Development and validation of CYP26A1 inhibition assay for high‐throughput screening

All‐trans retinoic acid (atRA) is an endogenous ligand of the retinoic acid receptors, which heterodimerize with retinoid X receptors. AtRA is generated in tissues from vitamin A (retinol) metabolism to form a paracrine signal and is locally degraded by cytochrome P450 family 26 (CYP26) enzymes. The CYP26 family consists of three subtypes: A1, B1, and C1, which are differentially expressed during development. This study aims to develop and validate a high throughput screening assay to identify CYP26A1 inhibitors in a cell‐free system using a luminescent P450‐Glo assay technology. The assay performed well with a signal to background ratio of 25.7, a coefficient of variation of 8.9%, and a Z‐factor of 0.7. To validate the assay, we tested a subset of 39 compounds that included known CYP26 inhibitors and retinoids, as well as positive and negative control compounds selected from the literature and/or the ToxCast/Tox21 portfolio. Known CYP26A1 inhibitors were confirmed, and predicted CYP26A1 inhibitors, such as chlorothalonil, prochloraz, and SSR126768, were identified, demonstrating the reliability and robustness of the assay. Given the general importance of atRA as a morphogenetic signal and the localized expression of Cyp26a1 in embryonic tissues, a validated CYP26A1 assay has important implications for evaluating the potential developmental toxicity of chemicals.


INTRODUCTION
All-trans retinoic acid (atRA), an active metabolite of vitamin A (retinol), plays a crucial role in regulating cell growth and differentiation in many tissues. [1]It acts as an endogenous ligand, along with its isomer, 9cis-RA, for two families of ligand-activated nuclear receptors: retinoic acid receptors (RARα, β, and γ) and retinoid X receptors (RXRα, β, and γ), respectively.RAR/RXR heterodimers, together with co-activators or co-repressors, regulate gene expression through retinoic acid response elements (RARE) in many genes that function in metabolic This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.Published 2024.This article is a U.S. Government work and is in the public domain in the USA.Biotechnology Journal published by Wiley-VCH GmbH.and developmental pathways (Figure 1A). [2,3]The atRA signaling pathway plays a crucial role in developmental processes and maintenance of cellular phenotype. [4]Disrupting the atRA signaling pathway can lead to altered retinoid concentrations in target tissues, resulting in heightened or diminished RAR-mediated signaling.The main pathway for atRA elimination occurs through oxidative metabolism via 4-hydroxylation, which is mediated by a range of cytochrome P450 enzymes (CYP). [3,5]e first CYP to be identified as a major contributor to atRA 4-hydroxylation in human liver was CYP2C8, with CYP3A F I G U R E 1 (A) General scheme of retinoic acid signaling pathway and metabolism by CYP26 enzymes.Retinol is converted to all-trans retinoic acid (atRA) via two oxidative steps, first to retinal mainly by retinol dehydrogenase 10 (RDH10) and then to atRA by retinal dehydrogenases 1 to 3 (RALDH1-3).atRA is further metabolized by the CYP26 and other CYP family of enzymes to primary metabolites, including 4-OH-RA, 18-OH-RA, and 5,6-epoxy-RA and are further metabolized to polar metabolites such as 4-oxo-RA.atRA is the activating ligand for the retinoic acid receptors (RARs), which heterodimerize with retinoid X receptors (RXRs) on the retinoic acid response element (RARE).In the inactive state, nuclear receptor co-repressor (CoR) complexes tighten and prevent the transcription of genes.Recruitment of co-activator (CoA) promotes transcription of genes downstream of the RARE; (B) Conversion of P450-Glo™ substrate by CYP26A1.CYP enzyme acts on a luminogenic P450-Glo™ substrate (first reaction) to produce a luciferin product that generates light with the luciferin detection reagent (second reaction).The amount of light produced is proportional to CYP activity.The red arrow indicates the site of modification by CYP enzyme.
