Development of potent μ and δ opioid agonists with high lipophilicity

Abstract

An SAR study on the Dmt-substituted enkephalin-like tetrapeptide with a N-phenyl-N-piperidin-4-yl propionamide moiety at C-terminal was performed, and has resulted in highly potent ligands at μ and δ opioid receptors. In general, ligands with the substitution of D-Nle² and halogenation of the aromatic ring of Phe⁴ showed highly increased opioid activities. Ligand 6 with good biological activities in vitro demonstrated potent in vivo antihyperalgesic and antiallodynic effects in the tail-flick assay.

Introduction

Many studies have shown that co-administration of a small amount of a δ opioid agonist can reduce serious side effects of morphine, a μ agonist, such as tolerance, while it increases the potency and efficacy.^1-3 Furthermore several studies have observed synergistic antinociceptive effects between μ agonists and δ agonists.⁴ These observations imply that the modulation of both μ and δ receptors can allow treatment of pain with lower doses of μ agonists, therefore lessening μ receptor-mediated side effects. Our group previously sought to design nonselective bifunctional ligands possessing μ and δ agonist activities and as a result developed the very potent μ and δ opioid ligand 12 (K_i = 0.36 nM and 0.38 nM at hDOR and rMOR, respectively; IC₅₀ = 1.8 nM and 8.5 nM in MVD and GPI assays, respectively) with the C-terminus of an enkephalin-like tetrapeptide (H-Dmt-DAla-Gly-Phe-OH) linked to a N-phenyl-N-piperidin-4-yl propionamide moiety which is a part of the fentanyl structure.⁵ We found that the propionamide moiety in the ligand resulted in a great increase in lipophilicity (aLogP = 2.96) which can improve the bioavailability of the ligand.

Despite their high potency and important role in combating pain, natural opioid peptides are limited as clinically viable drugs because of poor bioavailability, mainly due to their inability to penetrate the blood-brain barrier (BBB) and rapid degradation in vivo by peptidases. Increasing the lipophilicity can improve analgesia with a concomitant enhancement of bioavailability since the lipophilicity is a key factor in determining the rate at which a drug crosses the BBB. Methylation and halogenation enhance the lipophilicity and cell permeability by reducing overall hydrogen bonding. For example, addition of a chlorine on the Phe⁴ residue of [D-Pen²,D-Pen⁵]enkephalin (DPDPE) led to a significant increase in cell permeability in both in vivo and in vitro studies. ^6-7

Pursuing enhanced bioavailability at the μ and δ receptors by increasing lipophilicity, a systematic Structure-Activity Relationship (SAR) study on the ligand 12 was performed (Figure 1). The propionamide moiety and the Dmt¹ residue were conserved to increase cell permeability and potency, respectively. As a result, a series of highly lipophilic (aLogP: 3.01 - 4.74 in Table 1) nonselective μ and δ opioid agonists were obtained.

Figure 1. Design of opioid ligands.

Table 1. Structure and analytical data of opioid ligands.

no	1	2	3	4	aLogP11	molecular formula	HRMSa (M-TFA + H)+		HPLCb (tR, min)
							Observed	calcd	A	B
1	Dmt	DNle	Gly	Phe	3.66	C42H56N6O6	741.4325	741.4340	19.3	10.8
2	Dmt	DNle		Phe	4.20	C40H53N5O5	684.4109	684.4125	19.8	11.9
3	Dmt	DAla		Phe	3.52	C37H47N5O5	642.3665	642.3655	18.3	8.9
4	Dmt	DNle	Gly	Phe(Cl)	4.18	C42H55ClN6O6	775.3995	775.3950	20.7	12.5
5	Dmt	DNle	Gly	Phe(F)	3.74	C42H55FN6O6	759.4247	759.4245	20.0	12.0
6	Dmt	DAla	Gly	Phe(Cl)	3.47	C39H49ClN6O6	733.3494	733.3480	19.2	10.3
7	Dmt	DNle		Phe(Cl)	4.74	C40H52ClN5O5	718.3713	718.3735	22.2	14.0
8	Dmt	DAla		Phe(Cl)	3.97	C37H46ClN5O5	676.3272	676.3266	19.8	11.0
9	Dmt	DAla	Gly	Phe(F)	3.01	C39H49FN6O6	717.3797	717.3776	17.7	9.5
10	Dmt	DTic	Gly	Phe(F)	4.02	C46H53FN6O6	805.4055	805.4089	20.3	12.0
11	Dmt	DTic	Gly	Phe(Cl)	4.46	C46H53ClN6O6	821.3799	821.3793	21.2	13.1
12	Dmt	DAla	Gly	Phe	2.96	C39H50N6O6	699.3852	699.3870	20.1	12.7

