Antinociceptive Activity of Thiazole-containing Cyclized DAMGO and Leu-(Met) Enkephalin Analogs

Abstract

Numerous studies demonstrate the promise of opioid peptides as analgesics, but poor oral bioavailability has limited their therapeutic development. This study sought to increase the oral bioavailability of opioid peptides by cyclization, using Hantzsch-based macrocyclization strategies to produce two new series of cyclized DAMGO and Leu/Met-enkephalin analogs. Opioid receptor affinity and selectivity for compounds in each series were assessed in vitro with radioligand competition binding assays. Compounds demonstrated modest affinity but high selectivity for the mu, delta, and kappa opioid receptors (MOR, DOR and KOR), while selectivity for mu opioid receptors varied by structure. Antinociceptive activity of each compound was initially screened in vivo following intracerebroventricular (i.c.v.) administration and testing in the mouse 55°C warm-water tail-withdrawal test. The four most active compounds were then evaluated for dose- and time-dependent antinociception, and opioid receptor selectivity in vivo. Cyclic compounds 1924–10, 1936–1, 1936–7, and 1936–9 produced robust and long- lasting antinociception with ED₅₀ values ranging from 0.32–0.75 nmol following i.c.v. administration mediated primarily by mu- and delta-opioid receptor agonism. Compounds 1924–10, 1936–1 and 1936–9 further displayed significant time-dependent antinociception after oral (10 mg/kg, p.o.) administration. A higher oral dose (30 mg/kg. p.o.) of all four cyclic peptides also reduced centrally-mediated respiration, suggesting successful penitration into the CNS. Overall, these data suggest cyclized opioid peptides synthesized by a Hantzsch-based macrocyclization strategy can retain opioid agonist activity to produce potent antinociception in vivo while conveying improved bioavailability following oral administration.

Graphical Abstract

Introduction

Opioid peptides, such as the endogenous endorphins like the enkephalins and synthetic peptides such as DAMGO, are attractive subjects for analgesic drug development (1–6). Opioid peptides have potential advantages of high receptor affinity and efficacy, low toxicity, and minimal drug-drug interactions (7). Unfortunately, endorphins are readily degraded by proteolytic enzymes within minutes of release, reducing their therapeutic efficacy.

Conformational constraint by cyclization is a common approach used to restrict the flexibility of peptides and is therefore a valuable approach to study topographical requirements of receptors (8–16). Cyclization of peptides can provide potent and selective ligands for receptors when appropriate conformational constraints are incorporated. Furthermore, cyclic peptides and peptidomimetics are often more stable to peptidases, and therefore they can have improved pharmacokinetic profiles and represent promising lead compounds for further development (16–18).

Macrocyclic compounds that have demonstrated improved pharmacokinetic properties include ligands for opioid receptors (19–23). For example, the cyclic enkephalin analogue, Tyr–c[-D-A2bu-Gly-Phe-Leu] was first reported by Schiller et al. to have enhanced activities compared to linear sequences (30), and a number of enkephalin analogues were subsequently synthesized using different cyclization approaches (24–26). In most enkephalin derivatives, cyclization was achieved by means of a disulfide bond. In the sigma-receptor selective Tyr-c[-D-Pen-Gly-Phe-D-Pen]-]OH and Tyr-c-[-D-Pen-Gly-Phe-L-Pen-]-OH, the two D-penicillamine residues are connected by a disulfide bond. The incorporation of two constrained Pen residues in a D-configuration in place of natural amino acids also form compounds with excellent resistance against enzymatic degradation, while still maintaining low toxicity (26–28). Likewise, cyclic analogues of the linear pentapeptide β-casomorphin were reported to have prolonged analgesic potency comparable to both the enkephalin analog DAMGO and to the alkaloid morphine (29–31).

We have developed a unique approach to the generation of macrocyclic opioid peptides through the intramolecular SN2 substitution of chloro methyl thiazoles with the thiol side chain of cysteine (Scheme 1). (32, 33) Using this new intramolecular thioether cyclative approach, we presently performed the parallel synthesis of a variety of cyclic analogs of DAMGO and cyclized analogs of Leu-enkephalin and Met-Enkephalins to produce the 1924 series (Figure 1). In addition, Leu-Enkephalin and Met-Enkephalin analogs were also synthesized in which cysteine was positioned systematically through the peptide sequence other than at the C-terminal, allowing for extension of the peptide beyond the C-side of the cyclic link, producing the 1936 series (Figure 1). We performed comparative computational studies of the chemical distribution of different cyclic peptides in the chemical space. The studies show clearly that the proposed thiazole containing cyclic peptides occupy a different region in chemical space when compared to other common known cyclic forms. (43)

Scheme 1:

a) FmocNCS (6 equiv) in DMF (0.3 M), RT; b) 20% piperidine/DMF; c) 1,3-dichloroacetone (10 equiv) in DMF (0.3 M), 70°C, overnight; d) TFA/(But)₃SiH/DCM (5:5:90), 30 min; e) Cs2CO3 in DMF overnight; f) HF/anisole, 0°C, 90 min

Fig 1:

Cyclic peptidomimetics analogs of Damgo and Enkephalin (Brackets indicate cyclized regions withing analog structure). L-Ala was used instead of D-Ala.

