µ Agonists with Potent Antinociceptive Activity in Mice

Transcription

1 JPET Fast This article Forward. has not Published been copyedited on and January formatted. 12, The final 2004 version as DOI: /jpet may differ from this version. Novel 2,6 -Dimethyl-L-Tyrosine-Containing Pyrazinone Opioid Mimetic µ Agonists with Potent Antinociceptive Activity in Mice YUNDEN JINSMAA, YOSHIO OKADA, YUKO TSUDA, KIMITAKA SHIOTANI, YUSUKE SASAKI, AKIHIRO AMBO, SHARON D. BRYANT and LAWRENCE H. LAZARUS National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina (Y.J., S.D.B., L.H.L.); Kobe Gakuin University, Nishi-ku, Kobe, Japan (Y.O., Y.T., K.S.); and Tohoku Pharmaceutical University, Aoba-ku, Sendai, Japan (Y.S., A.A.) 1 Copyright 2004 by the American Society for Pharmacology and Experimental Therapeutics.

2 a) Running title: Pyrazinone Ring-Containing Potent Opioid Mimetics b) Address correspondence to: Yunden Jinsmaa, Ph.D. Medicinal Chemistry Group, LCBRA, NIEHS, P.O. Box 12233, Research Triangle Park, NC 27709, USA. Tel: (919) Fax: (919) c) Text pages-45 pages Table-1, Figures-9, References- 40, Abstract- 236 words, Introduction- 470 words, Discussion-1057 words d) Abbreviations: TF, tail-flick; HP, hot-plate; TFL, tail-flick latency; HPL, hot-plate latency; NTI, naltrindole hydrochloride; Dmt, 2,6 -dimethyl-l-tyrosine; i.c.v., intracerebroventricular; s.c., subcutaneous; p.o., per oral; MED, Minimum Effective Dose; NAZ, naloxonazine; β-fna, β-funaltrexamine; GPI, guinea pig ileum; MVD, mouse vas deferens; AUC, area 2

3 under the curve; Boc, tert-butyloxycarbonyl; Orn, ornithine; Lys, lysine. e) Behavioral Pharmacology 3

4 ABSTRACT Novel bioactive opioid mimetic agonists containing 2,6 -dimethyl-l-tyrosine (Dmt) and a pyrazinone ring interact with µ- and δ-opioid receptors. Compound 1 [3-(4 -Dmt-aminobutyl)-6-(3 -Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone] exhibited high µ-opioid receptor affinity and selectivity (K i µ = nm and K i δ/ K i µ = 1,519, respectively), and agonist activity on guinea pig ileum (GPI, IC 50 = 1.7 nm) with weaker δ bioactivity on mouse vas deferens (MVD, IC 50 = 25.8 nm). Other compounds (2-4) had µ-opioid receptor affinities and selectivities 2- to 5-fold and 4- to 7-fold less than 1, respectively. Intracerebroventricular (i.c.v.) administration of 1 in mice exhibited potent naloxone reversible antinociception (65- to 71-times greater than morphine) in both tail-flick (TF) and hot-plate (HP) tests. Distinct opioid antagonists had differential effects on antinociception: naltrindole (δ-antagonist) partially blocked antinociception in the TF, but was ineffective in the HP test, while β-funaltrexamine (irreversible antagonist, µ 1 /µ 2 -subtypes) but not naloxonazine (µ 1 -subtype) inhibited TF test antinociception, yet both blocked antinociception in the HP test. Our data indicated that 1 acted through µ- and δ-opioid receptors to produce spinal antinociception, although primarily through the µ 2 -receptor subtype; however, the µ 1 -receptor subtype dominates supraspinally. 4

5 Subcutaneous and oral administration indicated that 1 crossed gastrointestinal and blood-brain barriers to produce CNS-mediated antinociception. Furthermore, daily s.c. dosing of mice with 1 for week developed tolerance in a similar manner to that of morphine in TF and HP tests implicating that 1 also acts through a similar mechanism analogous to morphine at µ-opioid receptors. 5

6 Since the discovery of the endogenous opioid pentapeptides, [Met 5 ]- and [Leu 5 ]-enkephalin (Hughes et al., 1975), many peptide and nonpeptide compounds were synthesized in a search for greater receptor selectivity, stability and potent bioactivity. Although all known endogenous and exogenous µ-, δ- and κ-opioid receptor peptides have distinct structural diversity, they share a common message domain, which is considered important for recognition by opioid receptors. This message region consists of a tyrosine residue, an N-terminal amino group and a spacer (D-Ala, D-Met, Pro or Gly-Gly) between the tyrosine and second aromatic ring, usually Phe, or Trp in the endomorphins (Zadina et al., 1997). The address domain, which is responsible for the biological response, contains the second aromatic ring and residues C-terminal thereafter. Message and address domains of opioid peptides represent distinct starting points for the design and development of novel opioid mimetics. For example, studies on opioid peptides showed that the substitution of the N-terminal tyrosine by 2,6 -dimethyl-l-tyrosine (Dmt) dramatically increased receptor affinity in numerous peptides and enhanced their antinociceptive effect (Chandrakumar et al., 1992; Hansen et al., 1992; Pitzele et al., 1994; Guerrini et al., 1996; Sasaki et al., 1999; Schiller et al., 2000; Bryant et al., 2003). Moreover, Dmt played a key role in the formation of the Dmt-Tic pharmacophore family 6