^aFAB-MS (JEOL HX110 sector instrument) or MALDI-TOF.

^bPerformed on a Hewlett Packard 1100 [C-18, Microsorb-MV™, 4.6 mm × 250 mm, 5 μm]. (A)10-90% of acetonitrile containing 0.1% TFA within 40 min and up to 100% within an additional 5 min, 1 mL/min, (B) 25-65% of acetonitrile containing 0.1% TFA within 20 min and up to 100% within an additional 5 min, 1 mL/min.

Results and Discussion

As published before, the ligands designed were prepared by stepwise solution-phase peptide syntheses using N^α-Boc-chemistry and the crude products were isolated by preparative RP-HPLC to afford > 98% pure compounds in 40-50% overall yields.⁵ During the chain elongations, pure peptide intermediates were obtained by simple precipitation from appropriate organic solvents, usually diethyl ether. Their in vitro bioassays were carried out as previously described.⁸

Opioid binding affinities at the human δ opioid receptor (hDOR) and the rat μ opioid receptor (rMOR) were determined by competition analyses against [³H]DPDPE (δ) and [³H]DAMGO (μ) using membrane preparations from transfected HN9.10 cells that constitutively express the respective receptors. Opioid agonist efficacy was examined by monitoring [³⁵S]GTP-γ-S binding. Functional assays to evaluate their opioid agonist activities were performed using the stimulated isolated mouse vas deferens (MVD, δ) and guinea pig ileum (GPI, μ) bioassays. In general, the functional assay results correlated with those from the [³⁵S]GTP-γ-S binding assay.

Our SAR study showed that replacement of DAla² with DNle², and halogenation at the 4 position of the aromatic ring of Phe⁴, increased opioid activitiy at both receptors in vitro, which was in agreement with the other observations, ^9-10 along with increased lipophilicity (aLogP increase: 0.7 for DNle, 0.5 for Cl).¹¹ First of all, replacement of DAla in 12 with DNle led to ligand 1, which showed increased opioid agonist activities [IC₅₀ = 0.69 nM (3-fold increase) in the MVD assay; IC₅₀ = 1.6 nM (5-fold increase) in the GPI assay] and efficacies [EC₅₀ = 0.10 nM (8-fold increase) with Emax = 47% at hDOR; EC₅₀ = 0.11 nM (8-fold increase) with Emax = 77% at rMOR in the [³²S]GTP-γ–S assay], whereas its binding affinities at both receptors were not affected distinctly. The same trend was observed in the other ligands in which the DAla residue was replaced by DNle. The DNle substitution caused greater effects on the functional activities and efficacies than on the binding affinities, suggesting that the change of functional efficiency was due to increased lipophilicity.

