Assessment of μ-Opioid Receptor Signaling Responses for Fentanyl Analogs Using Luciferase Complementation Assay

Abstract

To study the toxicity of fentanyl analogs that damage the liver and kidneys in rats, these analogs were evaluated by examining two types of μ-opioid receptor (MOR) signaling responses using HEK293 cells. The results indicated that, in the MOR-mini-Gi recruitment assay, the 50% effective concentration (EC₅₀) ranked as iBF < DAMGO ≈ 4F-iBF < 4Cl-iBF, and the maximum response (E_max) ranked as iBF ≈ DAMGO > 4F-iBF > 4Cl-iBF. In the MOR-β-arrestin 2 recruitment assay, the EC₅₀ ranking was DAMGO < iBF < 4F-iBF < 4Cl-iBF, and the E_max ranking was DAMGO > iBF > 4F-iBF. In addition, each of the desphenethylated metabolites, likely the major metabolites of these analogs, showed no MOR signaling responses.

INTRODUCTION

Fentanyl analogs, classified as new psychoactive substances, have been increasingly distributed illegally in recent years, posing a serious threat to public health, including their involvement in fatal accidents.¹⁾ We have been investigating the toxicity of fentanyl analogs, which have become a societal concern, and previously found histological lesions in the liver and kidneys of mice administered 4-fluoro-isobutyrylfentanyl (4F-iBF) (Fig. 1A) via intraperitoneal injection. To further investigate, we administered 4F-iBF, along with 4-chloro-isobutyrylfentanyl (4Cl-iBF) and isobutyrylfentanyl (iBF) (Fig. 1A), which have similar chemical structures as 4F-iBF, intraperitoneally to rats and examined their toxicity to the liver and kidneys, as well as their metabolism, in detail. 4F-iBF or iBF caused substantial ischemic-like damage to the liver and kidneys, with iBF also resulting in deaths. By contrast, 4Cl-iBF caused only mild kidney damage. This organ damage may have been induced by respiratory depression mediated through μ-opioid receptors (MORs). Furthermore, nor-4F-iBF, nor-4Cl-iBF, or nor-iBF (Fig. 1B), the likely major desphenethylated metabolites of these analogs,^2,3) were detected in serum.⁴⁾

Fig. 1. Structures of (A) Fentanyl, 4F-iBF, 4Cl-iBF, iBF, and (B) Their Respective Desphenethyl Metabolites

Generally, agonist-bound G protein-coupled receptors (GPCRs) including MORs, activate two primary signaling pathways: the G protein signaling pathway, which alters the receptor’s binding mode to G proteins, and the β-arrestin signaling pathway, in which GPCRs are phosphorylated by GPCR kinases (GRKs) and subsequently recruit β-arrestin.^5,6) Because these two pathways are targets in exploring the effects of MOR agonists, cell-based assays have been developed to monitor the recruitment of mini-Gi,^7,8) an engineered GTPase domain of the Gαi subunit, and β-arrestin 2 (βarr2) to MOR using NanoBiT^® technology⁹⁾.^10–12) Based on these assays, we developed HEK293 cell lines stably expressed either the MOR-mini-Gi system (HEK-MOR-mini-Gi cells (Fig. 2A)) or the MOR-βarr2 system (HEK-MOR-βarr2 cells (Fig. 2B)), and investigated MOR-mediated signaling responses to better understand the relationship between the agonist effects of 4F-iBF, 4Cl-iBF, and iBF on MOR and their toxicity to the liver and kidneys in rats. We also investigated the MOR-mediated signaling responses of nor-4F-iBF, nor-4Cl-iBF, and nor-iBF.

Fig. 2. Schematic Representations of MOR-mini-Gi and MOR-βarr2 Recruitment Assays Applied in This Study

Recruitment of (A) mini-Gi (an engineered GTPase domain of the Gαi subunit) or (B) βarr2 to MOR upon agonist binding is detected in each cell line. NanoBiT^® luciferase, composed of LgBiT fused to the C-terminus of MOR and SmBiT fused to the N-terminus of (A) mini-Gi or (B) βarr2, produces a luminescent signal upon addition of the substrate furimazine, through structural and functional complementation between the two luciferase subunits.