subfamily as minor participants. [6]Several other human CYPs were subsequently shown to catalyze 4-hydroxylation of atRA, including CYP1A1, CYP1A2, CYP2C9, CYP3A4, CYP3A7, and CYP26. [3]The three isoforms of CYP26 family, CYP26A1, CYP26B1, and CYP26C1 are considered the primary mammalian atRA hydroxylases. [7,8]The primary atRA metabolites include 4-OH-RA, 18-OH-RA, and 5,6epoxy-RA, among which 4-OH-RA appeared to be the most common metabolite formed by CYP26, which can then be further metabolized to polar secondary metabolites such as diols and 4-oxo-RA. [9,10]All three isoforms are subject to cell-specific regulation in developing and adult tissues. [11]CYP26A1 is enriched in the human liver and is mainly responsible for the hepatic clearance of atRA. [12]CYP26B1 is enriched in placenta, ovary, testes, and intestine, while CYP26C1 is enriched in brain and liver. [9]Knockout of these three genes was shown to be severely detrimental [9] ; a post-natal inducible global knockout of Cyp26a1 and Cyp26b1 in adult mice or juvenile resulted in severe dermatitis, hyperkeratosis and hyperplasia, blepharitis, splenomegaly, and inflammation, [13] whereas the loss of both Cyp26a1 and Cyp26c1 in mice blocked the production of cranial neural crest cells in the forebrain and midbrain. [14]7] Currently, there is no in vitro CYP26A1 inhibition assay available in a high-throughput screening format.Furthermore, the Organization of Economic Cooperation and Development (OECD) in their detailed 2021 review of the retinoid system emphasized the development of a CYP26 inhibition assay as a high priority [OECD, Detailed review paper on the retinoid system Series on Testing and Assessment, (2021) No. 343 https://one.oecd.org/document/ENV/CBC/MONO(2021)20/en/pdfref].Here, we developed a high-throughput enzyme-based CYP26A1 assay.The assay relies on a probe substrate that engages the enzyme active site and is converted to a readily detected product (e.g., via an optical property).0] We validated the assay against a subset of 39 compounds, which included known CYP26 inhibitors and retinoids, as well as positive and negative control compounds selected from the literature and/or the ToxCast/Tox21 portfolio.The compounds' activities on CYP26A1 enzyme were then compared with their effects in CYP26B1, CYP1A2, CYP3A4, CYP2C9, CYP2C19, CYP2D6, RAR, and RXR in vitro assays to evaluate their selectivity toward CYP26A1.The successful development and validation of a high-throughput screening assay for CYP26A1 enzyme activity reported here paves the way for identifying novel CYP26A1 inhibitors from chemical libraries.

P450-Glo™ CYP26A1 assay
P450-Glo™ CYP26A1 screening system was purchased as a custom product from Promega Corp. (Madison, WI).The system includes recombinant wild-type human CYP26A1 enzyme co-expressed with P450 oxidoreductase (POR) in Escherichia coli expression system and prepared as Bactosomes (purchased from Cypex/BioIVT, Dundee, Scotland, UK; catalog # CYP070), Luciferin-benzyl ether (Luciferin-BE) synthesized as described, [21] KPO4 (pH 7.4) and NADPH regeneration solution consisting of 1.3 mM NADP + , 3.3 mM Glucose-6-P, 3.3 mM MgCl 2 , 0.4 U mL −1 G-6-P dehydrogenase, and 0.05 mM NaCitrate (concentrations in the final reaction volume).The regenerating system ensures that NADPH remains non-limiting throughout the reaction time course. [22]P450-Glo assays rely on CYP-dependent conversion of a proluciferin probe substrate to D-luciferin that is detected as relative luminescence units (RLU) in a second reaction with luciferase. [23]o microliters enzyme (0.02 pmol µL  Following compound treatment and incubation of the assay plates for 10 min at room temperature, 2 µL NADPH-regeneration solution was added using an FRD to start the reaction.The reaction was continued by incubating the assay plates at 37 • C for 30 min.The reaction was stopped by adding 4 µL luciferin detection reagent.The luminescence signal was measured using a ViewLux plate reader (Perkin Elmer, Waltham, MA).The raw data were obtained as relative luminescence units.