Individual compounds from the 1924- and 1936-series were screened for opioid receptor affinity through radioligand competition binding assays. (30) The antinociceptive activity of individual compounds in each series was concurrently screened in vivo after intracerebroventricular (i.c.v.) administration using the mouse 55°C warm water tail-withdrawal test. (39–41) The most active antinociceptive compounds so identified (1924–10, 1936–1, 1936–7 and 1936–10) were then characterized for dose response and opioid receptor selectivity. Finally, testing potentially improved bioavailability, the antinociceptive and respiratory effects of the lead cyclized compounds were evaluated after oral (per os, p.o.) administration.

Results and discussion

Synthesis of 1924- and 1936-series cyclized compounds.

As outlined in scheme 1, we performed the parallel synthesis of the thiazolyl containing DAMGO and Leu-(Met) Enkephalin Analogs (L-Alanine was used instead of the D-Alanine for DAMGO analogs). Starting from p-methylbenzhydrylamine hydrochloride (MBHA·HCl) resin-bound orthogonally protected Fmoc-Cys-(Trt)-OH, the thiazolyl macrocyclic peptidomimetics were synthesized following stepwise Fmoc deprotection and standard repetitive Fmoc-amino-acid couplings yielding the linear DAMGO and Leu-(Met) Enkephalin analogs. The resulting N-terminal free amine was treated with Fmoc-isothiocyanate. Following Fmoc deprotection, the thiourea was treated with 1,3-dichloroacetone to afford following Hantzsch’s cyclocondensation the resulting resin-bound chloromethyl thiazolyl peptide. The Trt group was removed with 5%TFA in DCM and the resin-bound peptide was treated with a solution of Cs₂CO₃ in DMF to undergo an SN₂ intramolecular thioalkylation reaction. The resin was treated with HF/anisole and the desired thiazolyl thioether desired cyclic peptides were obtained in good yield and good purity. The purity of the purified compounds is higher than 90% for all the compounds with yields rangimg from 57 to 63%.

Opioid-receptor binding affinity and selectivity to murine brain membrane protein by 1924- and 1936-series compounds.

The individual compounds from the 1924- and 1936-series showed modest (micromolar) affinity for the three opioid receptors when tested in vitro with radioligand competition binding assays with [³H]DAMGO, [³H]U69,593 and [³H]DPDPE (Table 1). Overall, compounds from both series displayed higher affinity for delta-opioid receptors (DOR), and the lowest (if any) affinity for the kappa-opioid receptors (KOR). In contrast, compounds displayed varying affinity for Mu-opioid receptors (MOR), albeit generally more selective for the DOR.

Table 1.

K_i values for the inhibition of μ-, δ-, and κ-opioid receptor binding to Murine Brain Membrane protein by 1924- and 1936- series compounds^b

Sample:	MOR	KOR	DOR
1924–1	50.7 ± 4.4	116.3 ± 85.1	0.95 ± 0.05
1924–4	60.3 ± 1.7	18.2 ± 5.7	1.65 ±0.05
1924–7	35.4 ± 3.8	> 1mM	1.33 ± 0.03
1924–10	27.8 ± 1.2	23.5 ± 10.8	1.05 ± 0.13
1924–13	28.0 ± 8.5	> 1mM	0.47 ± 0.002
1924–16	62.6 ± 2.5	79.7 ± 14.9	0.54 ± 0.21
1924–19	59.0 ± 3.1	69.3 ± 48.3	1.34 ± 0.17
1924–22	22.4 ± 0.5	32.3 ± 5.6	0.56 ± 0.06
1924–25	5.0 ± 0.4	11.4 ± 0.4	0.52 ± 0.02
1924–28	28.2 ± 3.8	32.1 ± 19.4	0.67 ± 0.05
1936–1	14.9 ± 0.8	43.9 ± 2.3	3.67 ± 0.09
1936–2	50.6 ± 10.0	> 1mM	> 1mM
1936–3	26.7 ± 5.6	64.6 ± 19.1	8.08 ± 2.34
1936–4	31.5 ± 4.5	32.6 ± 0.9	3.48 ± 0.11
1936–5	23.4 ± 3.5	20.3 ± 3.3	27.72 ± 3.86
1936–6	82.9 ± 5.0	40.0 ± 2.6	3.57 ± 0.72
1936–7	49.7 ± 14.7	53.3 ± 15.2	24.21 ± 16.41
1936–8	29.8 ± 0.9	21.7 ± 2.1	8.86 ± 2.48
1936–9	13.2 ± 0.0	57.9 ± 13.4	2.90 ± 0.38
1936–10	22.2 ± 4.4	46.3 ± 9.3	4.63 ± 0.54

