Fluorogenic chemical tools to shed light on CES1-mediated adverse drug interactions

Abstract

Studying factors that cause interindividual variability of carboxylesterase 1 (CES1) activity is currently difficult due to limited methods. Here, fluorogenic tools for measuring CES1 activity are developed and demonstrated to report on these factors in living cells. These tools enable experiments that will develop a deeper understanding of CES1 metabolism.

---

Approximately 20% of all drugs on the market undergo hydrolytic metabolism.¹ Human carboxylesterase (CES1) is a serine hydrolase that is highly expressed in the liver and is reported to be responsible for over 80% of the liver's hydrolytic activity.^2–5 This makes CES1 particularly important in the metabolism of orally delivered drugs absorbed by the digestive system as they will pass through the liver via the hepatic portal system before reaching their therapeutic target (first-pass metabolism).^6,7 Many commonly prescribed drugs that contain functional groups susceptible to hydrolysis have been reported to have their pharmacological properties mediated by CES1 including methylphenidate (Ritalin®),⁸ enalapril (Vasotec®),⁹ and oseltamivir (Tamiflu®).¹⁰

The metabolism of drugs by CES1 has classically been assumed to be the same between individuals.³ This assumption, however, is not correct as CES activity is known to vary person to person due to a variety of factors including drug–drug interactions (DDIs), where drugs requiring metabolism by CES1 may compete for binding causing an overall decrease in hydrolysis rate of both drugs.^2,5,11 This can lead to less effective treatments and serious side effects. Furthermore, CES1 found in cells outside of the liver and digestive tract are thought to play endogenous roles in lipid metabolism unrelated to xenobiotic metabolism.^12–15 It is possible that drugs could interfere with the CES1-mediated metabolism of endogenous substrates and disrupt these pathways causing adverse effects. Despite the importance of CES1 in human metabolism, the factors that influence CES1 activity are understudied.^2,16,17,18

Current methods to study CES1 activity are quite limited likely due to the complicated nature of human CES1.²In vitro methods of exploring CES1 activity are hampered by activity differences between CES1 isolated through different methods.^19,20 As an alternative, CES1 is often studied in complex mixtures in the form of cultured cell lysates and human liver microsomes.^2,21 However, the results from these approaches are restricted by not being a complete system and studies in model organisms are complicated by differences in substrate specificities among CES orthologs.²

To enable study in living cells and tissues, fluorogenic chemical tools have been developed.^2,19,20,22 These tools have several advantages over other methods of studying CES1 including the ability to study endogenously expressed CES1 in live cells which overcomes the challenges in studying CES1 in vitro.² However, nearly all currently reported tools have not been completely characterized for use in live cells which limits their utility in studying CES1 metabolism.² To address the need for better characterized chemical tools for CES1 we previously developed FCP-1.²⁰ FCP-1 uses a carbonate substrate group that mimics a CES1 preferred substrate that upon hydrolysis releases the fluorescent compound fluorescein. This tool demonstrated that carbonate-based fluorogenic chemical tools can be used to report on CES1 activity in live cells.

Here we sought to expand the toolkit of well-characterized chemical tools by adapting our approach to an easily modifiable xanthene fluorophore scaffold. To achieve this, we adapted our carbonate-based substrate approach to a new fluorophore, 3-O-methylfluorescein (MOF; Fig. 1). We chose this framework for several reasons. First, MOF-based fluorogenic molecules have been successfully used in live cells.^23,24 Second, there is only one substrate site which simplifies in vitro characterization. Third, the 3-O position, while methylated here, could be used as a future site for incorporation of subcellular targeting groups,²⁵ PEGylation to improve water solubility,²⁶ or groups to enhance cellular uptake.²⁷ Finally, MOF-based fluorogenic tools have a straightforward synthesis from commercially available starting materials which makes these probes more accessible for researchers to study CES1 in their systems of interest.

Fig. 1 Synthesis of nonfluorescent MCP-Me and MCP-Et from fluorescein and proposed mechanism of hydrolysis of the carbonate group by CES1 to release fluorescent 3-O-methylfluoroscein (MOF).

