Characterization of a Nonselective Opioid Receptor Functional Antagonist: Implications for Development as a Novel Opioid Dependence Medication

Abstract

Opioids represent the most extensive category of abused substances in the United States, and the number of fatalities caused by these drugs exceeds those associated with all other drug overdoses combined. The administration of naltrexone, a potent pan-opioid receptor antagonist, to an individual dependent on opioids can trigger opioid withdrawal and induce severe side effects. There is a pressing demand for opioid antagonists free of opioid withdrawal effects. In our laboratory, we have identified a compound with affinity to mu, delta, and kappa opioid receptors in the range of 150–250 nM. This blood–brain barrier (BBB)-permeant compound was metabolically stable in vitro and in vivo. Our in vivo work demonstrated that 1–10 mg/kg intraperitoneal administration of our compound produces moderate efficacy in antagonizing morphine-induced antiallodynia effects in the chemotherapy-induced peripheral neuropathy (CIPN) model. The treatment was well-tolerated and did not cause behavioral changes. We have observed a fast elimination rate of this metabolically stable molecule. Furthermore, no organ toxicity was observed during the chronic administration of the compound over a 14-day period. Overall, we report a novel functional opioid antagonist holds promise for developing an opioid withdrawal therapeutic.

Drug addiction is a chronic and relapsing disease of the central nervous system (CNS), resulting in compulsive seeking and use despite physical, social, and psychological harm to the user. It directly contributes to an estimated annual cost of over $190 billion from lost work productivity, health care costs, and crime in the United States.¹ Prescription opioid painkillers and heroin are the major class of abused agents in the U.S., resulting in the highest death toll when compared to other drug overdose deaths.¹⁻³ The three approved medications for opioid addiction, methadone, buprenorphine, and naltrexone, are opioid receptor ligands.⁴ Methadone and buprenorphine, full and partial mu opioid receptor (MOR) agonists, respectively, mainly due to their relatively long-lasting half-life activity, can produce dependence by themselves, and discontinuation of these therapies can promote withdrawal symptoms leading to opioid relapse. Naltrexone, a potent nonselective opioid receptor antagonist, when administered to an opioid-dependent individual, can precipitate opioid withdrawal, inducing multiple side effects, such as agitation, nausea, vomiting, pain, and diarrhea.⁵ Thus, a nonselective short-acting antagonist may be therapeutically beneficial as an opioid withdrawal medication.

As part of our ongoing program focused on developing kappa opioid receptor (KOR) ligands,⁶ we have identified a scaffold (compound 1, Table 1) characterized as a nonselective opioid receptor ligand with a similar affinity profile to MOR, delta opioid receptor (DOR), and KOR. The core structure of this scaffold is based on the tricyclic system containing a diketopiperazine (DKP) moiety. To understand structural requirements for the observed affinity of the initial compound 1, we have designed a set of analogs with modifications around the DKP scaffold. Because of the structural similarities between compound 1 and the bicyclic DKP-based modulators of opioid signaling previously characterized in our laboratory,⁶ we have hypothesized that compound 1 can effectively antagonize morphine, a nonselective opioid receptor agonist,⁷⁻⁹ in in vivo experiments. Opioids, including morphine, are often recommended to treat chemotherapy-induced peripheral neuropathy (CIPN).⁶ A hallmark symptom of CIPN is mechanical allodynia, also known as light touch mechanical sensitivity. We also hypothesized that compound 1 would antagonize the antiallodynic effects of morphine and U69,593, a well-characterized KOR agonist, in a well-characterized CIPN model in vivo. This Article describes an initial assessment of the structural requirements contributing to the observed affinity of compound 1. Additionally, we report the evaluation of the pharmacokinetic and pharmacological profile of this novel nonselective opioid ligand using various in vitro and in vivo assays.

Table 1. Structures and Binding Affinities of Key Tricyclic DKP-Based Analogs.

			Ki (nM) ± SEM
compound	R1	R2	MORa	KORa	DORa	μ/κ/δ
1	–OMe	–Me	304 ± 28	365 ± 96	200 ± 58	1.5/1.8/1
2	–OMe	–H	>10,000	>10,000	>10,000
3	–OMe	–Bn	>10,000	>10,000	>10,000
4	–H	–Me	>10,000	>10,000	>10,000

^aK_i (nM) versus [³H] DAMGO (MOR, K_i 3.38 ± 0.44 nM), [³H] U69,593 (KOR, K_i 2.30 ± 0.34 nM), and [³H] DADLE (DOR, K_i 5.56 ± 1.97 nM). All values are mean ± SEM from three separate experiments. In primary screening assays, compounds were tested in triplicate or quadruplicate at a final concentration of 10 μM.

Results and Discussion

Chemistry

We have prepared a set of tricyclic DKP analogs 1–4 using three main steps. First, bicyclic DKP intermediates were prepared as previously reported⁶ via synthesis of corresponding dipeptide molecules (Scheme S2) followed by their intramolecular cyclization (Scheme S2) under basic conditions (NH₄OH, Arrhenius base). In the third step, final compounds 1–4 were prepared using a Ullmann cross-coupling reaction (Scheme 1).¹⁰

Scheme 1. Synthesis of Compounds 1–4 Starting from Bicyclic DKP-Based Analogs^6,

Reagents and conditions: CuI, K₃PO₄, DMEDA, toluene, 110 °C, 20 h.

Receptor binding profile and K_i values were determined at the National Institute of Mental Health’s (NIMH) Psychoactive Drug Screening Program.^11,12 Compounds exhibiting a substantial inhibitory effect of ≥50% in the primary binding assay underwent additional testing in a radioligand binding assay to calculate K_i values (Table 1).

First, we were focused on identifying the role of the N-methyl substitute on 2-N of the DKP ring. In compound 2, the N-methyl group was substituted with a secondary amine, whereas in compound 3, we increased the size of the tertiary substituent, moving from a methyl group to a benzyl group. The binding results indicated that only N-methyl substitution retains affinity to opioid receptors, indicating the importance of only hydrogen-bond accepting (HBA) on 2-N of the DKP ring. In addition, these results highlight the limited available space for compound 1 within the binding pocket of opioid receptors. Next, we assessed the role of the methoxy group on the phenyl ring by substituting this motif with a −H atom (4) (Table 1). In the case of compound 4, we observed a loss of affinity, suggesting that either the electron-donating or HBA properties of the methoxy group are preferred to maintain affinity at the opioid receptors. Overall, in the case of DKP-based tricyclic compounds, we have observed a limited capacity for the compounds to tolerate structural modifications. However, the SAR studies are to be continued.

Biology

Based on our binding assay results, the racemate active compound 1 (Supporting Information) was selected and evaluated in different in vitro and in vivo assays to further understand its properties.