^bMurine brain membrane protein was incubated with six different concentrations of each compound in the A) 1924-series or B) 1936-series in the presence of 0.87 nM [³H]DAMGO, 2 nM [³H]U69,593, or 2 nM [³H]DPDPE in 50 mM Tris–HCl, pH 7.5, at 25°C. Data are expressed as the mean K_i value ± SEM for 2–4 experiments each performed in duplicates. Note: The desired cyclic peptides derived from the 1924 series were purified by preparative reverse-phase HPLC. The cyclic compounds derived from the 1936 series were tested as crude material.

Characterization of Antinociception of Cyclized Enkephalin Analogs Following i.c.v. Administration.

Compounds in the 1924-and 1936- were tested concurrently. Antinociceptive properties of the 1924- and 1936-series were screened at 1 nmol, i.c.v. doses in the 55°C warm-water tail-withdrawal assay. Whereas this dose of morphine displayed withdrawal latencies equivalent to vehicle treated animals, compounds in the 1924- (Figure 2A) and 1936-series (Figure 2B) showed varying and significant magnitudes of antinociception F_(11,92)=83.35, p<0.001 and F_(11,92)=38.58, p<0.001, respectively; each one-way ANOVA with Tukey post hoc test. Among these compounds, 1924–10, 1936–1, 1936–7, and 1936–9 (Figure 3) were the most active of these samples producing significantly higher withdrawal thresholds than vehicle treated animals (p≤0.05.).

Fig 2:

Screening of compounds in the (A) 1924-series and (B) 1936-series: mean summed tail-withdrawal latency in the 55° warm-water assay. C57BL/6J mice were administered vehicle (represented by dashed line) or 1 nmol i.c.v. of either morphine, or a novel cyclized peptide compound. Values represent mean summed tail-withdrawal latency ± SEM (n= 8–16 per group). *Greater than all other samples; P<0.05; One-way ANOVA and Tukey post hoc-test. Note: The desired cyclic peptides derived from the 1924 series were purified by preparative reverse-phase HPLC. The cyclic compounds derived from the 1936 series were tested as crude material.

Fig. 3:

Individual chemical structures of cyclic enkephalin analogs and their affinities for DOR (delta), MOR (mu) and KOR (kappa) opioid receptors

The antinociception of the top four cyclized peptides was further characterized in the 55°C warm-water tail-withdrawal assay. Perhaps surprisingly given the poor binding affinity, each compound produced time- and dose-dependent antinociception, with 1 and 10 nmol doses (i.c.v.) significantly increasing tail-withdrawal latencies (p<0.05) over vehicle response lasting no more than 2 h (Figure 4). Moreover, all compounds proved significantly more potent than morphine (F_(3,176)=9.45; P<0.0001, nonlinear regression modelling). 1924–10 produced antinociception with an ED₅₀ (and 95% confidence interval) values of 0.39 (0.20–0.77) nmol i.c.v., 6-fold more potent than that of morphine (itself 2.34(1.13–5.03) nmol, i.c.v.). Likewise, cyclic enkephalin analogs 1936–1, 1936–7 and 1936–9 all proved significantly potent, with ED₅₀ values of 0.32 (0.06–1.69), 0.40 (0.20–0.88), 0.75 (0.34–1.69) nmol i.c.v., respectively (Table 2).Together, these results suggest these cyclized encephalin analogs produce a long-lasting, dose dependent antinociception.

Fig. 4:

Antinociception produced by A) 1924–10, B) 1936–1, C) 1936–7, and D) 1936–9 were assessed repeatedly over 2 hrs following i.c.v. administration of compounds (0.01–10 nmol) in the 55°C warm-water tail-withdrawal assay. Compounds were compared with a minimally effective dose of morphine (1 nmol) and vehicle. Values represent mean % antinociception ± SEM (n= 8–16 per group).

Table 2.

Antinociceptive Activity of Morphine and Individual Selected compounds from the 1924- and 1936- Series^c

Compound:	ED50 and 95% C.I. (nmol,i.c.v.)
Morphine	2.35 (1.13–5.03)
1924–10	0.39 (0.20–0.77)
1936–1	0.32 (0.06–1.69)
1936–7	0.40 (0.20–0.88)
1936–9	0.74 (0.34–1.69)

^cMice were administered a graded dose of morphine (as a positive control) or Cyclized enkephalin analogs (i.c.v.) in the 55° warm-water tail-withdrawal assay 20–30 min later. ED₅₀ and 95% confidence interval values (nmol) are reported.