To reduce pharmacophore size but increase lipophilicity, the Gly³ residue in ligands 1, 4, 6 and 12 was truncated, yielding ligands 2, 7, 8 and 3, respectively. The resulting ligands had decreased bioactivities at both receptors but raised μ selectivity because δ receptor activity was further reduced. Even though 3 had increased binding affinity (K_i = 0.15 nM) and agonist activity (IC₅₀ = 3.6 nM) at the μ opioid receptor 2-fold by the truncation, in general, the truncation of Gly was not tolerated for both opioid receptors. It has been known that the relative proximity of the Phe⁴ aromatic ring and Tyr¹ aromatic ring is critical for δ versus μ receptor selectivity.¹¹ In our case, shortening their proximity by the truncation of Gly³ turned out to reduce δ receptor activity much more than μ receptor activity and subsequently to increase μ selectivity. All the Gly truncated ligands showed μ opioid receptor selectivity in the binding and functional assays.

All of the ligands with Cl (or F) at the 4′ position of the aromatic ring of Phe⁴ showed highly increased opioid activity in both binding and functional assays. Ligand 5, in which DNle² and Phe(F)⁴ were substituted, showed the most potent subnanomolar opioid agonist activity (IC₅₀ = 0.37 nM and 0.26 nM in the MVD and GPI assays, respectively) with excellent efficacy (EC₅₀ = 0.07 nM, Emax = 48% at hDOR; EC₅₀ = 0.29 nM, Emax = 98% at rMOR) at both receptors. The binding affinity at the μ receptor was subnanomolar (Ki = 0.02 nM) and the Emax at rMOR in the [³⁵S]GTP-γ-S binding assay was 98% which represents high efficacy at the receptor. These results suggest that Cl (or F) in the opioid ligands could increase cell permeability and enhance potency which is more pronounced for the μ receptor, thus producing very well balanced (IC₅₀^δ/IC₅₀^μ < 2 in ligands 4, 5, and 9 mixed agonist activities for both receptors, respectively. On the contrary, the p-Cl substitution in the Gly truncated ligands 7 and 8 decreased opioid activity in all of the assays, especially the functional assays, up to 47-fold (IC₅₀ = 250 nM and 220 nM for 7, IC₅₀ = 180 nM and 93 nM for 8 at MVD and GPI, respectively), but retained μ opioid receptor selectivity in ligands 2 and 3.

Dmt-Tic is a highly potent selective δ opioid antagonist, and this modification has been recognized as leading to significant alterations in opioid activities and selectivities.¹² In utilizing this potent pharmacophore to alter agonist functions at both opioid receptors, the Tic residue was modified to D-Tic in ligands 10 and 11. These ligands showed biological activities analogous to the other ligands, which confirmed that the constrained and very hydrophobic D-Tic residue could be replaced for the flexible D-Nle and D-Ala residues in the enkephalin structure. Both ligands showed high δ opioid receptor selectivities (δ/μ=0.03 for 9, and 0.04 for 10) in the functional assays, while ligand 10 gave the most potent agonist activity at the δ opioid receptor (IC₅₀ = 0.029 nM in the MVD). No δ opioid antagonist activity was observed for either of the D-Tic-containing ligands.

Based on the in vitro bioassay results showing balanced high μ and δ opioid activities (Ki = 0.14 nM, EC₅₀ = 0.16 nM at hDOR; IC₅₀ = 0.70 nM in MVD, Ki = 0.14 nM, EC₅₀ = 0.31 nM at rDOR; IC₅₀ = 2.6 nM in GPI), ligand 6 was chosen for in vivo assays. Neuropathic pain was induced by L5/L6 spinal nerve ligation (SNL) in male Sprague Dawley rats.¹³ Thermal hypersensitivity and tactile allodynia were determined by paw withdrawal latencies to infrared radiant heat and paw withdrawal thresholds from von Frey filaments, respectively, which were applied to the plantar aspect of the hind paw. The time course of paw withdrawal latency or threshold after intrathecal microinjections of 6 at 3μg/5μL, 10μg/5μL, and 30μg/5μL were examined (Figure 2 and Figure 3). A dose-response curve was generated at the time of peak effect (30 min). The results showed that 6 attenuates tactile allodynia and thermal hyperalgesia when administered via i.th. injection. The L5/L6 SNL model was used to evaluate efficacy against tactile allodynia or thermal hyperalgesia as compared to baseline. Significant efficacy for 6 was observed at doses of 3, 10 and 30μg. A dose response curve was generated at peak effect which occurred 30 min after i.th. administration. The A₅₀ values at this time point were 11.0 (6.5-21.1 μg; 95% CI for allodynia) and 1.04 (1.0-1.2 μg/5μL; 95% CI for hyperalgesia). No effect was seen in vehicle treated animals.

