were centrifuged twice (each time to remove the supernatant) and the obtained bacterial biomass was added to the “digestions” at the beginning of the “large intestine” stage as described below. 2.4.2. In vitro digestion The method was adopted from Brodkorb et al. [16] with modifications described below. The equipment used was described in detail in Baranowska-Wójcik et al. [17]. Five grams of the 50% ethanolic extract was mixed in the “digestion” chamber (3 jacketed glass tanks, each 1 L in volume, Supplementary Information (SI), Fig. S3-5 (https://doi.org/10.38212/2224-6614.3602)) with 45 mL of distilled water. Then, 2 mL of simulated salivary fluid (containing 113 mg KCl (Sigma P9541), 50.3 mg KH2PO4 (Sigma P0662), 114.2 mg NaHCO3 (Sigma S5761), 3.045 mg MgCl2(H2O)6 (Sigma M2670), and 0.58 mg (NH4)2CO3) (Sigma 207861), 2 mL of an α-amylase solution (containing 451 units of the enzyme; human saliva Type IX-A, 1000–3000 u/mg protein, Sigma-Aldrich A0521), 0.25 mL of a 0.3 M/L CaCl2 (Sigma C1016) solution and 9.75 mL of distilled water were added. The “digested” samples were kept at 37 °C for 2 min with the intermittent stirring set at 10 rpm for the whole in vitro digestion (30 s mixing vs 30 s pause) and “oral “sample (4 mL) was taken for the analyses. Next, simulated gastric fluid (SGF) (60 mL) and pepsin solution (40 mg/~136000u, Sigma P6887 dissolved in 6.4 mL SGF) were added, the pH was corrected to 2.0 using HCl solution (Sigma H1758), and 0.04 mL of 0.3 M/L CaCl2 was added. After 10 min, the second portion of pepsin (40 mg/~136000u in 6.4 mL SGF, Sigma P6887) was added. The volume of all samples was adjusted to 160 mL using distilled water to allow quantitative analysis. The “gastric phase” lasted for 70 min and samples (4 mL) were taken after 10 and 70 min. Then, 44 mL of simulated intestinal fluid (SIF) was added, followed by the addition of 1.224 g pancreatin (from porcine pancreas, Sigma P1625) and 7.33 g bile salts (Sigma 48305) (both dissolved in 37.5 mL of SIF). The concentrations of pancreatin and bile salts were adopted after [18]. Next, 0.16 mL of a 0.3 M/L CaCl2 solution, 1 M/L NaOH (Sigma S8045) was added to reach pH 7.0 and the “intestinal” volume in all samples was made equal using distilled water. The “small intestine” phase lasted for 140 min (collection of the sample, 4 mL). Then, the inoculum (Lactiplantibacillus plantarum B 4496 strain) was added to obtain 108 CFU/g of “digested” sample (start of the “large intestine” stage). The “large intestine” phase lasted for 24 h and samples (4 mL) were taken after 5 and 24h. Collected samples were stored at −80 °C until analysis (maximum for one week). During the whole “digestion”, sterile CO2 (analytical grade, Linde Gas Poland, 5–8 mL/min, syringe PTFE 0.45 μm filter) was run individually through each “digested” sample. Reagent (“control”) samples (containing the “digestion” fluids in distilled water, inoculated with LAB) were run in the same manner as “digestions” containing tested plant extracts. All “digestions” were triplicated. 2.4.3. Preparation of samples after in vitro digestions for qualitative analysis Each “digested” sample was deproteinated by mixing with acetonitrile (1:1 v/v) followed by centrifugation for 5 min at 10000 RPM. The supernatant was collected and filtered through a nylon syringe filter (0.1 μm 13 mm Labex Ltd, FilterBio NY, Nantong City, China) and then analyzed by HPLC-ESI-QTOF-MS/MS for their qualitative and quantitative composition. 2.5. Statistical analysis Chemometric analyses and visualizations were performed using R v4.1.2 [19] programming language in RStudio [20] software with pracma [21], factoextra [22], matlib [23], and corrplot [24] packages installed. The Bioscreen C data was analyzed using the PYTHON script according to Hoeflinger et al. to obtain lag times, doubling time, maximum and minimum OD, maximum specific growth rate, and more [25]. 3. Results and discussion 3.1. The qualitative analysis of the 50% ethanol extract from the red phenotype of L. meyenii through HPLC-ESI-QTOF-MS/MS The main varieties of Lepidium meyenii include black, yellow, gray and red phenotypes. They are not distinguished from each other by new compounds [26]. However, they differ in the concentration of secondary metabolites from different groups, which affects their total pharmacological profile. In the performed study, the presence of several groups of compounds was confirmed in the analyzed extract from the red maca through the HPLC-MS methodology in an optimized method. Among them, amino acids, β-carboline alkaloids, imidazole alkaloids, macaene, macamides and glucosinolates were assigned based on the existing scientific literature data and MS/MS fragmentation pattern (also presented in Table S4 (https://doi.org/10.38212/2224-6614.3602) of the Supplementary File), as shown in Table 1 [27]. Table 1The chemical profile of the 50% ethanolic extract from Lepidium meyenii using HPLC-ESI-QTOF-MS/MS in the positive and negative ionization modes. (Ion - ionization mode: positive (+) or negative (−), Rt - retention time, delta - error of mass measurement [ppm], RDB - rings and double bond number, compounds marked in bold are subjected to the quantitative analysis). The numbered compounds correspond to the numbers of the tested compounds from the table. No. Ion.