The antinociceptive effect of intrathecal kynurenic acid and its interaction with endomorphin-1 in rats

Interactions between the Kynurenine and the Endocannabinoid System with Special Emphasis on Migraine

Both the kynurenine and the endocannabinoid systems are involved in several neurological disorders, such as migraine and there are increasing number of reports demonstrating that there are interactions of two systems. Although their cooperation has not yet been implicated in migraine, there are reports suggesting this possibility. Additionally, the individual role of the endocannabinoid and kynurenine system in migraine is reviewed here first, focusing on endocannabinoids, kynurenine metabolites, in particular kynurenic acid. Finally, the function of NMDA and cannabinoid receptors in the trigeminal system-which has a crucial role in the pathomechanisms of migraine-will also be discussed. The interaction of the endocannabinoid and kynurenine system has been demonstrated to be therapeutically relevant in a number of pathological conditions, such as cannabis addiction, psychosis, schizophrenia and epilepsy. Accordingly, the cross-talk of these two systems may imply potential mechanisms related to migraine, and may offer new approaches to manage the treatment of this neurological disorder.

Kynurenic acid and its analogue can alter the opioid receptor G-protein signaling after acute treatment via NMDA receptor in rat cortex and striatum

Previously, we have shown that the N-methyl d-aspartate (NMDA)-receptor antagonist kynurenic acid (KYNA) and its analogue KYNA1 do not bind directly to mu, kappa and delta opioid receptors in vitro. On the other hand, chronic administration of KYNA and KYNA1 resulted in region (cortex vs striatum) and opioid receptor-type specific alterations in G-protein activation of mouse brain homogenates. Here we describe for the first time the acute effect of KYNA and KYNA1 on opioid receptor function with the possible involvement of the NMDA receptor. The acute 30minute in vivo KYNA1 and KYNA treatments altered opioid receptor G-protein signaling or ligand potency depending on the opioid receptor type and brain region (rat cortex vs striatum) using [³⁵S]GTPγS binding assays. Pretreatment with the NMDA receptor antagonist MK-801 impaired or reversed the effects of KYNA1 and KYNA. These results suggest an NMDA receptor mediated effect. After acute 30minute treatment HPLC measurements revealed a similar KYNA1 and a higher KYNA plasma concentration compared to cerebrospinal fluid concentrations. Finally, KYNA, KYNA1 and MK-801 showed comparable results in opioid receptor G-protein activity and ligand potency with acute in vivo treatments when they were administered in vitro for 30min on isolated cortex and striatum slices. We previously demonstrated that KYNA1 and KYNA acutely altered opioid receptor function in vivo and in vitro through the NMDA receptor depending on the opioid receptor type and brain region. This study may lead to a new, indirect approach to influence opioid receptor signaling.

Prolonged Subdural Infusion of Kynurenic Acid Is Associated with Dose-Dependent Myelin Damage in the Rat Spinal Cord

BACKGROUND

Kynurenic acid (KYNA) is the end stage metabolite of tryptophan produced mainly by astrocytes in the central nervous system (CNS). It has neuroprotective activities but can be elevated in the neuropsychiatric disorders. Toxic effects of KYNA in the CNS are unknown. The aim of this study was to assess the effect of the subdural KYNA infusion on the spinal cord in adult rats.

METHODS

A total of 42 healthy adult rats were randomly assigned into six groups and were infused for 7 days with PBS (control) or 0.0002 pmol/min, 0.01 nmol/min, 0.1 nmol/min, 1 nmol/min, and 10 nmol/min of KYNA per 7 days. The effect of KYNA on spinal cord was determined using histological and electron microscopy examination. Myelin oligodendrocyte glycoprotein (MOG) was measured in the blood serum to assess a degree of myelin damage.