MATERIALS AND METHODS

Reagents and Plasmid Constructs

4F-iBF, 4Cl-iBF, iBF, and their respective desphenethyl metabolites (nor-4F-iBF, nor-4Cl-iBF, and nor-iBF) as monohydrochloride were obtained as described previously.⁴⁾ [d-Ala², N-Me-Phe⁴, Gly⁵-ol]-enkephalin acetate salt (DAMGO) (Selleck Biotechnology, Kanagawa, Japan) was used as a positive control for MOR agonist. Among these compounds, those that precipitated were dispersed during preparation and application.

To utilize the NanoBiT^® System (Promega, WI, U.S.A.), which consists of Large BiT (LgBiT) and Small BiT (SmBiT), the generation of the following expression vectors was entrusted to Kazusa Genome Technologies (Chiba, Japan). The coding region of human OPRM1 (NM_000914.5) was inserted into pBiT1.3-C_LgBiT vector (Promega) along with the linker sequence between OPRM1 and LgBiT as determined from a published procedure,¹⁰⁾ resulting in the pBiT1.3-C_OPRM1-LgBiT vector. Based on another published procedure,¹³⁾ the linker sequence of SmBiT and human ARRB2 (NM_004313.4) in the SmBiT-ARRB2 fusion vector (Promega) was modified, yielding the pBiT2.3-N_SmBiT-ARRB2 vector. The amino acid sequence of mini-Gi was obtained from references,^7,14) then converted to nucleotide sequences using the GenSmart™ Codon Optimization web tool (GenScript Biotech, NJ, U.S.A.). The ARRB2 sequence in the pBiT2.3-N_SmBiT-ARRB2 vector was replaced with the mini-Gi sequence, resulting in the pBiT2.3-N_SmBiT-mini-Gi vector. Finally, the OPRM1-LgBiT sequence from the pBiT1.3-C_OPRM1-LgBiT vector was cloned into either the pBiT2.3-N_SmBiT-mini-Gi or pBiT2.3-N_SmBiT-ARRB2 vector, generating the BiBiT_OPRM1-LgBiT_SmBiT-mini-Gi vector and BiBiT_OPRM1-LgBiT_SmBiT-ARRB2 vectors, respectively.

Cell Culture and Electroporation

HEK293 cells (ATCC, VA, U.S.A.) were maintained at 37°C in a humidified atmosphere with 5% CO₂ in Dulbecco’s modified Eagle’s medium containing 4.5 g/L glucose (Thermo Fisher Scientific, MA, U.S.A.), supplemented with 10% fetal bovine serum (Thermo Fisher Scientific) and 4 mM l-glutamine (Thermo Fisher Scientific).

The BiBiT_OPRM1-LgBiT_SmBiT-mini-Gi or BiBiT_OPRM1-LgBiT_SmBiT-ARRB2 vector was introduced into HEK293 cells, pre-washed with Opti-MEM™ I reduced-serum medium (Thermo Fisher Scientific), using an NEPA21 Type II electroporator (Nepa Gene, Chiba, Japan) according to the manufacturer’s protocol. The cells were maintained under Geneticin™ (Thermo Fisher Scientific) selection. Stably expressing cells were not sorted, as test compounds, including the positive control, were assessed within the same 96-well plate. It was assumed that there would be no significant variability in relative light units (RLUs) between identical concentrations of the same compound.