(+/−)Rt [min]Molecular formulam/z calculatedm/z experimentalDeltaRDBMS/MS fragmentsProposed compound based on the fragmentation patternRef.1+2.11C6H14N4O2175.1190175.1182+4.322158.0912, 130.0959, 116.0701, 70.0653Arginine(Ibrahim et al., 2022)2+2.227C5H9NO2116.0706116.0710−3.43270.0647Proline(Ibrahim et al., 2022)3+2.945C5H11NO2118.0863118.0860+2.18172.0802Valine(Ibrahim et al., 2022)4+4.729C6H13NO2132.1019132.1015−2.25186.0953, 44.0506Leucine(Ibrahim et al., 2022)5+5.529C9H11NO3182.0812182.0805+3.75165.0533, 136.0749, 123.0429, 97.9681, 91.0542Tyrosine(Ibrahim et al., 2022)6+6.930C7H9N108.0808108.0806+1.64491.0530Benzylamine(Zhou et al., 2017)7+8.448C9H11NO150.0913150.0913+0.275135.0300, 110.0129, 91.0543N-acetylbenzylamine(Zhou et al., 2017)8+8.682C19H11NO2166.0863166.0859+2.155120.0789, 91.0542Phenylalanine(Bai et al., 2015) 9 + 9.649 C 13 H 16 N 2 O 2 233.1285 233.1278 +2.82 7 218.1140, 190.0696, 166.0822, 116.0695, 91.0537 1-methyltetrahydro-β-5,6-hydride-carboline-3-carboxylic acid (CC) (Zhou et al., 2017)10+9.866C11H12N2O2205.0972205.0981−4.637188.0692, 146.0588, 121.0635Tryptophan(Ibrahim et al., 2022)11+10.533C13H14N2O2231.1128231.1135−3.028214.0844, 158.0950, 130.0639, 91.05341-methyltetrahydro-β-carboline-3-carboxylic acid(Zhou et al., 2017)12+11.150C14H18N2215.1543215.155−3.857123.0902, 91.0530Lepidiline G(Le et al., 2021) 13 + 11.233 C 15 H 18 N 2 227.1543 227.1521 +1.66 8 135.0907, 121.0656, 109.0755, 91.0533 Lepidiline E (LEE) (Le et al., 2021)14+11.572C16H20N2O257.1648257.1640+3.288121.0640, 91.0545Lepidiline F(Le et al., 2021) 15 + 13.818 C 19 H 20 N 2 277.1699 277.1701 −0.63 11 198.9383, 185.1058, 160.1116, 109.0759, 91.0530 Lepidiline A (LEA) (Le et al., 2021) 16 + 14.235 C 20 H 22 N 2 291.1856 291.1859 −1.12 11 250.1551, 200.1287, 181.1057, 158.0992, 118.0771, 91.0541 Lepidiline B (LEB) (Le et al., 2021) 17 + 14.402 C 20 H 22 N 2 O 307.1805 307.1813 −2.65 11 292.1581, 229.1309, 215.1159, 185.1063, 121.0643, 106.0414, 91.0539 Lepidiline C (LEC) (Le et al., 2021) 18 + 14.786 C 21 H 24 N 2 O 321.1961 321.1966 −1.44 11 306.1702, 280.1677, 230.1391, 199.1220, 175.1110, 158.0965, 121.0636, 91.0541 Lepidiline D (LED) (Le et al., 2021)19+15.036C13H13NO2216.1019216.109+4.678198.0899, 170.0947, 138.0560, 108.0434, 91.0537N-benzyl-1,2-dihydro-Nhydroxypyridine-4-carbaldehyde (macaridine)(Zhou et al., 2017)20+15.486C21H24N2305.2012305.2005+2.3811213.1372, 123.0913, 91.05421,3-Dibenzyl-2(R)-ethyl-4,5-dimethylimidazilium(Zhou et al., 2017)21+17.404C25H24N2353.2012353.2014−0.515262.1443, 171.0895, 91.05401,3-Dibenzyl-2-phenyl-4,5-dimethylimidazilium(Zhou et al., 2017)22+17.571C23H28N2333.2325333.2321+1.2811241.1687, 200.1295, 91.05381,3-Dibenzyl-2-butyl-4,5-dimethylimidazilium(Zhou et al., 2017)23+18.405C24H30N2347.2482347.2486−1.2311255.1843,200.1289, 91.05381,3-Dibenzyl-2-pentyl-4,5-dimethylimidazilium(Zhou et al., 2017)24+22.575C18H30O3295.2268295.2273−1.84277.2139, 151.11135-Oxo-6E,8E-octadecadienoic acid (macaene)(Zhou et al., 2017)25+22.958C25H37NO368.2948368.2936+3.248355.0705, 266.9959, 246.1831 108.0804, 91.0539N-benzyl-(9Z, 12Z, 15Z)-octadecatrienamide(Zhou et al., 2017)26+22.992C25H37NO2384.2897384.2912−3.98277.2144, 260.1645, 170.1164, 108.0807, 91.0534N-benzyl-5-oxo-6E,8Eoctadecadienamide(Zhou et al., 2017)27+23.375C25H37NO2384.2897384.2906−2.338349.2722, 260.1634, 152.1074, 108.0807, 91.0531N-Benzyl-13-oxooctadeca-9E,11E-dienamide(Zhou et al., 2017)28+23.742C25H37NO2384.2897384.2902−1.298260.1638, 170.1163, 108.0807, 91.0531N-Benzyl-9-oxo-12E,15Zoctadecadienamide(Zhou et al., 2017)29+24.076C25H39NO2386.3054386.3061−1.937279.230, 261.2175, 170.1173, 108.0807, 91.0543N-Benzyl-9-oxo-12Zoctadecenamide(Zhou et al., 2017)30−8.848C14H19NO10S2424.0378424.0380−0.566380.0229, 274.9855, 240.9940, 218.0608, 182.0271, 148.0398 96.9600Glucosinalbin (GS)(Zhou et al., 2017) 31 − 9.515 C 14 H 19 NO 9 S 2 408.0428 408.0409 +4.76 6 328.0877, 290.9805, 259.0121, 241.0013, 195.0329, 166.0326, 96.9603 Glucotropaeolin (GTP) (Zhou et al., 2017) 32 − 9.899 C 15 H 21 NO 10 S 2 438.0534 438.0535 −0.2 6 358.0919, 290.9800, 259.0127, 196.0423, 96.9600 Glucolimnanthin (GL) (Zhou et al., 2017) The chemical profile of the 50% ethanolic extract from Lepidium meyenii using HPLC-ESI-QTOF-MS/MS in the positive and negative ionization modes. (Ion - ionization mode: positive (+) or negative (−), Rt - retention time, delta - error of mass measurement [ppm], RDB - rings and double bond number, compounds marked in bold are subjected to the quantitative analysis). The numbered compounds correspond to the numbers of the tested compounds from the table. Thanks to the application of a gradient of acetonitrile in water, with an addition of formic acid, a clear fingerprint was obtained. The number of metabolites present in the positive ionization mode was markedly higher than in the negative ionization mode. The constituents that were marked in bold in Table 1 as the leading components of the extract from different subgroups of metabolites were used for the quantitative analysis to demonstrate the differences that occur during the