We designed two different carbonates to mimic methyl and ethyl esters present in known CES1-substrate drugs:²⁸ 3-O-Methylfluorescein CES1-specific Probe methyl (MCP-Me) and 3-O-Methylfluorescein CES1-specific Probe ethyl (MCP-Et). The addition of the carbonate groups lock MOF in a lactone configuration that has limited fluorescence. Upon hydrolysis of the alkyl group, the carbonate rapidly rearranges to leave as carbon dioxide, allowing the opening of the lactone ring to produce the highly fluorescent carboxylate form of MOF (Fig. 1). Before synthesis, we confirmed that our design of MCP-Me and MCP-Et could interact with CES1 by carrying out rigid receptor-flexible ligand docking with AutoDock Vina^29,30 and a reported crystal structure of human CES1 (PDB ID: 2DR0).³¹ Analysis of the top five scoring poses of MCP-Me (Fig. S1, ESI†) and MCP-Et (Fig. S2, ESI†) suggested that both compounds can bind in the active site of CES1 with MCP-Me and MCP-Et well aligned for nucleophilic attack by Ser221 in two out of five poses. This suggested that both compounds could be substrates of CES1.

MCP-Me and MCP-Et were synthesized in two steps from commercially available fluorescein and the appropriate chloroformate (Fig. 1). Following purification and chemical characterization, we evaluated the fluorescence properties of MOF and the MCP series. MOF displayed fluorescent properties with a broad excitation peak from ca. 400 nm to 500 nm (Fig. S3, ESI†). Excitation of this band results in an emission with a maxima of ca. 525 nm. Exciting MCP-Me (Fig. S3B, ESI†) and MCP-Et (Fig. S3D, ESI†) at these wavelengths produced very limited fluorescence indicating that the fluorescence of these chemical tools is dependent on the hydrolysis of the carbonate group.

Next, we determined if MCP-Me and MCP-Et would undergo hydrolysis in the absence of CES1 in solution. Both MCP-Me and MCP-Et were found to be stable across physiological pH (ca. 6–8) having very little fluorescence compared to an equal amount of MOF (Fig. S4, ESI†). We then characterized the stability of MCP-Me and MCP-Et over time in a simple aqueous solution (1X PBS, pH 7.4) and in a more complex aqueous solution containing amino acids and other simple biomolecules such as glucose (Gibco FluoroBrite DMEM supplemented with 20 mM HEPES, pH 7.4). In PBS, both MCP-Me and MCP-Et show little fluorescence indicating they are stable in this solution for one hour with less than 5% hydrolysis after 3 hours (Fig. S5, ESI†). In the more complex solution, MCP-Et was found to be stable to hydrolysis with less than 5% hydrolysis after an hour (Fig. S6B, ESI†). MCP-Me, however, was unstable with 14% hydrolysis occurring after an hour (Fig. S6A, ESI†).

Following this, we studied the specificity of MCP-Me and MCP-Et for CES1 over the closely related CES2 hydrolase also present in the liver.^2,5,11 Mixing CES1 with MCP-Me and MCP-Et produces rapid fluorescence with MCP-Me being hydrolyzed approximately two times faster by CES1 than CES2 and MCP-Et four times faster (Fig. S7, ESI†). Furthermore, 0.5 μM of MCP-Et was able to show a linear response to CES1 down to 0.0031 units (Fig. S8, ESI†). To better characterize the interaction of MCP-Et with CES1 and CES2, we performed Michaelis–Menten kinetics (Fig. 2). These studies showed MCP-Et has micromolar K_m values for both CES1 and CES2, but with ca. six times greater affinity towards CES1. The V_max values of the hydrolysis of MCP-Et by CES1 and CES2 were found to be similar. This results in CES1 having a greater catalytic efficiency (K_cat/K_m) for the hydrolysis of MCP-Et and suggests that the observed in vitro specificity for CES1 over CES2 of MCP-Et is primarily determined by CES1's higher affinity for MCP-Et.

Fig. 2 Michaelis–Menten kinetics of the interaction of MCP-Et with CES1 and CES2.