In Vitro Blood–Brain Barrier (BBB) Permeability

Previous studies have indicated that some DKP-containing compounds can cross the BBB, particularly via the passive diffusion process as ideal candidates for new therapeutic agents for brain diseases.¹³ With this knowledge, we evaluated compound 1 in different in vitro assays for its ability to cross the BBB. In these experiments, we used diazepam as the positive control and NaFl as the negative control and a marker of the monolayer’s integrity.¹⁴ To evaluate the BBB permeability, we applied the test compounds at a concentration of 1 μM to the apical chamber of the transwell. The samples were collected from the basolateral chamber at different time points and quantified using an LC-MS/MS system. To investigate the ability of compound 1 to cross the BBB via passive diffusion, we performed a parallel artificial membrane permeability assay (PAMPA).¹⁵ Our results (Figure 1A) demonstrated that compound 1 can cross the BBB via passive diffusion at an approximately five times lower rate than the one observed for diazepam. Next, we utilized a more advanced bEnd-3 monolayer model of BBB permeability to assess the involvement of paracellular and active transport in the BBB permeability of compound 1.¹⁶ After confirming the nontoxic profile of compound 1 on bEnd-3 cells (Figure S1), the experiment was performed. In these experimental settings, the BBB permeability for compound 1 is comparable to diazepam (Figure 1B), suggesting that compound 1 can also utilize paracellular and active transport mechanisms to cross the BBB. Organic anion transporting polypeptides (OATPs) and organic cation transporters (OCTs) are each responsible for an uptake of compounds with negative (OATPs) and positive (OCTs) charges, respectively.¹⁷ These are known to be one of the major brain uptake transporters. Since compound 1 contains the DKP ring with two amide functional groups, we hypothesized that our compound is a substrate for OCTs. To further investigate this, we evaluated the effect of corticosterone (150 μM), a nonselective inhibitor of OCT-1, -2, and -3, on the ability of 1 to cross the bEnd-3 monolayer, expressing all three subtypes of OCTs.^17,18 Our data (Figure 1B) confirmed the initial hypothesis of compound 1 being an OCT’s substrate, as the presence of corticosterone has significantly reduced the ability of compound 1 to permeate the cell monolayer. Our future work will be focused on identifying the specific OCT subtype responsible for the observed effect.

Figure 1.

Transport of compound 1 from apical to basolateral chamber at 1 μM in (A) the PAMPA and across (B) the bEnd-3 monolayer model of the BBB at 37 °C. The apparent permeability coefficient (P_e) was calculated considering the amount of compound crossing the membrane over time. Bar histograms show the mean ± SEM of 3–7 measurements. Statistics: One-way ANOVA followed by Dunnett’s test for comparison of multiple groups vs NaFl or unpaired t test for comparison of diazepam/compound 1 + corticosterone vs compound 1 (ns = not significant, ***p < 0.001, ****p < 0.0001).

In Vitro Bidirectional Transport Assay (P-Glycoprotein)

P-Glycoprotein (P-gp), or ATP-binding cassette subfamily B member 1 (ABCB1), is a cell membrane protein that functions as a biological barrier by expelling toxins and xenobiotics from cells.¹⁹ To ensure that compound 1 is not subject to P-gp efflux, we performed an in vitro efflux transport assay using the Caco-2 monolayer model. The Caco-2 cells are colorectal cancerous epithelial cells overexpressed with P-gp, widely used for the bidirectional (efflux) transport assay.²⁰ In this study, we used verapamil (100 μM) as an inhibitor of P-gp^21,22 and NaFl as a marker of the monolayer’s integrity.¹⁴ The test compound 1 was assessed at a concentration of 1 μM. Our data (Figure 2) demonstrated that compound 1 is not a P-gp substrate or is a poor P-gp substrate, as the calculated mean efflux ratio values were below two (<2).²³

Figure 2.

Bidirectional transport of compound 1 (1 μM) in Caco-2 cell monolayers. The apparent permeability coefficient (P_e) was calculated based on the amount of compound crossing the membrane after 120 min. Bar histograms show the mean ± SEM (N = 5–6). Experiments were performed in the presence and the absence of verapamil (100 μM), a P-gp inhibitor.

In Vitro Metabolic Stability Studies

To predict the metabolic stability of compound 1 in animals, we conducted in vitro assessments using plasma and liver microsomes derived from four species (human, dog, rat, and mouse). The stability of compound 1 (1 μM) was monitored for 24 h, using LCMS/MS to detect the compound’s concentration. We observed that compound 1 is metabolically stable in the plasma and liver microsome of all four species with the predicted half-lives being longer than 24 h (Table 2), suggesting that this compound can be administered intravenously (i.v.) and intraperitoneally (i.p.). The liver microsomal stability results suggest that this compound might also be suitable for oral route administration, but more studies are needed to confirm this hypothesis.

Table 2. In Vitro Metabolic Stability of Compound 1 in Plasma and Liver Microsomes (Human, Dog, Rat, and Mouse)^a.

	plasma				liver microsome
compound 1	human	dog	rat	mouse	human	dog	rat	mouse
t1/2 (min)	>1440	>1440	>1440	>1440	>1440	>1440	>1440	>1440
Clint (μL/min/mg)					<0.48	<0.48	<0.48	<0.48

^aData expressed as the mean of N = 4 experiments.

In Vitro Plasma Protein and Tissue Protein Binding

Following the characterization of compound 1 for its BBB permeability and metabolic stability, we evaluated its plasma protein and tissue protein binding. Using a rapid equilibrium device (RED), we performed the assessment at the concentration of 1 μM using the plasma from four different species and mouse liver and brain homogenates. All samples were quantified by using LCMS/MS. Our results (Table 3) demonstrated that compound 1 has a metformin-like behavior as it did not bind to the plasma proteins or bound to a small extent.²⁴ Compound 1 was 100% in an unbound form, except in the case of human plasma, where we observed the highest plasma protein binding of 25%. Similarly, only the unbound form of compound 1 was identified in the mouse brain and liver homogenates.

Table 3. In Vitro Plasma Protein Binding (Human, Dog, Rat, and Mouse) and Tissue Protein Binding (Mouse Brain and Liver) of Compound 1^a.

	fu
compound 52	human	dog	rat	mouse
plasma	0.75 ± 0.01	1	0.98 ± 0.02	1
brain homogenate	ND	ND	ND	1
liver homogenate	ND	ND	ND	1

^aData expressed as the mean ± SEM of n = 3, N = 4 experiments. ND= not detected.

In Vivo Toxicity Studies of Compound 1

The toxicity studies in the animal models were carried out to assess potential adverse effects following repeated daily exposure to the tested compound 1.²⁵ We used C57BL/6J mice divided into a control group and two treatment groups, each comprised ten mice (n = 10). The treatment groups received a daily dose of either 10 or 20 mg/kg, administered through the i.p. route. Upon completion of the study period (14 days), the animals were sacrificed between 5–10 h postlast injection, and the major body organs, including liver, lung, kidney, heart, and brain, were collected for analysis. The samples were prepared from the organ homogenates (dilution factor (DF) = 4) and quantified using LCMS/MS. As expected from the results of plasma protein and tissue protein binding experiments, we did not observe accumulation of compound 1 in any of the organs analyzed at 10 or 20 mg/kg doses. Of note, throughout the study duration, we did not observe any side effects associated with compound 1, such as sedation or respiratory depression.

