Bifunctional Peptide-Based Opioid Agonist–Nociceptin Antagonist Ligands for Dual Treatment of Acute and Neuropathic Pain

Abstract

Herein, the opioid pharmacophore H-Dmt-d-Arg-Aba-β-Ala-NH₂ (7) was linked to peptide ligands for the nociceptin receptor. Combination of 7 and NOP ligands (e.g., H-Arg-Tyr-Tyr-Arg-Ile-Lys-NH₂) led to binding affinities in the low nanomolar domain. In vitro, the hybrids behaved as agonists at the opioid receptors and antagonists at the nociceptin receptor. Intravenous administration of hybrid 13a (H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-Tyr-Arg-Ile-Lys-NH₂) to mice resulted in potent and long lasting antinociception in the tail-flick test, indicating that 13a was able to permeate the BBB. This was further supported by a cell-based BBB model. All hybrids alleviated allodynia and hyperalgesia in neuropathic pain models. Especially with respect to hyperalgesia, they showed to be more effective than the parent compounds. Hybrid 13a did not result in significant respiratory depression, in contrast to an equipotent analgesic dose of morphine. These hybrids hence represent a promising avenue toward analgesics for the dual treatment of acute and neuropathic pain.

Graphical Abstract

INTRODUCTION

The efficient treatment of pain remains of paramount importance to our society. Not only does it influence the individual’s quality of life, it also significantly impacts healthcare systems around the world. Pain affects almost one in five adults across Europe (ca. 160 million people) and in the U.S., and one-third of all European households are affected by chronic pain.¹ To overcome pain, potent opioids like morphine, oxycodone, buprenorphine (1), and fentanyl are widely prescribed. However, these drugs are not effective in all patients and may even be not well-tolerated as a consequence of deleterious effects such as constipation,² nausea and vomiting,³ physical dependence,^4,5 tolerance,⁶ and respiratory depression.⁷ During the last 10 years, increasing prescriptions, availability, and misuse resulted in a rising number of fatalities due to respiratory depression in relation to unintentional opioid overdoses.⁸ Opioids represent about 8.6% of all drug-induced fatalities reported by the American Association of Poison Control Centers.⁹

To eliminate or reduce opioid-related side effects, other receptors involved in pain can be targeted. Examples of this strategy include the use of neurokinin-1 antagonists,¹⁰ neurotensin agonists,^11,12 or nociceptin (ant)agonists¹³ in combination with opioids. The nociceptin receptor (NOP), previously called the opioid-receptor like-1 (ORL-1) receptor, is a G protein-coupled receptor that was discovered in 1994.^14,15 The endogenous ligand of this receptor is the heptadecapeptide nociceptin/orphanin FQ (N/OFQ, H-Phe-Gly-Gly-Phe-Thr-Gly-Ala-Arg-Lys-Ser-Ala-Arg-Lys-Leu-Ala-Asn-Gln-OH). Activation of the nociceptin receptor induces several biological responses including hypotension and immunodepression, and it strongly modulates nociceptive transmission and potentiates the effect of morphine in neuropathic pain.^16,17 The nociceptin/orphanin FQ system seems to be involved in the modulation of acute nociceptive stimulation, as well as in chronic pain processes, for example, in inflammation^18,19 and neuropathic pain.^20–23 It is well established that N/OFQ may perform both pro- and antinociceptive actions, depending on the administration route²⁴ or dosage,²⁵ and both agonists and antagonists of NOP appear to be useful.^16,24,26

NOP antagonists on their own do not demonstrate antinociceptive potency, neither in acute²⁷ nor in neuropathic pain models, although they are reported to potentiate the antiallodynic effect of morphine under neuropathic conditions when coadministered iv,²⁸ and to enhance DAMGO ([d-Ala,²NMePhe,⁴Gly-ol]-enkephalin, MOP agonist) induced analgesia after microinjection to the ventrolateral periaqueductal gray, being the main output pathway involved in descending pain control.²⁹ Natural N/OFQ administered intrathecally, but not supraspinally, induces naloxone insensitive antinociceptive actions in rhesus monkeys.³⁰

NOP agonists in contrast may demonstrate powerful antinociceptive potency in acute pain but only in low doses (0.001–1 nmol it vs morphine >1 nmol it), as was demonstrated by Micheli and co-workers in rats.³¹ In the range of 0.001-1 nmol, a dose of 0.3 nmol it administered NOP agonist [pF-Phe,⁴Aib,⁷Arg,¹⁴Lys¹⁵]N/OFQ-NH₂ (UFP-112) afforded the highest antinociceptive effect. In addition, NOP agonists have not only been reported to be effective in acute pain but also not to lose their potency in neuropathic pain, as opioids do.^32,33 Interestingly, morphine’s effect in acute pain was significantly greater in naïve than in NOP(−/−) animals which suggests that the endogenous NOP system contributes to acute pain processing, at least in rats. Other authors report that this effect does not occur in mice models.³⁴ NOP agonists proved to be especially potent in nonhuman primates, which suggests a promising outlook for human patients therapies.³⁵ There is also evidence substantiating the use of NOP/MOP agonist hybrids in pain: intrathecal administration of a combination of inactive doses of UFP-112 and morphine attenuated hyperalgesia in rhesus monkeys.³⁶ Consequently, targeting two distinct pathways to produce analgesia may potentiate the overall analgesic effect, while ameliorating side effects, for example by requiring lower doses.

Activation or blockage of distinct targets can be achieved by combination therapy (i.e., using drug cocktails) or by the use of designed multiple ligands (DMLs).³⁷ The latter are single chemical entities able to bind two or more well-chosen receptor types. Although both strategies proved to be very useful, major advantages of DMLs consist of an early stage, thus less expensive optimization in the drug discovery process, and less complex pharmacokinetics.³⁷ Several small molecules with nanomolar binding affinities at both the nociceptin and the opioid systems have been reported (Figure 1).^28,38–44 Depending on their structural architectures, they behave as agonists, partial agonists, or antagonists on one or both of these systems. Several of these small molecule hybrids demonstrated interesting analgesic properties when tested in vivo. The high affinity, nonselective NOP/MOP partial agonist SR16435 (2) produced an antiallodynic response in the rat chronic constriction injury (CCI) model of neuropathic pain after intraperitoneal administration.⁴⁵ In a thermal antinociception test, the development of analgesic tolerance was reduced compared with morphine.⁴³ Co-administration of a NOP antagonist potentiated the antiallodynic effect of 2 and of morphine. This led the authors to suggest that NOP/MOP ligands with a dual profile of NOP antagonism and MOP agonism may be particularly useful for the treatment of chronic pain.⁴⁵

Figure 1.

Examples of reported opioid–NOP bifunctional ligands: 1^38,47–492^28,423⁴³4^44,46 a chimeric peptide 5^50,51 and 6^52,53

Compound 3, a full NOP agonist and weak partial (<20%) MOP agonist, did not have an effect in an acute pain model, but after subcutaneous administration to neuropathic sciatic nerve ligation (SNL) mice, it proved to possess an antiallodynic activity which could be suppressed by co-administration of a NOP, but not MOP, antagonist.⁴³ However, compound 3 was less potent than morphine in the SNL model. This may implicate the need to find a proper balance between activation of MOP and NOP receptors in allodynia alleviation.²⁸ Interestingly, cebranopadol 4, a highly potent agonist at MOP, DOP, KOP, and NOP, produced potent antinociception in rat models of acute and chronic pain and presented a clearly delayed development of tolerance.^44,46

Also peptide and peptidomimetic bifunctional compounds have been reported. Kawano⁵⁰ linked the opioid agonist dermorphin sequence to the Ac-RYYRIK-NH₂ sequence that was shown to have high NOP affinity and to behave as an agonist, partial agonist, or antagonist depending on the assay.^33,54,55 A synergistic and improved affinity for both the MOP and NOP was observed, leading to subnanomolar affinities when a long spacer -Gly-Gly-Gly-Lys(Gly-Gly-) was incorporated between the pharmacophores (5). In the mouse tail flick test, hybrid 5 showed an antinociceptive effect comparable to that of the isolated dermorphin sequence after it administration but less potent after icv administration. The antinociception was lower for the hybrid than for the coadministered opioid and nociceptin monomeric components despite the higher receptor affinities of the hybrid. No studies in neuropathic pain models were reported for hybrid 5.⁵¹ Becker⁵² screened a library of peptides, all incorporating β-turn inducing motifs for their binding at the opioid and nociceptin receptors and identified peptide III-BTD 6 with nanomolar binding affinities, and agonist properties at the opioid receptors and antagonist properties for NOP.^52,53

