(T1530-01-05) Evaluating Açaí Cytochrome P450 Inhibitors Using High-Resolution Mass Spectrometry and Machine Learning Based Biochemometric Analysis

Purpose: The use of natural products in the form of botanical dietary supplements (BDS) is rapidly increasing in the U.S., as public awareness of their potential health benefits continues to grow. Despite their widespread use, BDS are not required to undergo FDA approval to assess safety and efficacy, leading to variations in chemical composition across brands and unpredictable effects. Studies have shown that ̴ 50% of patients with cancer, diabetes, and cardiovascular disease use BDS, emphasizing the need to understand their safety profiles, particularly regarding ADME properties and potential pharmacokinetic interactions with other botanicals or prescribed medications. A major concern is the inhibition of drug-metabolizing enzymes, especially cytochrome P450 (CYP). This study focuses on Euterpe oleracea (açaí), a South American berry with a market value of $1.89 billion in 2024. Açaí is widely used for its anti-inflammatory, antioxidant, and anti-cancer properties, which are attributed to its rich content of polyphenols and anthocyanins. Despite these therapeutic benefits, açaí consumption has been associated with vascular adverse events, particularly when used alongside CYP-interacting anticancer drugs. These findings underscore the critical need to evaluate açaí’s potential to cause clinically relevant pharmacokinetic interactions. Addressing this gap, the present study focuses on a comprehensive analysis of key CYP enzymes using physiologically relevant in vitro models, examining multiple açaí extract types sourced from widely consumed commercial products, and identifying the bioactive constituents in açaí through high-resolution mass spectrometry and machine learning tools. The study aims to understand the pharmacokinetic interactions of BDS consumed by patients with polypharmacy for its safe and efficacious use.

Methods: To evaluate botanical-drug interaction, an in-house developed CYP cocktail enzyme assay was employed to simultaneously assess the activity of seven key CYP enzymes (CYP1A2, 2A6, 2C8, 2C9, 2D6, 3A4/5) using pooled human liver microsomes. The assay mixture included phosphate buffer (as negative control), known inhibitors or açaí extracts (as positive controls), a cocktail of probe substrates, and a quench solution containing internal standards. Details of the assay components are provided in Figure 1. Metabolites of probe substrates and internal standards were quantified using an LC-MS-QTOF system, with chromatographic separation performed on a Poroshell 120 CS-C₁₈ column (2.1 × 50 mm, 2.7 µm). A 3-minute linear gradient was applied using mobile phase A (water with 0.1% FA) and mobile phase B (acetonitrile with 0.1% FA) at a flow rate of 0.6 mL/min. Positive and negative electrospray ionization technique was used. Extracts showing significant CYP inhibition were further fractionated using solvents of varying polarity, including hexane, dichloromethane, ethyl acetate, and water. Chemometric analysis of the active fraction was performed based on LC-MS profiles which was conducted using a Poroshell 120 EC-C₁₈ column (2.1 × 100 mm, 2.7 µm) with a 50-minute gradient employing mobile phases A (water with 0.1% FA) and B (methanol with 0.1% FA). A machine learning model (random forest) was used to identify bioactive constituents associated with CYP inhibitory activity, which were subsequently characterized using tandem mass spectrometry (MS/MS) in conjunction with publicly available spectral databases.