In Vivo Pharmacokinetic Studies of Compound 1

Taking into account that no trace of the compound was identified in organs collected from animals sacrificed within 10 h after their last injection, it became evident that the compound was rapidly eliminated from the body. To understand the rate of elimination, we performed in vivo pharmacokinetic analysis. For this purpose, C57BL/6J mice were administered a dose of 10 mg/kg compound 1 through the i.p. route. The animals were sacrificed at different time points (15, 30, 60, 90, and 180 min), and the animal brains and plasma were collected for analysis using LCMS/MS. The brain homogenate was prepared with DF = 4.

Our results from the in vivo pharmacokinetic studies (Figure 3, Table S1) indicated that the T_max for compound 1 occurs at 15 min in both brain and plasma with a mean C_max of 305 ng/g and 1330 ng/mL, respectively. These results were in line with our previous observations. As expected, the compound is rapidly eliminated from the brain and plasma with elimination half-lives of 6.9 and 11 min, respectively, and this is explained by the inability of compound 1 to bind to the tissue proteins. Additionally, simultaneous T_max for compound 1 at 15 min in both plasma and brain further suggests that the elimination rate is higher than the BBB-permeability rate.

Figure 3.

In vivo pharmacokinetic profile of compound 1 (10 mg/kg, i.p.) in mice (N = 3).

Toxicity Profile of Compound 1

We have established the nontoxic nature of compound 1 through various assessments. Initially, we conducted viability assessments in bEnd-3 and Caco-2 cell lines (Figure S1). Subsequently, we extended our evaluation to SH-SY5Y cells, a human-derived neuroblastoma cell line, differentiated to form human neuronal cultures.²⁶ Compound 1 showed no signs of toxicity in any of these settings (Figure S1). To further validate the safety profile of compound 1, we have performed in vivo toxicity studies. In this experiment, three groups of C57BL/6J mice were subjected to daily intraperitoneal injections with either a vehicle or 10 or 20 mg/kg compound 1. Following a 14-day dosing regimen, animals were sacrificed, and organs were collected for subsequent histopathology analysis. The examination of the collected data revealed no evidence of necrosis, inflammation, or neoplasia present across the liver, lung, kidney, and brain sections (Figure 4).

Figure 4.

Representative images of histopathology analysis. The images were obtained from organ slices stained with hematoxylin and eosin (H&E) after the daily intraperitoneal administration of vehicle and 10 and 20 mg/kg compound 1 over 14 days.

Behavioral Pharmacology

As opioid receptor agonists are well-characterized to block acute pain signaling, compound 1 was tested alone and in combination with the prototypical nonselective opioid receptor agonist morphine in the mouse hot plate assay. Mice exposed to a 52 °C hot plate demonstrated a 16.3 ± 2.42 s response latency to the thermal stimulus. Intraperitoneal (i.p.) vehicle was without effect (p = 0.76, student’s t test, pre- vs postvehicle injection). Compound 1 (1–10 mg/kg, i.p.) was devoid of antinociceptive activity in the acute antinociception hot plate assay (p = 0.68). The nonselective opioid receptor agonist, morphine sulfate (1.78–10 mg/kg, i.p.) (F (4, 25) = 36.44, p < 0.0001, one-way ANOVA), dose-dependently produced acute antinociception in the 52 °C hot plate assay (Figure 5).

Figure 5.

Effects of intraperitoneal (i.p.) morphine alone and in combination with either naltrexone or compound 1 on hot plate antinociception in mice. Abscissae: compound concentrations in mg/kg. Ordinates: percentage of maximum possible antinociceptive effect. Vehicle (i.p., green circle) does not produce appreciable antinociception. The nonselective opioid receptor agonist morphine (1.78–10 mg/kg, i.p.) produced significant antinociception. Compound 1 (1–10 mg/kg, i.p., upward-facing blue triangles) and naltrexone (0.032 mg/kg, i.p., blue circle) alone did not alter antinociceptive thresholds. When administered 5 min before morphine, naltrexone (0.032 mg/kg, i.p., upward-facing purple triangles) produced a rightward shift of the dose–response curve. When administered 10 min before morphine (1.78–10 mg/kg, i.p.), compound 1 produced dose-related antagonism of morphine antinociception, as indicated by a rightward shift of the dose–response curve. Each symbol represents the mean ± SEM (n = 6 mice).

To examine the functional opioid receptor antagonist activity of compound 1 compared to a gold-standard opioid receptor antagonist, we next administered the opioid receptor antagonist naltrexone (i.p.) 5 min before morphine. When compared to vehicle, naltrexone (0.032 mg/kg, i.p.) alone did not significantly alter nociceptive response thresholds (p = 0.16, student’s t test). Administration of naltrexone (0.032 mg/kg, i.p.) significantly antagonized morphine-induced antinociception (F (3, 40) = 6.35, p < 0.01, two-way ANOVA; Figure 5). At a low dose of 1 mg/kg, compound 1 pretreatment had minimal effect on morphine-induced hot plate antinociception (p = 0.78). Pretreatment with 3.2 mg/kg compound 1 produced a small but significant right-ward shift of the morphine dose–response curve (F (3, 44) = 10.96, p < 0.05, two-way ANOVA; Figure 5). Pretreatment with 10 mg/kg compound 1 robustly antagonized morphine-induced antinociception (F (3,40) = 4.31, p < 0.01, two-way ANOVA; Figure 5).

Next, we examined compound 1 in the paclitaxel CIPN model. Compound 1 was tested alone and in combination with the prototypical nonselective opioid receptor agonist morphine as well as the KOR-selective agonist U69,593 in the mouse von Frey assay. Mice that received paclitaxel developed significant mechanical allodynia (p < 0.0001, student’s t test, pre- vs postpaclitaxel injection). Intraperitoneal (i.p.) vehicle was without effect (p = 0.19, student’s t test, pre- vs postvehicle injection). Compound 1 (1–10 mg/kg, i.p.) was devoid of antiallodynic activity in CIPN mice (p = 0.31). The nonselective opioid receptor agonist, morphine sulfate (1.78–10 mg/kg, i.p.) (F (4, 55) = 65.06, p < 0.0001, one-way ANOVA), dose-dependently attenuated paclitaxel-induced mechanical allodynia (Figure 6A).

Figure 6.

Effects of intraperitoneal (i.p.) morphine and U69593 alone and in combination with either naltrexone or compound 1 on paclitaxel-induced mechanical allodynia. Abscissae: compound concentrations in mg/kg. Ordinates: stimulus intensity in grams required to produce a response. (A) Vehicle + vehicle mice (i.p., red diamond) do not display mechanical allodynia, whereas paclitaxel + vehicle mice (i.p., blue diamond) do display mechanical allodynia. The nonselective opioid receptor agonist morphine (1.78–10 mg/kg, i.p., dark green circle) produced dose-related mechanical allodynia attenuation. Compound 1 (1–10 mg/kg, i.p., upward-facing purple triangles) and naltrexone (0.032 mg/kg, i.p., green circle) alone did not alter antinociceptive thresholds. When administered 15 min before morphine, naltrexone (0.032 mg/kg, i.p., pink upward-facing triangles) produced a rightward shift of the dose–response curve. When administered 15 min before morphine (1.78–10 mg/kg, i.p.), compound 1 produced dose-related antagonism of morphine antiallodynia, as indicated by a rightward shift of the dose–response curve. (B) The KOR-selective agonist U69,593 (0.32–5.6 mg/kg, i.p., purple circles) produced dose-related mechanical allodynia attenuation. When administered 15 min before morphine (3.2–10 mg/kg, i.p.), compound 1 produced significant antagonism of U69,593-induced allodynia attenuation, as indicated by a rightward shift of the dose–response curve. Each symbol represents the mean ± SEM (n = 6 mice).