in vitro digestion process of maca extract in vitro. Among them, glucotropaeolin and glucolimnanthin – two main glucosinolates, five compounds from the group of imidazole alkaloids - lepidilines A, B, C, D and E, as well as 1-methyltetrahydro-β-5,6-hydride-carboline-3-carboxylic acid – a β-carboline alkaloid were analyzed for their behavior under the influence of digestive enzymes and microbiota in an in vitro digestion model. To move towards the analysis of digestion impact on the metabolic profile, a careful assignment of the leading molecules was performed. Glucotropaeolin > glucolimnanthin > glucosinalbin were the three compounds occurring in order from the highest to the lowest concentrations. In the extracted ion chromatograms, GSLs are detectable in the negative ionization mode because they lose a proton from the typical form to the GSLs sulfonated oxime moiety [28], which gives them a negative charge. Glucotropaeolin and glucolimnanthin have a very similar fragmentation profile, as they differ only slightly in the aglycone part. GL has an acetyl group built into the meta-position of an aromatic ring. The fragmentation pattern of imidazole alkaloids is less complicated than in the case of glucosinolates. They produce [M+H]+ ions in the positive ionization mode (See SI table S4 (https://doi.org/10.38212/2224-6614.3602) for their MS/MS spectra and for detailed fragmentation description), whereas the 1-methyltetrahydro-β-5,6-hydride-carboline-3-carboxylic acid - an alkaloid from the β-carboline group – shows a simple MS/MS fragmentation profile in the positive ionization mode coming from the breakdown of the molecular ion [M+H]+ at m/z 233.1277 (C13H17N2O2+). The detailed analysis of the MS/MS fragmentation pattern of the aforementioned molecules is presented at the beginning of Supplementary File and provided evidence for their identification and enabled a further analysis on their behavior in the digestion model. As a subsequent stage of the study, quantitative analysis was performed on the total extract obtained from the red phenotype of maca. The contents of glucotropaeolin, glucolimnanthin, lepidilines A–E, and a β-carboline alkaloid were determined in the crude extract. Quantification was based on triplicate injections of the total extract, which enabled the calculation of the relative contribution of each compound. Glucotropaeolin constituted 1.178% of the total extract, corresponding to 11.78 mg/g of extract, while glucolimnanthin accounted for 0.313% (3.13 mg/g). Lepidiline A was present at 0.117% (1.17 mg/g), lepidiline B at 0.124% (1.24 mg/g), lepidiline C at 0.064% (0.64 mg/g), lepidiline D at 0.077% (0.77 mg/g), and lepidiline E at 0.069% (0.69 mg/g of extract). The content of lepidilines was determined using lepidiline A as an external reference standard, and the results were expressed as lepidiline A equivalents. The obtained data indicate that glucotropaeolin represents the predominant constituent of the extract; however, the remaining metabolites also contribute substantially to the overall chemical profile of the analyzed sample. All quantitative values were recorded in the linear range of corresponding injections of authentic standard. These quantitative values provide a robust reference point for subsequent in vitro digestion studies, enabling the calculation of compound losses at individual stages of the gastrointestinal tract model. 3.2. The impact of in vitro digestion on the quantitative fingerprint of the extract in HPLC-ESI-QTOF-MS/MS based studies In this study, we present the first evidence of the digestive stability of maca extract with particular emphasis on the behaviour of three groups of secondary metabolites, i.e. glucosinolates, imidazole alkaloids and a β-carboline alkaloid derivative, that are the leading components of the total extract, were analyzed for their content at every stage of in vitro digestion. Fig. 1A illustrates the changes in the content of major compounds from the red phenotype of L. meyenii observed during the digestion process (after rescaling). For each compound, the initial content (RM10 – 10 min after the start of digestion) was normalized to a value of 1, and subsequent values were recalculated relative to the peaks obtained at the RM10 stage. As observed, the highest concentrations of glucosinolates (G), β-carboline alkaloid (CC), and two imidazole alkaloids (LE) were recorded when the extract was present in the stomach (RS70). This phenomenon is likely associated with the complete dissolution of the dry maca extract at this stage. As a result of the transformation occurring during the digestion process, changes in the content of the selected secondary metabolites were noticed. After 140 min, as a result of the processes taking place in the small intestine, the behavior of the analyzed compounds can be divided into two groups (RS140). A significant decrease in content at this stage was observed in the case of glucosinolates (G) and a β-carboline alkaloid derivative (CC). On the other hand, the digestion of the imidazole alkaloids – lepidilines (LE) showed a greater durability and only after bacteria were introduced to the system (at the large intestine level) they