Effective and adequate management of pain, particularly chronic pain, is still an area of unmet medical need at the beginning of the third millennium. Opioids are the gold standard for the treatment of moderate to severe acute and chronic pain [1]. However, strong opioid analgesics, such as morphine, oxycodone and fentanyl, are not effective for pain relief in all patients, nor are they well-tolerated by all patients, because of an array of severe side effects, including respiratory depression, constipation, sedation, nausea and dizziness [2]. With prolonged use, opioid safety is dramatically reduced because of analgesic tolerance, and risks of physical dependence and addiction, promoting the development of opioid use disorders and overdose deaths [2]. Moreover, the misuse of prescription opioids (i.e., oxycodone, hydrocodone and fentanyl) [3], as well as over-the-counter opioids (i.e., codeine, hydrocodeine and loperamide) [4], has led to the current opioid epidemic, particularly in North America and Europe. In the USA, the opioid-involved overdose deaths had increased to 80,411 in 2021 [5]. The recent explosion in fatalities resulting from overdose of prescription and synthetic opioids, especially fentanyl and its various analogs ^[3][5][3,5], has dramatically increased the need for safer analgesics.

Opioid receptors represent the most important players in pain modulation and are the molecular targets of clinically used opioids. There are three main opioid receptor types, mu (MOR), delta (DOR) and kappa (KOR), and the non-classical receptor, nociceptin/orphanin FQ (NOP) receptor [1]. All opioid receptor types belong to the family of G protein-coupled receptors (GPCRs) with seven transmembrane domains, and are expressed throughout the central and peripheral nervous systems (CNS and PNS, respectively), and in various non-neuronal tissues ^[1][6][7][1,6,7]. Because of their therapeutic relevance, opioid receptors are among the few GPCRs determined in different activation states, providing important information on the type-specific binding characteristics of ligands [8]. Although opioid receptors contribute to pain inhibition, the MOR is recognized as the dominant type for its pain-relieving effects ^[2][9][10][2,9,10]. The major drawback of targeting the MOR for clinical analgesia is that it is also responsible for the undesirable side effects. Most of the detrimental side effects, including respiratory depression, sedation, analgesic tolerance, reward and dependence, are mediated by the MOR in the CNS, whereas constipation is mainly mediated by the MOR in the intestinal myenteric plexus ^[2][11][12][2,11,12].

Among therapeutically valuable opioids, morphinans are of the utmost clinical importance as analgesic drugs because of their agonistic actions to the MOR. They include powerful pain relieving agents, such as naturally occurring alkaloids (e.g., morphine and codeine), semisynthetic analogues (e.g., hydrocodone, hydromorphone, oxycodone, oxymorphone and buprenorphine), and synthetic derivatives (e.g., levorphanol) (Figure 1) ^{[9][13][14][15]}[9,13,14,15]. Morphine and structurally related MOR agonists also share similar side effects, including addictive liability. Because of their outstanding medicinal relevance in combination with an attractive chemical scaffold, morphinan alkaloids represent attractive synthetic targets. Therefore, diverse research approaches toward morphine and its congeners have been devised for mitigating their deleterious side effects and limiting abuse and misuse (for reviews, see ^{[9][13][14][15][16][17][18][19][20]}[9,13,14,15,16,17,18,19,20]).

Experimental and clinical studies provide substantial evidence that opioid analgesia is not exclusively mediated via the central opioid receptors (for reviews, see ^{[21][22][23][24]}[21,22,23,24]). Pharmacological, neuroanatomical, molecular and electrophysiological studies have shown that peripheral opioid receptors are expressed on the peripheral terminals of sensory neurons, where they can modulate both afferent and efferent neuronal functions, resulting in potent and clinically measurable analgesia. Early clinical studies using intra-articular morphine administration in conjunction with arthroscopy in the knee joint supported the notion that the activation of peripheral opioid receptors induced pain relief by a peripheral mechanism and did so without side effects ^[25][26][25,26]. These findings have led to new research directions aiming on targeting the peripheral opioid receptors for superior pain management.