7 of potent δ-opioid receptor antagonists (Salvadori et al., 1997; 1999; Bryant et al., 1998) which were transformed into potent δ-opioid receptor agonists by subtle changes in the C-terminal link to a third aromatic nucleus (Balboni et al., 2002). Recently we reported that symmetric opioidmimetic substances, which contain two identical Dmt residues separated by simple unbranched alkyl chain, are able to serve as both the message and address domains for binding to µ-opioid receptors (Okada et al., 2003). The cyclization of dipeptidyl chloromethyl ketones yielded pyrazinone derivatives, which can be easily inserted into an opioid peptide sequence (Okada et al., 1998; 1999) leading to our synthesis of a new series of opioidmimetic substances containing Dmt. As we report herein, these Dmt-containing pyrazinone opioid mimetic compounds exhibit high affinity and high selectivity for the µ-opioid receptor in rat brain membrane preparations. Furthermore, they were biologically potent toward the µ-opioid receptor in functional guinea pig ileum bioassays and general had considerably weaker to insignificant activity for the δ-opioid receptor based on the mouse vas deferens bioassay. In this study, we focused on an examination of their in vivo activity using TF and HP tests as the analgesic endpoints in mice for their centrally mediated activity (intracerebroventricular administration) and systemic injections (subcutaneous and per oral) in comparison to the 7

8 effects of morphine. In order to determine the specificity of those interactions, we utilized a series of specific opiate antagonists, β-funaltrexamine in order to differentiate between µ 1 - and µ 2 -receptor subtypes, and the µ 1 -opioid receptor antagonist naloxonazine, in addition to the δ-opiate antagonist naltrindole. 8

9 Materials and Methods Animals. Male Swiss-Webster mice (20-25 g, Taconic, Germantown, NY) were used. Animals were housed in plastic cages and maintained on a 12-h light/dark cycle with free access to food and water. Guinea pigs were housed for several days before use. All experiments with animals were carried out according to protocols approved by and on file with the NIEHS Animal Care and Use Committee. Opioid Receptor Competitive Binding Assays. The receptor binding affinities were determined under equilibrium conditions (2.5 h at 22 o C) using rat brain membranes. The synaptosomal preparation (P2) was preincubated to remove endogenous opioid ligands in a solution containing 0.1 M NaCl, 0.4 mm GDP, 50 mm HEPES, ph 7.5, and 50 µg/ml soybean trypsin inhibitor for 60 min at room temperature (22 o C). The preparation was exhaustively washed with ice-cold buffer (50 mm HEPES, ph 7.5, containing protease inhibitor); after the final wash, the synaptosomes were stored in buffer with protease inhibitor and 20% (v/v) glycerol and stored at -80 o C. δ- and µ-opioid receptors were radiolabeled with [ 3 H]DPDPE (Perkin Elmer, Boston, MA; 34.0 Ci/mmol) and [ 3 H]DAMGO (Amersham, Buckinghamshire, UK; 50.0 Ci/mmol), respectively, as described in detail previously (Salvadori et al., 1997; Okada et al., 2003). Excess 9

10 unlabeled peptide (2 µm) established the level of non-specific binding. Radiolabeled membranes were rapidly filtered onto Whatman GF/C glass fibre filters soaked in 0.1% polyethenimine to enhance the signal to noise ratio, washed 3 times with 2 ml ice cold buffer (50 mm Tris HCl, ph 7.5 containing 0.1% BSA), dried at 75 o C for an hour and radioactivity determined using CytoScint (ICN). All analogues were analyzed in duplicate using 5-8 dosages and 3-5 independent repetitions; different synaptosomal preparations were frequently used to ensure statistical significance (n values are listed in Table 1 in parenthesis and results are mean ± S.E.M.) The affinity constants (K i ) were calculated according to Cheng and Prusoff. Functional Bioactivity in Isolated Tissue Preparations. Preparations of the myenteric plexus-longitudinal muscle (2-3 cm segments) were obtained from the small intestine of guinea pigs (GPI) for the study of µ-opioid activity, while a single mouse vas deferens (MVD) was used as a source of biological δ-opioid receptors. Both tissues, suspended in a balanced salt solution, were used for field stimulation with bipolar rectangular pulses of supramaximal voltage. Agonists were tested for their inhibition of the electrically evoked twitch and the results are expressed as the IC 50 values obtained from dose-response curves and represent the mean ± SE of 5-6 separate assays. 10

11 [D-Ala 2 ]Deltorphin I and dermorphin were used as internal δ- and µ-opioid peptide standards for MVD and GPI, respectively (Okada et al., 2003). Drugs. Morphine and naloxone HCl were obtained from Sigma (Louis, MO, USA), naltrindole hydrochloride (NTI), naloxonazine dihydrochloride (NAZ) and β-funaltrexamine hydrochloride (β-fna) were purchased from Tocris (Ellsville, MO, USA). The opiate antagonists were injected s.c.: naloxone (2 mg/kg) and NTI (3 mg/kg) were administered 30 min before the test compounds while with NAZ (35 mg/kg) and β-fna (40 mg/kg) the mice were treated 24 h prior to testing (Paul et al., 1989). Under this experimental paradigm, NAZ antagonizes µ 1 - (Ling et al., 1986) and β-fna affects both µ 1 - and µ 2 -opioid receptor mediated effects. The dose of naloxone and NTI were selected by their ability completely block the effect of morphine (0.5 µg/mouse, i.c.v.) and deltorphin B (8 µg/mouse, i.c.v.), respectively. Opioid Mimetic Substances. Briefly, the optical purity (> 98%) of 2,6 -dimethyl-l-tyrosine (Dmt) prepared as previously described (Dygos et al., 1992) was determined by HPLC using a chiral column [CROWNPAC CR(+)] and reaction with D-amino acid and L-amino acid oxidases followed by amino acid analysis. Boc-Dmt-OH was prepared as described (Okada et al., 1999). The 3,6-bis-(Tyr-aminoalkyl)- 11