were degraded by ca. 10–15%. When averaged, glucosinolates (G) and a β-carboline alkaloid derivative (CC) behaved similarly, yet differently in comparison to lepidilines (LE) – they were degraded by about 40% by the enzymes of the small intestine (Fig. 1B). Interestingly, at the last stages of digestion, an increasing concentration trend was observed for alkaloids (LE and CC) and glucotropaeolin (GTP), and exceptionally, a decrease in the content was observed for glucolimnanthin (GL, Fig. 1A). During the last stage of digestion, in the large intestine, the selected strain of probiotic bacteria was added and its effect on the content of compounds was examined. The chart uses data obtained after 5 h (RLI300) and 24 h (RLI1440). Panel A: Differentiation of the content of the main compounds from the red variety of Lepidium meyenii during the in vitro digestion process, based on the data obtained from each stage of “digestion”. For the standard deviation value see SI Table S3 (https://doi.org/10.38212/2224-6614.3602). RM10 (stomach 10 min after the beginning of the “digestion” process), RS70 (“stomach” 70 min after the beginning of the “digestion” process), RSI140 (“small intestine” after 140 min of “digestion” process), RLI300 (“large intestine” after 300 min of “digestion” with the addition of Lactiplantibacillus plantarum B 4496), RLI1440 (“large intestine” after 1440 min of “digestion” with the addition of Lactiplantibacillus plantarum B 4496). Panel B: Differentiation of the averaged content of the main compounds from the red variety of Lepidium meyenii during the in vitro digestion process, based on the data obtained from each stage of “digestion”. Glucosinolates (GTP and GL) were averaged to yield group G, lepidilines from LEA to LEE were averaged to yield group LE. β-carboline alkaloid derivative (CC), as a sole representative of its class, constituted the group CC. The decrease in GSLs content in our study, in the in vitro gastrointestinal model system, can be explained in two ways. First, it can be induced by a low stomach pH [29]. Secondly, the reason can be the conversion to nitrile derivatives due to the presence of lactic acid bacteria [14,30]. Also, degradation of GSLs can be observed between the small intestine and large intestine levels. As previously demonstrated in the scientific literature, GLT can have three possible hydrolysis products: benzyl isothiocyanate, benzyl thiocyanate and phenylacetonitrile. Unfortunately, no traces of ITCs or TCs derivatives from glucotropaeolin or glucolimnanthin were detected. This may be associated with the lack of activity of the gut bacterial consortium. The biotransformation of GSL into ITC occurs with the support of intestinal bacteria through hydrolysis of the S-glycosidic bond [31]. Although a strain of L. plantarum B 4496 in an in vitro digestion model was used, the authors were not able to reflect the entire activity of intestinal bacteria that form a chain of compounds’ transformation. Only trace amounts of a nitrile derivative of GTL, i.e. phenylacetonitrile (PhACN), were identified in the mass spectrometry based analysis at each stage of the digestion process. Detected trace amounts of PhACN may be due to using the total extract in which the GTL content is nearly 5%. An almost twice as high a signal, although of an insignificant intensity, was observed between the stomach and the intestine. The presence of bacteria did not affect the PhACN signal level. All this suggests that the nitrile derivative was created independently of the activity of bacteria, and the low pH in the stomach may promote the degradation of glucosinolates to nitrile compounds. Lepidilines, forming another group of the investigated metabolites of maca, appear to be stable in the small intestine and don’t undergo degradation. A visible decrease in the content of imidazole alkaloids was noted only at the large intestine level. It is worth noting that after 19 h of in vitro co-incubation, the levels of 1-methyltetrahydro-β-5,6-hydride-carboline-3-carboxylic acid (CC) increased in the colon, probably due to the activity of Lactiplantibacillus plantarum B 4496. This compound is not only present in the maca extract, but also described as an ingredient of fermented garlic extract [32]. This is particularly important because, according to the authors’ current knowledge, there is no information on LAB strains capable of producing CC. However, CC is a very eminent compound and an ingredient of various plants which, when consumed as a component of a given plant, loses its undesirable effects. There are studies showing its antioxidant and UV protective effects [33]. These new observations on the behavior of maca constituents in an in vitro digestion model distinguish maca from other glucosinolate-rich plants, highlighting its potential in maintaining bioactive compound stability through the digestive process. By demonstrating the stability of lepidilines, this research sheds light on broader impact of maca extract on digestive system and underscore its potential application in functional foods and dietary supplements contributing to gut health. 