MOR-mini-G_i and MOR-βarr2 Recruitment Assays

To assess the potency and efficacy of 4F-iBF, 4Cl-iBF, iBF, their respective desphenethyl metabolites, and DAMGO, the Nano-Glo^® Live Cell Assay System (Promega) was used. HEK-MOR-mini-Gi or HEK-MOR-βarr2 cells were seeded in 96-well plates pre-coated with poly-d-lysine at a density of 1.0 × 10⁵ cells/well and 1.2 × 10⁵ cells/well, respectively. Following overnight incubation, the medium in each well was removed, and 100 µL of Opti-MEM™ I reduced-serum medium along with 25 µL of Nano-Glo^® Live Cell Reagent (prepared by combining 1 volume of Nano-Glo^® Live Cell Substrate with 19 volumes of Nano-Glo^® LCS Dilution Buffer) was added to each well. Luminescence was immediately measured for a single time point using a GloMax^® Explorer microplate reader (Promega). Next, 10 µL of a 13.5-fold concentrated stock solution of each test compound (prepared in Opti-MEM™ I reduced-serum medium with a final dimethyl sulfoxide (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan)) concentration of 1% or less was added per well (n = 4 or 3), and luminescence was monitored for 120 min. The final concentrations of each test compound were 10⁻¹⁰, 10⁻⁹, 10⁻⁸, 10⁻⁷, 10⁻⁶, and 10⁻⁵ M. However, 10⁻¹⁰ M was not set for 4Cl-iBF and nor-4Cl-iBF, as these compounds were assumed to have weak effects even at high concentrations. Following luminescence measurement, it was confirmed that none of the test compounds exhibited cytotoxicity within the applied concentration range, as assessed by the CellTiter-Fluor™ Cell Viability Assay (Promega).

Data Analysis

Based on a published procedure,¹⁵⁾ RLU profiles obtained from MOR-mini-Gi and MOR-βarr2 recruitment assays were used to calculate the areas under the curve (AUCs) (n = 4, except for nor-iBF (MOR-mini-Gi, 10⁻⁸ M), where n = 3). The AUC values for test compounds were corrected using the corresponding AUCs of solvent controls and normalized to the maximum response of DAMGO (arbitrarily set to 100%). Various calculations, curve fitting, and statistical analyses were performed using Microsoft Excel (Microsoft, WA, U.S.A.) and GraphPad Prism 9 (Dotmatics, MA, U.S.A.). A nonlinear regression model (three parameters with the Hill slope fixed at 1) was used to determine the 50% effective concentration (EC₅₀) and maximum response (E_max).

To quantify the signaling bias of each test compound, we used the web tool Propagation of Uncertainty Calculator (Nicolas Gnyra) to compute the following: the intrinsic relative activity (RA_i), bias factor (β), and the standard error of the mean (S.E.M.) for β. DAMGO was selected as the reference agonist due to its frequent use in bias quantification studies.¹⁶⁾ Calculations were performed according to the following equations:

  

$${\mathrm{R}\mathrm{A}}_{\mathrm{i},\text{ DAMGO }}^{\text{assay}}=\left(\frac{{\mathrm{E}}_{max,\mathrm{i}}}{{\mathrm{E}\mathrm{C}}_{50,\mathrm{i}}}\right)/\left(\frac{{\mathrm{E}}_{max,\text{ DAMGO}}}{{\mathrm{E}\mathrm{C}}_{50,\text{ DAMGO}}}\right)$$

  

$$\mathrm{\beta }=\log \left(\frac{{\mathrm{R}\mathrm{A}}_{\mathrm{i},\text{ DAMGO}}^{\mathrm{\beta }\mathrm{a}\mathrm{r}\mathrm{r}2}}{{\mathrm{R}\mathrm{A}}_{\mathrm{i},\text{ DAMGO}}^{\text{mini-Gi}}}\right)$$