To examine the functional opioid receptor antagonist activity of compound 1 in the CIPN model compared to a gold-standard opioid receptor antagonist, we next administered the opioid receptor antagonist naltrexone (i.p.) 5 min before morphine. When compared to vehicle, naltrexone (0.032 mg/kg, i.p.) alone did not significantly alter nociceptive response thresholds (p = 0.67, student’s t test). Administration of naltrexone (0.032 mg/kg, i.p.) significantly antagonized morphine-induced antinociception (F (2, 66) = 8.27, p < 0.001, two-way ANOVA; Figure 6A). At a low dose of 1 mg/kg, compound 1 pretreatment produced a nonsignificant trend of morphine-induced antiallodynia antagonism (p = 0.35). Pretreatment with 3.2 mg/kg compound 1 produced a significant right-ward shift of the morphine dose–response curve (F (2, 66) = 3.32, p < 0.05, two-way ANOVA; Figure 6A). Pretreatment with 10 mg/kg compound 1 robustly antagonized morphine-induced mechanical allodynia attenuation (F (2, 66) = 9.50, p < 0.01, two-way ANOVA; Figure 6A).

To assess the KOR-selective antagonist activity of compound 1, we examined it in combination with U69,593, a known KOR agonist.²⁷ U69,593 (0.32–5.6 mg/kg, i.p.) (F (6, 77) = 25.42, p < 0.0001, one-way ANOVA) dose-dependently attenuated paclitaxel-induced mechanical allodynia (Figure 6B). Pretreatment with 3.2 mg/kg compound 1 produced a robust and significant right-ward shift of the U69,593 dose–response curve (F (2, 66) = 18.08, p < 0.0001, two-way ANOVA; Figure 6B). Pretreatment with 10 mg/kg compound 1 also robustly antagonized U69,593-induced mechanical allodynia attenuation (F (2, 66) = 8.71, p < 0.01, two-way ANOVA; Figure 6B).

To summarize the findings of our behavioral pharmacology studies, we employed morphine as a nonselective agonist of MOR, DOR, and KOR⁷⁻⁹ to investigate the functional characteristics of compound 1. The outcomes of our experiments established that compound 1 can block the signaling nonselectively induced across all three opioid receptor subtypes. It is noteworthy, however, that morphine exhibits a higher degree of selectivity toward MOR in comparison to DOR and KOR.^28,29 To provide a more comprehensive understanding of compound 1 pharmacology, we observed the ability of our compound to inhibit U69,593-induced²⁷ responses, confirming its KOR antagonistic properties. Our future studies will include an assessment of its effect on signaling induced by DOR agonists, such as SNC80 or DPDPE.^30,31 This comprehensive approach will provide a nuanced understanding of the receptor-specific actions of compound 1, ensuring a thorough exploration of its pharmacological profile across all three opioid receptor subtypes.

Conclusion

In this paper, we report an initial evaluation of a novel nonselective opioid receptor ligand, compound 1, with approximately equal affinities to three subtypes, MOR, DOR, and KOR. Compound 1 demonstrates metabolic stability and the ability to cross the BBB via passive diffusion and OCT transport mechanisms. We have observed a rapid elimination of compound 1 from the body of C57BL/6J mice, probably due to its minimal protein binding. Pharmacokinetic analysis revealed a plasma-to-brain ratio of 1:0.25, with the complete elimination at 60 min postinjection. However, the ratio can be potentially optimized by reducing the compound’s elimination rate. Compound 1 behaves as a functional antagonist, blocking the morphine-induced effect in mice, similarly to the action of naltrexone. In addition, we have observed that compound 1 can antagonize U69593-induced attenuation of allodynia in the CIPN model.

Given the in vivo functionality of compound 1 as a nonselective functional opioid receptor antagonist, we see the potential for further optimization of this compound to create an effective therapeutic for opioid withdrawal. Notably, buprenorphine, an FDA-approved drug for treating opioid dependence,³² carries drawbacks such as respiratory depression as well as some withdrawal symptoms attributed partly to its prolonged half-life (3–44 h), large volume of distribution, and extensive protein binding (up to 96%).³²⁻³⁴ Recognizing these limitations, we propose that a more optimized compound characterized by reduced protein binding, a faster elimination profile, and a diminished side effect profile may offer therapeutic advantages over buprenorphine. This suggests a promising avenue for the development of a novel opioid withdrawal therapeutic with improved safety and efficacy profiles compared to existing options.

Although outside the scope of the present study, additional work is needed to assess compound 1 to determine its potential to precipitate opioid withdrawal symptoms and its relative abuse potential. We will continue our work on identifying structural features of compound 1 that contribute to the observed activity and on refining its elimination rate via the structural optimization route.

Experimental Section

Chemistry

In all experiments, reagent-grade solvents and chemicals were employed, unless specified otherwise. These high-quality reagents and solvents were procured from commercial vendors and utilized in their as-received state. The compounds were assessed for their purity and characterization through a comprehensive set of methods, incorporating thin-layer chromatography (TLC), ultra-performance liquid chromatography (UPLC) with detection at 280 nm, high-resolution mass spectrometry (HRMS), and nuclear magnetic resonance (NMR) analysis. The ¹H and ¹³C NMR spectra were acquired using a Bruker 400 MHz Advance III HD spectrometer, with CDCl₃ or DMSO-d₆ as solvents. The chemical shifts (δ) are documented in ppm while coupling constant (J) values are recorded in hertz (Hz). The multiplicity is denoted as s (singlet), br s (broad singlet), d (doublet), dd (doublet of doublets), t (triplet), and m (multiplet). Thin-layer chromatography (TLC) was conducted using Silicycle UltraPure SILICA GELS F-254 as the stationary phase. Visualization of spots was achieved through UV light exposure or ninhydrin staining. Flash column chromatography was executed on a Teledyne Isco (CombiFlash NEXTGEN 300+). High-resolution mass spectra were acquired employing a Q-Exactive HF Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermofisher Scientific). Prior to analysis, the Q-Exactive instrument underwent calibration for mass accuracy, achieving a remarkable mass error of <1 ppm across all samples. Spectra were recorded through direct infusion from a syringe pump, maintaining a consistent 5 μL/min flow. Formula determination relied on the observation and analysis of singly charged ions. All key compounds were determined to be at least 95% pure, utilizing Waters Acquity UPLC with a C18 column (Kinetex, 50 mm × 2.1 mm, 2.6 μm 100A). The gradient elution method employed LC–MS/MS-grade water with 0.1% trifluoroacetic acid (A) and LC–MS/MS-grade acetonitrile (ACN) with 0.1% trifluoroacetic acid (B). The injection volume was 2 μL, and the flow rate was maintained at 0.3 mL/min. The gradient sequence included 0:0 min; 5% B, 0–3:0 min; reaching 95% B, 3:0–7:0 min; holding at 95% B, 7:0–7:10 min; returning to 5% B, 7:10–15:0 min; and concluding at 5% B. For chiral screening of the active analog, Waters ACQUITY UPLC H-Class was employed with a Lux Cellulose 2 column (100 mm × 4.6 mm, 3 μm). The gradient elution involved (A) 20 mM ammonium bicarbonate in LC–MS/MS-grade water with 0.1% diethyl amine and (B) LC–MS/MS-grade acetonitrile (ACN) with 0.1% diethyl amine. The injection volume was 2 μL, and the flow rate was set at 1.0 mL/min. The gradient sequence comprised 0–20:0 min; 60% B.