3.3. Statistical analysis of the impact of individual stages of digestion on the composition of red maca extracts The presented digestion profiles of red maca extracts (Fig. 1) indicate that the content of individual components changes during the subsequent stages of digestion. While analyzing the digestion profiles of the red maca, two patterns can be distinguished, one for lepidilines (LEA to LEE) and another one for glucosinolates and β-carboline alkaloid derivative (GTP, GL and CC). In order to confirm the observed relationships, the coefficients of linear correlation (R), determination (R2) and the tangent distances (DT) between individual compounds were analyzed and summarized on a heat map (see also Fig. S1 (https://doi.org/10.38212/2224-6614.3602) in SI). Coefficient of linear correlation (R), along with its derivatives, i.e. R2 and DT, can be considered a direct measure of similarity between the shapes of the digestion profiles (Fig. 1). Analyzing the data derived from the heat map (Fig. 2), it can be concluded that all lepidilines manifest the same common nature of transformations across all stages of digestion (light to dark red color in the heat map indicates a high degree of correlation). This is also evidenced by the small tangent distances between individual lepidilines on the cluster diagram included on the upper and left side of the heat map. Glucosinolates exhibit a distinctive pattern in terms of digestion. Interestingly, this pattern is also shared by a β-carboline alkaloid derivative, a compound that is not a glucosinolate. In fact, its concentration changes at individual stages of digestion are more similar to GL than both glucosinolates towards each other (GL and GTP), as it can be seen on the cluster diagram below the heat map and is also confirmed by the determination coefficient values (see SI Table S2 (https://doi.org/10.38212/2224-6614.3602)). Heatmap of glucosinolates, lepidilines and β-carboline alkaloid derivative on the basis of the LC-MS dataset obtained for red maca (see SI table S1 (https://doi.org/10.38212/2224-6614.3602)). For more detailed data upon linear determination coefficients for individual pairs of compounds, please see the SI Fig. S1 (https://doi.org/10.38212/2224-6614.3602). To confirm the differences between lepidilines, glucosinolates, and CC, as shown in Fig. 1, an ANOVA test was performed to determine whether the digestion process affects the content of individual compounds in a statistically different manner. The obtained results unequivocally demonstrate that a uniform “digestion” profile within each major chemical class, while no statistically significant differences in the digestion profiles can be established among the five lepidilines (from LEA to LEE). Conversely, no differences in the digestion profiles could be detected within the glucosinolates (GTP vs. GL) or between the glucosinolates and the β-carboline alkaloid derivative (GTP vs. CC and GL vs. CC). These findings confirm that within each chemical class, the compounds behave uniformly during “digestion”. Subsequently, all possible pairs of compounds were analyzed using post-hoc HSD Tukey’s test. The results and detailed description of these analyses are presented in SI and Fig. S2 (https://doi.org/10.38212/2224-6614.3602). Overall, the results of Tukey’s test align are consistent with the data presented in Figs. 1 and 2. This limited statistical distinction, particularly where visual trends (Fig. 1) suggest differences, is likely due to the relatively large variances in the measured content combined with the limited sample size (only four measurements), which reduces the statistical power of the Tukey’s test. The statistically confirmed uniformity of behavior within each chemical class and the significant divergence of degradation profiles between the classes underscore the compound-specific bioavailability dynamics of red maca during in vitro digestion. Taking into account all the above data, it can be concluded that the composition of the red maca extracts varies depending on the stage of digestion, and these differences are in fact dependent on the class of the compounds tested. The lepidilines were statistically confirmed to be the most stable group during the digestion process. Additionally, in order to determine whether the addition of bacteria (in RLI1440) affects the overall content of glucosinolates and lepidilines, the Mann-Whitney U test was applied. The test showed statistically significant differences (p < 0.05) between the last two stages of digestion (RLI300 to RLI1440) only for glucolimnanthin (GL) and lepidiline E (LEE) (see SI Table S2 (https://doi.org/10.38212/2224-6614.3602)). 