The thawing and assay media were same as the culture medium except Pintool station.The assay plates were incubated with the compounds for 6 h at 37 • C, 99% humidity, and 5% CO 2 , followed by the addition of 5 µL One-Glo™ luciferase assay reagent (Promega US) using an FRD.
After incubation of assay plates for 30 min at room temperature, the luminescence signal was measured using a ViewLux plate reader.The raw data were obtained as relative luminescence units.
For thawing cells, culture medium without hygromycin and zeocin was used and the assay medium contains phenol-red free DMEM supplemented with 2% charcoal-stripped FBS, 0.1 mM NEAA, 1 mM sodium pyruvate and 100 U mL −1 penicillin -100 µg mL −1 streptomycin.The RXRα-bla cells were seeded at 2000 in 6 µL assay medium per well in black-clear 1536-well tissue culture plates (Greiner Bio-One North America Inc.) using an FRD.After 5 h incubation of the assay plates at 37 • C, 99% humidity, and 5% CO 2 , the positive control (9-cis-RA) along with DMSO-only wells and test compounds, were transferred to the assay plates at 23 nL using a Wako Pintool station.The assay plates were incubated with the compounds for 16 h at 37 • C, 99% humidity, and 5% CO 2 .After addition of 1 µL LiveBLAzer™-FRET B/G (CCF4-AM) substrate (Thermo Fisher Scientific Inc.) using an FRD, the assay plates were incubated for 2 h at room temperature.Fluorescence signal (405 nm excitation, 460 and 530 nm emissions) in each well was measured using an Envision plate reader (PerkinElmer).The raw data were expressed as the ratio of 460/530 emissions.

Data analysis
The experimental data (relative luminescence or fluorescence units) from each assay well were first normalized relative to the positive con-

Development of enzyme-based CYP26A1 inhibition assay
A P450-Glo™ assay technology has been used to develop enzymebased CYP26A1 assays.In P450-Glo™ assay, CYP enzyme converts a luminogenic probe substrate to D-luciferin that is detected as light output from a second reaction with luciferase (Figure 1B).Light output is reduced by test compounds that inhibit the CYP enzyme.To find the optimal substrate for the assay, several luminogenic substrates [23,25,26] were evaluated for their activities in the assay plus and minus a recombinant human CYP26A1 enzyme co-expressed with POR.Given the superficial structural similarities between retinoids and the D-luciferin derivatives (e.g., a COOH group on one end of the molecules at sim-ilar spacing from sites of CYP-dependent oxidation on the opposite end) along with known CYP26 substrate promiscuity, it was not unreasonable to expect to find a CYP26 luciferin derivative substrate. [11,26]ong the substrates screened, the luciferin-BE substrate yielded the highest signal with CYP26A1 (Figure 2A).This identified luciferin-BE as an active CYP26A1 substrate with utility for identifying compounds that inhibit CYP26A1 activity.
To optimize the assay conditions of the luciferin-BE reaction with CYP26A1, seven concentrations of luciferin-BE substrate ranging from 0.39 to 25 µM plus a vehicle control were tested in a 96-well plate in triplicate, with 1.0 pmol CYP26A1 enzyme per 50 µL reaction volume for 30 min at 37 • C. To derive reaction rates (pmol D-luciferin/pmol CYP26A1/min) from experimental data, relative luminescent units (RLU) were converted to CYP reaction product concentrations by interpolation from a D-luciferin standard curve (data not shown).
Enzyme kinetic parameters (K m and V max ) were estimated through nonlinear regression analysis using the Michaelis-Menten model in GraphPad Prism (Figure 2B).Based on the determined K m value, the luciferin-BE substrate concentration at 2 µM was selected for the assay.The assay was linear over time at 2 µM substrate for at least 40 min (linear regression: R 2 = 0.988) (Figure 2C) with a close correlation between time and substrate conversion as indicated by a slope of 0.92 ± 0.08 from the log-log plot of time (10-40 min) versus net signal (data not shown).The assay showed a linear response (R 2 = 0.998) with the increasing concentrations of CYP26A1, from 0.0 to 2.0 pmol per 50 µL reaction volume (Figure 2D).The final concentration of enzyme selected for the CYP26A1 assay was 1 pmol per 50 µL reaction volume.