On the basis of the potential of dual opioid–nociceptin ligands for the treatment of chronic and neuropathic pain, which remain an unmet medical need, we now report the design and synthesis of new DMLs.³⁷ In contrast to the work of Kawano, the opioid part is directly connected to the nociceptin pharmacophore.⁵⁰ The opioid agonist part consists of the earlier reported and optimized H-Dmt-d-Arg-Aba-β-Ala-NH₂ tetrapeptide 7,⁵⁶ which is characterized by subnanomolar affinity and agonist activity for the opioid receptors. The nociceptin part is based on hexapeptides that were discovered by Dooley⁵⁷ (i.e., Ac-Arg-Tyr-Tyr-Arg-Ile-Lys-NH₂ (8) and analogues). Substitution of the N-terminal Ile-Lys dipeptide led to the Ac-Arg-Tyr-Tyr-Arg-Trp-Arg-NH₂ sequence which has a K_i of 0.1 nM for the human NOP and also behaves as a partial agonist in the [³⁵S]GTPγS assay.⁵⁴ The hexapeptides behave as partial agonists in cells expressing NOP,^54,57 but antagonism has been observed in rat brain preparations and in acute and chronic pain models in mice and rats.^33,55 Replacement of the Tyr³ residue in this hexapeptide with a para-fluorophenyl-alanine (pF-Phe) results in an even higher affinity (K_i = 0.05 nM).⁵⁴ The latter product behaves as a partial agonist (70% stimulation) similar to the parent compound.⁵⁴ Introduction of a 2′,6′-dimethyl-l-tyrosine (Dmt) residue at position 3 also gives a high affinity (K_i = 0.04 nM) partial agonist.⁵⁴

The N-terminus of these four peptides was linked to the C-terminus of our opioid pharmacophore in order to obtain bifunctional peptides. These new bifunctional peptide ligands were evaluated in vitro for their affinity and activity on both the opioid and the nociceptin receptors, and additionally these compounds were tested in vivo for acute and neuropathic pain alleviation. One of the hybrids was examined for effects on respiratory depression.

RESULTS AND DISCUSSION

Synthesis

The targeted bifunctional peptides all contain a tetrahydro-4-amino-2-benzazepinone (Aba) subunit which serves as a conformationally constrained Phe analogue. Incorporation of this building block, which was previously developed by us,^58,59 into several bioactive peptide sequences has successfully led to more stable and GPCR-subtype selective ligands and has provided highly potent opioid agonists.^60–63 In the current work, its insertion into the desired opioid–nociceptin hybrids was realized in two ways (indicated as pathways A and B in Scheme 1).

Scheme 1.

Solid Phase Synthesis of the Bifunctional Opioid–Nociceptin Peptides

Both strategies start with the synthesis of the resin-bound hexapeptides 10 by use of standard Fmoc-solid phase peptide synthesis (SPPS) on Rink amide AM resin. In pathway A, a segment condensation was performed between resin-bound 10 (with Xxx, Tyr(OtBu); Yyy, Ile; and Zzz, Lys(Boc)) and a preassembled protected tetrapeptide analogue Boc-Dmt-d-Arg(Pbf)-Aba-β-Ala-OH (11), which was prepared on 2-chlorotrityl resin, to obtain hybrid 12. The Aba-β-Ala component was formed directly on the solid support as previously reported.⁵⁶ After full protecting group removal of 12 and simultaneous cleavage from the resin, the product was purified by means of preparative RP-HPLC to obtain the pure compound 13a. In route B, dipeptide Fmoc-Aba-β-Ala-OH (14), was prepared by solution synthesis (Scheme 2) and coupled to the hexapeptides 10 to give 15. Fmoc-Aba-β-Ala-OH (14) was obtained, starting from phthaloyl protected phenylalanine 16 (Scheme 2), which was coupled to β-Ala-OEt·HCl using TBTU as coupling reagent and triethylamine as base. After crystallization of the resulting dipeptide 17, formation of the desired azepinone ring was achieved by use of an adapted literature procedure.⁶⁴ Dipeptide 17 was mixed with trioxane, AcOH, and H₃PO₄ in benzene at reflux in a Dean–Stark apparatus. P₂O₅ was added to avoid hydrolysis of the ester in the starting material due to the presence of traces of water. After refluxing for 4 h, ring closure proved to be complete. Following workup and flash chromatography purification of the crude mixture, the ester in 18 was hydrolyzed with a 1 N HCl in water/acetone (1:1) mixture at 90 °C for 16 h to isolate 19. The phthaloyl protecting group was then removed by hydrazinolysis. Subsequently, the free dipeptide was N-protected with a Fmoc group and the pure and SPPS-compatible building block 14 was isolated after final purification with flash chromatography in 22% yield (over four steps). After coupling of dipeptide mimetic 14 onto resin-bound 10 (Scheme 1), further peptide elongation via SPPS, full deprotection, and cleavage from the resin, the final peptides of type 13 were obtained in high purity (>95%) after preparative HPLC purification. The overall yield of pathway B was slightly higher and this pathway was easier to perform on larger scale (0.2 mmol). Only the first sequence 13a was synthesized via both methods; the rest of the series (i.e., 13b–13d, Table 1) was synthesized by use of pathway B. To allow comparison with the parent nociceptin receptor ligands, compounds 8 (Ac-Arg-Tyr-Tyr-Arg-Ile-Lys-NH₂) and 20 (H-Arg-Tyr-Tyr-Arg-Ile-Lys-NH₂) were also prepared via SPPS and evaluated in vitro and in vivo.

Scheme 2.

Synthesis of Dipeptidomimetic Fmoc-Aba-β-Ala-OH (14)

Table 1.

Compound Numbers, Sequences, and Yields of the Bifunctional Peptides

compd	sequence	yield (%)
13a	H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-Tyr-Arg-Ile-Lys-NH2	18(A)/25(B)
13b	H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-(pF-Phe)-Arg-Trp-Arg-NH2	17
13c	H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-Dmt-Arg-Trp-Arg-NH2	23
13d	H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-Tyr-Arg-Trp-Arg-NH2	22

In Vitro Biological Evaluation

In vitro binding and functional activity at the MOP, DOP, and KOP as well as at the nociceptin receptor were established for the parent (7, 8, 20) and hybrid compounds 13a–d (Table 2). In contrast to the parent structures, all hybrids maintained affinities within the moderate to low nanomolar range at both the opioid receptors and NOP. As expected, the parent structures possessed good binding only at the opioid or NOP receptors.

Table 2.

Binding Affinity and Activity at Nociceptin and Opioid Receptors

	binding (Ki, nM)				opioid activitya (IC50, nM)		activity at NOP
compd	MOP	DOP	KOP	NOP	GPI	MVD	PA2	EC50 (nM)
7	0.11 ± 0.01	4.2 ± 1.4	29.9 ± 4.2	>10000	0.8 ± 0.1	0.24 ± 0.02	nd	nd
8	321 ± 71	3100 ± 800	732 ± 45	0.53 ± 0.06	nd	nd	ndb	29 ± 1
20	3150 ± 150	>10000	1900 ± 800	39 ± 17	nd	nd	b	b
13a	5.0 ± 1.7	99 ± 4	33 ± 15	42 ± 6	6.1 ± 0.1	5.3 ± 1.4	5.39 ± 0.20	>10000
13b	15 ± 0.1	194 ± 19	84 ± 34	27 ± 6	2.0 ± 0.4	1.4 ± 0.2	6.00 ± 0.05	>10000
13c	15 ± 1.4	117 ± 12	22 ± 6	38 ± 4	5.5 ± 0.4	12.6 ± 1.9	6.25 ± 0.25	>10000
13d	9.6 ± 0.7	85 ± 19	47 ± 18	47 ± 9	3.6 ± 0.4	5.1 ± 0.8	6.01 ± 0.02	>10000

^aValues represent means of 3–4 experiments. The GPI functional assay is representative of MOP activation, whereas the MVD is a DOP representative assay.

^bReported as NOP agonists⁵⁷ and antagonists.^33,55 Here, no agonist activity up to 10 µM nor antagonist activity up to 20 µM. K_i, pA₂, and EC₅₀ values represent means ± SEM for at least two experiments performed in duplicates. nd: not determined IC₅₀.