3.4. Effect of maca extracts on lactic acid bacteria The Bioscreen C experiment confirmed the effects of varying concentrations of the selected extracts on five LAB strains. The purpose was to determine the point of growth inhibition of microorganisms at particular concentrations of the maca extract. Importantly, this study also represents the first investigation of the stability of lepidilines under simulated gastrointestinal conditions, providing new insights into the behavior of these metabolites during digestion. As can be seen in Fig. 3, maca extracts influenced the growth of LAB in three ways. In the first way observed, strains of L. casei, L. intermedius and L. rhamnosus showed no difference in growth relative to the control at concentrations of 3% and 5%. Inoculation of listed strains to the medium with maca at 3% and 5% gives a slight prolongation of the lag phase as compared with the control. However, at 10%, maca extract caused reduction in the growth of these bacteria but had little effect on the ability of maximum specific growth. At above discussed concentrations, maca had no effect on the growth of L. plantarum strain. At higher concentrations (15, 20, and 25%), a gradual inhibition was observed as the concentration of maca was increased. As a result (Table 2), with the addition of 25% maca, the reduction of L. plantarum was observed. The time required to double the number of cells was 3 h for 10% of maca addition, 12 h for 15%, and 15 h for 20% maca. By contrast, L. lactis strain exhibited a slight linear growth enhancement with the maca addition with higher values of max. OD for all tested concentrations compared to the control. Therefore, our results suggest that the lowest concentrations used in this study had little effect on altering the bacterial growth curve relative to the control medium. It is worth noting that L. lactis typically exhibits limited growth under Bioscreen C conditions, as consistently observed in our previous experiments; therefore, the growth stimulation observed in the presence of maca extract is particularly significant for this strain. Growth curves drawn for the tested microorganisms on MRS medium supplemented with 3% 5% 10% and 15% 20% 25% (for selected) red maca. Table 2Growth kinetics for Bioscreen C results. OD values were rescaled in relation to Max OD of MRS (control) of a given strain and expressed in %.StrainBrothCodenameLag Time (hours)aMax Specific Growth Rate (1/hours)aDoubling Time (hours)aMax OD (rescaled)aDelta OD (Median Filtered Data, rescaled)aMax ODMin ODDelta OD (Median Filtered Data) L. casei MRS LC- 4.0810.1793.863100.0%94.7%1.7710.0911.6783%maca LC3 00.0749.42273.6%68.8%1.3040.0831.2195%maca LC5 3.4470.1664.18199.4%93.5%1.7600.1021.65610%maca LC10 3.3080.1624.27399.4%92.9%1.7610.1101.646 L. rhamnosus MRS LR- 2.9390.2023.425100.0%96.7%1.7150.0551.6583%maca LR3 4.0880.1295.36589.8%86.5%1.5400.0571.4835%maca LR5 2.9330.2023.42996.0%91.6%1.6470.0731.57110%maca LR10 3.1550.1943.57793.5%89.3%1.6030.0681.532 L. intermedius MRS LI- 2.1120.2772.5100.0%95.6%1.5130.0651.4463%maca LI3 3.1090.2792.48698.3%94.8%1.4880.0481.4355%maca LI5 2.360.2762.508100.3%95.0%1.5170.0751.43810%maca LI10 2.1720.3042.279103.4%97.2%1.5650.0941.470 L. lactis MRS LL- 2.1530.1086.418100.0%91.5%0.6450.0540.5903%maca LL3 1.1850.0729.648121.7%117.1%0.7850.0300.7555%maca LL5 2.1170.1146.084100.9%92.2%0.6510.0510.59510%maca LL10 2.2130.1126.163104.0%94.0%0.6710.0600.606 L. plantarum MRS LP- 3.4350.2372.927100.0%97.3%1.6110.0441.5673%maca LP3 2.7210.1993.4993.2%91.0%1.5020.0311.4665%maca LP5 3.1250.2362.937104.0%100.2%1.6760.0621.61410%maca LP10 3.0230.2263.062100.4%98.0%1.6180.0391.57915%maca LP15 00.05612.38968.2%67.5%1.0990.0001.08720%maca LP20 00.04615.12720.6%20.2%0.3320.0040.32625%maca LP25 12.2%9.5%0.1960.0290.153aVariables that were subject to principal component analysis (PCA). Growth kinetics for Bioscreen C results. OD values were rescaled in relation to Max OD of MRS (control) of a given strain and expressed in %. Variables that were subject to principal component analysis (PCA). An excellent example of how maca can be used as an energy source by lactic acid bacteria is shown in the study by Yang’s team in which it was fermented [34]. The above results reflect previous studies demonstrating that a number of compounds present in maca can affect bacterial growth; of the three groups of compounds that were highlighted, only one – GSLs – has a proven effect on lactic acid bacteria. There is scientific evidence that some bacteria can use GSLs, by decomposition to isothiocyanates, nitriles or other hydrolysis products by gut bacteria [35]. Glucosinolate hydrolysis products, such as sulforaphane, erucine exhibit strong antimicrobial properties against a variety of pathogens, such as Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa, Listeria monocytogenes [36], Salmonella typhimurium, Staphylococcus aureus, Streptococcus lactis and Vibrio parahaemolyticus Helicobacter pylori [29]. The second factor important for human health is the metabolism of GSL by gut bacteria, which are mainly responsible for the health-desirable properties of ITCs after GSL ingestion. Moreover, it was found that among lactic acid bacteria, Lactobacillus plantarum strain KW30 and Lactococcus lactis subsp. lactis KF147 were able to transform glucosinolates to nitriles [14]. In maca extracts the authors detected glucotropaeolin and glucolimnanthin, and little is known on the mechanism of action of these compounds on bacteria. Cheng et al. described that three strains of the genus Bifidobacterium are capable of converting sinigrin and glucotropaeolin into nitrile derivatives [13]. In a paper by Liou et al. on glucosinolates from broccoli, it was shown that the addition of glucotropaeolin