Optimization of CYP26A1 inhibition assay in a 1536-well plate
The assay was miniaturized into 1536-well plates, with a final assay volume of 4 µL.To minimize well-to-well variations resulting from enzyme adhesion on the plate dispenser's tubing, we assessed the impact of 0.1% Bovine Serum Albumin (BSA), along with various detergents such as triton X-100, tween-20, and CHAPS in our assay.The concentrations of triton X-100 and tween-20 were tested at 0.01%, a common concentration used in biochemical assays.Notably, the 0.025% CHAPS condition exhibited the highest potency, as indicated by the lowest IC 50 values (Figure 3A-C).The assay was performed in three independent runs, and its performance was assessed using the following listed parameters, which are expressed as mean ± SD values from three runs.Signal to Background (S/B) ratios are 48.6 ± 2.4 and 25.7 ± 1.8 with atRA at 58 and 29 µM, respectively; coefficient of variation (CV) is 8.9% ± 2.6; Z′ is 0.7 ± 0.08.We validated additional isomers of atRA, including 9-cis-RA and 13-cis-RA, in our assay to assess the selectivity of these substrates for the CYP26A1 enzyme.The concentration-response curves representing the inhibitory effects of these three isomers are presented in Figure 3D.The IC 50 values are: 1.1, 2.8, and 6.1 µM for atRA, 9-cis-RA, and 13-cis-RA, respectively.These inhibitory effects with 2-and 6-fold differences in IC 50 values between atRA and its isomers, supports the fact that both these isomers still undergo phase I metabolism to several oxidized metabolites including 4-oxo-RA, 4-OH-RA, 18-OH-RA, and 5,6-epoxy-RA. [3,27]The results further demonstrated that this assay is robust and suitable for high-throughput screening.

Evaluation of CYP26A1 assay using a subset of compounds
To validate the CYP26A1 assay, a group of 39 compounds (Table 1) including known CYP26 and/or CYP inhibitors, retinoids and their analogues, and other diverse structural sets of compounds were selected.These compounds were previously predicted to be potential developmental toxicants associated with atRA signaling pathway disruptions. [28,29]The assay was performed in three independent runs (n = 3).The highest concentration of each compound tested in the CYP26A1 assay was 115 µM, except for propargite (1.7 mM), and SSR126768 (109 µM).The compounds with maximum response greater than 40% are considered actives in the CYP26A1 assay.The concentration-response curves of the actives from the CYP26A1 assay are shown in Figure 4.The potencies (IC 50 ) and maximum responses (%) of compounds tested in the CYP26A1 assay are given in Table 1.
Reliability of the assay for sensitive and specific detection of CYP26A1 LC-MS analysis. [27,30,31]In contrast, R115866 inhibition was 10 times less potent in the luciferin-BE assay compared to an assay that used atRA as the probe substrate (50 vs. 5 nM). [27,32]Furthermore, the IC 50 of 1.1 µM for atRA inhibition of the luciferin-BE reaction (Figure 3D) is higher than would be expected if the two substrates are perfectly exclusive in the CYP26A1 active site given that CYP26A1 atRA K m values have been reported in a range of about 9 to 50 nM. [11]In aggregate, these results are consistent with the previous observation that CYP26A1 inhibition is substrate dependent. [31]For example, the fold IC 50 differences from assays that used either 9-cis-RA or tazarotenic acid as CYP26A1 probe substrate, ranged from unity to from 3-to 10-fold. [31]Nevertheless, the best fit of our atRA inhibition curve has a Hill slope of −1, which is consistent with simple, non-cooperative atRA and luciferin-BE binding at a single active site on CYP26A1.