Since MCP-Et was stable in solution and responded to CES1 activity, we proceeded to explore if MCP-Et could report on CES1 activity in live HepG2 cells. This cell line was chosen as it expresses high levels of both CES1 and CES2 in a ratio similar to normal human liver tissue.^32–36 Treating HepG2 cells with MCP-Et resulted in the cells fluorescing, indicating that MCP-Et was being hydrolyzed in live cells (Fig. 3A and B and Fig. S9, ESI†). To confirm this was caused by CES1, we knocked down CES1 with shRNA. This resulted in a ca. 60% reduction in fluorescence compared to HepG2 cells treated with scrambled shRNA (control). Importantly, knocking down CES2, did not result in a statistically significant difference in MCP-Et fluorescence (Fig. 3C and D and Fig. S10, ESI†). Notably, fluorescence was observed to be diffuse across the cell suggesting that MOF, after being released from MCP-Et, diffuses out of the endoplasmic reticulum where CES1 is known to be localized. Additionally, treatment of HepG2 cells with MCP-Et for one hour demonstrated no change in cell viability in all concentrations tested up to 100 μM (Fig. S11, ESI†). This indicates that MCP-Et is suitable for use in HepG2 cells and is specific for CES1 in the presence of other hydrolytic enzymes including CES2.

Fig. 3 Epifluorescence images of (A) CES1 and (C) CES2 knockdown in live HepG2 cells. Cells were transfected with scrambled shRNA plasmid (control) or CES1 or CES2 shRNA. Before imaging cells were incubated with 25 μM MCP-Et for 30 min. (B) and (D) Quantification of the fluorescence signal. Error bars are standard errors of the mean (n = 3). Statistical analysis performed with a two-tailed Student's t-test with unequal variance. Scale bars are 20 μm.

With a chemical tool that can report on CES1 activity in live cells in hand, we investigated if MCP-Et could be used to evaluate CES1-mediated DDIs using well-established drugs that are known to interfere with CES1 and CES2 metabolism: troglitazone^37,38 and loperamide,^38,39 respectively. Treating HepG2 cells with troglitazone led to a significant reduction in CES1 activity, as evidenced by a pronounced decrease in MCP-Et's fluorescence intensity (Fig. S12, ESI†). Conversely, treatment with loperamide did not elicit a significant change in MCP-Et's fluorescence (Fig. S13, ESI†). These results demonstrate MCP-Et's utility in reporting on CES1-specific DDIs in HepG2 cells.

To further explore the potential of MCP-Et as a chemical tool, we investigated whether MCP-Et could be used in models of CES1's role in endogenous metabolism. THP-1 macrophages are THP-1 monocytic leukaemia cells differentiated with phorbol-12-myristate-13-acetate to generate cells that display an M0 macrophage phenotype.¹⁴ These cells express CES1 with no detectable expression of CES2 and have been used by other groups to elucidate the role of CES1 in endogenous processes.^13,14 Thus, this cell line is a good model to test if MCP-Et could report on changes in CES1 endogenous metabolism. Additionally, we are unaware of any reports to date that have used fluorogenic chemical tools to study CES1 in this cell line. Treatment of THP-1 macrophages cells with MCP-Et for one hour demonstrated no change in cell viability in all concentrations tested up to 100 μM (Fig. S14, ESI†). Deploying MCP-Et in THP-1 macrophages produced a bright fluorescence signal while treatment with a pan-CES inhibitor bis-p-nitrophenyl phosphate (BNPP; Fig. 4A and B and Fig. S15, ESI†)²⁸ or knocking down CES1 with shRNA (Fig. 4C and D and Fig. S16, ESI†) resulted in a reduction of fluorescence by approximately half indicating that MCP-Et can specifically report on CES1 activity in THP-1 macrophages. We then investigated if troglitazone could interfere with CES1 activity in cells known to have endogenous CES1 metabolic processes using THP-1 macrophages as a model. Treatment of THP-1 macrophages with troglitazone resulted in a ca. 70% decrease in MCP-Et activity indicating that troglitazone could interfere with endogenous CES1 metabolism (Fig. 4E and F and Fig. S17, ESI†). As expected, treatment with loperamide, which is known to not interfere with CES1 metabolism, does not cause a change in fluorescence from MCP-Et in THP-1 cells (Fig. 4G and H and Fig. S18, ESI†). This demonstrates that fluorogenic chemical tools like MCP-Et can be used in THP-1 macrophages to study endogenous CES1 metabolism.