RESULTS

The coupling of mini-Gi to activated MOR by 4F-iBF, 4Cl-iBF, iBF, or the positive control DAMGO was examined using HEK-MOR-mini-Gi cells. The potency (EC₅₀) of iBF was the lowest at 0.151 µM, lower than that of DAMGO at 0.336 µM, while 4F-iBF had an EC₅₀ of 0.365 µM, similar to that of DAMGO, and 4Cl-iBF had an EC₅₀ of 1.40 µM (Figs. 3A, 3C). The efficacy (E_max, relative to DAMGO, which was arbitrarily set at 100%) of iBF was 107%, comparable to that of DAMGO, whereas 4F-iBF and 4Cl-iBF had E_max values of 69.6 and 27.9%, respectively (Figs. 3A, 3C). On the contrary, βarr2 recruitment to MOR activated by 4F-iBF, 4Cl-iBF, iBF, or DAMGO was examined using HEK-MOR-βarr2 cells. The EC₅₀ of iBF was the lowest at 0.529 µM, slightly higher than that of DAMGO at 0.386 µM, whereas 4F-iBF and 4Cl-iBF had EC₅₀ values of 2.14 µM and greater than 10 µM, respectively (Figs. 3A, 3C). The E_max values were 79.6% for iBF and 33.3% for 4F-iBF; whereas that of 4Cl-iBF could not be determined because its EC₅₀ exceeded the tested concentration range (Figs. 3A, 3C). Furthermore, nor-4F-iBF, nor-4Cl-iBF, and nor-iBF, the respective desphenethylated metabolites, exhibited neither mini-Gi binding nor βarr2 recruitment to MOR (Fig. 3B). The EC₅₀ and E_max values for 4F-iBF and iBF were successfully calculated in both MOR-mini-Gi and MOR-βarr2 recruitment assays and were graphically compared between these assays (Fig. 4). In the MOR-βarr2 assay, the EC₅₀ of both 4F-iBF and iBF were higher than in the MOR-mini-Gi assay, whereas their E_max values were higher in the MOR-mini-Gi assay than in the MOR-βarr2 assay. To estimate biased agonism in G protein and β-arrestin signaling at MOR, β was calculated for 4F-iBF and iBF, compounds for which both EC₅₀ and E_max could be determined in both assays, using DAMGO as the reference agonist. The resulting β values were −1.03 for 4F-iBF and −0.614 for iBF (Figs. 3C and 5).

Fig. 3. MOR Activation Profiles

Concentration–response curves (n = 4 ^a, ± S.E.M.) are shown for (A) 4F-iBF, 4Cl-iBF, and iBF; (B) their respective desphenethyl metabolites; alongside the positive control DAMGO. Data are presented for both MOR-mini-G_i (circles, solid lines) and MOR-βarr2 (squares, dashed lines). The Y-axis represents AUC values normalized to the maximal response of DAMGO. (C) An overview of EC₅₀, E_max (with DAMGO arbitrarily set to 100%), and β for 4F-iBF, 4Cl-iBF, iBF, and the reference agonist DAMGO (n = 4) is summarized. Nor-iBF (MOR-mini-G_i, 10⁻⁸ M) is n = 3. ^a: The normalized AUCs at the highest tested concentration were assumed to represent the E_max values, as they were considered not to have clearly reached the true maximum response. The calculated EC₅₀, E_max, and β values should be interpreted with caution. CI: confidential intervals. ND: not determined.

Fig. 4. Inter-Assay Comparison of Test Compounds for (A) EC₅₀ and (B) E_max (DAMGO Arbitrarily Set to 100%)

EC₅₀ and E_max values for 4F-iBF, iBF, and the reference agonist DAMGO, which were calculated in both MOR-mini-Gi and MOR-βarr2 recruitment assays, are represented by circles (solid lines), squares (dashed lines), and triangles (dotted lines), respectively.

Fig. 5. β Values for MOR, Plotted for 4F-iBF, iBF, and the Reference Agonist DAMGO

A negative or a positive value (n = 4, ± S.E.M.) relative to DAMGO would imply a preference for mini-G_i or βarr2 recruitment, respectively.¹⁵⁾

DISCUSSION

Fentanyl is a relatively selective agonist of the MORs,¹⁷⁾ and MORs mediate respiratory depression by fentanyl.^18–20) It is possible that the opioids linked to cases of acute liver injury are not directly toxic to the liver, but cause ischemic liver injury by means of respiratory failure, cardiovascular collapse, shock, and anoxia that can occur with severe opioid overdose.²¹⁾ It is clear that the susceptibility of the kidney to hypoxia is a major factor in the development of both acute kidney injury and chronic kidney disease.²²⁾ Long-term use of several opioids, especially morphine, can lead to hepatic and renal damage in rats.²³⁾ Therefore, we investigated the effects of the fentanyl analogs 4F-iBF, 4Cl-iBF, and iBF on the liver and kidneys in rats, as well as their MOR agonist activity. The lesions induced by 4F-iBF or iBF may have resembled those found in ischemic injury that can occur with some opioids.⁴⁾ It was found that 4F-iBF was more toxic to the liver and kidneys and had an approximately 7.2-fold higher EC₅₀ on MORs than 4Cl-iBF.⁴⁾ Although iBF exhibited MOR agonist activity equivalent to that of 4F-iBF,²⁴⁾ it showed greater lethal toxicity than 4F-iBF, suggesting that additional toxic mechanisms may be involved. Based on the known effects of fentanyl and other opioids, it was hypothesized that the liver and kidney injuries caused by 4F-iBF and iBF were likely due to ischemia-like damage induced by MOR-mediated respiratory depression.