Hybrid 13a combines the opioid ligand 7 with the peptide Ac-RYYRIK-NH₂ 8. This fusion led to a decrease in nociceptin and opioid receptor binding compared to the parent compounds, which is in contrast to the synergistic effect reported for hybrid 5.⁵¹ In comparison to NOP, the effect was less pronounced at the opioid receptors. Changing the last two amino acids (Ile-Lys to Trp-Arg, see 13d) did not have a beneficial influence on affinity in contrast to the improvement observed in the parent compound (going from 1.5 to 0.6 nM).^54,57 Substitution of Tyr with para-fluoro phenylalanine gave a small decrease in opioid affinity for 13b, but the NOP binding improved almost 2-fold (47 to 27 nM). Introduction of a 2′,6′-dimethyltyrosine (Dmt) at this position (13c) improved DOP and KOP binding compared to 13b but with a slight drop in NOP affinity. With respect to nociceptin receptor binding, no significant differences were observed between the four investigated hybrids because the affinity varied maximally with a factor of 2.

The in vitro opioid functional activity of the hybrids was verified by means of the guinea pig ileum (GPI) and mouse vas deferens (MVD) assays, representative of µ- and δ-opioid receptor activity, respectively. In agreement with the affinity data of these compounds, low nanomolar activities were determined for all analogues, 13b being the most potent of the series. Some discrepancies between DOP receptor binding data and agonist potencies in the MVD assay were observed. For example, compound 13b showed quite low DOP binding affinity (K_i^δ = 194 nM), but turned out to be a potent agonist in the MVD assay (IC₅₀ = 1.4 nM). There is evidence to indicate that in the vas preparation a cooperative effect between MOP and DOP receptors may be produced by some compounds,⁶⁵ and such cooperativity may be the cause for the unexpected high agonist potency of compound 13b. An alternative explanation would be that this compound may interact with another unknown receptor in the vas, as was also suggested to be the case with the NOP agonist/MOP partial agonist SR16476.⁶⁶

The activity of nociceptin parent NOP ligands 8 and 20, as well as hybrid molecules 13a–13d, was then assessed in the forskolin (FSK)-stimulated 3′,5′-cyclic adenosine monophosphate (cAMP) assay in HEK293 cells stably expressing the human NOP as described in the experimental section. As expected, nociceptin efficiently inhibited FSK-stimulated cAMP production (Figure 2) with an EC₅₀ of 0.23 ± 0.02 nM. Compound 8 displayed partial agonist activity at NOP (70 ± 1% of nociceptin maximum activity) with an EC₅₀ of 29 ± 1 nM, while 20 displayed neither agonist activity up to 10 µM nor antagonist activity up to 20 µM (Table 2). Hybrids 13a–13d did not display agonist activity at NOP up to 10 µM. Moreover, 20 µM of each compound significantly shifted the dose-response curve of nociceptin to the right (Figure 2), indicating that they displayed antagonist properties at this receptor. We then determined pA₂ values for each hybrid (Table 2). Our results show that 13a was the less potent NOP antagonist (pA₂ = 5.39), while the performed structural modifications led to improved antagonism for 13b–d, with pA₂ values of 6.00, 6.25 and 6.01, respectively.

Figure 2.

Inhibition of nociceptin effect on forskolin-stimulated cAMP accumulation by hybrids 13a–13d in HEK293 stably expressing human NOP receptor. Dose–response curves of nociceptin were performed in absence or presence of 2 and 20 µM of each compound. This figure shows a representative experiment. Evaluation were performed at least two times in duplicate.

In Vitro BBB Permeation Assays

To validate BBB permeation, a selected hybrid (13a) and its parent opioid (7) and NOP parent structures (8 and 20) were assayed in a PAMPA assay using porcine brain polar lipids as a model to study their passive diffusion transport through the blood-brain barrier (BBB)⁶⁷ at a concentration of 500 µM in the donor compartment. Propranolol, a β-adrenergic receptor blocker with high brain penetration, was used as a positive control. Permeability (P_e) is considered excellent if values >4.0 × 10⁻⁶ cm/s, uncertain between 2.0 × 10⁻⁶ and 4.0 × 10⁻⁶ cm/s, and poor with values below 2.0 × 10⁻⁶ cm/s.⁶⁷ Thus, looking at the results depicted in Table 3, none of the peptide analogues show a significant passive diffusion transport, as all of them had permeabilities below 2.0 × 10⁻⁶ cm/s (see Table 3).

Table 3.

Transport in BBB-PAMPA Assay

no.	Pe (cm/s) × 10−6	SD Pe (cm/s) × 10−6	transport (%)	SD transport (%)
7	0.9	0.6	1.8	1.3
20	nda	nda	nda	nda
8	0.07	0.05	0.15	0.10
13a	0.11	0.02	0.23	0.04
propranolol	8.59	0.8	15.2	1.4

^and: not detected.

Next, to shed light on potential transport of these compounds through the BBB and study whether they are able to cross by active transport, an in vitro cell-based model of the BBB was used (Figure 3).⁶⁸ This human in vitro model of the BBB uses brain-like endothelial cells, generated from human cord blood-derived hematopoietic stem cells cocultured with perycites. Peptide quantification and integrity were analyzed by UPLC and MALDI-TOF, respectively. In this model, Lucifer Yellow lithium salt was used as internal control, whose permeability (P_app) should stay below 17 × 10⁻⁶ cm/s. In all cases, the integrity of the model was preserved.

Figure 3.

Peptide transport in an in vitro cell-based model of the BBB.

Excellent transports were obtained in all cases (ranging from 7.0 to 13.7 × 10⁻⁶ cm/s), except for the peptide 20 (2.7 × 10⁻⁶ cm/s). This compound, in spite of its high degree of similarity with 8 (the N-terminal acetylated analogue of 20), showed a 3-fold reduction in transport, likely to be caused by the action of aminopeptidases (half the peptide was hydrolyzed after 2 h, see mass balance in Table 4 and Supporting Information). Compound stability was preserved in compound 13a, which also contains moiety 20 at the C-erminus.

Table 4.

Peptide Transport in an in Vitro Cell-Based Model of the BBB⁶⁸

no.	Papp (cm/s) × 10−6	Papp SD (cm/s) × 10−6	transport (%)	transport SD (%)	mass balance (%)	mass balance SD (%)
7	13.7	0.6	22.0	0.9	102	14
20	2.7	0.6	4.4	0.9	45	2
8	8.4	1.4	14	2	95	3
13a	7.0	0.2	11.3	0.3	105	6

Thus, the in vitro cell-based human model of the BBB, together with the BBB-PAMPA assay, enabled us to determine that all these compounds are transported by active mechanisms and not through passive diffusion. Moreover, as it is known, the BBB is also an enzymatic barrier,^69,70 and in this regard we observed an increased resistance to proteases for compounds 8 and 13a thanks to the N-terminal capping either with an acetyl or the opioid pharmacophore, respectively.

Behavioral Studies in Mice and Rats

Acute Pain in Mice (Tail-Flick Test)

Evaluation of the antinociceptive potency of hybrid 13a by the tail-flick test in mice after iv administration revealed a maximal effect at 60 min post administration while the maximal effect of morphine was typically reached after 15–30 min (Figure 4). Compared to morphine, an equipotent effect was measured at much lower doses (6, 31, 61 nmol vs morphine in 390 nmol dose). The %MPE of morphine started to decline from 60 min postadministration, while the nociceptive effect of 13a maintained high %MPE values for longer periods of time. Longer lasting and maximal effects were measured at 180 and 240 min for both the 31 and 61 nmol doses (100 %MPE for both doses at both time points and 79.8 %MPE for 6 nmol dose at 180 min after administration (data not recorded for morphine at these time-points)). The above results suggested that potent CNS-mediated effects can be obtained by iv injection of hybrid 13a and, hence, BBB permeation did not seem to represent a limitation for this compound as indicated above (cf., in vitro BBB permeability assay). This peptide was modified at the N- and C-terminus and internally by the Aba lactam, which should increase its resistance to enzymatic degradation. The stability of 13a in plasma was determined and, to our satisfaction, revealed a half-life time of 585 min at 37 °C in human plasma (see Supporting Information).

Figure 4.

Time and dose dependent analgesic effect of 13a after iv administration compared to morphine. Ordinal values represent tail-flick latency measurements which were normalized as %MPE (mean ± SEM).

When hybrid 13a was administered orally to mice at doses of 61, 122, and 196 nmol, no analgesic effects were observed (see Supporting Information, no %MPE > 5% was observed), even at such high doses.