2. Peripheralization Strategies Applied to Morphinans

Different chemical strategies have been developed to limit the ability of opioids to cross the BBB, including (a) chemical modifications to the morphinan skeleton to increase hydrophilicity of known and new opioids, and (b) nanocarrier-based approaches to selectively deliver opioids, such as morphine to the peripheral tissue. The following sections discuss significant representatives, including design strategies, synthetical procedures, pharmacology and structure–activity relationships (SAR).

2.1. Quaternization of the Morphinan Nitrogen

The first effort to minimize the CNS effects of opioids while retaining their actions in peripheral tissue was the quaternization of the nitrogen in the clinically used morphine, oxymorphone, nalorphine, naloxone and naltrexone (Figure 23) ^[27][28][27,28]. Peripheral selectivity of the quaternary derivative of morphine, N-methylmorphine, was described over 50 years ago [29]. The systemic intravenous (i.v.) administration of N-methylmorphine caused the inhibition of gastrointestinal transit because of its action on the opioid receptors in the gut. In the hot-plate test, centrally mediated antinociception was produced by 15 mg/kg morphine in mice after intraperitoneal (i.p.) administration but not by N-methylmorphine at the same dose [29]. Furthermore, N-methylmorphine proved to be ineffective in the hot-plate test even in a dose 100 mg/kg [30]. In an acetic-acid-induced writhing assay, N-methylmorphine produced antinociceptive effects in mice after i.p. administration in a dose of 45 mg/kg, being 30-fold less potent than morphine [30]. N-methylmorphine was also shown to selectively inhibited phase II in the formalin test following systemic i.p. administration [31]. The antinociceptive effect of N-methylmorphine after central intracerebroventricular (i.c.v.) administration was antagonized by systemically applied naloxone but not by peripheral antagonist N-methylnaloxone, showing the peripheral site of action of N-methylmorphine [31]. It is important to note that quaternization of nitrogen on morphine-based structure derivatives has been reported to have negative impact on both the affinity to the opioid receptor and the agonist activity of generated analogues ^[28][32][28,32]. Quaternization also trends to reduce potency in vivo. Therefore, alternative strategies to limit BBB penetration have been pursued.

2.2. Introduction of Hydrophilic Substituents at Position 6

Polar or ionizable substitutions are able to increase polarity and inhibit the crossing of the BBB. Therefore, opioids with hydrophilic groups attached to the C-6 position of the morphinan skeleton were designed. The first examples of morphinans having ionizable residues at position 6 were reported more than 30 years ago. They were synthesized from β-oxymorphamine [33], β-naltrexamine [33] and β-funaltrexamine [34]. Such compounds with zwitterionic moieties showed significantly reduced access to the CNS without substantially decreased opioid receptor in vitro and in vivo activity ^[33][34][33,34]. Noteworthy are the 6-amide derivatives of β-oxymorphamine (a–e, Figure 34) reported as the first peripherally selective opioid agonists and effective antinociceptives [33]. All compounds have C-6 moieties that are ionized at the pH of the gut or at physiologic pH, accounting for a more restricted capability to enter the CNS than the unionized molecules. Compounds a, c and d were synthesized from β-oxymorphamine with the appropriate anhydride [33]. The fumaramic acid b was prepared by coupling the half-ester of fumaric acid with β-oxymorphamine and then subjecting the fumaramate esters to hydrolysis. The aspartyl derivative e was obtained through coupling BocAsp γ-benzyl ester with β-oxymorphamine, followed by deprotection with acid to remove the Boc group and hydrogenolysis of the benzyl function [33]. As regards biological activities, the ß-oxymorphamine derivatives a–e were all full agonists in the guinea pig ileum (GPI) bioassay with potencies that were 1.5- to 6-fold higher than the potency of morphine (Table 1) [33]. In a mouse model of acute thermal nociception, the tail-flick assay, all compounds possessed potent antinociceptive activity when administered by the i.c.v route (Table 1). They also were active in inducing antinociception when given systemically by i.v. administration to mice. When compared on a body weight basis, the i.v. ED₅₀ doses were about 1000-fold higher than the i.c.v. ED₅₀ values. Derivatives a and c were also active when given orally (p.o.) (Table 1) [33]. The attachment of polar groups, particularly zwitterionic moieties, at the C-6 position of the morphinan structure is effective in excluding such ligands from the CNS, thereby affording peripheral selectivity. Other more recent examples of morphinans with ionizable groups at position 6 emerging as peripheral opioid antinociceptives with restricted penetration into the CNS are described in the following sections.