Compound 1924–10 was also found to be stable in mouse blood in vitro for at least 90 min (Fig. 5A), and was detected in blood harvested up to 90 min after oral administration in vivo (Fig. 5B), suggesting the possibility of identifying new orally absorbed and stable analgesics. Planned studies will determine levels in blood and brain and metabolic stabilities.

Fig 5:

Pharmacokinetic testing of 1924–10 in mouse blood with LC-MS/MS analysis. A: Incubation of 1924–10 (1 μM) in vitro with isolated C57BL/6J blood at 37oC shows stability of the compound over time. B: Analysis of blood samples taken from mice (n=2/time point) after administration of 1924–10 (30 mg/kg, p.o.) shows detectable levels of sample up to 90 min. Note: blank samples (left most bar) show no detectable peak, denoted as “0±0”.

MOR-, KOR- and DOR-selective agonist activity

Potential opioid receptor agonist selectivity of the four cyclic peptides was determined by pre-treating mice with one of three established opioid receptor-selective antagonists: the MOR-antagonist, β-FNA, the KOR antagonist nor-BNI or the DOR-antagonist, naltrindole. Antinociception produced by the cyclized enkephalin analogs (10 nmol, i.c.v.) was then measured 40 min after administration in the 55°C warm-water tail-withdrawal assay (Figure 6). (39–41) Pre-treatment with β-FNA and naltrindole significantly decreased the antinociceptive effects of all compounds (Figure 6). In contrast, pre-treatment with nor-BNI partially reduced antinociception of all compounds except 1936–9. These data suggest compounds 1924–10, 1936–1, and 1936–7 are ligands that have MOR-DOR activity with little to no KOR agonist activity.

Fig 6:

Antinociceptive effects of a 10 nmol, i.c.v. dose of A) 1924–10, B) 1936–1, C) 1936–7, and D) 1936–9 in mice with or without pretreatment with the MOR antagonist β-FNA (5 mg/kg, s.c., −24 h), the KOR-antagonist, nor-BNI (10 mg/kg, i.p., −24 h), or the DOR-antagonist, naltrindole (20 mg/kg, i.p., −20 min). Tail-withdrawal latencies were measured in the mouse 55°C warm-water tail-withdrawal test 30 minutes after injection of the test compound. Mean % antinociception ± S.E.M. from 8–12 mice are represented in the data. * denotes statistical significance difference from baseline response p<0.05; † denotes statistical difference from the response of the matching cyclic peptide (p<0.05).

Demonstration of Oral Activity by the Cyclized Enkephalin Analogs Following Oral Administration

To determine if the cyclized analogs possessed greater bioavailability following oral administration, animals were first administered vehicle (p.o.), 1924–10, 1936–1, 1936–7, and 1936–9 p.o. (10 mg/kg, each) or morphine (5 mg/kg, p.o.), and antinociception measured repeatedly at least 80 min in the 55° warm-water tail-flick assay (Figure 7). There was a significant effect of treatment over time (F_(40,280)=7.24, p<0.0001; two-way repeated measures ANOVA with Tukey post-hoc test) with morphine, 1924–10, 1936–1 and 1936–9 (but not 1936–7) all producing significantly higher antinociceptive effects compared to vehicle treated animals at various time points (p <0.05). Notably, 1936–9 produced the most robust and long-lasting antinociception comparable to the effect of morphine.

Fig. 7:

Antinociceptive activity of 1924–10, 1936–1, 1936–7, and 1936–9 following p.o. (10 mg/kg) administration in the 55°C warm-water tail-withdrawal assay with repeated measurement over time. Comparisons were made to the effect of a p.o. administration of vehicle or morphine (5 mg/kg). Values represent mean % antinociception ± SEM (n= 8 per group).