In this study, to better understand this hypothesis, we investigated the MOR-mediated signaling responses of 4F-iBF, 4Cl-iBF, and iBF using MOR-mini-Gi and MOR-βarr2 recruitment assays. Previous research²⁴⁾ has shown that 4F-iBF and iBF decrease cAMP levels in MOR expressing cells. In the MOR-mini-Gi recruitment assay, the EC₅₀ ranking was iBF < DAMGO ≈ 4F-iBF < 4Cl-iBF, and the E_max ranking was iBF ≈ DAMGO > 4F-iBF > 4Cl-iBF. This indicates that 4F-iBF and iBF exhibit strong potency and efficacy relative to the positive control DAMGO. In the MOR-βarr2 recruitment assay, the EC₅₀ ranking was DAMGO < iBF < 4F-iBF < 4Cl-iBF, and the E_max ranking was DAMGO > iBF > 4F-iBF. Based on DAMGO as the reference, iBF demonstrated strong potency and efficacy, 4F-iBF showed weaker potency and efficacy than iBF, and 4Cl-iBF exhibited the weakest potency. Notably, as the size of the atom in the para position of the anilino group of iBF increased, potency and efficacy appeared to decrease, at least in the G protein signaling pathway at MOR.

Understanding the effects of metabolites on MOR agonistic activity is crucial for 4F-iBF, 4Cl-iBF, and iBF. Some desphenethylated nor-metabolites of fentanyl analogs have been reported to lack MOR agonistic activity, as they do not lower cAMP levels in MOR expressing cells.²⁵⁾ In this study, since the nor-metabolites of 4F-iBF, 4Cl-iBF, and iBF did not exhibit either type of MOR-mediated signaling responses, it was suggested that the phenethyl group plays an important role in MOR activation for these analogs as well.

Approximately 20 years ago, a series of studies led to the hypothesis that the G protein signaling pathway mediates analgesic effects, while the β-arrestin signaling pathway is responsible for respiratory depression, constipation, and analgesic tolerance.^26,27) However, several recent studies have challenged this hypothesis.^26,27) Although the relationship between the MOR signaling bias and pharmacological effects remains unclear, assessing MOR signaling bias is meaningful for understanding complex pharmacological effects. When evaluating the MOR signaling bias for 4F-iBF and iBF, for which EC₅₀ and E_max could be calculated, both compounds appeared to favor activation of the G protein signaling pathway over the β-arrestin signaling pathway. A comparison between MOR-mini-Gi and MOR-βarr2 recruitment assays for 4F-iBF has been previously reported by Vasudevan et al.,¹²⁾ and although our E_max findings showed a similar trend, EC₅₀ and β showed different tendencies. These discrepancies may be attributable to differences in experimental conditions, including reference agonist selection, GRK2 overexpression, and target protein expression levels. Regarding GRK2 overexpression, it has now become evident that GRK2 overexpression facilitates βarr2 recruitment to MOR for low-efficacy ligands such as morphine,^10,28,29) and that the primary GRK subtype responsible for MOR phosphorylation varies depending on the ligand.²⁸⁾ Therefore, GRK2 overexpression may be a factor to consider in bias assessment. In this study, GRK2 was not overexpressed in HEK-MOR-βarr2 cells because we focused on high-efficacy ligands such as DAMGO or fentanyl, and the RLU of DAMGO demonstrated a strong dynamic range in the MOR-βarr2 recruitment assay. Furthermore, given that various studies on the bias of MOR signaling by fentanyl each have their distinct experimental characteristics and have yet to reach a consensus,²⁷⁾ caution should be exercised when comparing MOR activity across different signaling pathways. Therefore, it would be inappropriate to draw definitive conclusions regarding inter-pathway comparisons for 4F-iBF and iBF based solely on the results of the present study.