12 5-methyl-2(1H)-pyrazinones were prepared by published procedures (Okada et al., 1999). Boc-R(Z)-OH (R: Orn or Lys) coupled with H-R (Z)-CH 2 Cl (R : Orn or Lys) by a mixed anhydride procedure to yield Boc-R(Z)-R (Z)-CH 2 Cl. The Boc groups were removed by hydrogen chloride in dioxane and the amine HCl in MeOH or THF was refluxed for 1 h to give the protected pyrazinone derivatives, which were converted to the corresponding amines (Okada et al., 1999) by catalytic hydrogenation or by HBr/AcOH. The resulting amines were then coupled with Boc-Dmt-OH by using PyBop, [benzotriazolyloxytris(pyrrolidino)-phosphonium hexafluorophosphate] reagent to produce the protected compounds 1-4. They were treated with TFA to give crude 1-4. The final products were purified by semi-preparative reverse-phase HPLC and each compound exhibited a single peak on analytical HPLC (Okada et al., 2003). Analyses by MALDI-TOF mass spectrometry (MS), 1 H and 13 C NMR and by HPLC revealed that they were the desired compounds with greater than 98% purity. These compounds were dissolved in physiological saline adjusted to ph 7 for i.c.v. or s.c. injections and dissolved in water for oral administration. Intracerebroventricular Injection. I.c.v. injection was performed with Hamilton microsyringe fitted with disposable 26-gauge needle; it was inserted mm deep as 12

13 described by Laursen and Belknap (Laursen and Belknap, 1986). Briefly, the bregma was found by lightly rubbing the point of the needle over the skull until the suture was felt through the skin (about 1-3 mm rostral to a line drawn through the anterior base of the ears). The needle was inserted about 2 mm lateral to the midline and the total volume injected was 4 µl. Control mice received the same volume of physiological saline. Shortly after testing, the animals were sacrificed according ACUC protocols: a slit was made along the midline of the scalp and mice having needle tract 2 mm lateral from the bregma were counted as having been injected correctly. Tail-Flick Test for Spinal Antinociception. Radiant heat on the dorsal surface of the tail measured the reaction time in a tail-flick instrument (Columbus Instruments, Columbus, OH). The latency for removal of the tail from the onset of the radiant heat is defined as the tail-flick latency (TFL). The baseline TFL was adjusted between 2 and 3 sec (pre-response time) and a cutoff time was set at 8 sec to avoid external heat-related damage. The analgesic response was measured 10, 15 or 30 min following i.c.v, s.c. or p.o. injections, respectively. Duration time for i.c.v. and s.c. injections, and p.o. administration were 10 and 15 min, respectively, and the test was terminated when TFL was close to the pre-response time. Mice were injected with either saline or different 13

14 doses of the compounds. Morphine served as a positive control. Minimum Effective Dose (MED) is the minimum dose of compound showing statistically significant antinociceptive effect expressed as area under the curve (AUC) value in comparison with saline-treated group. Hot-Plate Test for Supraspinal Antinociception. The animals were placed on an electrically heated plate (IITC MODEL 39D Hot Plate analgesia meter, IITC Inc, Woodland Hills, CA) at 55 ± 0.1 C after 10, 15 or 30 min following i.c.v., s.c. or p.o. administration of compounds, respectively. Hot-plate latency (HPL) was measured as the interval between placement of mice onto the hot plate and observing movement consisting of either jumping, licking or shaking their hind paws with a baseline latency of 15 sec and maximal cut off time of 30 sec. The measurement of time was the same as described for the TF test. Statistical Analysis. Statistical analysis between two independent groups was performed with Student's test. ANOVA was used for multiple comparisons. When a significant difference among the treatments was obtained in the ANOVA, a post hoc Bonferroni s test was used. The AUC was obtained by plotting the response time (sec) on the ordinate and time (min) on the abscissa after administration of the compounds. 14

15 Results Receptor Binding Affinity and Selectivity. Table 1 lists data from the competitive binding and in vitro bioassays. All the pyrazinone-containing compounds (1-4), 1 [3-(4 -Dmt-aminobutyl)-6-(3'-Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone], 2 [3,6-bis- (4 -Dmt-aminobutyl)-5-methyl-2(1H)pyrazinone], 3 [3,6-bis(3 -Dmt-aminopropyl)-5- methyl-2(1h)pyrazinone] and 4 [3-(3'-Dmt-aminopropyl)-6-(4 -Dmt-aminobutyl)-5- methyl-2(1h)pyrazinone], interacted to a high degree with both µ- and δ-opioid receptors in rat brain membrane preparations with selectivity for µ- over δ-opioid receptors. Compound 1 (Fig.1) displayed µ-opioid receptor affinity that was 2 to 5 times greater than that of compounds 2-4. Functional Bioactivity in vitro. All compounds exhibited potent opioid agonist activity to the µ-opioid receptor in guinea pig ileum (Table 1); however, 1 and 2 also showed modest δ-opioid agonist activity in the mouse vas deferens bioassay system. On the other hand, 3 exhibited the best bioactivity toward the µ-opioid receptor (GPI) and had no activity toward the δ-opioid receptor up to 10 µm in the MVD bioassay. Antinociception of Opioid Mimetic Compounds in Mice by Intracerebroventricular, Subcutaneous and Oral Administration. A dose-dependent 15