Figure 2.

Antihyperalgesic effects of 6 (i.th.) using radiant heat in L₅/L₆ SNL-operated male SD rats. Thermal hypersensitivity by paw withdrawal (left) Antinociceptive dose-response curve at 30 min (right)

Figure 3.

Antiallodynic effects of 6 (i.th.) using von Frey filaments in L5/L6 SNL-operated male SD rats. Tactile allodynia by paw withdrawal thresholds (left) Antinociceptive dose-response curve at30 min (right)

Conclusions

In the present study, a systematic SAR study on the enkephalin analogue 12 with conservation of the propionamide moiety and Dmt¹ residue resulted in the development of nonselective potent opioid agonists for μ and δ receptors with high lipophilicity. These analogues can produce the desired physiological effects without many of the undesirable side effects of selective μ opioid agonist, for example, 5 (K_i = 0.40 nM and 0.02 nM at hDOR and rMOR; IC₅₀ = 0.37 nM and 0.26 nM in MVD and GPI, respectively). In summary, the substitution of D-Nle² and the halogenation of the aromatic ring of Phe⁴ increased opioid activities at both receptors along with the lipophilicity. Ligand 6 which possessed highly potent opioid activities at both receptors (K_i = 0.14 nM at hDOR and rMOR; IC₅₀ = 0.70 nM and 2.6 nM in MVD and GPI assays, respectively) showed potent antihyperalgesic and antiallodynic effects in the in vivo assays, demonstrating that increase of lipophilicity can be a good approach to develop a potent ligand with good bioavailability.

Experimental Section

All amino acid derivatives were purchased from Novabiochem, ChemImpex International, and RSP. N-phenyl-N-piperidin-4-yl-propionamide was prepared as previously described.¹⁴ Coupling reactions were monitored by TLC using the following solvent systems: (1) CHCl₃:MeOH:AcOH = 90:10:3 with ninhydrin spray used for detection. Analytical HPLC was performed on a Hewlett Packard 1090 [C-18, Vydac, 4.6 × 250 mm, 5 μm] and preparative RP-HPLC on Hewlett Packard 1100 [C-18, Microsorb-MV™, 10 × 250 mm, 10 μm]. ¹H NMR spectra were recorded on a Bruker Advance-300 spectrometer. Mass spectra were taken in the positive ion mode under ESI methods.

Ligands 1-12

These ligands were prepared by stepwise synthesis using N^α-Boc chemistry starting from N-phenyl-N-piperidin-4-yl-propionamide. N-phenyl-N-piperidin-4-yl-propionamide (1 eq) and N^α–Boc-Phe-OH (1.1 eq) were dissolved in DMF and cooled in an ice bath for 10 min. Bop (1.1 eq), HOBt (1.1 eq), and NMM (2 eq) were added to the reaction mixture and stirred for 2 h at rt. After checking for the disappearance of the starting amine by TLC, the reaction mixture was concentrated under reduced pressure. The concentrated mixture was diluted with EtOAC and washed with 5% NaHCO3, 5% citric acid, brine, and water consecutively. The organic layer was dried under anhydrous Na₂SO4 and filtered. The solution was concentrated under reduced pressure and triturated with diethyl ether to give a solid. The N^α-Boc group was deprotected by 100% TFA at 0 °C for 20 mins. After completion of the chain elongation by subsequent coupling and deprotection, the mixtures were evaporated under reduced pressure. The residues were solidified by diethyl ether and purified by preparative RP-HPLC (20 – 50% acetonitrile within 20 min for 1, 6 and 7, 20 - 60% acetonitrile within 15 min for 5 and 10) to give pure ligands as white powders in overall yields of 20 - 42%. The purity of the ligands was determined as ≥ 95% by analytical HPLC (10-90% acetonitrile in 40 min, 25-65% acetonitrile in 20 min). For analytical data, see Supporting Information.