Acknowledgments

This work was supported by the Pharmaceutical Affairs Section, Health and Safety Division, Bureau of Public Health, Tokyo Metropolitan Government, Tokyo, Japan.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1)  Vandeputte MM, Vasudevan L, Stove CP. In vitro functional assays as a tool to study new synthetic opioids at the μ-opioid receptor: potential, pitfalls and progress. Pharmacol. Ther., 235, 108161 (2022).
• 2)  Watanabe S, Vikingsson S, Roman M, Green H, Kronstrand R, Wohlfarth A. In vitro and in vivo metabolite identification studies for the new synthetic opioids acetylfentanyl, acrylfentanyl, furanylfentanyl, and 4-fluoro-isobutyrylfentanyl. AAPS J., 19, 1102–1122 (2017).
• 3)  Kanamori T, Togawa Iwata Y, Segawa H, Yamamuro T, Kuwayama K, Tsujikawa K, Inoue H. Metabolism of fentanyl and acetylfentanyl in human-induced pluripotent stem cell-derived hepatocytes. Biol. Pharm. Bull., 41, 106–114 (2018).
• 4)  Ono Y, Sakamoto M, Makino K, Tayama K, Tada Y, Nakagawa Y, Nakajima J, Suzuki J, Suzuki T, Takahashi H, Inomata A, Moriyasu T. Hepatic and renal toxicities and metabolism of fentanyl analogues in rats. Naunyn Schmiedebergs Arch. Pharmacol., 396, 149–159 (2023).
• 5)  De Neve J, Barlow TMA, Tourwé D, Bihel F, Simonin F, Ballet S. Comprehensive overview of biased pharmacology at the opioid receptors: biased ligands and bias factors. RSC Med. Chem., 12, 828–870 (2021).
• 6)  Kawakami K, Yanagawa M, Hiratsuka S, Yoshida M, Ono Y, Hiroshima M, Ueda M, Aoki J, Sako Y, Inoue A. Heterotrimeric Gq proteins act as a switch for GRK5/6 selectivity underlying β-arrestin transducer bias. Nat. Commun., 13, 487 (2022).
• 7)  Nehmé R, Carpenter B, Singhal A, Strege A, Edwards PC, White CF, Du H, Grisshammer R, Tate CG. Mini-G proteins: novel tools for studying GPCRs in their active conformation. PLOS ONE, 12, e0175642 (2017).
• 8)  Wan Q, Okashah N, Inoue A, Nehmé R, Carpenter B, Tate CG, Lambert NA. Mini G protein probes for active G protein-coupled receptors (GPCRs) in live cells. J. Biol. Chem., 293, 7466–7473 (2018).
• 9)  Dixon AS, Schwinn MK, Hall MP, Zimmerman K, Otto P, Lubben TH, Butler BL, Binkowski BF, Machleidt T, Kirkland TA, Wood MG, Eggers CT, Encell LP, Wood KV. NanoLuc complementation reporter optimized for accurate measurement of protein interactions in cells. ACS Chem. Biol., 11, 400–408 (2016).
• 10)  Cannaert A, Vasudevan L, Friscia M, Mohr ALA, Wille SMR, Stove CP. Activity-based concept to screen biological matrices for opiates and (synthetic) opioids. Clin. Chem., 64, 1221–1229 (2018).
• 11)  Cannaert A, Deventer M, Fogarty M, Mohr ALA, Stove CP. Hide and seek: overcoming the masking effect of opioid antagonists in activity-based screening tests. Clin. Chem., 65, 1604–1605 (2019).
• 12)  Vasudevan L, Vandeputte M, Deventer M, Wouters E, Cannaert A, Stove CP. Assessment of structure-activity relationships and biased agonism at the Mu opioid receptor of novel synthetic opioids using a novel, stable bio-assay platform. Biochem. Pharmacol., 177, 113910 (2020).
• 13)  Cannaert A, Storme J, Franz F, Auwärter V, Stove CP. Detection and activity profiling of synthetic cannabinoids and their metabolites with a newly developed bioassay. Anal. Chem., 88, 11476–11485 (2016).