For the synthetic procedures and characterization data for the intermediates, see Schemes S1 and S2. The starting materials 3-methoxy benzyl alcohol (a1) and 2-iodobenzyl bromide (c2) (Scheme S1) and intermediates N-Boc-glycine (i1), Boc-sarcosine (i2), and 2-(benzyl(tert-butoxycarbonyl)amino)acetic acid (i3) (Scheme S2) were purchased commercially.

General Procedure for the Synthesis of Key Compounds 1–4 (Intramolecular Cyclization, Scheme 1)

The intermediate bicyclic compounds m1 or m2 or m3 or m4 (1.3 mM) were dissolved in toluene. CuI (0.13 mM) and K₃PO₄ (2.6 mM) and, then, DMEDA (0.26 mM) were added at room temperature. The reaction mixture was refluxed for 20 h at 110 °C. Reaction progress was monitored by TLC. Upon completion of the reaction, the reaction mixture was filtered, and the solvent was removed in vacuo. The desired compound was then isolated by Combi flash (mobile phase: 2% MeOH in DCM) to afford pure tricyclic compounds 1–4.

6-Methoxy-2-methyl-2,3,10,10a-tetrahydropyrazino[1,2-a]indole-1,4-dione (1)

White powder (480 mg, Yield: 75%). ¹H NMR (400 MHz, DMSO-d₆): δ_H 7.16 (1H, dd, J = 8.25, 7.40 Hz), 6.91–6.99 (2H, m), 4.85 (1H, t, J = 10.03 Hz), 4.45 (1H, dd, J = 16.08, 1.41 Hz), 3.82 (1H, d, J = 16.26 Hz), 3.78 (3H, s), 3.14–3.29 (2H, m), 2.91 (3H, s); ¹³C NMR (101 MHz, DMSO-d₆): δ_c 165.5, 160.2, 150.3, 134.7, 130.3, 127.5, 117.4, 113.1, 62.4, 56.4, 53.6, 35.1, 33.2. HRMS (ESI): calculated for C₁₃H₁₄N₂O₃ [M + H]⁺, 247.1004; found, 247.1076. UPLC purity: 100%.

6-Methoxy-2,3,10,10a-tetrahydropyrazino[1,2-a]indole-1,4-dione (2)

Yellow powder (128 mg, Yield: 71%). ¹H NMR (400 MHz, DMSO-d₆): δ_H 8.22–8.47 (1H, m), 7.09–7.20 (1H, m), 6.91–6.98 (2H, m), 4.85 (1H, t, J = 10.09 Hz), 4.27 (1H, d, J = 15.77 Hz), 3.78 (3H, s), 3.63 (1H, d, J = 15.89), 3.09–3.31 (2H, m); ¹³C NMR (101 MHz, DMSO-d₆): δ_c 167.5, 160.84, 150.2, 134.2, 130.4, 127.3, 117.4, 113.2, 62.2, 56.4, 46.8, 34.6. HRMS (ESI): calculated for C₁₂H₁₂N₂O₃ [M + H]⁺, 233.0848; found, 233.0915. UPLC purity: 99%.

2-Benzyl-6-methoxy-2,3,10,10a-tetrahydropyrazino[1,2-a]indole-1,4-dione (3)

White powder (102 mg, Yield: 78%). ¹H NMR (400 MHz, CDCl₃): δ_H 7.18–7.29 (5H, m), 7.04–7.09 (1H, m), 6.81 (2H, t, J = 8.13 Hz), 4.77 (1H, t, J = 10.03 Hz), 4.70 (1H, d, J = 14.43 Hz), 4.45 (1H, d, J = 14.55 Hz), 4.17 (1H, d, J = 16.14), 3.80 (3H, s), 3.75 (1H, d, J = 16.14 Hz), 3.31 (2H, d, J = 9.90 Hz); ¹³C NMR (101 MHz, CDCl₃): δ_c 165.5, 159.5, 150.2, 135.0, 133.8, 129.8, 128.9 (C × 2), 128.3 (C × 2), 128.2, 127.6, 117.1, 112.7, 62.6, 56.5, 51.3, 49.5, 35.3. HRMS (ESI): calculated for C₁₉H₁₈N₂O₃ [M + H]⁺, 323.1317; found, 323.1392. UPLC purity: 98%.

2-Methyl-2,3,10,10a-tetrahydropyrazino[1,2-a]indole-1,4-dione (4)

White powder (121 mg, Yield: 84%). ¹H NMR (400 MHz, CDCl₃): δ_H 7.88 (1H, d, J = 6.97 Hz), 7.04–7.12 (2H, m), 6.97 (1H, d, J = 7.34 Hz), 4.61 (1H, t, J = 9.90 Hz), 4.19 (1H, d, J = 16.87 Hz), 3.74 (1H, d, J = 16.90 Hz), 3.30–3.41 (1H, m), 3.12–3.24 (1H, m), 2.94 (3H, s); ¹³C NMR (101 MHz, CDCl₃): δ_c 166.9, 161.9, 140.7, 129.8, 127.6, 124.9, 124.8, 115.6, 59.6, 54.0, 33.7, 31.2. HRMS (ESI): calculated for C₁₂H₁₂N₂O₂ [M + H]⁺, 217.0899; found, 217.0970. UPLC purity: 99%.

Biology

All studies involving animals were preapproved and carried out in accordance with the Institutional Animal Care and Use Committees at the Texas Tech University Health Sciences Center as specified by the 2008 National Institutes of Health Guide for the Care and Use of Laboratory Animals. Consistent with these guidelines, ongoing statistical testing of data collected was used to minimize the number of animals used within the constraints of necessary statistical power.