Respiratory effects induced by the analogs following oral administration

Morphine produces respiratory depression through activation of MOR in the brain’s respiratory centers in the medulla. To further evaluate oral bioavailability and central nervous system (CNS) penetration of the cyclized peptides, mice were administered 1924–10, 1936–1, 1936–7, and 1936–9 p.o. (30 mg/kg, each) and respiratory rates were measured in 20 minute intervals over 2 hrs using an Oxymax/CLAMS assay. (40) There was a significant interaction of treatment and time (F_(25,285)=3.87; p<0.0001, two-way ANOVA repeated measures with Tukey’s post hoc test; Figure 8). As expected, the positive control morphine suppressed respiration up to an hour (p<0.05). Each of the compounds except AN1936–9 significantly reduced respiration after a 20 min pre-treatments, with ccompound 1924–10 robustly depressing respiration for over two hours (Figure 7). Perhaps surprisingly given it’s oral activity in the tail-withdrawal test, AN1936–9 treatment did not significantly reduce respiration from vehicle-treated levels (F_(1,21)=2.39; p=0.14, two-way ANOVA repeated measures). However, the respiratory activity of the remaining analogs after oral administration confirms their MOR agonist effects, and although detailed study is needed, further suggests a rapid ability of these cyclized peptides to cross the blood brain barrier and penetrate the CNS.

Fig 8:

Respiratory effects of compound 1924–10, 1936–1, 1936–7, and 1936–9 (30 mg/kg, p.o.) in C57BL/6J mice. Morphine (30 mg/kg, i.p.) served as a positive control; additional mice received vehicle (10% DMSO/10% Solutol/80% saline, p.o.). Mice were treated 5 min prior to the start of testing. All points represent respiration as compared to %vehicle-treated response (grey hex) from 8–10 mice/compound (13 mice for morphine and vehicle). *P<0.05 from vehicle response, Two-way RM ANOVA followed by Tukey’s post hoc test.

Discussion

The current study sought to determine if structural modifications of opioid peptides using thiolklyation approaches or Hantzsch-based macrocyclization strategy can increase the oral bioavailability while enhancing their systemic activity. The principal findings from this study demonstrate structural modifications of Thiazole based cyclized analogs of DAMGO and Leu-(Met-) enkephalins opioid peptides can increase the bioavailability while retaining antinociceptive activity when administered orally. These analogs, forming the 1924- and 1936- series, displayed low affinity but good selectivity for opioid receptors, especially for DOR and somewhat for MOR in vitro. Surprisingly, when administered i.c.v. compounds from both series showed a variety of antinociceptive effects in the 55°C warm-water tail-withdrawal assay, with compounds 1924–10, 1936–1, 1936–7, and 1936–9 the most active of these samples. Further testing with these active compounds demonstrated each analog produced long-lasting and dose dependent antinociception equivalent to morphine and mediated by the MOR and DOR, and less so the KOR that lasted up to 2 hours. Following oral administration all but 1936–7 produced antinociception, and all but 1936–9 produced respiratory depression, suggesting both oral bioavailability, systemic activity, and potential CNS penetration for at least three of the peptides tested.

Evaluating opioid peptides in vivo for analgesic properties has been a challenge as peptides are rapidly degraded in harsh environments such as the digestive tract (34). This limitation prevents characterization of the antinociceptive activity for opioid peptides following therapeutic routes such as oral administration. By improving the pharmacokinetic profile of opioid peptides through thioalklyation approaches and Hantzsch-based macrocyclization strategy to produce newly synthesized cyclized DAMGO and Leu—(Met)-enkephalin analogs that are more stable to peptidases. Macrocyclic compounds provide diverse functionality and stereochemical complexity in a conformationally pre-organized ring structure. This can result in high affinity and selectivity for protein targets, while preserving sufficient bioavailability to reach intracellular locations. This improved pharmacokinetic profile allows characterization of antinociceptive properties produced by opioid peptides, a challenge researchers have faced for over 20 years (27). Indeed, this research demonstrates macrocyclization of opioid peptides can produce robust and long-lasting antinociception following oral administration.

The development of analgesics with mixed opioid agonist activity could have a significant impact on opioid analgesic development. As compounds 1924–10, 1936–1, and 1936–7 demonstrate mixed MOR-DOR activity, suggesting these compounds could provide increased analgesia accompanied with lowered side effect profiles. Further experimentation would be required to define the limits of these side effects including respiratory depression and sedation. A surprising and provocative aspect of this study is that the in vitro binding profile of compounds 1924–10, 1936–1, 1936–7, and 1936–9 did not align with their in vivo antinociceptive responses. While determining the cause of this discrepancy is beyond the scope of this initial synthesis and screening, it is possible these compounds could be a pro-drug as in vitro binding does not take into account modifications such as metabolism or pharmacokinetics that occur in vivo. It is also possible that their additive effects of binding to MOR and DOR receptors enables ligands with lowered affinity to achieve synergistic efficacy in vivo. Further studies will evaluate these possibilities.

Conclusions

Taken together, the newly developed macrocyclization strategy provided cyclic DAMGO and Enkephalin analogs with improved oral bioavailability while retaining significant agonist activity. These approaches may advance the development of novel opioid peptides with demonstrated in vivo efficacy and potentially novel treatments for a variety of pain conditions (35, 36).