Substrate-dependent CYP enzyme inhibition is well known, CYP3A4 being the best studied example. [33,34]Additionally, the performance metrics demonstrated the assay's sensitivity, robustness, and suitability for high throughput screening.Therefore, the identification of CYP26A1 inhibitors through this assay from large chemical libraries represents a significant advancement in the field of developmental toxicity assessment.
In the next step, we aim to use this assay for screening the Tox21's collection of approximately 10,000 chemicals to identify CYP26A1 inhibitors.This collection was established for the Tox-Cast/Tox21 program, a collaborative effort among federal agencies aimed at developing innovative methods for assessing the potential risks of approved drugs and environmental chemicals to human health. [35,36]The Tox21 chemical library has been extensively screened against various cellular targets and pathways using in vitro cell-and biochemical-based assays on a fully automated high throughput platform to prioritize compounds for further toxicological evaluation. [37] date, the Tox21 assays have enabled high-throughput screenings for a panel of nuclear receptors, stress response pathways, GPCRs, various CYP biochemical activities, enzyme targets, developmentalrelated cellular pathways such as TGF-β, SHH/GLI1, and the retinol signaling pathway (RAR), [36,38] although the CYP26 family is not currently part of this portfolio.The CYP26 family plays a pivotal role in maintaining proper atRA levels within tissues, with their expression believed to be autoregulated based on atRA concentrations.[9] Dysregulation of atRA metabolism and CYP26A1 activity can have detrimental effects on normal embryonic development, leading to developmental abnormalities.[4] In the context of regulatory developmental toxicity for drugs and chemicals, it remains unclear whether any observed developmental defects can be attributed exclusively to disruption of retinoid signaling.[39] To support recommendations regarding assay development for the retinoid system and the use of resulting data in the regulatory context for developmental toxicity, a CYP26 HTS assay could be considered as a top priority for addressing missing information in the retinoid system for New Approach Methodologies (NAMs).Therefore, the development and validation of robust screening assays to identify CYP26A1 inhibitors is significant.
Here, we assessed the inhibitory capabilities of the chemicals for several other CYPs (e.g., CYP1A2, CYP2C9, CYP3A4, and CYP26B1) known to be involved in the in vitro metabolism of atRA. [3]Although the CYP26B1 assay demonstrated a lower S/B ratio with positive control compared to the CYP26A1 assay, we evaluated selectivity toward screening for CYP26A1 inhibition.As might be anticipated in a cell-free bioassay, the active compounds identified in the CYP26A1 assay had efficacy across a wider spectrum of other CYP assays, including CYP3A4, CYP2C9, CYP2C19, and CYP26B1 (Figure 5).This finding supports the involvement of CYP26, CYP3A, and CYP2C families, which are commonly implicated in atRA metabolism via 4-hydroxylation. [3,9]However, in P450-Glo™ assays, different CYP enzymes can react with more than one P450-Glo™ substrate.The selectivity of a particular P450-Glo™ substrate in each CYP assay has been demonstrated by showing the sensitivity of the reactions [26] and selectivity of the respective reactions is further supported by their insensitivity to compounds that inhibit other CYPs.
In this study, the chemicals were evaluated and categorized based on their structure and mode of actions, as depicted in the heatmap (Figure 5).The chemical groups demonstrating high potency in the CYP26A1 and other CYP assays were ranked in descending order as follows: RAMBAs, conazole fungicides, organochlorine pesticides, rexinoids, and retinoids.RAMBAs specifically target the CYP enzymes responsible for mediating retinoic acid metabolism. [17]Among the RAMBAs investigated in this study, R115866 (talarazole), liarozole, and ketoconazole exhibited significant potency in both CYP26A1/B1 and CYP3A4 assays.R115866 is widely recognized as a highly potent and specific inhibitor of human CYP26A1 and CYP26B1, and other known inhibitors of CYP26A1 include liarozole and ketoconazole were confirmed. [10,17,40]Most compounds tested here displayed inhibitory activity across a wide range of CYP assays, except for retinoids.