16 antinociceptive effect in both TF and HP tests was observed after i.c.v. administration of 1. The effect was potent with a MED of 0.7 and 6.9 ng/mouse in the TF and HP tests, respectively, and times greater than with morphine (Fig. 2A, B and 3A, B). Compounds 2 and 3 also exhibited potent antinociception in the TF test; they were stronger than morphine by 39- to 41- and 50- to 63-fold, respectively. However, 2 and 3 were only 22-89% as effective as 1 depending on the method of measurement (Fig. 4). Compound 4 was not tested in vivo because of its weak bioactivity in the GPI and MVD bioassays (Table 1). In the TF test, the spinal antinociceptive effect induced by 1 was completely blocked by the non-selective opioid antagonist naloxone to same degree as morphine-treated mice (data not shown), and partially (47%) blocked by NTI, a selective δ-opioid antagonist (Fig. 5A). The irreversible µ 1 /µ 2 -receptor antagonist β-fna reduced the antinociception of 1 by 74%; however, the µ 1 antagonist NAZ had no significant effect (Fig. 5B). The supraspinal analgesic effect of 1, determined in the HP test, was completely blocked by naloxone; the same results were obtained with morphine-treated animals (data not shown). However, NTI was ineffective in the HP test (Fig. 6A). On the other hand, β-fna and NAZ strongly reduced the antinociception of 1 to a similar degree, namely 87% and 82%, 16

17 respectively (Fig. 6B). Injection of 1 s.c. revealed a dose-dependent antinociception in both the TF and HP tests (Fig. 7); the effect remained significant for at least 2 to 3 h depending on the dose and measurement paradigm (data not shown). The antinociception measured by compound 1 was equivalent to that produced by morphine in the TF test; however, it was only 89% as effective as morphine in the HP test (Fig. 7). The oral administration of 1 exhibited antinociception with a MED at 10 and 30 mg/kg in the TF and HP tests, respectively (Fig. 8). Although 1 was 65% as effective as morphine in the TF test, it was not comparable to morphine even at 30 mg/kg; in the HP test, 1 demonstrated only 16% of the activity of morphine. Tolerance by the Opioid Mimetic Compound 1 and Morphine. To assess the development of tolerance, mice were injected s.c. with 3 mg/kg of 1 or morphine daily for 7 days, and TF and HP latencies were measured. On day one of the injection regime, both compound 1 and morphine showed antinociception to a similar degree in the TF test with an AUC of 342 and 350, respectively; both were significantly reduced to 56 and 51%, respectively, after a daily injection scheme (Fig. 9A, B). At the end of a 7 day dosing period, antinociception measured by the HP test was reduced 78 and 98% with 1 and 17

18 morphine-treated mice, respectively (Fig. 9 C, D). 18

19 Discussion The goal of developing highly potent opioid mimetics with strong antinociception and lacking tolerance, physical and psychological dependence led many researchers to design new opioid mimetics by initially modifying the address and message domains of naturally occurring mammalian ligands. Later alterations in opioid structure, in particular the substitution of the N-terminal residue Tyr by Dmt (Chandrakumar et al., 1992; Hansen et al., 1992; Pitzele et al., 1994; Guerrini et al., 1996; Salvadori et al., 1997; Sasaki et al., 1999; Schiller et al., 2000; Bryant et al., 2003) and dimerization of Dmt through alkyldiamine (Okada et al., 2003), led to increases in the binding to opioid receptors and bioactivities in vitro and in vivo (Bryant et al., 2003). This study provided evidence that a pyrazinone ring connecting to two symmetric Dmt N-termini by alkyl chains might serve a dual role in opioid peptides as both message and address domains. Further, the data revealed interesting differences in the bioactivity in vitro and in vivo to known opioid peptides. Compounds of the family of 3-[Dmt-NH-(CH 2 ) n ]-6-[Dmt-NH-(CH 2 ) m ]-5-methyl- 2(1H)-pyrazinones (where n, m = 3 or 4) interacted at µ-opioid receptors with high affinity and bioactivity similar to that of the bis-(dmt-nh)-alkyl compounds (Okada et al., 2003). In particular, 1 showed the highest 19

20 affinity, excellent selectivity for µ-opioid receptors, and potent bioactivity in vitro and in vivo. The in vivo studies revealed that 1 was a potent analgesic acting at µ-opioid receptors through supraspinal and spinal mechanisms, and the antinociception induced by 1 was reversible by the general opioid antagonist naloxone, which confirmed that it acted through opioid receptors (Sawynok et al., 1979). Although 1 is more selective for µ- than δ-opioid receptors (Table 1), the fact that the selective δ-opioid antagonist NTI (Portoghese et al., 1988) partially blocked the i.c.v. antinociceptive effect measured by TF, not by HP tests, suggests that the spinal effect of 1 is mediated by both µ- and δ-opioid receptors; however only µ-opioid receptors were responsible for the supraspinal mechanism, which is defined by the HP test. Similarly, Heyman et al. (1988) reported that the δ-opioid receptor antagonist ICI-154,129 is a more potent inhibitor of spinal DPDPE- than spinal DAMGO-induced antinociception, and that the irreversible µ-antagonist β-fna blocks the antinociceptive effects of spinally injected DAMGO, but not DPDPE. Those data supported the involvement of µ- and δ-opioid receptors on spinal antinociception. Other reports also provided substantial evidence to validate our observations that both δ- and µ-receptors are involved in the appearance of antinociception at the level of the spinal cord 20