Experimental

Synthesis of the 1924 and 1936 series cyclic enkephalin analogs

All the reagents, amino acids and solvents were commercial grade. LC-MS (ESI) traces were recorded on samples with concentrations of 1 mg/ mL in 50:50 MeCN/ water at both 214 nm and 254 nm using a reverse phase Vydac column with a gradient of 5 % to 95 % formic acid in MeCN. The purity of the crude samples was estimated based on the UV traces recorded. Hydrofluoric acid cleaves were performed in specially equipped and ventilated hoods with full personal protective equipment. All synthesized compounds were purified by RP-HPLC. The purity of all final compounds was >95%.

General procedure for the solid-phase synthesis of resin-bound linear cyclic peptides

We previously reported the synthesis of a variety of thiazole containing cyclic peptides. (32, 33) For each cyclic peptide from the 1924 and 1936 series, one bag of prepared resin-bound Fmoc-cysteine (100 mg, 0.115 mmol) or resin-bound N-terminal Fmoc-cysteine short peptide was put into a small polyethylene bottle and the Fmoc group was deprotected with 15 ml 20% piperidine in DMF (2×10min). The 20 bags (each containing 100 mg of resin) (42) were then washed with 15 ml DMF (3×) and 15 ml DCM (3×). The first amino acid Fmoc-Xaa-OH (6 equiv, 0.267 g, 11.04 mmol) was coupled in the presence of hydroxybenzotriazole (HOBt, 6 equiv, 0.094 g, 11.04 mmol) and diisopropylcarbodiimide (DIC, 6 equiv, 0.101 ml, 11.04 mmol) in 15 ml anhydrous DMF for 2h at room temperature. The resin-bound dipeptide was washed with DMF (3x) and DCM (3x). Completion of the coupling was monitored by the ninhydrin test. The Fmoc group was deprotected with 15 ml 20% piperidine in DMF (2×10min) and followed by standard repetitive Fmoc-amino-acid couplings yielding the desired linear peptides.

The N-terminal free amine of resin-bound linear peptide was treated with Fmoc-isothiocyanate (6 equiv, 0.193 g, 11.04 mmol) in 15 ml DMF anhydrous overnight at room temperature. Following Fmoc deprotection with a solution of 20% piperidine in DMF, the resin-bound N-terminal thiourea was treated with 1,3-dichloroacetone (10 equiv, 0.145 g, 18.4 mmol) in DMF anhydrous overnight at 70 °C to afford following Hantzsch’s cyclocondensation the resulting resin-bound chloro methyl thiazolyl peptide. The Trt group was deprotected in the presence of TFA/(But)₃SiH/DCM (5:5:90) for 30 min. The resin was washed with DCM (5x) and DIEA/DCM (5:95) and was treated overnight with a solution of Cs₂CO₃ (10 eq, 0.325 g) in 15 ml DMF at room temperature to undergo an SN2 intramolecular thioalkylation. The resin was cleaved with HF/anisole for 90 min at 0 °C, and the desired thiazolyl thioether cyclic peptides was obtained following extraction with 95% acetic acid in water and lyophilization as a white powder. All the peptides were analyzed by HPLC and LCMS.

The desired cyclic peptides derived from the 1924 series were purified by preparative reverse-phase HPLC. The cyclic compounds derived from the 1936 series were tested as crude material.

Physical characterization data of representative compounds Compound 1924–10:

¹H NMR (400 MHz, DMSO-d6) δppm 9.1 (s, 1H) 7.09 (t, J= 8.96 Hz, 1H), 7.47 (d, J= 2.44 Hz, 1H), 7.19–7.26 (m, 2H), 7.10–7.15 (m, 2H), 6.95 (d, J= 8.24 Hz, 2H), 6.58 (d, J= 8.16 Hz, 1H), 6.31 (s, 1H), 5.23 (d, J= 11.56 Hz, 1H), 4.30–3-42 (m, 2H), 4.15–4.25 (m, 2H), 3.95–4.07 (m, 2H), 3.52 (d, J= 14.48 Hz, 1H), 3.33–3.40 (m, 1H), 3.15–3.30 (m, 2H), 2.95–3.05 (m, 2H), 2.81–2.86 (m, 4H), 2.75–2.79 (m, 2H), 2.68–2.73 (m, 2H), 2.57 (s, 3H), 0.27 (d, J= 6.4 Hz, 3H). ¹³C NMR (100 MHz, DMSO-d6) δppm 173.59, 172.89, 172.60, 167.54, 156.48, 150.18, 138.44, 130.39, 129.82, 129.00, 127.72, 126.99, 115.57, 103.68, 62.32, 62.09, 56.96, 42.32, 37.04, 33.92, 33.10, 29.86, 16.57. MS (ESI): m/z calcd for C₂₉H₃₄N₆O₅S₂ [M + H]⁺ : 611.2, found: 611.0. HRMS MALDI-TOF: M⁺= 610.8237.