Results: Preliminary screening conducted at human-relevant concentration (2.321 ng/mL of cyanidin-3-glucoside) showed three extracts: F3AC (acidic methanol extract of Nature’s Way capsule), MRME & MRET (methanolic and ethanolic extract of Mountain Rose powder) to be active. These three extracts were further tested to determine IC₅₀ values (depicted in Figure 2). CYP2C9 inhibition was most pronounced, with IC₅₀ values of 0.3 µg/mL, 68.86 µg/mL, and 42.24 µg/mL for F3AC, MRME, and MRET extracts, indicating potent and moderate inhibitors, respectively. Fractionation of these extracts revealed that hexane, dichloromethane, and ethyl acetate fractions of F3AC, as well as dichloromethane and ethyl acetate fractions of MRME and MRET, retained CYP2C9 inhibitory activity. LC-MS and LC-MS/MS analysis of these fractions in conjunction with spectral databases led to the detection of compounds such as vitexin, kaempferol 7-O-glucoside, epigallocatechin, dihydrokaempferol, etc. Figure 3 represents chemometric analysis, such as PCA  and correlation matrix, demonstrating chemical similarities among the active fractions. Further random forest modelling generated heatmaps highlighting the top 50 most important molecular features associated with CYP2C9 inhibition. Based on these features, MS/MS spectral analysis was conducted, which led to tentative identification of potential compounds to be responsible for interactions: oleic acid (m/z 283.2294) and oleoylglycerolphosphocholine (m/z 522.3537).

Conclusion: This study underscores the potential of açaí fruit extracts to inhibit key CYP enzymes, particularly CYP2C9, thereby raising concerns about clinically relevant botanical–drug interactions. This is of particular significance given that several widely used anticancer drugs, such as cyclophosphamide, ifosfamide, and tamoxifen, as well as supportive therapies like warfarin, are primarily metabolized by CYP2C9. Inhibition of this enzyme may lead to reduced activation of prodrugs or impaired drug clearance, potentially resulting in diminished efficacy or increased toxicity. Considering the growing consumption of BDS like açaí and the narrow therapeutic index of many CYP2C9-metabolized drugs, these findings highlight the need for clinical studies to assess the relevance of botanical–drug interactions.

References: [1] Information for Consumers on Using Dietary Supplements, Food and Drug Administration (2024).
[2] Acai Berry Market Growth Analysis Report, Future Market Insights (2024).
[3] S.M. Fahim et al; Expert Opinion on Drug Safety (2018), 18(2), 145-152.
[4] L.F. Laurindo et al; Nutrients (2023), 15(4), 989.

Acknowledgements: This project was funded by NCCIH/ODS R15AT011047 & OID Exploratory Grant program from Auburn University.


Figure 1. Illustration of the in-house developed CYP enzymatic assay, outlining assay conditions and detailing the specific probe substrates, known inhibitors, and açaí extract types used for CYP inhibition screening.


Figure 2. (A) IC₅₀ values of three active açaí extracts—F3AC (acidic methanol extract of Nature’s Way capsule), MRME, and MRET (methanolic and ethanolic extract of Mountain Rose powder)—against CYP2C9. Based on established criteria for CYP inhibition by botanical extracts, F3AC was classified as a potent inhibitor (IC₅₀ < 10 µg/mL), while MRME and MRET demonstrated moderate inhibition (IC₅₀ between 10–100 µg/mL). (B) Fractionation of these extracts revealed that hexane, dichloromethane (DCM), and ethyl acetate (EA) fractions of F3AC, as well as DCM and EA fractions of MRME and MRET, retained CYP2C9 inhibitory activity at 100 µg/mL and 50 µg/mL concentrations.


Figure 3. Chemometric analysis of açaí extract fractions exhibiting CYP2C9 inhibitory activity. (A) Principal component analysis (PCA) plot showing chemical similarities among active fractions across different extract types, based on features detected in both ESI+ and ESI− modes. (B) Correlation matrix displaying pairwise similarity scores among active fractions; higher scores (red, 0 to 1) indicate greater chemical similarity, while lower scores (blue, -1 to 0) indicate greater dissimilarity. (C) Heatmap generated by random forest model, highlighting the top 50 most discriminative molecular features (m/z and retention times) detected in both ESI+ and ESI- modes, associated with CYP2C9 inhibition. Feature abundance is represented by colour intensity, with higher values (green to yellow, 0.5–1.0) indicating greater abundance, and lower values (blue, 0.0–0.5) indicating reduced abundance in the respective fractions.