21 (Heyman et al., 1987; Porreca et al., 1984; Paul et al., 1989). µ-opioid receptors are important in both spinal and supraspinal antinociception (Ling and Pasternak, 1983; Bodnar et al., 1988; Heyman et al., 1988; Porreca et al., 1984) and are divided into two distinct pharmacological subtypes. The µ 1 -type binds morphine and most enkephalin analogues, and the µ 2 -type interacts primarily with morphine (Goodman and Pasternak, 1985, Moskowitz and Goodman, 1985a,b; Nishimura et al., 1984; Pasternak and Wood, 1986; Wolozin and Pasternak, 1981). Heyman et al. (1988) and Paul et al. (1989) suggest that different µ-opioid receptor subtypes mediate antinociception at the spinal and supraspinal levels: whereas the µ 1 -opioid receptor subtype is more important for supraspinal, the µ 2 -subtype is involved in spinal antinociception (Ling and Pasternak, 1983; Bodnar et al., 1988; Porreca et al., 1984, Neyman et al., 1988; Paul et al 1989). However, recent studies with µ-receptor knockout mice show that the µ-receptor agonists, such as heroin and morphine-6-β-glucuronide (M6G), induce antinociception by a mechanism different from that of morphine, which had no effect on CXBK mice lacking µ 1 -opioid receptors, while heroin and M6G showed potent antinociception (Rossi et al. (1996). Narita et al. (2002) also showed that antinociception induced by fentanyl, the anilidopiperidine class of opioids, may be mediated predominantly through µ 1 -opioid 21

22 receptors at both supraspinal and spinal sites in mice. Therefore, to examine the relative roles of µ 1 - and µ 2 -receptors in spinal and supraspinal antinociception of 1, we assessed the effects of NAZ, a selective µ 1 -opioid receptor antagonist and β-fna, an irreversible µ 1 /µ 2 -opioid antagonist. β-fna reduced the antinociception of 1 in the TF test; however, NAZ had no significant effect, which suggesting that µ 2 -opioid receptors are mainly involved in the spinal mechanism of 1. On the other hand, β-fna and NAZ reduced the antinociception of 1 to a similar degree in the HP test, indicating that supraspinal antinociception is mediated through µ 1 -opioid receptors. Nor-BNI, a selective κ-opioid antagonist, did not block the effect of 1 in TF and HP tests (data not shown). Thus, we conclude that 1 is an opioid mimetic with µ- and δ-opioid receptor agonist activities acting spinally and supraspinally. Spinal effects occurred through δ- and the combined action of both µ 1 - and µ 2 -opioid receptors with a preference of µ 2 - over µ 1 -opioid receptor subtypes, whereas supraspinal effects arise primarily through the action of µ 1 -opioid receptors. Administration of 1 either by s.c. or p.o. exhibited a different degree of antinociception depending on measurement test, which suggested that 1 crossed both the gastrointestinal epithelial and blood-brain barriers into the brain to produce a CNS-mediated antinociception. Although 1 was 65- to 71-times more potent than morphine following 22

23 i.c.v. administration, it did not show the same degree of antinociception after peripheral injection, which may be explained by diverse factors, such as lipophilicity, molecular weight, metabolic and protease stabilities, and transport mechanisms, which might limit transit of 1 through membrane barriers (Banks and Kastin, 1990; Smith et al., 1992; Ermisch et al., 1993; Samii et al., 1994). Morphine is the one of the most potent opiates that is widely used to relieve pain. However, the development of tolerance, physical and psychological dependence during long-term therapy, and its respiratory depressant effects, sedation and withdrawal symptoms are serious side effects (Yaksh et al., 1977). When we investigated the tolerance of 1 following s.c. injection, it produced an effect similar to morphine, which suggests that both compounds appeared to act by the same mechanism; however, since 1 is more potent, lower concentrations could be used for post-operative pain, or in the treatment of acute or chronic cancer neuropathies. Thus, the pyrazinone ring may be an ideal platform on which to develop stable opioids with sufficient lipophilicity suitable for clinical and therapeutic applications, albeit other biological endpoints must be studied to determine if these compounds have a similar spectrum of detrimental effects as morphine. 23

24 References Balboni G, Salvadori S, Guerrini R, Negri L, Giannini E, Jinsmaa Y, Bryant SD, and Lazarus LH (2002) Potent δ-opioid receptor agonists containing the Dmt-Tic pharmacophore. J Med Chem 45: Banks WA and Kastin AJ (1990) Peptide transport systems for opiates across the blood-brain barrier. Am J Physiol 259: E1-E10. Bodnar RJ, Williams CL, Lee SJ, and Pasternak GW (1988) Role of mu 1-opiate receptors in supraspinal opiate analgesia: a microinjection study. Brain Res 447: Bryant SD, Salvadori S, Cooper PS, and Lazarus LH (1998) New δ-opioid antagonists as pharmacological probes. Trends Pharmacol Sci 19: Bryant SD, Jinsmaa Y, Salvadori S, Okada Y, and Lazarus LH (2003) Dmt and opioid peptides: a potent alliance. Biopolymers (Peptide Sci) 71: Chandrakumar NS, Stapelfeld A, Beardsley PM, Lopez OT, Drury B, Anthony E, Savage MA, Williamson LN, and Reichman M (1992) Analogs of the delta opioid receptor selective cyclic peptide [2-D-penicillamine,5-D-penicillamine]-enkephalin: 2', 6' - dimethyltyrosine and Gly 3 -Phe 4 amide bond isostere substitutions. J Med Chem 32:

25 Dygos JH, Yonan EE, Scaros MG, Goodmonson OJ, Getman DP, Periana RA, and Beck GR (1992) A convenient asymmetric synthesis of the unnatural amino acid 2,6-dimethyl-L-tyrosine. Synthesis 8: Ermisch A, Brust P, Kretzschmar R, and Ruhle HJ (1993) Peptides and blood-brain barrier transport. Physiol Rev 73: Goodman RR and Pasternak GW (1985) Visualization of mu1 opiate receptors in rat brain by using a computerized autoradiographic subtraction technique. Proc Natl Acad Sci U S A 82: Guerrini R, Capasso A, Sorrentino L, Anacardio R, Bryant SD, Lazarus LH, Attila M, and Salvadori S (1996) Opioid receptor selectivity alteration by single residue replacement: synthesis and activity profile of [Dmt 1 ]deltorphin B. Eur J Pharmacol 302: Hansen JDW, Stapelfeld A, Savage MA, Reichman M, Hammond DL, Haaseth RC, and Mosberg HI (1992) Systemic analgesic activity and delta-opioid selectivity in [2,6-dimethyl-Tyr 1,D-Pen 2,D-Pen 5 ]enkephalin. J Med Chem 35: Heyman JS, Mulvaney SA, Mosberg HI, and Porreca F (1987) Opioid delta-receptor involvement in supraspinal and spinal antinociception in mice. Brain Res 420:

26 Heyman JS, Williams CL, Burks TF, Mosberg HI, and Porreca F (1988) Dissociation of opioid antinociception and central gastrointestinal propulsion in the mouse: studies with naloxonazine. J Pharmacol Exp Ther 245: Hughes J, Smith TW, Kosterlitz HW, Fothergill LA, Morgan BA, and Morris HR (1975) Identification of two related pentapeptides from the brain with potent opiate agonist activity. Nature 258: Laursen SE and Belknap JK (1986) Intracerebroventricular injections in mice. J Pharmacol Meth 16: Ling GS and Pasternak GW (1983) Spinal and supraspinal opioid analgesia in the mouse: the role of subpopulations of opioid binding sites. Brain Res 271: Ling GS, Simantov R, Clark JA, and Pasternak GW (1986). Naloxonazine actions in vivo. Eur J Pharmacol 129: Moskowitz AS and Goodman RR (1985a) Autoradiographic distribution of mu1 and mu2 opioid binding in the mouse central nervous system. Brain Res 360: Moskowitz AS and Goodman RR (1985b) Autoradiographic analysis of mu1, mu2, and delta opioid binding in the central nervous system of C57BL/6BY and CXBK (opioid 26

27 receptor-deficient) mice. Brain Res 360: Narita M, Imai S, Itou Y, Yajima Y, and Suzuki T (2002) Possible involvement of m1-opioid receptors in the fentanyl- or morphine-induced antinociception at supraspinal and spinal sites. Life Sci 70: Nishimura SL, Recht LD, and Pasternak GW (1984) Biochemical characterization of high-affinity 3H-opioid binding. Further evidence for Mu1 sites. Mol Pharmacol 25: Okada Y, Tsukatani M, Taguchi H, Yokoi T, Bryant SD, and Lazarus LH (1998) Amino acids and peptides. LII. Design and synthesis of opioid mimetics containing a pyrazinone ring and examination of their opioid receptor binding activity. Chem Pharm Bull (Tokyo) 46: Okada Y, Fukumizu A, Takahasi M, Yokoi T, Tsuda Y, Bryant SD, and Lazarus LH (1999) Synthesis of pyrazinone ring-containing opioid mimetics and examination of their opioid receptor binding activity. Chem Pharm Bull (Tokyo) 47: Okada Y, Tsuda Y, Fujita Y, Yokoi T, Sasaki Y, Ambo A, Konishi R, Nagata M, Salvadori S, Jinsmaa Y, Bryant SD, and Lazarus LH (2003) Unique high-affinity synthetic 27

28 mu-opioid receptor agonists with central- and systemic-mediated analgesia. J Med Chem 46: Pasternak GW and Wood PJ (1986) Multiple mu opiate receptors. Life Sci 38: Paul D, Bodnar RJ, Gistrak MA, and Pasternak GW (1989) Different mu receptor subtypes mediate spinal and supraspinal analgesia in mice. Eur J Pharmacol 168: Porreca F, Mosberg HI, Hurst R, Hruby VJ, and Burks TF (1984) Roles of mu, delta and kappa opioid receptors in spinal and supraspinal mediation of gastrointestinal transit effects and hot-plate analgesia in the mouse. J Pharmacol Exp Ther 230: Portoghese PS, Sultana M, and Takemori AE (1988) Naltrindole, a highly selective and potent non-peptide delta opioid receptor antagonist. Eur J Pharmacol 146: Pitzele BS, Hamilton RW, Kudla KD, Tsymbalov S, Stapelfeld A, Savage MA, Clare M, Hammond DL, and Hansen DW Jr (1994) Enkephalin analogs as systemically active antinociceptive agents: O- and N-alkylated derivatives of the dipeptide amide L-2,6-dimethyltyrosyl-N-(3-phenylpropyl)-D-alaninamide. J Med Chem 37: Rossi GC, Brown GP, Leventhal L, Yang K, and Pasternak GW (1996) Novel receptor 28