Compound 1936–9:

¹H NMR (400 MHz, DMSO-d6) δppm 9.08 (s, 1H), 8.23 (d, J= 7.72 Hz, 1H), 7.78 (d, J= 6.72 Hz, 1H), 7.69 (d, J= 7.60 Hz, 1H), 7.07–7.11 (m, 3H), 7.00–7.06 (m, 2H), 6.56 (d, J= 6.84 Hz, 2H), 6.28 (s, 1H), 4.39–4.44 (m, 1H), 4.24–4.26 (m, 1H), 4.21–4.24 (m, 1H), 4.20 (d, J=6.40 Hz, 1H), 4.12 (d, J= 8.28 Hz, 1H), 3.90 (d, J=6.40 Hz, 1H), 3.43 (bs, 1H), 2.90–2.94 (m, 2H), 2,85–2.87 (m, 2H), 2.74–2.77 (m, 2H), 2.70–273 (m, 2H), 1.50–1.53 (m, 2H), 1.39 (t, J=6.48 Hz, 3H), 0.77 (dd, J= 20.00 Hz, J=5.20 Hz, 3H). ¹³C NMR (100 MHz, DMSO-d6) δppm 172.77, 171.75, 171.67, 169.96, 167.88, 155.79, 148.99, 137.76, 137.48, 129.73, 129.05, 128.85, 114.90, 103.42, 61.45, 56.20, 54.55, 54.32, 51.59, 36.83, 36.50, 35.88, 32.68, 30.30, 23.51, 22.03. MS (ESI): m/z calcd for C₃₃H₄₁N₇O₆S₂ [M + H]+ : 696.3, found: 696.0.

Opioid-receptor binding affinity and selectivity to rodent brain membranes

All radioligand competition binding assays were performed as described previously (30). Specifically, for each opioid receptor:

Mu opioid receptor (MOR) binding assay: Rat cortices were homogenized using 50 mM Tris, pH 7.4, and centrifuged at 16,500 rpm for 10 min. The pellets were resuspended in fresh buffer and incubated at 37°C for 30 min. Following incubation, the suspensions were centrifuged as before, the resulting pellets resuspended in 100 volumes of 50 mM Tris, pH 7.4 plus 2 mg/ml bovine serum and the suspensions combined. Each assay tube contained one of eight different concentrations of the competitor compound and 0.5 ml of membrane suspension at a 0.02 mg/ml concentration with 2 nM [³H]-DAMGO in a total volume of 0.65 ml for a 1 h incubation at 25°C. Nonspecific binding was determined by addition of 10 μM naloxone.

Kappa opioid receptor (KOR) binding assay: Guinea pig cortices and cerebella were homogenized using 50 mM Tris, pH 7.4, 10 mM MgCl₂-6H₂O, 200 μM and centrifuged and incubated as above. Each assay tube contained 2 nM [³H]U69,593. Assay tubes were incubated for 2 h at 25°C. Unlabeled U50,488 was used as a competitor to generate a standard curve and determine nonspecific binding.

Delta opioid receptor (DOR) binding assay: Rat cortices were homogenized using 50 mM Tris, pH 7.4, 10 mM MgCl₂-6H₂O, 200 μM PMSF, centrifuged and incubated as above. Each assay tube contained 2 nM [³H]-DPDPE, Assay tubes were incubated for 2.5 h at 25°C. Unlabeled DPDPE was used as a competitor to generate a standard curve and determine nonspecific binding.

All three binding assays were terminated by filtration through GF/B filters, soaked in 5 mg/ml bovine serum albumin, 50 mM Tris, pH 7.4, on a Tomtec Mach II Harvester 96 (37). The filters were subsequently washed with 6 ml of assay buffer. Bound radioactivity was counted on a Wallac Betaplate Liquid Scintillation Counter. Experiments were conducted using 2 replicates and repeated at least twice.

Animal studies

Male C57BL/6J mice (20–25 g; Jackson Labs; Bar Harbor, ME, USA) were housed 4 per cage in a temperature-controlled vivarium. Mice were maintained under a 12-hour light/dark cycle with lights on at 6. Food and water were available ad libitum.

All experiments were performed in compliance with the Torrey Pines institute’s policy on animal use and ethics. All animal studies were preapproved by the Torrey Pines Institute for Molecular Studies (Port St. Lucie, FL. USA), in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals. All results of animal testing are reported in accordance with ARRIVE guidelines. Sample sizes (i.e. number of animals) were not predetermined by a statistical method, and animals were assigned to groups randomly. Drug treatment experiments were conducted in a blinded fashion. No animals were excluded from statistical analysis.