Cytotoxicity Assay (MTT)

The MTT assay was utilized to assess the cytotoxicity of the test compound following the previously established procedure.^6,35,36 In brief, bEnd-3 cells (passage 25–28) (ATCC, Manassas, VA) or Caco-2 cells (passage 21–25) were seeded into 96-well plates (8 × 10³ cells per well) in 100 μL of complete medium. The complete medium included a mixture of Dulbecco’s Modified Eagle Medium (DMEM) (Fisher Scientific, Waltham, MA, USA), 10% Fetal Bovine Serum (FBS) (Thermo Fisher Scientific, Waltham, MA, USA) in the case of bEnd-3 cells and 20% FBS in the case of Caco-2 cells, 1% penicillin–streptomycin (Fisher Scientific, Waltham, MA, USA), and 1% Non-Essential Amino Acids (NEAA) (Fisher Scientific, Waltham, MA, USA). The cells were incubated at 37 °C in a 5% CO₂ humidified incubator for 60 h. Then, the medium was removed, and the wells were washed with 150 μL of Phosphate-Buffered Saline (PBS) (Fisher Scientific, Waltham, MA, USA). Subsequently, the cells underwent a 24-h treatment with the test compound. The dimethyl sulfoxide (DMSO) stock solution of the test compound was diluted in DMEM containing 2% FBS, 1% penicillin–streptomycin, and 1% NEAA to achieve the concentrations of 1, 2.5, 5, 10, 15, and 20 μM, with the final concentration of DMSO maintained below 0.1% in the treatment medium in all wells. Following the 24-h incubation period, the wells were treated with 10 μL of 5 mg/mL MTT stock solution in PBS. After an additional 2 h of incubation, the medium with MTT was removed to obtain crystals. 100 μL of DMSO was added to dissolve the crystals. The absorbance of the formazan product was measured at a wavelength of 570 nm using a spectrophotometric plate reader (Epoch, BioTek Instruments, Vermont, United States). The experiment was performed in triplicate (N = 3, n = 15) for the test compound. The cell viability was normalized to the control group (taken as 100%), and the graphs were prepared using GraphPad Prism 9 Software. SH-SY5Y neuroblastoma cells (10 × 10³ cells per well) were seeded in 96-well plates in 100 μL of DMEM-F12 (Thermo Fisher Scientific, Waltham, MA, USA) medium containing 10% and 1% penicillin–streptomycin and incubated overnight. To promote differentiation, the serum-containing medium was replaced with Neurobasal medium (Thermo Fisher Scientific, Waltham, MA, USA) containing B27 supplement (Thermo Fisher Scientific, Waltham, MA, USA), GlutaMAX (Thermo Fisher Scientific, Waltham, MA, USA), and 10 μM all-trans-retinoic acid (ATRA) (Fisher Healthcare, Waltham, MA, USA). Then, the cells were allowed to differentiate for 4 days, after which the cells were treated with different concentrations of the compounds (1, 2.5, 5, 10, 15, and 20 μM). After 24 h, MTT solution was added at a final concentration of 0.5 mg/mL and incubated at 37 °C for 2 h. Then, the media were replaced with 100 μL of DMSO. The optical density was recorded at the wavelength of 570 nm.

In Vitro Blood–Brain Barrier (BBB) Permeability

Parallel Artificial Membrane Permeability Assays (PAMPAs)

For early stage BBB permeability studies, the PAMPA kit was purchased from BioAssay Systems (Hayward, CA, USA). NaFl was used as a marker to confirm the integrity of the monolayer. For each compound, the permeability was measured over 5 h, and the experiment was performed in 4 replicates following the protocol provided by BioAssay Systems, with minor modifications. BBB lipid solution was prepared in dodecane and was directly added to the well membranes of the donor plate without puncturing the membranes. Then, in each well, 200 μL of each test compound (1 μM) in HBSS buffer containing 1 μM NaFl was introduced to the apical chamber, and 300 μL of blank HBSS buffer was added to the basolateral chamber. After 5 h of incubation at 37 °C, 100 μL of assay buffer was collected from the basolateral chamber for concentration determination and NaFl absorbance detection. The concentration of compounds and equilibrium standards was determined by LC–MS/MS. The apparent permeability coefficient (P_e in cm/min) was calculated based on the cleared volume of the test compound after 5 h, using the equations:

wherein

where donor volume is 0.2 cm³, acceptor volume is 0.3 cm³, membrane area (area) is 0.24 cm², and time is 5 h (18 000 s).

bEnd-3 Monolayer Model of BBB Permeability

The model was developed, and the experiment was performed as described previously.⁶ Briefly, bEnd3 cells were cultured in a complete medium composed of DMEM supplemented with 10% FBS, 1% penicillin–streptomycin, and 1% NEAA in a T75 Corning cell culture flask (Sigma, St. Louis, MO, USA). The cells were incubated in a 5% CO₂ humidified incubator at 37 °C for 8 days or until confluency. Meanwhile, the growth medium was refreshed every other day. For assessment of BBB permeability in vitro, 12-well Corning transwell inserts (0.4 μm pore size) (Fisher Scientific, Waltham, MA, USA) were used. The bEnd-3 cells (5 × 10⁴ cells per filter) in 500 μL of complete medium were seeded onto the apical chamber of the transwell inserts, and 1.5 mL of blank medium was added to the basolateral chamber. The transwell inserts underwent an incubation period of 8 days, during which the growth medium in both compartments was refreshed every other day. To assess and validate the integrity of the developed monolayers, NaFl was employed as a marker. Apparent permeability coefficient (P_e, in cm/min) for each compound was determined by calculating the cleared volume at each time point, following an established procedure.¹⁶

A permeability measurement for each compound was performed in 3–7 replicates. After the removal of the medium, the transwells were washed with fresh HBSS. Next, 500 μL of test compound (1 μM) in HBSS buffer containing 1 μM NaFl was added to the apical chamber. At the same time, we added 1.5 mL of blank HBSS buffer to the basolateral chamber. Following the addition of compounds, 100 μL of assay buffer was collected from the basolateral chamber in duplicate for concentration determination and NaFl absorbance detection. The collection was performed at specific time points (0, 30, 60, and 120 min). To avoid the back diffusion of the tested compound, the removed volume was replaced with an equivalent volume of fresh buffer. The concentration of compounds in samples was determined using LC–MS/MS. To account for resistance induced by the transwell insert membrane, we have measured the permeability of each compound in a blank insert lacking cells. The final P_e for each compound was calculated considering the permeability of each compound in a blank insert.