had a stimulatory effect on the growth of commensal Bacteroides thetaiotaomicron, and simultaneously reduced the growth of opportunistic strains of E. coli and Salmonella typhimurium [15]. Significant amounts of native forms of GSLs in maca and the fact that maca extracts did not adversely affect the LAB strains used may prove that it is necessary to create a starter consortium for the fermentation of this vegetable. On the example of the fermentation of cruciferous vegetables, cabbage, broccoli or cauliflower, it was proved that an increased amount of their hydrolysis products may be observed [37]. Since glucosinolates can be converted to other compounds, adding bacteria capable of this conversion to maca products would definitely increase the availability of active compounds to consumers. Based on this information, studies on the biotransformation of maca should be undertaken in the near future to study the potential of added bacterial strains on the biotransformation of glucosinolates. Building on this, these findings have direct implications for the development of maca-based functional foods and synbiotic formulations. The dose-dependent and strain-specific effects observed are critical. On one hand, the selective growth stimulation of beneficial strains like L. lactis highlights maca’s prebiotic potential. On the other hand, the inhibitory effects on key probiotics, such as L. plantarum at high concentrations, underscore a significant challenge. This dual activity suggests that for maca to be effectively used as a functional ingredient, careful formulation considerations are essential and further studies in both in vitro and in vivo models. The dosage must be precisely optimized to harness its prebiotic benefits while avoiding the antimicrobial properties that could compromise the viability of probiotic cultures in a fermented food or a dietary supplement. After excluding LP25 (missing data), the dataset from Table 2 was subjected to PCA in R. The first three components (Figs. 4 and 5) explained ~99% of the total variance and defined three orthogonal dimensions: Relative OD (vs. MRS; driven by nearly parallel Max OD and Delta OD), Growth Ratio (defined by Max Specific Growth Rate and Lag Time, which were strongly inversely correlated), and LagTime (dominated by the original Lag Time variable). In the initial model (Fig. 3), the score distribution was strongly influenced by LP15, LP20 and LC3, which were located far from the remaining samples; LP15 and LP20 (L. plantarum, 15% and 20% maca) showed below-average Growth Ratio with low Relative OD and the shortest lag times, while LC3 (L. casei, 3% maca) exhibited a similar profile. The diagonal histograms were left-skewed (and bimodal for LagTime), indicating dataset heterogeneity and compression of differences among the remaining samples. Three-dimensional principal component analysis (PCA) of the dataset from Table 2 performed in R after exclusion of LP25 (missing data). The first three principal components (PC1–PC3; upper-right panels) explain ~99% of the total variance. Loadings indicate that Relative OD (vs. MRS control) is mainly described by Max OD and Delta OD (nearly parallel vectors), Growth Ratio by Max Specific Growth Rate and Lag Time (strong inverse correlation), and LagTime primarily by the original Lag Time variable. Sample scores (lower-left panels) identify LC3, LP15 and LP20 as markedly separated observations, corresponding to L. casei with 3% maca extract and L. plantarum with 15% and 20% maca extracts, characterized by below-average growth ratio, low relative OD, and short lag time. Diagonal histograms show left-skewed distributions (bimodal for LagTime), indicating dataset heterogeneity and compression of the remaining samples by these extreme observations. Three-dimensional principal component analysis (PCA) of the dataset from Table 2 performed in R after excluding LP15 and LP20 (extreme observations) and LP25 (missing data). The first three principal components (PC1–PC3; upper-right panels) reproduce the same three orthogonal dimensions as in Fig. 4: Relative OD (vs. MRS control; Max OD and Delta OD), Growth Ratio (Max Specific Growth Rate inversely correlated with Lag Time), and LagTime (dominated by the Lag Time variable). Diagonal histograms indicate improved coherence of the dataset. Sample scores (lower-left panels) separate LC3 and the LL series from the remaining samples: L. lactis shows slower growth with the shortest lag times, and LL3 exhibits the highest Relative OD, while most other samples cluster tightly and LC3 yields the lowest scores across all factors. To better resolve the structure of the remaining data, LP15 and LP20 were removed and PCA was repeated (Fig. 4), yielding the same component definitions but more coherent distributions. In this refined space, LC3 and all LL samples separated from the rest: L. lactis displayed slower growth in the presence of maca extracts while having the shortest lag times, with Relative OD comparable to other strains except LL3, which showed the highest OD and the strongest displacement along the Relative OD axis; most other samples clustered tightly, whereas LC3 retained the lowest scores