Injection Techniques

All intracerebroventricular (i.c.v.) injections were made directly into the lateral ventricle according to the method of Haley and McCormick (38). Compounds were delivered via the i.c.v. route to initially assess antinociceptive activity in vivo, yet prevent complications associated with distribution (i.e., blood-brain barrier penetration) that could affect activity following systemic administration. The mice were lightly anesthetized with isoflurane, an incision made in the scalp, and the injection made 2 mm lateral and 2 mm caudal to bregma at a depth of 3 mm. A volume of 5 μL was injected, using a 10 μL Hamilton microliter syringe. A vehicle of 50% DMSO/50% saline (0.9%) was used.

Oral (per os, or p.o.) administration was made with standard methods as demonstrated previously (REF) in a volume of 0.25 ml/26 g mouse body weight. A vehicle of 10% DMSO/10% Solutol/80% saline (0.9%) was used for the p.o. studies.

Antinociceptive Assay

The 55°C warm-water withdrawal assay was performed to quantify antinociception as previously described (39–41). Briefly, for testing, each mouse was gently restrained and its tail submerged 2 cm into the warm water bath. The latency to tail withdrawal was recorded and served as the dependent measure. After determining baseline tail-withdrawal latencies, mice were administered vehicle or compounds and tail-withdrawal latencies were measured every 10 min following administration for 1 h or until latencies returned to baseline values. A cut-off time of 15 s was used in this study; if the mouse failed to display a tail-withdrawal response during that time, the tail was removed from the water and the animal was assigned a maximal antinociceptive score of 100%. Mice that did not respond within 5 s during baseline testing were excluded from the experiment.

For screening studies, results are presented as the sum of average responses at every 10 min across 90 min of testing. For more detailed analysis across time and dose (1924–10, 1936–1, 1936–7 and 1936–9), antinociception was calculated according to the formula: % antinociception = 100 × (test latency – control latency)/ (15 – control latency).

To determine agonist selectivity, mice were pretreated with one of three established opioid receptor selective antagonists as described earlier (40): the KOR-selective antagonist, nor-binaltorphimine (nor-BNI, 10 mg/kg, i.p., administered 24 h prior), the MOR-selective antagonist, β-funaltrexamine (β-FNA, 5 mg/kg, s.c., administered 24 h prior), or the DOR-selective antagonist, naltrindole (20 mg/kg, i.p., 20 min prior) prior to administration of a cyclized peptide. Antinociception produced by the cyclized analogs was then measured 40 min later.

Assessment of Respiration Rate

Respiration rates (in breaths per minute) were assessed using the automated, computer-controlled Comprehensive Lab Animal Monitoring System (CLAMS; Columbus Instruments, Columbus, OH) as described previously (40). Before starting each experiment, the CLAMS system was calibrated 60 min, and mice habituated in the apparatus for 60 min prior to administration of vehicle, morphine (30 mg/kg, i.p.) or either compound 1924–10, 1936–1, 1936–7 or 1936–9 (30 mg/kg, p.o.). Five min after administration, respiration rate (breaths/min) of each occupant mice was measured with pressure transducers built into the sealed CLAMS cage for 120 min.

Statistics

IC₅₀ values were calculated by least squares fit to a logarithm-probit analysis. The K_i value for each compound was calculated from the equation K_i = IC₅₀/(1+S), where S = (concentration of radioligand)/ (K_D of radioligand) by Prism 5.0 software (GraphPad Software, LaJolla, CA). All dose–response lines were analyzed by regression and D₅₀ (dose producing 50% antinociception) values and 95% confidence limits determined using each individual data point with Prism 5.0 software. Tail-withdrawal testing formed a within- (repeated antinociceptive testing of the same group of mice over time) and between-subject (drug treatment between groups) factorial design. Data were analyzed using one-and two-way analysis of variance (ANOVA) using SPSS software. Significant main effects of treatment on tail-withdrawal latencies were further analyzed using Bonferroni multiple comparisons post hoc testing. Student’s t tests comparing baseline and post-treatment tail-withdrawal latencies were used to determine statistical significance for all tail-withdrawal data. Latency to withdraw tail, rather than percent antinociception, was used to determine within group effects and to allow comparison to baseline latency in tail-withdrawal experiments. Statistical significance of ED₅₀ values was determined by evaluation of the ED₅₀ value shift via nonlinear regression modelling using Prism 7.0. Respiration data collected with CLAMS were analyzed via two-way matching-samples ANOVA, with treatment and time as factors, and Tukey’s post-hoc test used to assess group differences. All data are presented as means ± SEM with significance set at p<0.05.