Neuropathic Pain Tests in Mice

All bifunctional compounds (13a–13d), the parent NOP sequence (20), as well as its N-acetylated form (8), were tested in neuropathic pain models in mice. The ligands were all examined as potential suppressors of allodynia and hyperalgesia, phenotypes which were caused by chronic constriction injury to the sciatic nerve (CCI) or by a streptozocin (STZ)-induced type 1 diabetes.^71,72 The results of these tests are presented in Figure 5 (von Frey test, allodynia) and Figure 6 (cold plate test, hyperalgesia). The calculated ED₅₀ values for the parent compounds and all four hybrids at one time point (30 min) are presented in Table 3. The results for the parent opioid sequence 7 in mice and rats were obtained during an earlier study.^56,73

Figure 5.

Effect of intrathecal (it) administration of hybrids 13a–13d and nociceptin parents 8 and 20 and morphine (all doses in nmol, 6–8 animals per group) on allodynia in CCI (A) and STZ (B) mice. As naive mice do not react to non-noxious stimuli, their reaction threshold may be established close to (5.9 ± 0.1 g) cutoff level (6 g). In comparison, the reaction threshold in CCI mice is 1.03 ± 0.07 g.

Figure 6.

Effect of intrathecal (it) administration of hybrids 13a–13d and nociceptin parents 8 and 20 and morphine (all doses in nmol, 6–8 animals per group) on hyperalgesia in CCI (A) and STZ (B) mice. The naive mice very poorly respond to noxious thermal stimuli in cold plate test (26.8 ± 0.45 s), and the value is close to cutoff level (30 s). In comparison, the reaction threshold in CCI mice is 5.4 ± 0.3 s.

As can be noticed, all bifunctional compounds proved to be significantly more effective than morphine, both in terms of antiallodynic and antihyperalgesic effects. Interestingly and with exception of hybrid 13b, the hybrids performed better than the nociceptin parent compounds corresponding to the NOP pharmacophore of 13a (Figures 5 and 6). The hybrids were highly effective at low doses, especially when compared to morphine [e.g., ED₅₀ (13a) = 0.003 nmol vs ED₅₀ (morphine) = 7 nmol, allodynia; ED₅₀ (13a) = 0.004 nmol vs ED₅₀ (morphine) = 7.46 nmol, hyperalgesia]. When compared to the opioid parent compound 7 with ED₅₀ values of 0.85 nmol (von Frey) and 1.60 nmol (cold plate), the hybrids were also much more potent.⁷³ These observations support the hypothesis that more potent effects can be induced by the presence of both pharmacophores in 13a–13d when compared to the effect of a single opioid or nociceptin unit. The effect of opioid–nociceptin hybrids, in spite of their low NOP antagonist potency, is profound because it originates from the simultaneous modulation of the transmitted signal in the same part of the nociceptive pathway and the influence concerns both endogenous opioid and nociceptin systems which are very important for nociceptive transmission. Moreover, the endogenous nociceptin/orphanin FQ system, apart from its analgesic action, is known to exhibit anti-opioid activity and may contribute to lower responsiveness to morphine in neuropathic pain.²¹ Therefore, simultaneous modulation can cause a much greater change of the input signal than a single compound can evoke. The hybrid strategy offers the advantage over the administration of the two substances which can give way to a different distribution profile.

Nevertheless, also the NOP parent compounds remained effective at low doses, especially with regard to allodynia [e.g., ED₅₀ (8) = 0.008 nmol, ED₅₀ (20) = 0.004 nmol]. The N-terminally acetyl-capped parent analogue 8 was clearly less potent than the noncapped 20, which was most pronounced in the cold plate assay [e.g., ED₅₀ (8) = 0.69 nmol, ED₅₀ (20) = 0.06 nmol]) despite its higher NOP affinity (IC₅₀ = 0.53 vs 39 nM, Table 2). Both NOP parents (8 and 20) were less potent in the cold plate test (thermal stimulus). Generally, thermal hyperalgesia and tactile allodynia do not correlate neither in mice nor in humans,⁷⁴ and these effects are mediated not only by distinct cellular mechanisms, which is corroborated by a huge amount of data of assorted nature from electrophysiological to clinical observations⁷⁵ but also by different afferent fibers types.⁷⁶

Hybrid 13a seems to be the best of all investigated hybrids, parent peptides, and morphine. The antiallodynic and antihyperalgesic effects of 13a were more prolonged than those produced by a classic analgesic like morphine and stayed extremely high even 180 min after it administration and injection of very low doses (e.g., 0.0005 nmol). The effect of the parent compound 20 is transient, while the acetylated parent 8 is efficient at higher doses only (0.8 vs 0.05). Hybrid 13a is clearly superior (even at very low doses) to the parents in the cold plate test as well as two neuropathy models. Hence, in contrast to common opioids such as morphine, the tested opioid–NOP DMLs exhibit high efficacy in neuropathic pain models. In mice, morphine performs dose-dependently in acute pain, with a dose of 10 µg/animal giving 63.8 ± 15.9 %MPE in the tail-flick test. The effect reaches 100 %MPE upon increased dosing.⁷³ On the contrary, under neuropathic conditions, morphine provides 36.5 ± 9.5 %MPE in von Frey test and 30.15 ± 8.2 %MPE in cold plate test at the respective dose of 10 µg/animal. The effect never reaches maximum values; it sustains at about 50 %MPE no matter how high the dose is.⁷³

Of all hybrids, compound 13b was the least effective [ED₅₀ of 0.18 nmol (von Frey) and 0.14 nmol (cold plate)] and it only showed a maximal effect, comparable to the other hybrids, at a high dose of 0.5 nmol (Figure 5A and 6A). Hybrids 13c–13d seemed also to be promising compounds in the investigated neuropathic pain models. Taking the results depicted in Figures 5 and 6 into consideration, these analogues presented the most potent and long lasting effects, especially with regard to antiallodynic (von Frey) efficacy.

Altogether, hybrid 13a and 13d seemed to be the most promising compounds in the investigated neuropathic pain models. Taking the results depicted in Figure 5 and 6 into consideration, these analogues presented the most potent and long lasting effects, especially with regard to antiallodynic (von Frey) efficacy (Table 5).

Table 5.

Calculated ED₅₀ Values for Effect of Hybrids (13a–13d), Opioid Parent Compound 7, Nociceptin Parent Compounds 8 and 20, and Morphine 30 min after Their Administration in Neuropathic Pain CCI Model in Mice, 7–14 Days after Surgical Procedure^a

compd	response at 30 min – ED50
	(dose in nmol it) with confidence interval 95%
	von Frey	cold plate
7	0.85 (0.22–11.30)	1.60 (0.51–5.07)
8	0.008 (0.0002–0.4)	0.69 (0.13–3.5)
20	0.004 (0.001–0.01)	0.06 (0.02–0.15)
13a	0.003 (0.001–0.005)	0.004 (0.001–0.01)
13b	0.18 (0.06–0.5)	0.14 (0.1–0.21)
13c	0.0009 (0.0001–0.03)	0.005 (0.001–0.02)
13d	0.0009 (0.0001–0.05)	0.002 (0.001–0.003)
morphine	7 (0.63–76.75)	7.46 (1.57–35)

^aAllodynia was measured by the von Frey test, and hyperalgesia was measured by the cold plate test.

The observations made in the injury-induced neuropathy models also applied to the diabetic neuropathy, although in the latter model the antiallodynic action of the compounds showed a maximum at 30 min which slightly decreased at 180 min (Figure 4B). This decrease was also present in the antihyperalgesic action (Figure 5B) except for 13a. This may result from intermodel differences in mechanisms underlying the neuropathic symptoms occurrence and maintenance, although it was still promising that 13a was able to alleviate the painful symptoms of both hyperalgesia and allodynia in distinct models. As the discovery of a multifunctional drug alleviating assorted symptoms of neuropathy remains of high interest, the above results provide a worthy outlook toward other novel therapies.

All compounds were also evaluated in the Rota-Rod test to examine motor function of mice. Only the 0.5 nmol dose of 13b gave way to impaired motor function. At the used doses, all other compounds did not impair motor functions (data not shown).

Respiratory Depression in Rats

Respiratory effects in rats were determined by plethysmography after iv administration of equipotent antinociceptive doses (in tail-flick test) of 13a (0.5 mg/kg, 6 nmol/kg) and morphine (5 mg/kg, 390 nmol/kg). On the basis of the measurement of the ventilation minute, administration of 13a did not result in a significant respiratory depression in comparison to saline, a result that stands in contrast to the deleterious effect of morphine (p < 0.01; Figure 7F). Interestingly, morphine induced a significant increase in inspiratory time in comparison to saline and 13a (p < 0.05; Figure 7B), while 13a induced a significant decrease in expiratory time in comparison to morphine (p < 0.05; Figure 7E). Combining both effects, morphine induced a decrease in respiratory frequency (p < 0.001, Figure 7D), while 13a slightly increased this frequency. The tidal volume was significantly reduced with 13a in comparison to saline and morphine (p < 0.05; Figure 7C).