In Vitro Bidirectional Transport Assay (P-Glycoprotein)

The monolayer model of Caco-2 cells was developed like the bEnd-3 monolayer model discussed earlier. Briefly, the Caco-2 cells (passage 21–25) were seeded at a concentration of 6 × 10⁴ cells per well in 500 μL of complete medium containing DMEM, 20% FBS, 1% penicillin–streptomycin, and 1% NEAA onto the apical chamber of the transwell inserts, and 1.5 mL of blank medium was added to the basolateral chamber. The transwell inserts were incubated for 8 days. Transport of compound 1 from the upper/donor to lower/receiver compartment (A > B) and lower/receiver to upper/donor (B > A) of the transwell inserts was evaluated in the presence/absence of verapamil (100 μM), a P-gp inhibitor.²² Approximately 30 min before performing the experiment, the medium in both compartments was replaced with respective volumes of blank HBSS with/without verapamil. The inhibitor was present on both sides of the membrane during the preincubation and transport period. The experiment was started by removing HBSS from both compartments and replacing it with HBSS containing 1 μM compound 1, upper/donor compartment for A > B, lower/receiver compartment for B > A, and blank HBSS in opposite compartment. To confirm the integrity of the cell monolayer in each well, NaFl was added with/without the test compound to the upper/donor compartment at the beginning of the experiment. Samples were collected from both chambers independently following 2 h of incubation at 37 °C. The concentration of test compounds in the transport buffer was quantified by LC-MS/MS. The apparent permeability coefficient (P_e, cm/min) and efflux ratios were calculated using the literature method.³⁷

In Vitro Metabolic Stability Studies

Liver Microsomal Stability

Following the previously established procedure,⁶ we utilized corning male rat (Sprague–Dawley), human, CD-1 mouse, and canine beagle liver microsomes (Fisher Scientific, Waltham, MA, USA; 20 mg protein/mL in 250 mM sucrose) to assess the liver microsomal stability of the test compound. Briefly, a 100 μM stock solution of the test compound (100× final concentration) was prepared in 0.1 M potassium phosphate buffer, pH 7.4. Concurrently, we prepared the NADPH generating system by mixing solution A and solution B (5:1 ratio) of the Corning NADPH regenerating system (Corning Incorporated, Corning, NY, USA). Next, 5 microcentrifuge tubes, each assigned for a specific time point (0, 3, 6, 12, and 24 h), and one tube for the negative control were spiked with 415 μL of blank 0.1 M potassium phosphate buffer, pH 7.4, prewarmed to 37 °C. It was followed by the addition of 30 μL of the NADPH generating system and 5 μL of 100× test compound. In the negative control, the NADPH generating system was replaced with an equivalent volume of prewarmed 0.1 M potassium phosphate buffer. The tubes were incubated at 37 °C for 5 min. Subsequently, the thawed microsome was diluted in prewarmed 0.1 M potassium phosphate buffer to achieve 5 mg protein/mL concentration (10× final concentration). The addition of a 50 μL aliquot of microsome to each tube initiated the reaction. The sample in each tube was briefly vortexed and then incubated at 37 °C in a shaking water bath. At specific time points, the addition of 250 μL of ice-cold acetonitrile to the respective tube and its vortexing stopped the reaction. The experiment was continued by centrifugation of the tube for 5 min at 10 000g at 4 °C. In the case of the 0 h sample, the reaction was stopped immediately after adding the microsome to the tube. The supernatant fraction was collected and subjected to LC–MS/MS analysis. The experiment was done in quadruplicate (N = 4). The data obtained from the analysis were then employed to compute the half-life and intrinsic clearance using the principles of pseudo-first-order kinetics, which describe the compound’s disappearance rate over time.

Plasma Stability

To assess the plasma stability of compound 1, human plasma with heparin (Oklahoma Blood Institute, Oklahoma City, OK, USA), rat (Sprague–Dawley) plasma with heparin (Innovative Research, Novi, MI, USA), CD-1 mouse plasma with heparin (Biochemed Services, Winchester, VA, USA), and canine beagle plasma with heparin (Biochemed Services, Winchester, VA, USA) were used, following a previously described procedure.⁶ Briefly, a stock solution of compound 1 was prepared at 100 μM concentration (100× final concentration) in LC–MS/MS-grade water. Then, a 1 μM final concentration of the test compound in the plasma was achieved by spiking the 100× compound 1 stock solution in plasma from each species, which was followed by incubation at 37 °C on a shaking water bath. At designated time points (0, 3, 6, 12, and 24 h), 50 μL sample aliquots were collected for quantification using LC–MS/MS. The experiment was performed in quadruplicate (N = 4). Subsequently, the acquired results were employed to compute the half-life using the principles of pseudo-first-order kinetics, elucidating the rate of disappearance of the compound over the specified time course.

In Vitro Plasma Protein and Tissue Protein Binding

Human plasma with heparin (Oklahoma Blood Institute, Oklahoma City, OK, USA), rat (Sprague–Dawley) plasma with heparin (Innovative Research, Novi, MI, USA), CD-1 mouse plasma with heparin (Biochemed Services, Winchester, VA, USA), and canine beagle plasma with heparin (Biochemed Services, Winchester, VA, USA) were used to determine the plasma protein binding of compound 1, as previously described.⁶ Additionally, mouse (C57BL/6J) liver and brain homogenate (DF = 4) was used to determine the tissue protein binding of this compound. For this purpose, RED Device Single-Use Plates with Inserts (membrane molecular weight cutoff = 8 kDa; Thermo Fisher Scientific, Waltham, MA, USA) were used, following the procedure described by the manufacturer. A stock solution of compound 1 was prepared at 100 μM concentration (100× final concentration) in LC–MS/MS-grade water. Then, the plasma from each species or the mouse liver and brain homogenates were spiked with the 100× compound 1 stock solution to obtain a final concentration of 1 μM. Then, 200 μL aliquot of plasma or tissue containing 1 μM of the test compound was added to the sample chamber of the inset (donor), and 400 μL of blank PBS was added to the buffer chamber of the dialysis membrane (receiver). The base plate was sealed using sealing tape. The plate was incubated at 37 °C on an orbital shaker at 250 rpm for 4 h to achieve equilibrium. Subsequently, 50 μL of the sample was withdrawn from each chamber and transferred into individual microcentrifuge tubes. To each buffer sample, 50 μL of blank plasma was added, and conversely, 50 μL of blank buffer was introduced to the collected plasma sample. The samples were quantified using LC–MS/MS. The experiment was performed in triplicate (N = 3, n = 15). The obtained results were utilized to determine the unbound fraction of the test compound. The calculations were performed using the equations specified by the manufacturer, as outlined below:

In Vivo Toxicity Studies

Toxicity profile of compound 1 was assessed in male C57BL/6J mice (n = 10), in three different groups, including vehicle, 10 mg/kg, and 20 mg/kg treatment groups. The animals were administered the respective injections once daily for 14 days. Upon termination of the study period, the animals were anesthetized and euthanized 5–10 h postlast injection, and the body organs, including liver, lung, kidney, heart, and brain, were collected. For assessment of the compound 1 accumulation, tissue homogenates were prepared (DF = 4) and the samples were quantified using LC–MS/MS. For histopathology analysis, a section of the liver lobe, a single kidney, one-half of the brain, and a single lung lobe was submitted in paraformaldehyde. The tissues were trimmed and processed routinely for histology. Sections were cut at 5 μm and stained with Hematoxylin and Eosin. Slides were evaluated with a Nikon Eclipse Ni microscope.

In Vivo Pharmacokinetic Studies

The PK studies were performed in male C57BL/6J (8–11 weeks old) mice who received a dose of 10 mg/kg compound 1 through the i.p. route. At each specific time point (15, 30, 60, 90, and 180 min), the animals were anesthetized and decapitated, and the plasma (from trunk) and brain were collected. Brain homogenates (DF = 4) and plasma samples were used for quantification using LCMS. The PK parameters were calculated using a standard noncompartmental analysis model.³⁸

Animals for Hot Plate Antinociceptive and Chemotherapy-Induced Peripheral Neuropathy

Adult (8–11 weeks old) male and female C57BL/6J were obtained from the Jackson Laboratory (Bar Harbor, Maine, USA). All mice were housed four a cage in a temperature- and humidity-controlled room at the Texas Tech University Health Sciences Center (Amarillo, TX, USA) vivarium on a 12:12-h light/dark cycle (lights off at 19:00 h) with free access to food and water except during experimental sessions.

Compounds and Preparations for in Vivo Studies

Morphine sulfate (morphine) and naltrexone were obtained from the National Institute on Drug Abuse Research Technology Branch (Rockville, MD). Morphine, naltrexone, and U69,593 were dissolved in sterile saline, which also served as the vehicle control for these injections. The dose of 0.032 mg/kg naltrexone was based on previously published studies. Paclitaxel (Taxol; Bio-Techne, Minneapolis, MN) was dissolved in a vehicle solution containing a 1:1:18 ratio of ethanol, emulphor (Rhodia, Cranbury, NJ), and saline (0.9% NaCl).³⁹ A paclitaxel cycle consisted of four i.p. paclitaxel injections (8 mg/kg per injection, given on alternating days).³⁹ Vehicle control, paclitaxel-naive mice were given four injections of vehicle i.p. for four total doses given every other day.

Hot Plate Assay

The hot plate assay was performed as previously described.⁴⁰ In brief, mice were positioned on a heated (52 °C) enclosed Hot plate Analgesia Meter (Columbus Instruments, Columbus, OH), and latency to jump or lick/shake the back paws was determined. A 30 s cutoff time was implemented to prevent potential tissue damage. To minimize the number of animals used for these studies, compounds were administered in a cumulative dosing paradigm over a total of 5 cycles spaced 30 min apart. As the hot plate experiments were conducted over 5 cycles separated by 30 min, we verified that intraperitoneal vehicle (1st injection) + vehicle (2nd–5th injections), 0.032 mg/kg naltrexone (1st injection) + vehicle (2nd–5th injections), 1 mg/kg compound 1 (1st injection) + vehicle (2nd–5th injections), 3.2 mg/kg compound 1 (1st injection) + vehicle (2nd–5th injections), or 10 mg/kg compound 1 (1st injection) + vehicle (2nd–5th injections) produced no observable antinociceptive activity throughout the duration of the experiments (Figure S2).

Mechanical Allodynia Assessment

Mice underwent a four-day habituation period to the von Frey testing environment, with daily 30 min sessions preceding the initial testing session. Baseline responses to light touch and mechanical touch were measured, as previously described, using von Frey monofilaments (North Coast Medical, Morgan Hills, CA) before administration of paclitaxel or the vehicle.^39,40 During testing, mice were positioned on a wire mesh screen with spaces 0.5 mm apart. The mice, left unrestrained, were individually placed beneath an inverted wire mesh basket (8 cm diameter, 15 cm height) and allowed a 30 min acclimation period to the apparatus before the commencement of testing.

The von Frey test employs a series of calibrated monofilaments with stimulus intensities ranging from 0.4 to 4.0 g. These filaments are applied to the left and right plantar surfaces of the hind paws using the “up-down” method. A response is considered when the mouse lifts, licks, or shakes the paw upon filament application. Positive responses were defined as three or more responses out of five monofilament stimulations. The reported measurements from the von Frey assay represent the mean response thresholds obtained from both the left and right hind paw of each subject. This approach was adopted as peripheral administration of paclitaxel induces bilateral allodynia of comparable magnitude in both paws. The assessment of mechanical allodynia was carried out by observers blinded to treatment conditions.

LC-MS/MS Quantification

The samples obtained from in vitro and in vivo experiments involving BBB permeability, microsomal and plasma metabolic stability, and plasma protein and tissue protein binding experiments, the bidirectional transport assay, and toxicity as well as PK studies were quantified by taking into account the peak area ratio of the test compound vs either the internal standard or the analyte concentration by preparing the standard curve for each test compound in the respective matrix to obtain the concentration range of 2–1000 nM and then subjected to the similar sample preparation methods used for experimental samples. Following a similar sample preparation method, we also prepared blank controls without adding the test compound.

For sample analysis, an AB SCIEX QTRAP 5500 triple quadrupole mass spectrometer, coupled with a Nexera UPLC system (Shimadzu Corporation), was employed. The UPLC system consisted of an autosampler (Sil-30AC), pumps (LC-30AD), a controller (CBM-20A), a degasser (DGA-20A5), and a column oven (CTO-30A). Data acquisition and quantification were performed using Analyst software. Further details regarding the chromatographic separation of the test compounds are provided below.

Column: Gemini-C18 (2 mm × 50 mm, 3 μm; Phenomenex, Torrance, CA, USA); Mobile Phases: (A) water + 0.1% formic acid (FA), (B) methanol + 0.1% FA; Flow Rate: 0.3 mL/min; Gradient: 0–0:10 min; 30% B, 0:10–1 min; reaching 95% B, 1:0–2:0 min; maintaining 95% B, 2:0–2:10 min; returning to 30% B, 2:10–3:0 min; 30% B.

Data and Statistical Analysis

The results are expressed as the mean ± SEM. The N-value denotes the number of independent experiments, while the n-value indicates the number of repeats within each independent experiment. GraphPad Prism 9 Software (GraphPad Software, San Diego, CA) was used to conduct the data analysis.

The data from in vitro experiments were subjected to statistical analysis using a one-way ANOVA, followed by Dunnett’s test for the comparison of multiple groups against the control. A significance level of p < 0.05 was employed to determine statistical significance.

Furthermore, an unpaired t test was employed for the comparison of the two groups. The hot plate data were transformed into percent maximum possible effect (% MPE) using the formula: ([(experimental test value – baseline value)/(maximum test value – baseline value)] × 100). Subsequently, the % MPE values were plotted against log dose values. Student t tests were used to compare pre- vs postvehicle injection thresholds, as well as vehicle vs morphine responses. A one-way ANOVA, with Dunnet’s multiple comparison’s post hoc test, was used to analyze compound dose effect, and a two-way ANOVA was used to analyze test morphine or U69,593 treatment in the presence or absence of naltrexone or compound 1. Outcomes were considered significant at p < 0.05.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.3c00262.

• Synthesis of intermediate compounds, NMR characterization of key compounds, purity and chiral screening data, cytotoxicity results, and in vivo PK parameters and results (PDF)

This work received support from the Department of Pharmaceutical Sciences, Texas Tech University Health Sciences Center startup funding (L.R.M., J.L.W.).

The authors declare the following competing financial interest(s): Compounds described herein are the subject of published patent applications; U.S. Pat. Appl. Publ. (2020), US 2020/0131183 A1.

Notes

This paper is part of a Ph.D. thesis to be submitted to Texas Tech University Health Sciences Center, TX, USA.