Conversely, retinoids demonstrated a higher level of selectivity by inhibiting CYP26A1 and CYP26B1 (Figure 5).These findings support that the inhibition of retinoic acid metabolism by RAMBAs has the potential to enhance the endogenous retinoic acid levels in vivo, while the inhibition of CYP26 by retinoids has the potential to elevate the atRA gene expression in vitro. [32,41] conclusion, we successfully developed a novel high-throughput CYP26A1 inhibition assay for identifying CYP26A1 inhibitors and validated its reliability using a diverse set of compounds.These findings clearly demonstrate the assay robustness and its suitability for highthroughput screening purposes.By elucidating the complexities of atRA metabolism and its regulation, this research could bridge the gaps in retinoid system, which can contribute to a more comprehensive understanding of developmental toxicology and the potential risks associated with chemical exposure.

F I G U R E 2
Development of CYP26A1 assay.(A) Luminogenic substrate reactivities with or without CYP26A1.(B) V max and K m determined for luciferin-BE.(C) Time course of the CYP26A1 reaction with 2 µM luciferin-BE.(D) The rates of luciferin-BE reaction at varied CYP26A1 enzyme amounts.Data shown for A are single determinants, data shown for B, C, and D are the mean ± SD from three experiments.puromycin addition.The C3RL4 cells were seeded at 1000 in 5 µL assay medium per well in white opaque 1536-well tissue culture plates (Greiner Bio-One North America Inc.) using an FRD.After an overnight incubation of the assay plates at 37 • C, 99% humidity, and 5% CO 2 , the positive control (retinol) along with DMSO-only wells and test compounds, were transferred to the assay plates at 23 nL using a Wako trol (agonist assays: 100% and inhibition assays: −100%) and DMSOonly wells (0%) within each plate as follows: % Activation = [(V compound F I G U R E 3 Concentration-response curves.Testing the positive control compounds in the CYP26A1 assay under various conditions, such as (A) 0.1% BSA + 0.025% CHAPS, (B) 0.1% BSA + 0.01% Triton X-100, and (C) 0.1% BSA + 0.01% Tween-20.(D) Comparison of the inhbitory effects of all-trans retinoic acid (atRA) with its isomers, including 9-cis-RA and 13-cis-RA with 0.1% BSA + 0.025% CHAPS condition.Each value represents the mean ± SD from three experiments.− V DMSO )/ (V pos − V DMSO )] *100, and % Inhibition = [(V compound − V DMSO )/ (V DMSO − V pos )] *100, where V compound represents the compound well reads, V pos and V DMSO represents the average values of the positive control and DMSO-only wells respectively.The half maximum effective concentration (EC 50 ) and half maximum inhibitory concentration (IC 50 ), and maximum response values were obtained by fitting the concentration-response curves of each compound to a four-parameter Hill equation by nonlinear regression using GraphPad Prism version 9.0.0 (GraphPad Software, Boston, MA).AC 50 (EC 50 or IC 50 ) is defined as the half maximum activity concentration.
inhibitors, various other CYP inhibitors, eight retinoids, two rexinoids (RXR-selective retinoids), and three RAMBAs.The assay performed well with S/B ratio exceeding 25-fold, CV of 8.9%, and a Z' of 0.7.Of TA B L E 1 A list of 39 compounds, their corresponding IC 50 values and maximum responses (in parenthesis) obtained from CYP26A1 assay and their mode of action.IC 50 and maximum response values were expressed as mean ± standard deviation from three experiments.± 0.0 (−101.03± 0.15)

F I G U R E 5
Compound activities in various assays.The heatmap is colored based on the half maximum activity, (AC 50 , µM) in the left portion and maximum response (%) values displayed on the right.Inhibitory assays include all CYPs, and the activation assays are RAR and RXR.The compounds are classified into different groups based on their chemical structures and functions, such as organochlorine pesticides and conazole fungicides, or their mechanism of action, including retinoids, rexinoids, and RAMBAs.