Figure 7.

Effects in Sprague–Dawley rats of intravenous 13a (0.5 mg/kg, open circles), morphine (5 mg/kg, gray triangles), and 0.9% NaCl (black triangles) on plethysmography parameters (N = 6/group). Areas under the effect-time curve (AUC) were determined. Results are expressed as mean ± SEM. Comparisons were performed using Kruskal–Wallis tests. *p < 0.05, **p < 0.01.

An increase in inspiratory time is a marker of opioid-related effects on the respiratory control centers in the brainstem.⁷⁷ Remarkably, 13a did not result in any significant increase in inspiratory time at the studied 0.5 mg/kg dose. Moreover, in comparison to morphine, 13a significantly reduced the expiratory time. Both effects show the limited respiratory effects related to 13a in comparison to the deleterious ones of morphine, studied only at a single therapeutic dose. A more elaborate study will be carried out to compare the ratio of the ED₅₀ % of the respiratory-to-analgesic effects of 13a versus morphine.

CONCLUSIONS

In vitro, the opioid–nociceptin hybrids behaved as potent opioid agonists and weak nociceptin antagonists. Although fusion of the two peptide fragments led to a slight loss in opioid and nociceptin receptor binding, low nanomolar affinities were maintained. On the basis of the in vitro evaluation, compound 13a was selected for further in vivo tests. After intravenous administration to naïve mice, it was more active than morphine and showed a prolonged analgesic action (>3 h) in an acute pain model. This result indicated that hybrid 13a was transported through the BBB, a hypothesis which was further supported by a cell-based in vitro BBB model. In addition, all hybrid compounds were administered intrathecally to CCI mice and were proven to be more effective than morphine and the parent NOP ligands in both antiallodynic and antihyperalgesic effects, with compounds 13a and 13d exhibiting the best in vivo profile. Compared to morphine, the effects of 13a were much more prolonged and extremely high at low doses even at 180 min postadministration. Because 13a proved to be very potent in both acute and neuropathic pain models, potential respiratory depression effects linked to the use of this product were evaluated. Inspiratory and expiratory times show that the administration of 13a did not result in a significant respiratory depression. Altogether, the opioid–NOP hybrids are highly effective analgesics in the investigated neuropathic pain models. The best compounds have ED₅₀ values about 1000 times lower than those of the opioid agonist/NK1 antagonist hybrid H-Dmt-d-Arg-Aba-β-Ala-NMe-Bn(CF₃)₂ in the von Frey test and about 20 times lower in the cold plate test.^56,73 This shows that opioid–nociceptin hybrids are highly promising compounds as analgesics for the dual treatment of acute and neuropathic pain, devoid of respiratory depression.

MATERIALS AND METHODS

General

Thin-layer chromatography (TLC) was performed on glass plates precoated with silica gel 60F₂₅₄ (Merck, Darmstadt, Germany) using the mentioned solvent systems. Purification of organic molecules was done with flash chromatography (Davisil LC60A, 40–63 µm). Mass spectrometry (MS) was performed on a Micromass Q-Tof Micro spectrometer with electrospray ionization (ESI). Data collection and spectrum analysis was done with Masslynx software. Analytical RP-HPLC was performed using a Waters 717plus autosampler, a Waters 1525 binary HPLC pump, and a Waters 2487 dual absorbance wavelength detector (Milford, MA) on a Grace (Deerfield, IL) Vydac RP C18 column (25 cm × 4.6 mm × 5 µm) using UV detection at 215 nm. The mobile phase was a mixture of water and acetonitrile both containing 0.1% TFA. The used gradient runs from 3 to 100% acetonitrile in 20 min at a flow rate of 1 mL/min. Preparative RP-HPLC purification was done on a Gilson (Middleton, WI) HPLC system with Gilson 322 pumps, controlled by the software package Unipoint and a reversed phase C18 column (DiscoveryBIO SUPELCO Wide Pore C18 column, 25 cm × 2.21 cm, 5 µm) using a gradient that increased by 1%/min of acetonitrile in water (both containing 0.1% TFA) until the product eluted. After purification, the purity of all compounds was evaluated as being more than 95% by analytical RP-HPLC. All fractions were lyophilized using a Flexy-Dry lyophilizer (FTS Systems, Warminster, PA). ¹H and ¹³C NMR spectra were recorded at 500 and 125 MHz on a Bruker Avance II 500 (Bruker Corp, Billerica, MA). Tetramethylsilane (TMS) or residual solvent signals were used as internal standard. The solvent used is mentioned in all cases, and the abbreviations used are as follows: s (singlet), d (doublet), dd (double doublet), t (triplet), q (quadruplet), and m (multiplet).

Synthesis

General Peptide Synthesis

All peptides were synthesized manually by Fmoc-based solid phase peptide synthesis (SPPS) on Rink amide AM resin. For standard couplings, a 3-fold excess of the Fmoc-protected amino acids and 3-fold excess of coupling reagent (HCTU) in 0.4 NMM in DMF was used for 1.5 h. For the coupling of Fmoc-Aba-β-Ala-OH, 1.5-fold excess for both the dipeptide and the coupling reagent was used and coupling was left for 3 h. Fmoc deprotection was carried out by treatment of the resin with 20% 4-methylpiperidine in DMF for 5 and 15 min. After every reaction step, the resin was washed with DMF (3 × 1 min), iPrOH (3 × 1 min), and CH₂Cl₂ (3 × 1 min).

When the sequence was complete, the peptide was cleaved from the resin and completely deprotected with TFA/TES/H₂O (95:2.5:2.5) for 3 h. The resin was filtered, and the filtrate was concentrated and added to cold ether. The precipitated peptide was then dissolved in acetonitrile/H₂O and lyophilized to get the crude peptides as a powder.

The crude peptides were dissolved in H₂O, and acetonitrile was added until complete dissolution was observed. The solution was injected on a Gilson preparative RP-HPLC. Fractions containing the pure peptide were collected, combined, and lyophilized. The peptides were obtained as white powders with a purity of >95% as determined by analytical HPLC. The structures were confirmed by high-resolution electrospray mass spectrometry.

Coupling of the Protected Peptide (11) to the Sixmer (10) (Pathway A)

The protected peptide 11 (1.5 equiv) was dissolved in DMF, and DIC (1.5 equiv) and HOBt (1.5 equiv) were added to the solution. This coupling solution was then added to the resin and shaken for 3 h. No base was added to avoid additional coupling at the unprotected Dmt side chain. After washing with DMF (3×), iPrOH (3×), and CH₂Cl₂ (3×), the peptide was cleaved from the resin and purified (See General peptide synthesis).

Peptide Characterization

Ac-Arg-Tyr-Tyr-Arg-Ile-Lys-NH₂ (8)

Preparative HPLC yielded the desired compound (white powder, 34%). HPLC: t_R = 9.4 min. TLC R_f 0.23 (EBAW). HRMS (ESP⁺) found m/z 939.5515 [M + H]⁺, C₄₄H₇₁N₁₄O₉⁺ requires 939.5523.

H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-Tyr-Arg-Ile-Lys-NH₂ (13a)

Preparative HPLC yielded the desired compound (white powder, 25%). HPLC: t_R = 10.0 min. TLC R_f 0.60 (EBAW). HRMS (ESP⁺) found m/z 1474.8403 [M + H]⁺, C₇₂H₁₀₇N₂₁O₁₃⁺ requires 1474.8430.

H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-(pF-Phe)-Arg-Trp-Arg-NH₂ (13b)

Preparative HPLC yielded the desired compound (white powder, 17%). HPLC: t_R = 11.2 min. TLC R_f 0.63 (EBAW). HRMS (ESP⁺) found m/z 1577.8538 [M + H]⁺, C₇₇H₁₀₆FN₂₄O₁₂⁺ requires 1577.8401.

H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-Dmt-Arg-Trp-Arg-NH₂ (13c)

Preparative HPLC yielded the desired compound (white powder, 23%). HPLC: t_R = 10.7 min. TLC R_f 0.61 (EBAW). HRMS (ESP⁺) found m/z 802.4427 [M/2 + H]⁺, C₇₉H₁₁₁N₂₄O₁₃⁺/2 requires 802.4418.

H-Dmt-d-Arg-Aba-β-Ala-Arg-Tyr-Tyr-Arg-Trp-Arg-NH₂ (13d)

Preparative HPLC yielded the desired compound (white powder, 22%). HPLC: t_R = 10.5 min. TLC R_f 0.60 (EBAW). HRMS (ESP⁺) found m/z 788.4251 [M/2 + H]⁺, C₇₇H₁₀₇N₂₄O₁₃⁺/2 requires 788.4261.