Radioligand Labeled Binding Assay, [³⁵S]GTP-γ-S binding Assay, GPI and MVD in Vitro Bioassay

The methods were carried out as previously described.⁸

In Vivo Assay

Animals

The experiments contained herein were carried out using male Sprague Dawley rats (250-350g; Harlan; Indianapolis, IN). All animals were maintained on a 12/12 h light/dark cycle (lights on at 07:00 am) and provided food and water ad libitum except as noted during the experimental procedures. All the experiments were performed under a protocol approved by Institutional Animal Care and Use Committee (IACUC) of the University of Arizona, and in accordance with policies and guidelines for the care and use of laboratory animals as adopted by International Association for the Study of Pain (IASP) and the National Institutes of Health (NIH). L₅/L₆spinal nerve ligation surgery. SNL injury was induced as described by Chung and colleagues.¹³ Intrathecal catheter surgery. All rats were prepared for intrathecal (i.th.) drug administration by placing anesthetized (ketamine/xylazine 100 mg/kg, i.p.) animals in a stereotaxic head holder. The cisternum magnum was exposed, an incision was made and animals were implanted with a catheter (PE: 10, 8 mm) that terminated in the lumbar region of the spinal cord. The animals were allowed to recover 5-7 days post-surgery before any pharmacological manipulations were made. Drug Administration. 6 was dissolved in 100% methanol. For i.th. drug administration, 5 μL of drug was injected followed by a 9 μL saline flush. Testing took place 15, 30, 45, 60, 75 and 90 min after drug injection and dose–response curves were generated from data gathered at the time of peak effect.

Behavioral Assessment

Thermal Hypersensitivity

Thermal hypersensitivity was assessed using the rat plantar test (Ugo Basile, Italy) following a modified method of Hargreaves et al.¹⁵ Paw withdrawal latencies were recorded in seconds. An automatic cut off point of 33 sec was set to prevent tissue damage. The apparatus was calibrated to give a paw withdrawal latency of approximately 20 sec on the uninjured paw. The radiant heat source was activated with a timer and focused onto the plantar surface of the hindpaw. Paw-withdrawal latency was determined by a motion detector that halted both heat source and timer when the paw was withdrawn.

Mechanical Hypersensitivity

The assessment of mechanical hypersensitivity consisted of measuring the withdrawal threshold of the paw ipsilateral to the site of nerve injury in response to probing with a series of calibrated von Frey filaments. Each filament was applied perpendicularly to the plantar surface of the ligated paw of rats kept in suspended wire-mesh cages. Measurements were taken both before and after administration of drug or vehicle. The withdrawal threshold was determined by sequentially increasing and decreasing the stimulus strength (‘up–down’ method) analyzed using a Dixon non-parametric test¹⁶ and expressed as the mean withdrawal threshold.

Table 2. Bioactivities of the Opioid Ligands.