Fig. 4 Epifluorescence microscopy of MCP-Et in live THP-1 macrophages. (A) and (B) Inhibition of CES1 by BNPP. THP-1 macrophages were treated with DMSO (control) or 100 μM BNPP before treatment with 25 μM MCP-Et for 30 min and subsequent imaging by epifluorescence microscopy. (C) and (D) Knockdown of CES1. Cells transfected with scrambled shRNA plasmid (control) or CES1 shRNA were incubated with 25 μM MCP-Et for 30 min before imaging by epifluorescence microscopy. (E)–(H) Influence of drugs on CES1 activity in THP-1 macrophages. THP-1 macrophages were treated with DMSO (control) or (E) 50 μM of troglitazone or (G) 50 μM loperamide for 30 min before treatment with MCP-Et (25 μM) for 30 min and subsequently imaged using epifluorescence microscopy. (B), (D), (F) and (H) Quantification of the fluorescence signal. Error bars are standard errors of the mean (n = 3). Statistical analysis performed with a two-tailed Student's t-test with unequal variance. Scale bars are 20 μm.

In conclusion, MCP-Et, a new CES1-specific fluorogenic tool built on a modifiable framework, was developed and validated in vitro and in live cells. Deploying MCP-Et in two different models of CES1 metabolism demonstrates the utility of well-characterized fluorogenic chemical tools like MCP-Et in identifying potentially harmful CES1-mediated drug interactions.