H-Arg-Tyr-Tyr-Arg-Ile-Lys-NH₂ (20)

Preparative HPLC yielded the desired compound (white powder, 33%). HPLC: t_R = 9.0 min. TLC R_f 0.07 (EBAW). HRMS (ESP⁺) found m/z 897.5428 [M + H]⁺, C₄₂H₆₉N₁₄O₈⁺ requires 897.5417.

Synthesis of the Dipeptide Building Blocks

Fmoc-Aba-β-Ala-OH (14)

Phth-Aba-β-Ala-OH 19 (3.91 g, 10.3 mmol, 1 equiv) was dissolved in 110 mL of ethanol. Hydrazine hydrate (3.0 mL, 62.0 mmol, 6 equiv) was added, and the solution was refluxed for 1.5 h, after which the solvent was evaporated. The residue was dissolved in 60 mL of water, and the pH was carefully adjusted to pH 5 by dropwise addition of acetic acid and monitoring with a pH meter. The suspension was stirred for 1 h at room temperature and the phthalhydrazide side-product precipitated. This was filtered off, and the filtrate was evaporated. The residue was dissolved in 100 mL of a water/acetone mixture. Sodium carbonate (261 mg, 11.4 mmol, 1.1 equiv) was added together with Fmoc-OSu (3.49 g, 10.3 mmol, 1 equiv) for the Fmoc protection. The reaction mixture was stirred for 16 h. The solvent was evaporated, and the residue was brought to pH 2 with 6 N HCl. Ethyl acetate was added, and the phases were separated. The organic phase was washed with a saturated NaHCO₃ solution (3×) and brine. The residue was purified by column chromatography with 1% methanol in CH₂Cl₂ (+1% AcOH) and a yellow solid was obtained in 54% yield.

Yield: 54% (2.66 g). Formula: C₂₈H₂₆N₂O₅. MW: 470.53 g/mol. R_f = 0.75 (EtOAc + 1% AcOH). HPLC: t_R = 17.3 min. HRMS (ESP⁺): found m/z 471.1909 [M + H]⁺, C₄₄H₇₁N₁₄O₉⁺ requires 471.1914; melting interval, 165.0–170.0 °C (decomposition). ¹H NMR (CDCl₃, 500 MHz): δ (ppm) = 2.32 (2H, m), 2.94 (1H, dd, J = 17.1 Hz, J = 13.5 Hz), 3.18 (1H, dd, J = 17.3 Hz, J = 4.5 Hz), 3.56 (2H, m), 4.14 (1H, d, J = 16.7 Hz), 4.24 (1H, m), 4.31 (2H, m), 5.08 (1H, m), 5.13 (1H, d). ¹³C NMR (CDCl₃, 125 MHz): δ (ppm) = 33.3, 35.6, 43.9, 47.1, 50.2, 51.4, 66.3, 120.6, 125.8, 126.4, 127.6, 128.1, 128.1, 129.2, 131.0, 135.2, 135.9, 141.2, 144.4, 156.1, 171.5, 173.0.

Phth-Phe-β-Ala-OEt (17)

Phthaloyl protected phenylalanine 16 (5 g, 16.9 mmol, 1 equiv) was dissolved in 80 mL of CH₂Cl₂. β-Alanine ethylester hydrochloride (β-Ala-OEt.HCl, 2.86 g, 18.6 mmol, 1.1 equiv) and coupling reagent TBTU (5.98 g, 18.6 mmol, 1.1 equiv) were added. Et₃N (7.08 mL, 50.8 mmol, 3 equiv) was added to the solution, and the pH was kept at pH 8 by use of Et₃N. The solution was then stirred for 1 h. The solvent was evaporated, and the residue was dissolved in ethyl acetate. The solution was then washed with 1 N HCl solution (3×), saturated NaHCO₃ solution (3×), and brine (3×). The organic phase was dried with MgSO₄, filtered, and evaporated. The resulting residue was crystallized from a minimum amount of hot ethanol. After cooling down and filtration, white crystals were obtained with a yield of 70% after two subsequent crystallizations.

Yield: 70% (4.649 g, white crystals). Formula: C₂₂H₂₂N₂O₅. MW: 394.43 g/mol. R_f = 0.56 (EtOAc/petroleum ether 1:1). HPLC: t_R = 15.9 min. MS (ES+): 395 [M + H]⁺, 349 [M – OEt]⁺; melting interval, 121.1–121.9 °C. ¹H NMR (CDCl₃, 500 MHz): δ (ppm) = 1.23 (3H, t, J = 7.1 Hz), 2.53 (2H, m), 3.54 (4H, m), 4.09 (2H, q, J = 7.1 Hz), 5.08 (1H, dd, J = 10.4 Hz, J = 6.3 Hz), 6.70 (1H, m), 7.04–7.23 (5H, m), 7.63–7.82 (4H, m). ¹³C NMR (CDCl₃, 125 MHz): δ (ppm) = 14.1, 33.7, 34.7, 35.2, 55.7, 60.8, 123.5, 126.9, 128.6, 128.9, 131.4, 134.2, 136.7, 167.8, 168.4, 172.5.

Phth-Aba-β-Ala-OEt (18)

In a two-neck flask, equipped with a Dean–Stark apparatus, P₂O₅ (10 g, 70.5 mmol, 11.1 equiv), 85% H₃PO₄ (6.5 mL, 95.0 mmol, 15 equiv), 100 mL of acetic acid, and 150 mL of benzene were mixed. The solution was refluxed for 30 min, then the starting material 17 (2.5 g, 6.34 mmol, 1 equiv) and trioxane (3.75 g, 41.6 mmol, 6.6 equiv) were added. The mixture was refluxed until completion of the reaction (4 h) and every 30 min, trioxane (3.75 g, 41.6 mmol, 6.6 equiv) was added. After 2 h, P₂O₅ (10.0 g, 70.5 mmol, 11.1 equiv) was added. After completion, benzene was evaporated and the residue was diluted with ether. This mixture was washed three times with 1 N HCl, saturated NaHCO₃, and brine. The organic phase was dried, filtered, and evaporated to obtain a yellow oil. This residue was then purified with flash chromatography with 30% EtOAc in petroleum ether to yield a yellow oil in 65%.

Yield: 65% (1.68 g). Formula: C₂₃H₂₂N₂O₅. MW: 406.44 g/mol. R_f = 0.61 (EtOAc/petroleum ether 1:1). HPLC: t_R = 16.6 min. MS (ES +): 407 [M + H]⁺, 429 [M + Na]⁺, 361 [M − OEt]⁺. ¹H NMR (CDCl₃, 500 MHz): δ (ppm) = 1.22 (3H, t, J = 7.2 Hz), 2.62 (2H, m), 3.12 (1H, dd, J = 15.6 Hz, ³J = 4.7 Hz), 3.74 (1H, m), 3.89 (1H, m), 4.11 (3H, m), 4.71 (1H, d, J = 15.8 Hz), 4.78 (1H, d, ²J = 15.9 Hz), 5.36 (1H, dd, J = 13.1 Hz, J = 4.9 Hz), 7.25–7.30 (4H, m), 7.75 (2H, m), 7.89 (2H, m). ¹³C NMR (CDCl₃, 125 MHz): δ (ppm) = 14.1, 33.2, 34.1, 46.6, 52.0, 53.0, 60.7, 123.5, 127.1, 128.5, 128.7, 130.0, 132.0, 134.1, 135.6, 135.9, 168.0, 168.6, 172.0.

Phth-Aba-β-Ala-OH (19)

Phth-Aba-β-Ala-OEt 18 (4.74 g, 11.7 mmol, 1 equiv) was dissolved in 60 mL of acetone. Then 60 mL of a 1 N HCl solution was slowly added. The mixture was refluxed in an oil bath at 90 °C for 16 h and then cooled to room temperature, and the solvent was evaporated. A white solid was obtained in 89% yield and was used in the next step without purification.