• 14)  Wouters E, Walraed J, Robertson MJ, Meyrath M, Szpakowska M, Chevigné A, Skiniotis G, Stove C. Assessment of biased agonism among distinct synthetic cannabinoid receptor agonist scaffolds. ACS Pharmacol. Transl. Sci., 3, 285–295 (2019).
• 15)  Pottie E, Tosh DK, Gao ZG, Jacobson KA, Stove CP. Assessment of biased agonism at the A₃ adenosine receptor using β-arrestin and miniGα_i recruitment assays. Biochem. Pharmacol., 177, 113934 (2020).
• 16)  Conibear AE, Kelly E. A biased view of μ-opioid receptors? Mol. Pharmacol., 96, 542–549 (2019).
• 17)  Toll L, Berzetei-Gurske IP, Polgar WE, Brandt SR, Adapa ID, Rodriguez L, Schwartz RW, Haggart D, O’Brien A, White A, Kennedy JM, Craymer K, Farrington L, Auh JS. Standard binding and functional assays related to medications development division testing for potential cocaine and opiate narcotic treatment medications. NIDA Res. Monogr., 178, 440–466 (1998).
• 18)  Hill R, Santhakumar R, Dewey W, Kelly E, Henderson G. Fentanyl depression of respiration: comparison with heroin and morphine. Br. J. Pharmacol., 177, 254–265 (2020).
• 19)  Jenkins MW, Khalid F, Baby SM, May WJ, Young AP, Bates JN, Cheng F, Seckler JM, Lewis SJ. Glutathione ethyl ester reverses the deleterious effects of fentanyl on ventilation and arterial blood-gas chemistry while prolonging fentanyl-induced analgesia. Sci. Rep., 11, 6985 (2021).
• 20)  Dahan A, Yassen A, Bijl H, Romberg R, Sarton E, Teppema L, Olofsen E, Danhof M. Comparison of the respiratory effects of intravenous buprenorphine and fentanyl in humans and rats. Br. J. Anaesth., 94, 825–834 (2005).
• 21)  National Institute of Diabetes and Digestive and Kidney Diseases. “LiverTox: clinical and research information on drug-induced liver injury [Internet], Opioids, Opioid Antagonists [Updated 2020 Nov 24]” (2012): <https://www.ncbi.nlm.nih.gov/books/NBK547864/pdf/Bookshelf_NBK547864.pdf>, accessed 2 July, 2025.
• 22)  Evans RG, Smith DW, Lee CJ, Ngo JP, Gardiner BS. What makes the kidney susceptible to hypoxia? Anat. Rec. (Hoboken), 303, 2544–2552 (2020).
• 23)  Atici S, Cinel I, Cinel L, Doruk N, Eskandari G, Oral U. Liver and kidney toxicity in chronic use of opioids: an experimental long term treatment model. J. Biosci., 30, 245–252 (2005).
• 24)  Kanamori T, Okada Y, Segawa H, Yamamuro T, Kuwayama K, Tsujikawa K, Iwata YT. Evaluation of agonistic activity of fluorinated and nonfluorinated fentanyl analogs on μ-opioid receptor using a cell-based assay system. Biol. Pharm. Bull., 44, 159–161 (2021).
• 25)  Kanamori T, Okada Y, Segawa H, Yamamuro T, Kuwayama K, Tsujikawa K, Iwata YT. Agonistic activity of fentanyl analogs and their metabolites on opioid receptors. Forensic Toxicol., 40, 156–162 (2022).
• 26)  Pineyro G, Nagi K. Signaling diversity of mu- and delta- opioid receptor ligands: re-evaluating the benefits of β-arrestin/G protein signaling bias. Cell. Signal., 80, 109906 (2021).
• 27)  Ramos-Gonzalez N, Paul B, Majumdar S. IUPHAR themed review: Opioid efficacy, bias, and selectivity. Pharmacol. Res., 197, 106961 (2023).
• 28)  Lemel L, Lane JR, Canals M. GRKs as key modulators of opioid receptor function. Cells, 9, 2400 (2020).
• 29)  Stahl EL, Bohn LM. Low intrinsic efficacy alone cannot explain the improved side effect profiles of new opioid agonists. Biochemistry, 61, 1923–1935 (2022).