2.2.1. 6-Amino-acid-substituted 14-Alkoxymorphinans

The first synthetic efforts directed towards the development of ionizable molecules in the class of 14-alkoxymorphinans as peripherally acting opioid analgesics started with the series of six 6-amino acids, i.e., Gly-, L-Ala- and L-Phe- substituted derivatives; 2a/b (HS-730/HS-731); 3a/b (HS-935/HS-936); and 4a/b (HS-937/HS-938), respectively, of the highly potent and centrally acting MOR agonist 14-O-methyloxymorphone (14-OMO, 1) (SchemeScheme 1) 1) ^[35][35]. A novel synthetic procedure for the synthesis of 6-amino-acid-substituted derivatives in the morphinan series was used. The tert-butyl ester derivatives 2aa/bb, 3aa/bb, and 4aa/bb were prepared from 14-OMO (1) by reductive amination with the respective tert-butyl ester hydrochlorides and sodium cyanoborohydride in ethanol. After separating the diastereoisomers by column chromatography, esters 2aa/bb, 3aa/bb, and 4aa/bb were treated with tetrafluoroboric acid in dichloromethane to afford 6-Gly (2a and 2b), 6-Ala (3a and 3b) and 6-Phe (4a and 4b) substituted derivatives, respectively (SchemeScheme 1) 1) ^[35][35]. The further targeted derivatization of 14-OMO (1) through introduction of other amino acid residues of the L- and/or D-series at position 6, including natural amino acids, i.e., Ser, Val, Lys, Tyr, Trp, Asn, Gln, Asp and Glu (5a–13b, Scheme 2), and unnatural amino acids, i.e., D-Ala, D-Val, D-Phe, L-Chg (L-cyclohexylglycine), L-Abu (L-2-aminobutyric acid), β-Ala and GABA (γ-aminobutyric acid) (14a–20b, Scheme 2) [36]. Additionally, three zwitterionic molecules with a dipeptide substitution at position 6, i.e., L-Val-L-Tyr and Gly-Gly in 14-OMO (21a–22a, Scheme 3) were synthesized [36]. The reductive amination of 14-OMO (1) was performed using amino acid tert-butyl ester hydrochlorides or dipeptide benzyl ester hydrochlorides, and NaBH₃CN in CH₃OH. Medium-pressure liquid chromatography (MPLC) was used to separate the diastereoisomers, providing ester derivatives 5aa–22aa (SchemeScheme 2 2 and Scheme 3). Typically, the ratio of 6β-amino to 6α amino epimers was between 4:1 and 2:1. The 6-amino-acid (5a–20b)-substituted derivatives were obtained through ester cleavage of the tert-butyl derivatives in dioxane/HCl (Scheme 2). Catalytic hydrogenation of the benzyl esters 21aa/bb and 22aa in CH₃OH using 10% Pd/C catalyst provided the 6-dipeptide-substituted 21a/b and 22a, respectively (Scheme 3) [36]. Synthetic work also targeted the combination of 6-amino amino (i.e., Gly) and 14-arylalkoxy (e.g., phenylpropoxy) substitutions in N-methyl-morphinans (Scheme 4) [37]. The reductive amination of the 14-phenylpropoxyoxymorphone (POMO, 23) was performed with Gly tert-butyl ester hydrochloride and NaCNBH₃ in DMF/MeOH at room temperature. The diastereoisomers were separated by column chromatography to obtain tert-butyl esters 24aa and 24bb. Ester cleavage in dioxane/HCl generated the amino acids 24a and 24b (Scheme 4) [37]. SAR studies on the series of amino acid and dipeptide substitution at position 6 in 14-OMO (1) as zwitterionic molecules explored their binding and activation of the opioid receptors and antinociceptive properties (Table 2, Table 3 and Table 4). The 6-amino acid groups included natural amino acids (i.e., Gly, Ala, Phe, Ser, Val, Lys, Tyr, Trp, Asn, Gln, Asp and Glu), unnatural amino acids (i.e., D-Ala, D-Val, D-Phe, L-Chg, L-Abu, β-Ala and GABA) (2a–20b) (Scheme 1, Scheme 2 and Scheme 4) ^{[35][36][37][38]}[35,36,37,38], and 6-dipeptide substitution (i.e., L-Val-L-Tyr and Gly-Gly) (22a/b and 23a) (Scheme 3) [36]. In vitro receptor binding (radioligand binding assays with membranes from rodent brain and CHO cells expressing the human opioid receptors) and functional assays (mouse vas deferens (MVD) bioassay and [³⁵S]GTPγS binding assay with membranes from CHO cells expressing the human opioid receptors) established the potent MOR/DOR agonist profile and reduced the binding and activation of the KOR for most compounds (Table 2). The replacement of the 14-methoxy group in N-methyl, 6-Gly substituted morphinans 2a (HS-730) and 2b (HS-731) with a 14-phenylpropoxy group (compounds 24a and 24b, respectively) resulted in a considerable increase in binding affinities to all three opioid receptor types, MOR, DOR and KOR, in rodent brain membranes (Table 2 and Table 3) [37]. Compared to the nonselective 14-phenylpropoxy-substituted POMO (23) [39], the 6-Gly analogues 24a and 24b showed a comparable binding profile to the opioid receptors, acting as mixed MOR/DOR/KOR ligands (Table 2) [37]. ^[41][42][43][36,37,40,41,42,43]. In the radiant heat tail-flick test in rats, they were up to 200-fold more potent than morphine after s.c. administration, and had similar potencies to fentanyl, with markedly longer duration of action (Table 4) ^[37][40][37,40]. A similar profile was reported following central i.c.v. administration of the 6-amino acid conjugates of 14-OMO (1), i.e., 2a/b (HS-730/HS-731), 3a/b (HS-935/HS-936) and 4a/b (HS-937/HS-938) [40]. All compounds, except 6β-L-Phe substituted 4b (HS-938), were more effective in producing an antinociceptive response than morphine, while they showed generally lower potencies compared to 14-OMO (1) in the acetic-acid-induced writhing assay after s.c. administration to mice (Table 4) ^[36][42][36,42]. Subcutaneous and local intraplantar (i.pl.) administration of 6-Gly-substituted 2a/b (HS-730/HS-731) and 6-L-Phe conjugates 4a/b (HS-937/HS-938) also produced antihyperalgesic effects in the formalin test in rats, with increased potencies compared to morphine (Table 4) ^[40][44][40,44]. In rats with neuropathic pain (i.e., sciatic nerve ligation), 2a/b and 4a/b compounds were equipotent or somewhat less active in producing antihyperalgesic and antiallodynic effects than morphine after i.pl. injection [44]. In carrageenan-induced inflammatory pain in rats, significant and long-lasting antihyperalgesic actions (up to 4 h) were demonstrated for the 6α- and 6β-Gly-substituted 14-phenylpropoxymorphinans 24a and 24b, respectively [37]. Among the developed 6-amino-acid-substituted N-methylmorphinans, the 6β-Gly-substituted analogue 3b (HS-731) was more extensively investigated for its antinociceptive effects in a multitude of diverse pain models, as summarized in Table 5. While it was shown to be very effective as an antinociceptive agent in rodents after systemic parenteral (s.c. and i.p.), central (i.c.v.) and local (i.pl.) application, its significant and prolonged duration of antinociceptive action (up to 4 h) after oral administration to rats with carrageenan-induced inflammatory pain was notable [41]. Furthermore, a recent study described the lack of binding to the human NOP receptor of 3b (HS-731) [45].