Yield: 89% (3.91 g). Formula: C₂₁H₁₈N₂O₅. MW: 378.38 g/mol. R_f = 0.62 (EtOAc + 1% AcOH). HPLC: t_R = 15.7 min. MS (ES+): 379 [M + H]⁺, 401 [M + H]⁺, 417 [M + K]⁺; melting interval, 201.5–203.0 °C. ¹H NMR (CDCl₃, 500 MHz): δ (ppm) = 2.68 (2H, m), 3.10 (1H, dd, J = 15.5 Hz, J = 4.8 Hz), 3.72 (1H, m), 3.88 (1H, m), 4.11 (1H, dd, J = 15.5 Hz, J = 13.7 Hz), 4.73 (2H, pseudo-s), 5.32 (1H, dd, J = 13.0 Hz, J = 4.9 Hz), 7.22–7.32 (4H, m), 7.73 (2H, m), 7.87 (2H, m). ¹³C NMR (CDCl₃, 125 MHz): δ (ppm) = 32.9, 34.2, 46.6, 52.1, 53.2, 123.6, 127.3, 128.4, 128.8, 130.0, 132.0, 134.2, 135.6, 135.8, 167.9, 168.8, 175.7.

In Vitro Biological Evaluation

In Vitro Affinity

Radioligand Binding Assays

Membranes from HEK293 cells transiently expressing human MOP, DOP, and KOP, as well as NOP, were obtained as previously reported. ^52,78 All membrane preparations were stored at −80 °C as aliquots (1 mg protein/mL) until use. Following their dilution in 50 mM Tris-HCl buffer (pH 7.4) supplemented with 0.1% BSA, opioid receptor-containing membranes were incubated for 1 h at 25 °C with [³H]Diprenorphine (0.6 nM; MOP, DOP, and KOP) or [³H]OFQ/nociceptin (0.15 nM; NOP) and increasing concentrations of compounds to be tested in a final volume of 0.2 mL. Nonspecific binding was defined in the presence of 10 µM naloxone (MOP, DOP, and KOP) or OFQ/nociceptin (NOP). Membrane-bound radioactivity was separated from free radioligand by rapid filtration through a 96-well GF/B unifilter apparatus (PerkinElmer Life and Analytical Sciences, Courtaboeuf, France) and quantified using a TopCount scintillation counter (PerkinElmer).

cAMP Accumulation

The agonist/antagonist nature of MOP/NOP hybrids at NOP was examined by using the GloSensor cAMP assay according to manufacturer recommendations (Promega, Madison WI, USA) with a few modifications. HEK293 cells selected for stable expression of human NOP and GloSensor were harvested from culture dishes and suspended (10⁶ cells per mL) in physiological Hepes buffer (10 mM HEPES, 0.4 mM NaH₂PO₄, 137.5 mM NaCl, 1.25 mM MgCl₂, 1.25 mM CaCl₂, 6 mM KCl, 10 mM glucose, and 1 mg/mL bovine serum albumin, pH 7.4) supplemented with 1 mM d-luciferin. Following pre-equilibration for 2 h at 25 °C, d-luciferin-loaded HEK cells were distributed (100000 cells per well) in white 96-well Greiner plates (Courtaboeuf, France) and kinetic recordings of their luminescence level were acquired using a FlexStation II. In the agonist mode, compounds to be tested were injected at various concentrations 15 min before forskolin addition (0.5 µM final concentration) and readings were pursued for 90 min. In the antagonist mode, different concentrations of compounds to be tested were preincubated with cells for 15 min before the addition of various concentrations of OFQ/nociceptin. Experiments were conducted at 25 °C in the presence of 0.5 mM IBMX to prevent the degradation of cAMP by phospodiesterases. pA₂ values were obtained by analyzing the antagonist-induced rightward shifts of dose–response curves according to the Arunlakshana and Schild equation:⁷⁹ log (DR-1) = s log [B] – log K, where dose ratios DR (EC_50,obs/EC_50,cont) serve to quantify midpoint shifts at each antagonist concentration [B], K is an estimate of antagonist potency, and s is the Schild slope. A pA₂ value for the antagonist was estimated via the pA₂ = pK/s relationship.

In Vitro Activity: Functional Guinea Pig Ileum (GPI) and Mouse Vas Deferens (MVD) Assays

The GPI and MVD bioassays were carried out as described in detail elsewhere. ^80,81 A dose–response curve was determined with [Leu⁵]enkephalin as standard for each ileum and vas preparation, and IC₅₀ values of the compounds being tested were normalized according to a published procedure.⁸²

Parallel Artificial Membrane Permeability Assay (PAMPA)

The PAMPA assay was used to determine passive diffusion capacity across the BBB. The assay was performed as described previously.⁸³ Compounds were dissolved with the System Solution (pION) containing 20% 1-propanol (1 mL) at a concentration of 500 µM. Propranolol was used as a positive control. Stirring magnets were added in the donor compartment, and 195 µL of each compound was added then to donor compartments. Subsequently, 4 µL of a phospholipid mixture (Porcine Brain Polar Lipid Extract from Avantis Polar Lipids; 20 mg/mL in dodecane) was added to the membrane, located at the bottom of the acceptor compartments. This phospholipid mixture comprised phosphatidylcholine (PC; 12.6%), phosphatidylethanolamine (PE; 33.1%), phosphatidylserine (PS; 18.5%), phosphatidylinositol (PI; 4.1%), phosphatidic acid (0.8%), and other compounds (30.9%). Acceptor wells were placed above the donor plate and filled with 200 µL of System Solution (20% 1-propanol). The PAMPA plate (pION) was placed into a GUTBOX (containing wet sponges) for 4 h at room temperature. Agitation was maintained in 25 µm of unstirred water layer (UWL). Compounds were then quantified by RP-HPLC and identified by MALDI-TOF. Effective permeability was calculated as shown in eq 1:

$$P_{\mathrm{e}}=\frac{-218.3}{t}\text{log}(1-\frac{2C_{\mathrm{A}}(t)}{C_{\mathrm{D}}(t_0)})\times {10}^{-6}\text{cm}/\mathrm{s}$$ (1)

In Vitro Human BBB Cell-Based Model Assay

This assay was performed using the model developed by Cecchelli and co-workers. ⁶⁸ Endothelial cells and pericytes were defrosted in gelatin-coated Petri dishes (Corning). Pericytes and endothelial cells were cultured in DMEM pH 6.8 or in supplemented endothelial cell growth medium (Sciencells), respectively. After 48 h, pericytes (50000 cells/well) and endothelial cells (80000 cells/well) were seeded in gelatin-coated 12-well plates or in Matrigel-coated 12-well Transwell inserts (Corning), respectively. Medium was changed every 2–3 days, and assays were performed 7–8 days after seeding. Lucifer Yellow (50 µM) was used as internal control (P_app <15 × 10⁻⁶ cm/s), and LY fluorescence was measured in a 96-well plate with a Fluoroskan Ascent microplate fluorometer (Thermo Fisher Scientific).

Compounds were dissolved in Ringer Hepes at the final concentration of 200 µM. Then, 500 µL of the compound and 1500 µL of Ringer HEPES alone were introduced in the apical or in the basolateral compartments, respectively. The plates were set on at 37 °C for 2 h. The solutions from both compartments were then recovered and quantified by UPLC and identified by UPLC-MS and MALDI-TOF. Apparent permeability was calculated using eq 2:

$$P_{\text{app}}=\frac{\mathrm{d}{Q}_{\mathrm{A}}(t)}{\mathrm{d}t}\frac{1}{A}\frac{1}{Q_{\mathrm{D}}(t_0)}{V}_{\mathrm{D}}$$ (2)

Behavioral Study in Mice

Animals

The male mice strain C57BL6 (20–25g), obtained from the Animal House Mossakowski Medical Research Centre, Polish Academy of Sciences, were housed 4–5 mice per cage in cages lined with sawdust. Animals was maintained under standard room temperature 22 ± 2 °C, humidity 50 ± 5%, 12/12 h light–dark cycle, with unlimited access to food and water. All experimental procedures were approved by The Local Committee for Ethics in Animal Experiments (permission number: 14/2013).

Acute Pain in Mice (Tail-Flick Test)

Analgesic activity of the compounds was measured on male mice strain C57BL6 in the tail-flick test. These experiments were conducted using a plantar test and Tail Flick Analgesia Meter apparatus (IITC Life Science Inc., USA), where a light beam is used as a thermal nociceptive stimulus. First, the mouse was restrained and either tested compound or saline (used as a control group) was administered by intravenous injection to the tail vain. During the measurements, the radiant heat was applied to the ventral side of animals’ tail and the time latency for withdrawal or shaking of the tail was recorded. Measurements were performed before drugs administration (baseline) and at following time-points of 5, 15, 30, 60, and 120 min after injection. A cutoff of 10 s was set in the tail flick test in order to avoid tissue damage. Responses are expressed as a percentage of the maximum possible effect (%MPE) which is calculated as [(T₁ - T₀)/(T₂ - T₀)] × 100, where T₀ and T₁ are the tail-flick latencies before and after drug injection respectively and T₂ is the cutoff time. Drug-treated and saline-treated (control) groups consisted of 8–10 mice each. The significance is given in the graphs (no significance P > 0.05, significant * P < 0.05, ** P < 0.01, *** P < 0.001).