no	hDORa [3H]DPDPE b		rMORa [3H]DAMGO c		[35S]GTP-γ-S binding						IC50 (nM)e
					hDORd			rMORd
	LogIC50f,g	(Ki, nM)h	LogIC50f,g	(Ki, nM)h	LogEC50f	EC50 (nM)i	Emax (%)j	LogEC50f	EC50(nM)i	Emax (%)j	MVD(δ)	GPI(μ)
1	-9.42±0.08	0.18	-9.06±0.06	0.39	-10.0±0.21	0.10	47	-9.95±0.20	0.11	77	0.69±0.09	1.6±0.1
2	-8.62±0.08	1.1	-9.16±0.10	0.36	-8.55±0.07	2.8	61	-8.73±0.16	1.9	29	8.5±1.7	4.7±1.2
3	-8.44±0.10	1.7	-9.47±0.20	0.15	-	nr	-	-9.03±0.19	0.94	33	15±1	3.6±0.8
4	-9.73±0.04	0.08	-9.71±0.03	0.10	-10.19±0.20	0.07	37	-9.85±0.43	0.14	58	1.9±0.5	1.3±0.3
5	-9.07±0.05	0.40	-10.50±0.10	0.02	-10.15±0.16	0.07	48	-9.53±0.33	0.29	98	0.37±0.12	0.26±0.14
6	-9.52±0.08	0.14	-9.41±0.06	0.14	-9.79±0.25	0.16	58	-9.46±0.13	0.31	55	0.70±0.38	2.6±0.8
7	-7.40±0.09	19	-7.97±0.11	5.0	-7.52±0.12	30	36	-7.72±0.25	19	15	250±90	220±40
8	-8.49±0.11	1.6	-9.13±0.04	0.33	-7.59±0.46	26	27	-8.33±0.33	4.7	29	180±50	93±30
9	-10.19±0.10	0.03	-10.81±0.07	0.01	-9.92±0.25	0.12	23	-9.16±0.21	0.69	50	1.8±0.8	1.0±0.1
10	-8.97±0.14	0.48	-9.15±0.13	0.35	-10.07±0.37	0.09	24	-9.14±0.23	0.72	51	0.029±0.007	0.96±0.13
11	-9.59±0.06	0.11	-9.55±0.04	0.15	-9.69±0.17	0.20	31	-10.23±0.27	0.06	37	0.21±0.06	4.8±0.7
12	-9.10±0.10	0.36	-9.08±0.22	0.38	-9.12±0.17	0.77	24	-9.05±0.18	0.88	50	1.8±0.2	8.5±3.3
YDAGF-NH2	-6.14±0.16	300	-8.24±0.13	2.8	-6.72±0.17	190	44	-7.98±0.22	13	99	120±10	47±9
DAMGO	-	-	-	-	-	-	-	-7.44± 0.19	37	150	-	-
DPDPE	-	-	-	-	8.80 ± 0.25	1.6	69	-	-	-	-	-

^aCompetition analyses were carried out using membrane preparations from transfected HN9.10 cells that constitutively expressed the respective receptor types.

^bK_d = 0.50 ± 0.1 nM.

^cK_d = 0.85 ± 0.2 nM.

^dExpressed in CHO cells.

^eConcentration at 50% inhibition of muscle contraction in electrically stimulated isolated tissues.

^fLogarithmic values determined from the non-linear regression analysis of data collected from at least two independent experiments.

^gCompetition against radiolabeled ligand.

^hAntilogarithmic value of the respective IC₅₀.

ⁱAntilogarithmic value of the respective EC₅₀.

^jNet total bound/basal binding × 100 ± SEM. nr: no response

Abbreviations

BBB

blood-brain barrier

Boc

tert-butyloxycarbonyl

BOP

(benzotriazole-1-yloxy)-tris(dimethylamino)-phosphonium hexafluorophosphate

CHO

Chinese hamster ovary

DMF

N,N-dimethylformamide

hDOR

human δ opioid receptor

DPDPE

c[D-Pen²,D-Pen⁵]enkephalin

DAMGO

[D-Ala²,NMePhe⁴,Gly⁵-ol]enkephalin

Dmt

2,6-dimethyltyrosine

GPI

guinea pig ileum

HOBt

1-hydroxybenzotriazole

IACUC

Institutional Animal Care and Use Committee

IASP

International Association for the Study of Pain

i.th

intrathecal

rMOR

rat μ opioid receptor

MVD

mouse vas deferens

NMM

N-methylmorpholine

RP-HPLC

reverse phase high performance liquid chromatography

SAR

structure-activity relationships

SNL

spinal nerve ligation

TFA

trifluoroacetic acid

Tic

1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid