Unlike most neuropeptides that are produced by the processing of precursor peptides, kyotorphin is produced from tyrosine and arginine by the action of a specific kyotorphin synthetase (16, 17). While studying kyotorphin release from brain slices, we found that kyotorphin accumulates in a time-dependent manner in the presence of bestatin, an aminopeptidase inhibitor (18). Extensive accumulation of kyotorphin was also observed in the synaptosome fraction. We discovered kyotorphin synthetase, which produces kyotorphin from tyrosine and arginine in the presence of ATP and MgCl₂. In a study on the purification of kyotorphin synthetase (17), we performed a radioimmunoassay using a specific kyotorphin antiserum. The starting materials, P2 crude synaptosomes, were lysed and the supernatant (synaptosol) was separated using Sephacryl S-300 gel-filtration, DE52 anion-exchange chromatography, and TSKgel G4000SW chromatography, in which active materials were observed at 240–250 kDa. From the enzymatic characterization, it was revealed that the optimal pH was 7.5–9.0, and enzymatic activities reached a plateau at 0.2–0.5 mM of tyrosine, 4–8 mM of arginine, 2–4 mM of ATP, and 2–4 mM of MgCl₂. The Km values for tyrosine, arginine, ATP, and MgCl₂ were 25.6, 926, 294, and 442 μM, respectively. The possible reaction mechanism of kyotorphin synthetase may be outlined as follows: tyrosine + arginine + ATP→ (synthetase, Mg²⁺)→ tyrosine-arginine (kyotorphin) + AMP + bisphosphate. The regional and subcellular distribution of kyotorphin synthetase in the brain (17) correlated well with that of kyotorphin content (12, 15). Enzyme activity was highest in the midbrain and medulla oblongata and in the synaptosome fractions. However, molecular characterization of kyotorphin synthetase has not been conclusively determined. A previous study demonstrated that Bacillus stearothermophilus tyrosyl-tRNA synthetase (TyrRS) could catalyze the in vitro synthesis of kyotorphin (17, 19), whereas Streptomyces septatus aminopeptidase could not synthesize kyotorphin from substrate amino acid; however, it resulted in several other dipeptides (20). Recently, we successfully demonstrated that tyrosyl-tRNA synthetase (TyrRS) is a potential kyotorphin synthetase in mammals (16). In this study, the Km of tyrosine and arginine for in vitro kyotorphin synthesis by recombinant human TyrRS (hTyrRS) was 200 and 1,400 μM, respectively, similar to that of partially purified kyotorphin synthetase from rat brains. Although the in vitro cell-free study showed that hTyrRS could produce several tyrosine-containing dipeptides (kyotorphin, tyrosine-tyrosine, tyrosine-proline, and tyrosine tryptophan), treatment of PC12 cells with siRNA antisense for TyrRS selectively blocked the in vitro biosynthesis of kyotorphin, but not that of tyrosine-tyrosine, tyrosine-proline, or tyrosine tryptophan. However, siRNA treatment did not affect cell survival or proliferation. In addition, TyrRS mRNA expression was higher in the midbrain and medulla oblongata than in other brain regions, which is consistent with the localization of kyotorphin synthetase activity in rat brains (17).

Regarding kyotorphin biosynthesis, there is a possibility that kyotorphin is also produced from precursor polypeptides. When the soluble fraction from synaptosomes was incubated in the presence of bestatin, an amino peptidase inhibitor, kyotorphin accumulated in a time-dependent manner, possibly through calcium-activated neutral proteases (μ- and m-CANP),calpain-1 or calpain-2 (21). Detailed studies further demonstrated that the enzyme purified using DE52 cellulose, Ultrogel AcA, thiopropyl-Sepharose 6B, second DE52 cellulose, Ultrogel AcA34, and blue Sepharose CL-6B was characterized as a 74 kDa protein, a novel type of calpain lacking caseinolytic activity, with calpastatin identified as the substrate for this enzyme or precursor of kyotorphin (22). As the activation of this new type of calpain and calpain-1 requires more than 1 μM Ca²⁺, which is higher than the concentration at resting state (0.1–0.3 μM), this calpain-calpastatin system may be driven when neurons are activated. However, the exact physiological mechanisms are unclear.

To detect the specific binding of kyotorphin to brain membranes, we prepared ³H-kyotorphin possessing high specific activity using B. stearothermophilus-derived TyrRS, which was immobilized on Sepharose 4B (19, 25). The enzymatic reaction was performed by incubation of ³H-tyrosine with high specific activity and high concentration of arginine in the presence of 20 mM ATP and MgCl₂ at 45°C for 48 h. After HPLC purification, ³H-kyotorphin was separated from other substrates. As ³H-kyotorphin non-specifically binds to glass-fiber filters, we adopted the centrifugation method to separate the membrane-bound form of ³H-kyotorphin from the free ³H-kyotorphin. From the Scatchard plot, it was revealed that brain membranes have high-affinity and low-affinity binding sites for ³H-kyotorphin. The Kd and Bmax were 0.34 nM and 36 fmol/mg protein for the high-affinity binding sites, whereas they were 9.07 nM and 1.93 pmol/mg protein for the low-affinity binding sites. The high-affinity binding using 0.5 nM ³H-kyotorphin in various brain regions was 121 (fmol/mg protein) in the hypothalamus, 104 in the amygdala, 77 in the occipital cortex, 67 in the frontal cortex, 65 in the thalamus, 58 in the medulla oblongata, 57 in the hippocampus, 53 in the midbrain, 48 in the striatum, and 39 in the cerebellum, indicating that there is some similarity in terms of regional distribution between ³H-kyotorphin binding and kyotorphin contents (12) or kyotorphin synthetase levels (17). In the competition study, the IC₅₀ value was calculated for 20.8 (nM) for kyotorphin, 11.2 for leucine-arginine, 12.7 for phenylalanine-arginine, 37.6 for tyrosine-leucine, 224 for tyrosine-lysine, and >1,000 for other dipeptides, suggesting that ³H-kyotorphin binding recognizes dipeptides containing tyrosine or arginine.

As ³H-kyotorphin binding was blocked in the presence of GTPγS + MgCl₂, the receptor was presumed to couple with GTP binding proteins (25). When the low Km GTPase activity in the brain membrane preparation was measured, kyotorphin showed a concentration-dependent enhancement of this activity in the range of 10 nM to 100 μM (25). As L-leucine-L-arginine (Leu-Arg) did not activate the low Km GTPase activity, but it showed a strong inhibition of ³H-kyotorphin binding, Leu-Arg was presumed to be an antagonist for the GTP-binding protein coupled receptor (GPCR) for kyotorphin. This speculation was proven by the finding that the addition of Leu-Arg eliminated kyotorphin-induced enhancement of low Km GTPase activity (25).

Since receptor-mediated enhancement of low Km GTPase activity is often reported in the case of G_i/o- or G_s-coupled receptors, but not for G_q/11-coupled receptors, we speculated that G_i/o could be a possible candidate for coupling with kyotorphin receptor. Based on the analogy of μ-opioid receptor-G_i/o coupling (26, 27), we first treated the brain membranes with pertussis toxin, which blocks the interaction of receptors to G_i/o by ADP-ribosylation of cysteine SH-residue at the fourth amino acid from the C-terminal of the α-subunit (25). Pertussis toxin-ablated kyotorphin-enhancement of low Km GTPase activity was recovered by reconstitution with purified G_i or G_o. However, as an effector system, we successfully observed that kyotorphin activated phospholipase C (PLC) in an experiment measuring inositol (1,4,5)-P₃/InsP₃ (25). The kyotorphin-induced and Leu-Arg-reversible PLC activation in the synaptosome membranes was also eliminated by pretreatment with pertussis toxin, and recovered by reconstitution with purified G_i, but not G_o. The pioneering finding that G_i-coupled receptor mediates PLC activation (Figure 2) was supported by successive studies, and the concept has been further refined by a recent study (28), where the Gα_i-Gβγ-PLCβ-Ca²⁺ signaling module was found to be entirely dependent on the presence of active Gα_q. In other words, G_i-mediated PLC activation was positively regulated by G_q-coupled receptor activation.

Based on the kyotorphin signaling through the activation of the G_i-PLC system, we concluded that kyotorphin caused a G_i-PLC-InsP₃ receptor (InsP₃R)-mediated transport of ⁴⁵Ca²⁺, which had been filled in the resealed vesicles made of synaptosomal membranes (29). In this experiment, lysed synaptosomes were incubated with ⁴⁵CaCl₂ to generate inside-out and outside-out types of resealed vesicles containing ⁴⁵CaCl₂. When the resealed vesicles were placed on the GF/C filter and perfused with buffer, the basal ⁴⁵Ca²⁺ release became stable after 10–20 min. The addition of InsP₃ or kyotorphin (+GppNHp) caused an increase in ⁴⁵Ca²⁺ release from inside-out or outside-out vesicles, respectively. The preceding treatment of synaptosomal membranes with pertussis toxin blocked the kyotorphin-induced ⁴⁵Ca²⁺ release, but this effect was reversed by reconstitution of pertussis toxin-treated membranes with purified G_i reversed. Although the experiment demonstrated ⁴⁵Ca²⁺ release from ⁴⁵Ca²⁺-preloaded vesicles, kyotorphin signaling through the activation of G_i-PLC-InsP₃R under physiological conditions could cause Ca²⁺ influx from the extracellular space (mM Ca²⁺) into the cytosol (sub-μM Ca²⁺). This unique mechanism of GPCR-PLC-InsP₃R-mediated Ca²⁺ influx was further explained by a working hypothesis that InsP₃-mediated ⁴⁵Ca²⁺ influx is driven by the conformational coupling of vesicular InsP₃R and transient receptor potential C1 (TRPC1) in the plasma membrane (24), as seen in Figure 2.

To measure Met-enkephalin release, guinea pig striatal slices of 500 μm thickness or 200 μm cubic slices of the whole spinal cord were placed in a 1.5 ml perfusion chamber (40). The perfusion from the bottom to the top was conducted at 37°C at a rate of 1 ml/min with Krebs-bicarbonate medium gassed with 95% O₂ and 5% CO₂, and perfusion samples were collected at 3-min intervals. The addition of 50 mM KCl caused an ~60-fold Met-enkephalin release, compared to the basal release, but it was completely eliminated by the omission of CaCl₂. The addition of 1 or 10 μM kyotorphin caused a 1.6- or 3.4-fold increase in Met-enkephalin release, respectively, and kyotorphin-induced release was also stopped by the removal of CaCl₂. The kyotorphin-induced Met-enkephalin release was also stopped by the addition of 2 μM tetrodotoxin, which itself slightly reduced the basal release, suggesting that striatal preparation may be spontaneously stimulated by excitatory transmitters such as glutamate leaked from damaged neurons during the slice preparation. When the striatal slices were electrically stimulated at 10 Hz by two disc electrodes fixed to both the top and bottom, a two-fold Met-enkephalin release was observed, and the further addition of 1 μM kyotorphin showed a 3.6-fold increase. However, it remains unclear whether kyotorphin potentiates the electrical stimulation-induced release of Met-enkephalin, since kyotorphin only shows additive effects (1.6 + 2.0 = 3.6-fold) to the electrical stimulation-induced Met-enkephalin release. In the experiments using spinal cord preparations, 10 μM kyotorphin caused a 2.2-fold increase in Met-enkephalin release (40), suggesting that the action of kyotorphin is also observed in the preparation of the spinal cord, while the potency of Met-enkephalin release is slightly weaker.

The addition of 10 μM L-Tyr-D-Arg (Tyr-D-Arg), an enzymatically stable kyotorphin derivative, caused an increase in Met-enkephalin release to the same level as kyotorphin (35). As Tyr-D-Arg showed more potent analgesia in the tail pinch test (ED50: 6.2 nmol/mouse, 60-min duration by 29.5 nmol) than kyotorphin (ED50: 15.7 nmol/mouse, 30-min duration by 59.2 nml) (36), the enhanced analgesic potency of Tyr-D-Arg seems to be attributed to the enzymatic stability, but not to the potency of Met-enkephalin release. Similar potency of Met-enkephalin release by kyotorphin and Tyr-D-Arg was reported by other investigators using the preparation of rat hypothalamus, in which both peptides showed no release of preloaded [³H]noradrenaline, [³H]GABA, or [³H]D-aspartate (43).

The mechanisms underlying kyotorphin-induced Met-enkephalin release or opioid analgesia remain unknown. To date, several studies have demonstrated kyotorphin-induced Met-enkephalin (or endogenous opioids) release from brain or spinal cord preparations (35, 40, 43). In addition to the direct evidence, electrophoretically charged kyotorphin caused an excitatory effect on the spontaneous activity of a neuron in the medulla NRPG, a site of morphine analgesia in a naloxone-reversible manner (44). However, in the peripheral system, kyotorphin and Tyr-D-Arg increased the amplitude and mean quantal content of the fast excitatory postsynaptic potentials in the isolated sympathetic ganglia with preganglionic nerves, and these responses were reversed by naloxone (45). It was also interesting to note that kyotorphin inhibited the isoprenaline-induced increase in the twitch tension of the ventricular papillary muscle of rats, and this effect was eliminated by Leu-Arg and naloxone (46).

In the receptor system, it was found that kyotorphin couples with G_i/o-coupled receptors (25), as seen in the case of the μ-opioid receptor (26). Although G_i/o is generally related to inhibitory functions through G_i-mediated adenylyl cyclase inhibition and G_o-mediated inhibition of voltage-dependent calcium channels (47), the kyotorphin receptor mediates the activation of G_i, PLC, and Ca²⁺ signaling in G_i reconstitution experiments (25). A recent report supported the G_i-mediated activation of PLCβ-Ca²⁺ signaling and also demonstrated that this module is dependent on the presence of active Gα_q (28). Furthermore, it has been reported that GPCR-PLC-InsP₃ activation mediates Ca²⁺ influx through the conformational coupling of vesicular InsP₃R and plasma membrane TRPC1 (24, 29), as seen in Figures 2, 3. Although the omission of Ca²⁺ in the perfusion medium abolished the kyotorphin-induced release of Met-enkephalin from brain slices (40), it remains unclear whether kyotorphin receptor-induced Ca²⁺ influx through a G_i-PLC-InsP₃R mechanism contributes to the Met-enkephalin release. As the kyotorphin-induced Met-enkephalin release was abolished by the presence of tetrodotoxin or the omission of Ca²⁺ in the perfusion medium (40), the activation of tetrodotoxin-sensitive voltage-dependent Na⁺ channels is presumably mediated by glutamate or other excitatory substances leaked from the damaged brain preparations, but not by kyotorphin through G_i-coupled receptors. However, it would be interesting to examine whether kyotorphin-induced Met-enkephalin release requires both the G_i-PLC-InsP₃R mechanism and physiologically occurring depolarization of nerve terminals.

Finally, it is not understood how kyotorphin induces the release of Met-enkephalin, but not other transmitters such as preloaded [³H]noradrenaline, [³H]GABA, or [³H]D-aspartate (43). Although no evidence is available, it has been speculated that kyotorphin is incorporated through unidentified transporters and subsequently displaces Met-enkephalin in synaptic vesicles, causing its release (Figure 3), as seen in the case of tyramine-induced noradrenaline release (48). Although kyotorphin is reported to be an endogenous substrate of the peptide transporter PEPT2, this transporter mechanism is unlikely to be involved in kyotorphin-induced Met-enkephalin release from neurons, since PEPT2 is involved in the clearance of enkephalin from the brain (49).