29 mechanisms for heroin and morphine-6 -glucuronide analgesia. Neurosci Lett 216:1-4 Salvadori S, Balboni G, Guerrini R, Tomatis R, Bianchi C, Bryant SD, Cooper PS, and Lazarus LH (1997) Evolution of the Dmt-Tic pharmacophore: N-terminal methylated derivatives with extraordinary delta opioid antagonist activity. J Med Chem 40: Salvadori S, Guerrini R, Balboni G, Bianchi C, Bryant SD, Cooper PS, and Lazarus LH (1999) Further studies on the Dmt-Tic pharmacophore: hydrophobic substituents at the C-terminus endow delta antagonists to manifest mu agonism or mu antagonism. J Med Chem 42: Samii A, Bickel U, Stroth U, and Pardridge WM (1994) Blood-brain barrier transport of neuropeptides: analysis with a metabolically stable dermorphin analogue. Am J Physiol 267: E124-E131. Sasaki Y, Suto T, Ambo A, Ouchi H, and Yamamoto Y (1999) Biological properties of opioid peptides replacing Tyr at position 1 by 2,6-dimethyl-Tyr. Chem Pharm Bull (Tokyo) 47: Sawynok J, Pinsky C, and LaBella FS (1979) On the specificity of naloxone as an opiate antagonist. Life Sci 25:

30 Schiller PW, Nguyen TM, Berezowska I, Dupuis S, Weltrowska G, Chung NN, and Lemieux C (2000) Synthesis and in vitro opioid activity profiles of DALDA analogues. Eur J Med Chem 35: Smith PL, Wall DA, Gochoco CH, and Wilson G (1992) Routes of delivery: case studies. Oral absorption of peptides and proteins. Ad Drug Delivery Rev 8: Wolozin BL and Pasternak GW (1981) Classification of multiple morphine and enkephalin binding sites in the central nervous system. Proc Natl Acad Sci U S A 78: Yaksh TL, Kohl RL, and Rudy TA (1977) Induction of tolerance and withdrawal in ratsreceiving morphine in the spinal subarachnoid space. Eur J Pharmacol 42: Zadina JE, Hackler L, Ge LJ, and Kastin AJ (1997) A potent and selective endogenous agonist for the mu-opiate receptor. Nature 386:

31 Footnotes Reprint reguests to: Yunden Jinsmaa, Ph.D, Medicinal Chemistry Group, LCBRA, NIEHS, P.O. Box 12233, Research Triangle Park, NC 27709, USA. Tel: (919) Fax: (919)

32 Legends for Figures Fig. 1. Chemical structure of 3-(4 -Dmt-aminobutyl)-6-(3'-Dmt-aminopropyl)- 5-methyl-2(1H)pyrazinone (1). Fig. 2. Effect of 3-(4 -Dmt-aminobutyl)-6-(3'-Dmt-aminopropyl)-5-methyl- 2(1H)pyrazinone (1) on the tail-flick latency after intracerebroventricular administration in mice. (A) Time course, (B) Area Under the Curve (AUC). Each value is the mean with S.E.M. of 5 to 6 mice. (*) Denotes values that are significantly different from saline-treated mice by Bonferroni s test (*** p < 0.001, ** p < 0.01, * p < 0.05). Fig. 3. Effect of 3-(4 -Dmt-aminobutyl)-6-(3'-Dmt-aminopropyl)- 5-methyl-2(1H)pyrazinone (1) on the hot-plate latency after intracerebroventricular administration in mice. (A) Time course, (B) Area Under the Curve (AUC). Each value is the mean with S.E.M. of 5 to 6 mice. (*) Denotes values that are significantly different from saline-treated mice by Bonferroni s test (*** p < 0.001, * p < 0.05). 32

33 Fig. 4. Effect of intracerebroventricularly administered 3-(4 -Dmt-aminobutyl)- 6-(3'-Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone (1), 3,6-bis-(4 -Dmt-aminobutyl)- 5-methyl-2(1H)pyrazinone (2), and 3,6-bis-(3 -Dmt-aminopropyl)-5-methyl- 2(1H)pyrazinone (3) on the tail-flick (A) and hot-plate (B) latencies in mice. Each compound was administered at a dose of 7 ng/mouse and morphine was at 0.5 µg/mouse. Each point represents the mean and S.E.M. determined by the use of 5 mice. (*) Denotes values that were significantly different than saline-treated mice by Bonferroni s test (*** p < 0.001, ** p < 0.01). Fig. 5. Effect of opioid receptor antagonists naloxone, naltrindole (NTI), naloxonazine (NAZ) and β-funaltrexamine (β-fna) on antinociception induced by 3-(4 -Dmt-aminobutyl)-6-(3'-Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone (1) (6.9 ng, i.c.v.) in the tail-flick test in mice. (A) Naloxone (2 mg/kg, s.c.) and NTI (3 mg/kg, s.c.). (B) NAZ (35 mg/kg, s.c.) and β-fna (40 mg/kg, s.c.). Each point presents the mean and S.E.M. determined using 5 to 6 mice in A and 7 to 10 mice in B. (*) Denotes values that were significantly different than saline-treated mice by Student`s test (*** p < 0.001) 33

34 and (#) denotes values that were significantly different from 1-treated mice by Bonferroni`s test (### p < 0.001, ## p < 0.01, # p < 0.05). Fig. 6. Effect of opioid antagonists on antinociception induced by 3-(4 -Dmt-aminobutyl)-6-(3'-Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone (1) (6.9 ng, i.c.v.) in the hot-plate test in mice. (A) Naloxone (2 mg/kg, s.c.) and NTI (3 mg/kg, s.c.). (B) NAZ (35 mg/kg, s.c.) and β-fna (40 mg/kg, s.c.). Each point presents the mean and S.E.M. determined using 5 to 6 mice in A and 7 to 10 mice in B. (*) Denotes values that were significantly different than saline-treated mice by Student`s test (** p < 0.01, * p < 0.05) and (#) denotes values that were significantly different from 1-treated mice by Bonferroni`s test (## p < 0.01). Fig. 7. Effect of subcutaneously injected 3-(4 -Dmt-aminobutyl)- 6-(3'-Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone (1) on the tail-flick (A) and hot-plate (B) latencies in mice. Each point represents the mean and S.E.M. computed using 5 mice. (*) Denotes values that were significantly different than saline-treated mice by Bonferroni s test (*** p < 0.001, ** p < 0.01) 34

35 Fig. 8. Effect of orally administered 3-(4 -Dmt-aminobutyl)- 6-(3'-Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone (1) on the tail-flick (A) and hot-plate (B) latencies in mice. Each point represents the mean and S.E.M. derived from the use of 5 to 8 mice. (*) Denote values that were significantly different than saline-treated mice by Bonferroni`s test (*** p < 0.001, ** p < 0.01, * p < 0.05) Fig. 9. Antinociceptive effect of subcutaneously injected 3-(4 -Dmt-aminobutyl)-6-(3'-Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone (1) (3 mg/kg) or morphine (3 mg/kg) for 7 days measured by the tail-flick (A, B) and hot-plate (C, D) tests. A and C represents the time course (open symbols: circles, morphine; squares, 1-treated mice. Closed symbols: day 1; open symbols, day 7). B and D represents area under the curve (AUC). Each point represents the mean and S.E.M. obtained by using 7 mice. (*) Denotes values that were significantly different from those on day one for 1- or morphine-treated mice by Student`s test (*** p < 0.001, ** p < 0.01, * p < 0.05). 35

36 Table 1. Opioid receptor interactions and functional bioactivity of 3-(4 -Dmt-aminobutyl)-6-(3'-Dmt-aminopropyl)-5-methyl-2(1H)pyrazinone (1), 3,6-bis- (4 -Dmt-aminobutyl)-5-methyl-2(1H)pyrazinone (2), 3,6-bis-(3 -Dmt-aminopropyl)-5- methyl-2(1h)pyrazinone (3) and 3-(3'-Dmt-aminopropyl)-6-(4'-Dmt-aminobutyl)-5- methyl-2(1h)pyrazinone (4). Cmpd K i µ, nm n K i δ, nm n δ/µ GPI (IC 50, nm) MVD (IC 50, nm) ± ± , ± ± ± ± >10, ± ± >10,000 36

37 HO CH 3 N ( )4 NH 2 N H H CH CH 3 3 N H N () 3 N O 2 H O O CH 3 HO Fig.1 Jinsmaa et al.

38 A TF latency (sec) B AUC pre Time after administration (min) saline * *** *** (1) ng/mouse *** morphine 0.5 µg/mouse saline 0.7 ng (1) 2.1 ng (1) 6.9 ng (1) 0.5 µg morph. Fig.2 Jinsmaa et al.

39 HP latency (sec) A 30 B AUC pre Time after administration (min) saline * *** (1) ng/mouse * saline 2.1 ng (1) 6.9 ng (1) 21 ng (1) 0.5 µg morph. 110 morphine 0.5 µg/mouse Fig.3 Jinsmaa et al.

40 AUC AUC A B saline saline ** ** *** (1) (2) (3) morphine * * (1) (2) (3) morphine Fig.4 Jinsmaa et al.

41 A B AUC AUC saline saline *** # ## naloxone NTI (1) *** ### ### naloxone β-fna ΝΑΖ (1) Fig.5 Jinsmaa et al.

42 A 700 ** AUC B 800 AUC saline saline * ## naloxone NTI (1) ## ## ## naloxone β-fna ΝΑΖ (1) Fig.6 Jinsmaa et al.

43 A 1000 *** AUC AUC B saline saline 1 1 *** ** (1) morphine 3 10 mg/kg 3 *** *** *** (1) morphine mg/kg Fig.7 Jinsmaa et al.

44 A AUC B AUC ** * saline (1) mg/kg * saline (1) mg/kg *** morphine 10 ** morphine 30 Fig.8 Jinsmaa et al.

45 TF latency (sec) HP latency (sec) A C * * * * *** *** *** *** pre Time after administration (min) pre Time after administration (min) * * AUC AUC B D ** (1) 3 mg/kg 1st day 7th day * (1) 3 mg/kg 1st day 7th day *** morphine 3 mg/kg 1st day 7th day * morphine 3 mg/kg 1st day 7th day Fig.9 Jinsmaa et al.