Research reported in this publication was supported by Eastern Illinois University and the National Institute of General Medical Sciences of the National Institutes of Health under award number R15GM152890. Graphical Abstract created with https://BioRender.com.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. B. Yan, in Encyclopedia of Drug Metabolism and Interactions, ed. A. V. Lyubimov, John Wiley & Sons, Inc., Hoboken, NJ, USA, 2012.
2. A. Singh, M. Gao and M. W. Beck, RSC Med. Chem., 2021, 12, 1142–1153.
3. L. Her and H. J. Zhu, Drug Metab. Dispos. Biol. Fate Chem., 2020, 48, 230–244.
4. T. Imai, M. Taketani, M. Shii, M. Hosokawa and K. Chiba, Drug Metab. Dispos., 2006, 34, 1734–1741.
5. Y. Liu, J. Li and H.-J. Zhu, Expert Opin. Drug Metab. Toxicol., 2024, 20, 377–397.
6. B. Homayun, X. Lin and H.-J. Choi, Pharmaceutics, 2019, 11, 129.
7. Foye's principles of medicinal chemistry, ed. W. O. Foye, T. L. Lemke and D. A. Williams, Wolters Kluwer Health/Lippincott Williams & Wilkins, Philadelphia, 7th edn, 2013.
8. Z. Sun, D. J. Murry, S. P. Sanghani, W. I. Davis, N. Y. Kedishvili, Q. Zou, T. D. Hurley and W. F. Bosron, J. Pharmacol. Exp. Ther., 2004, 310, 469–476.
9. R. Thomsen, H. B. Rasmussen and K. Linnet, Drug Metab. Dispos., 2014, 42, 126–133.
10. D. Shi, J. Yang, D. Yang, E. L. LeCluyse, C. Black, L. You, F. Akhlaghi and B. Yan, J. Pharmacol. Exp. Ther., 2006, 319, 1477–1484.
11. L. Di, Curr. Drug Metab., 2019, 20, 91–102.
12. M. E. Phillips, O. Adekanye, A. Borazjani, J. A. Crow and M. K. Ross, ACS Chem. Biol., 2023, 18, 1564–1581.
13. L. C. Mangum, X. Hou, A. Borazjani, J. H. Lee, M. K. Ross and J. A. Crow, Biochem. J., 2018, 475, 621–642.
14. J. A. Crow, B. L. Middleton, A. Borazjani, M. J. Hatfield, P. M. Potter and M. K. Ross, Biochim. Biophys. Acta, Mol. Cell Biol. Lipids, 2008, 1781, 643–654.
15. J. Lian, R. Nelson and R. Lehner, Protein Cell, 2018, 9, 178–195.
16. S. Casey Laizure, V. Herring, Z. Hu, K. Witbrodt and R. B. Parker, Pharmacotherapy, 2013, 33, 210–222.
17. D. L. Kroetz, F. J. Gonzalez and O. W. McBride, Biochemistry, 1993, 32, 11606–11617.
18. V. Arena de Souza, D. J. Scott, J. E. Nettleship, N. Rahman, M. H. Charlton, M. A. Walsh and R. J. Owens, PLoS One, 2015, 10, e0143919.
19. L. Lan, X. Ren, J. Yang, D. Liu and C. Zhang, Bioorganic Chem., 2020, 94, 103388.
20. A. Singh, M. Gao, C. J. Karns, T. P. Spidle and M. W. Beck, ChemBioChem, 2022, 23, e202200069.
21. M. K. Ross and A. Borazjani, Curr. Protoc. Toxicol., 2007, 33, 4.24.1–4.24.14.
22. J. Dai, Y. Hou, J. Wu and B. Shen, ChemistrySelect, 2020, 5, 11185–11196.
23. B. C. Dickinson, C. Huynh and C. J. Chang, J. Am. Chem. Soc., 2010, 132, 5906–5915.
24. J. Zhang, Y.-Q. Sun, J. Liu, Y. Shi and W. Guo, Chem. Commun., 2013, 49, 11305–11307.
25. P. Gao, W. Pan, N. Li and B. Tang, Chem. Sci., 2019, 10, 6035–6071.
26. S. S. Banerjee, N. Aher, R. Patil and J. Khandare, J. Drug Delivery, 2012, 2012, 103973.
27. R. Zhang, X. Qin, F. Kong, P. Chen and G. Pan, Drug Delivery, 2019, 26, 328–342.
28. D. Wang, L. Zou, Q. Jin, J. Hou, G. Ge and L. Yang, Acta Pharm. Sin. B, 2018, 8, 699–712.
29. J. Eberhardt, D. Santos-Martins, A. F. Tillack and S. Forli, J. Chem. Inf. Model., 2021, 61, 3891–3898.
30. O. Trott and A. J. Olson, J. Comput. Chem., 2009, 31, 455–461.
31. S. Bencharit, C. C. Edwards, C. L. Morton, E. L. Howard-Williams, P. Kuhn, P. M. Potter and M. R. Redinbo, J. Mol. Biol., 2006, 363, 201–214.
32. W. Shang, J. Liu, R. Chen, R. Ning, J. Xiong, W. Liu, Z. Mao, G. Hu and J. Yang, Xenobiotica, 2016, 46, 393–405.
33. X. Wang, B. He, J. Shi, Q. Li and H.-J. Zhu, Drug Metab. Dispos., 2020, 48, 31–40.
34. A. Basit, P. W. Fan, S. C. Khojasteh, B. P. Murray, B. J. Smith, S. Heyward and B. Prasad, Drug Metab. Dispos., 2022, 50, 197–203.
35. S. E. Pratt, S. Durland-Busbice, R. L. Shepard, K. Heinz-Taheny, P. W. Iversen and A. H. Dantzig, Clin. Cancer Res., 2013, 19, 1159–1168.
36. W. Luo, Y. Xin, X. Zhao, F. Zhang, C. Liu, H. Fan, T. Xi and J. Xiong, Br. J. Pharmacol., 2017, 174, 700–717.
37. T. Fukami, S. Takahashi, N. Nakagawa, T. Maruichi, M. Nakajima and T. Yokoi, Drug Metab. Dispos., 2010, 38, 2173–2178.
38. Z. Feng, H. Wang, R. Zhou, J. Li and B. Xu, J. Am. Chem. Soc., 2017, 139, 3950–3953.
39. S. K. Quinney, S. P. Sanghani, W. I. Davis, T. D. Hurley, Z. Sun, D. J. Murry and W. F. Bosron, J. Pharmacol. Exp. Ther., 2005, 313, 1011–1016.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cc03632h

‡ These authors contributed equally to this work.

---