across all three factors. Overall, the Bioscreen C experiments demonstrated a strain-dependent response of lactic acid bacteria to maca extract. L. lactis exhibited consistent growth stimulation across the tested concentration range, which may be related to its efficient utilization of readily available carbohydrates present in the extract. In contrast, L. plantarum showed reduced growth at higher extract concentrations, suggesting a dose-dependent inhibitory effect, potentially associated with increased exposure to bioactive secondary metabolites. These findings indicate that maca extract may exert selective, prebiotic-like effects on lactic acid bacteria; however, the observed inhibition at higher concentrations highlights the importance of concentration optimization in potential food or fermentation applications. 4. Conclusions This study provides novel insights into the behavior of red Lepidium meyenii (maca) 50% ethanol extracts in an in vitro model of the gastrointestinal tract. Glucosinolates, imidazole alkaloids and a β-carboline alkaloid derivative, which are the leading components of the extract, were analyzed for their content at every stage of digestion process to determine their susceptibility to different pH and enzymatic conditions. The findings demonstrate that while glucosinolates and β-carboline alkaloids are susceptible to degradation – especially at the small intestine stage, lepidilines exhibit high stability. A visual summary of the transformations of major metabolites and the resulting prebiotic effects during the in vitro passage through the digestive tract is presented in Fig. 6. This stability, coupled with observed prebiotic effects, suggests that maca could play a significant role in promoting gut health. The implications of these findings are substantial. Maca’s stability and prebiotic properties position it as a promising ingredient for dietary supplements and fermented foods. By enhancing gut microbiota balance, maca could contribute to reducing the risk of gastrointestinal disorders. This study shows that among the tested LAB strains, the L. lactis strain showed a particular predisposition to increased growth in the presence of red maca extract. Schematic representation of the relative content changes of major secondary metabolites from red maca (Lepidium meyenii) across five stages of in vitro digestion. Data are presented as percentages relative to the initial content (RM10). Group G represents the averaged content of glucotropaeolin (GTP) and glucolimnanthin (GL); group LE represents the averaged content of lepidilines A–E; CC refers to the β-carboline alkaloid derivative. Future applications of these findings could include the development of probiotic-maca formulations, dietary supplements and functional foods which may offer functional benefits as part of a balanced diet aimed at modulating microbiota composition. However, it is important to acknowledge the limitations of this study. The in vitro digestion model cannot fully replicate the complexity of human gastrointestinal conditions as it cannot fully mimic the physiological complexity of human digestion, including nutrient absorption, metabolic biotransformations, host–microbe interactions, and inter-individual variability of the gut microbiota. Therefore, further in vivo studies are essential to validate these findings and explore the full potential of maca in dietary applications. Specifically, clinical trials should be conducted to assess the real-world effects of maca consumption on human gut health. Future research should focus on controlled fermentation using defined microbial consortia, optimization of fermentation strategies to enhance metabolite stability and bioavailability, and evaluation of maca-based formulations within animal and human studies to better assess their functional and nutritional potential. These efforts will certainly provide indepth information on maca’s health benefits and applications in functional foods. Supplementary Information Acknowledgment We wish to thank Professor Jan Mazerski from the Department of Pharmaceutical Technology and Biochemistry, Faculty of Chemistry, Gdansk University of Technology for his invaluable suggestions regarding chemometric analysis. Funding Statement This study was funded by the Medical University of Lublin, Poland (grant no. UM-CTW-10/2021), the University of Life Sciences in Lublin, Poland, and Gdańsk University of Technology, Poland. The authors also gratefully acknowledge partial funding from Gdańsk University of Technology under the EUROPIUM –’Excellence Initiative –Research University’ program (grant no. DEC-2/1/2023/IDUB/IV.2a/Eu) Footnotes Declaration of competing interest: The authors declare no conflict of interest. Funding: This study was funded by the Medical University of Lublin, Poland (grant no. UM-CTW-10/2021), the University of Life Sciences in Lublin, Poland, and Gdańsk University of Technology, Poland. The authors also gratefully acknowledge partial funding from Gdańsk University of Technology under the EUROPIUM –’Excellence Initiative –Research University’ program (grant no. DEC-2/1/2023/IDUB/IV.2a/Eu). References Associated Data Supplementary Materials