Neuropathic Pain

Animals

Male Albino Swiss CD-1 IGS mice (30–35 g) obtained from Charles River Breeding Laboratories, Germany, were housed six per cage in cages lined with sawdust under a standard 12/12 h light/dark cycle (lights on at 6:00 am) with food and water available ad libitum. All experiments were performed according to the recommendations of IASP, the NIH Guide for Care and Use of Laboratory Animals, and were approved by the local Bioethics Committee (Krakow, Poland).

Chronic Constriction Injury in Mice

Chronic constriction injury (CCI) model was performed according to Bennet and Xie⁸⁴ and was modified for mice by Mika (2007).⁸⁵ Mice were preanaesthetized with chloral hydrate (500 mg/kg ip) to prolong and deepen anesthesia state. The surgical procedure was performed under isoflurane anesthesia. An incision was made below the mouse’s right hipbone, and the sciatic nerve was exposed. Three ligatures with 4/0 silk thread were made around the nerve distal to the sciatic notch with 1 mm spacing until a brief twitch in the respective hind limb was observed. After 7 days of recovery, mice were tested to assess neuropathy development. All CCI mice developed allodynia and hyperalgesia. Main experiments were conducted on days 7–16 after CCI surgical procedure.

Type 1 Diabetes Model in Mice

The type 1 diabetes model was obtained by a single intraperitoneal (ip) injection of streptozocin (STZ; 200 mg/kg; Sigma-Aldrich, USA) prepared in water for injection (Polpharma, Poland). The glucose concentration was measured in blood collected from the tail vein. Mice with serum glucose levels above 300 mg/dL were considered diabetic. Mice that did not develop diabetes, or allodynia and hyperalgesia, were excluded from the study. The behavioral tests were conducted on day 10 after STZ injection.

Drugs Administration

Drugs were dissolved in water for injection (Polpharma, Poland) and administered intrathecally (it) in 5 µL dose volume through lumbar puncture between L5 and L6 to non-anesthetized mice according to upgraded Hylden and Wilcox⁸⁶ model described with modifications by Fairbanks.⁸⁷ The it injections were performed with disposable 30 gauge ½ in. needles (Becton Dickinson and Company, Rutherford, NJ, U.S.A.) matched to a 25 µL syringe (Hamilton, Reno, NV, USA). The von Frey and cold plate tests were performed at 30, 90, and 180 min after drug administration, while Rota Rod test was performed at 15 and 120 min.

Behavioral Tests in Mice

Von Frey Test

Mechanical sensitivity to non-noxious stimuli was measured by applying a set of calibrated nylon monofilaments (0.6–6 g; Stoelting) in serial increments on a tested hind paw plantar surface until a behavioral response was observed. Response considered as pain behavior included paw withdrawal, shaking, and licking. In the von Frey test, results are expressed as pressure [g] applied with a calibrated plastic filament to the midplantar surface of the mouse’s injured hind paw (cutoff: 6 g), which elicited a foot withdrawal response.

Cold Plate Test

Sensitivity to noxious thermal stimuli was assessed with usage of Cold/Hot Plate Analgesia Meter, Columbus Instruments. The temperature of the plate was kept at 2 °C, and the cutoff latency was 30 s. The mice were placed on the cold plate, and the time until the hind paw was lifted is recorded. The lifting of the injured hind paw in the CCI model, or reaction of one of the hind paws in the STZ model were considered as a reaction to noxious cold stimulus.

Rota Rod Test

To assess motor coordination, Rota Rod 47600 for mice was used (Ugo Basile, Italy). Mice were placed on a horizontal rod which is rotating at accelerating speed, starting with 2 rpm and reaching 40 rpm within 300 s (cutoff). The experiment was conducted after three training sessions with minimum 15 min pause between each session. The time was recorded until a mouse fell off the rod.

Respiratory Depression

All the experimental protocols were carried out within the ethical guidelines established by the European Union Legislation. Protocols were approved by the local ethics committee for animal experimentation of Paris–Descartes University.

Animals

Male Sprague–Dawley rats (Janvier, France) weighing between 250 and 350 g at the time of experimentation were used. Animals were housed for 7 days before experimentation in an environment maintained at 21 ± 0.5 °C with controlled humidity and light-dark cycle (lights on between 08:00 and 20:00). Food and tap water were provided ad libitum.

Drugs

Morphine sulfate was purchased from Francopia, France, diluted in 0.9% NaCl to obtain a solution of 3.5 mg/mL.

Jugular Catheterization

One week before the study, the rat jugular vein was catheterized using 20 cm silastic tubing with external and internal diameters of 0.94 and 0.51 mm, respectively (Dow Corning Co., Midland, MI). Catheter was tunneled subcutaneously and fixed at the back of the neck. Heparinized saline was injected into the catheter to avoid thrombosis and catheter obstruction. Rats were then returned to their individual cages for a recovery period of 7 days, allowing complete anesthesia washout. On the day of experimentation, rats were placed in horizontal Plexiglas cylinders (6.5 cm internal diameter, up to 20 cm adjustable length) (Harvard Apparatus, Inc., Holliston, MA, USA), modified by the addition of several holes at the cephalic end to avoid CO₂ rebreathing. Before drug administration, the catheter was exteriorized, purged, and its permeability checked.

Respiratory Effects Measurement Using the Whole Body Plethysmography

Four days before the study, temperature transmitters (TA-F10, DSI, Chatillon, France) were implanted in the peritoneal cavity. Ventilatory parameters were recorded in a whole body plethysmograph by the barometric method described and validated in the rat.⁸⁸ The first measurement was performed after a 30 min period of accommodation, while the animal was quiet and not in deep or rapid eye movement sleep, as roughly estimated from their behavior, response to noise, and pattern of breathing. Then, the animal was gently removed from the chamber for iv injection of the tested drug at T₀ and replaced in the chamber for the remaining measurements. Ventilation was recorded at −30, −15, −5, 5, 10, 15, 30, 45, 60, 90, and 120 min, each record lasting about 60 s. The following parameters were measured: the tidal volume (V_T), the inspiratory time (T_I), the expiratory time (T_E), and the respiratory cycle duration (T_TOT = T_I + T_E). Additional parameters were calculated: the respiratory frequency (f) and the minute volume (V_E = V_T × f). T₀ values were the mean of the three baseline measurements.

Study Design

Rats were randomized into three groups receiving iv 0.9% NaCl, morphine (5 mg/kg) or 13a (0.5 mg/kg) to investigate antinociception (N = 7) and respiratory effects (N = 6).

Data Analysis

Results are expressed as mean ± SEM. To permit simultaneous analysis of the effect of time and treatments on %MPE and plethysmography parameters, for each animal and for each studied parameter, the area under the curve (AUC) from T₀ to the completion of the measurement (240 and 120 min, respectively) was calculated using the trapezoid method. For each parameter, we compared the AUCs using Kruskal–Wallis tests for comparisons between the three groups. All tests were performed using Prism version 6.0 (GraphPad Software, Inc., San Diego, CA), and P-values of less than 0.05 were considered as significant.

ABBREVIATIONS USED

%MPE

percent of maximal possible effect

Aba

4-amino-2-benzazepinone

cAMP

3’,5’-cyclic adenosine monophosphate

CCI

chronic constriction injury to the sciatic nerve

DIC

N,N’-dicyclohexylcarbodiimide

DMF

N,N-dimethylformamide

DML

designed multiple ligand

Dmt

2’,6’-dimethyl-l-tyrosine

DOP

δ-opioid receptor

EBAW

ethyl acetate/n-butanol, acetic acid/water

FSK

forskolin

GPI

guinea pig ileum

HOBt

1-hydroxybenzotriazole; i

it

intrathecal

iv

intravenous

KOP

κ-opioid receptor

MOP

µ-opioid receptor

MVD

mouse vas deferens

NMM

N-methylmorpholine

NOP

nociceptin receptor

PAMPA

parallel artificial membrane permeability assay

Pbf

2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl

PC

phosphatidylcholine

PE

phosphatidylethanolamine

pF-Phe

para-fluoro-phenylalanine

PI

phosphatidylinositol

RP-HPLC

reversed phase high pressure liquid chromatography

PS

phosphatidylserine

SD

standard deviation

SPPS

solid phase peptide synthesis

STZ

streptozotocin

HCTU

O-(6-chlorobenzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate

TES

triethylsilane