RESULT

In all rats continuous long-lasting subdural KYNA infusion was associated with myelin damage and myelin loss that was increasingly widespread in a dose-depended fashion in peripheral, sub-pial areas. Damage to myelin sheaths was uniquely related to the separation of lamellae at the intraperiod line. The damaged myelin sheaths and areas with complete loss of myelin were associated with limited loss of scattered axons while vast majority of axons in affected areas were morphologically intact. The myelin loss-causing effect of KYNA occurred with no necrosis of oligodendrocytes, with locally severe astrogliosis and no cellular inflammatory response. Additionally, subdural KYNA infusion increased blood MOG concentration. Moreover, the rats infused with the highest doses of KYNA (1 and 10 nmol/min) demonstrated adverse neurological signs including weakness and quadriplegia.

CONCLUSIONS

We suggest, that subdural infusion of high dose of KYNA can be used as an experimental tool for the study of mechanisms of myelin damage and regeneration. On the other hand, the administration of low, physiologically relevant doses of KYNA may help to discover the role of KYNA in control of physiological myelination process.

Telemetry monitoring for non-invasive assessment of changes in core temperature after spinal drug administration in freely moving rats

INTRODUCTION

There are no data available about the effects of spinally administered drugs on thermoregulation in freely moving animals. The first goal of the present study was to throw light on the consequences of intrathecally administered saline as a vehicle on core temperature and motor activity in unrestrained conditions. The second goal was to characterize the effects of morphine on these parameters as a widely used antinociceptive drug in spinal anesthesia, and reveal the potential role of the N-methyl-d-aspartate (NMDA) receptors in these processes.

METHODS

For these purposes, male Wistar rats were catheterized intrathecally, and E-Mitter battery-free transponders were implanted intraabdominally to continuously monitor core temperature and locomotor activity.

RESULTS

Saline induced a short-lasting hyperactivity accompanied by significant and prolonged hyperthermia that was blunted by systemic paracetamol administration. Morphine had no impact on motor activity; however, it caused high but equivalent degree hyperthermia in a wide dose range (1-15 μg), suggesting that it reached its peak effect. In the highest applied dose (25 μg), the NMDA receptor antagonist kynurenic acid blunted the saline-induced hyperthermia, and all doses caused higher hyperactivity compared to vehicle or morphine injections. In combination, kynurenic acid significantly inhibited the morphine-induced hyperthermia.

DISCUSSION

These data suggest that this method might be a valuable tool for investigating the thermoregulatory and locomotor effects of different drugs at spinal level; however, the prolonged effects of intrathecal vehicle injections should also be considered. The results point out that morphine is a very potent hyperthermic drug that may act primarily on the efferent limb of thermoregulation, at least partially, via an indirect NMDA-receptor mediated action mechanism.

Synergistic effect of the L-tryptophan and kynurenic acid with dipyrone or paracetamol in mice

PURPOSE

Our great interest in this work was study the synergism between l-tryptophan and dipyrone or paracetamol as well as the interaction of kynurenic acid (l-tryptophan metabolite) and these analgesics agents utilizing a robust methodology.

METHODS

We performed the writhing test induced by acetic acid in mice to evaluate the antinociceptive effect of the treatments isolated and combined (p.o. and i.p.). Dose-response curves were constructed and the values of ED50 for treatment alone and combined were statistically compared. In addition, isobolographic analysis was performed and the experimental values were compared with the theoretical values for simple additive effect.

RESULTS

The combined treatment with l-tryptophan and dipyrone or paracetamol reduced significantly the ED50 of these analgesics when compared to the isolated treatments. l-tryptophan alone has no antinociceptive effect. l-Tryptophan increases the central amount of 5-HT and the synergism with dipyrone is antagonized by the 5-HT depletion. The kyna has an antinociceptive dose-related effect and a synergistic interaction with dipyrone and paracetamol verified by isobolographic analyses and confirmed by experimental values of ED50 of combined treatments were statistically lower than theoretical calculated values for simple additive effect. Melatonin antagonist receptor attenuates the antinociceptive synergism between l-tryptophan and dipyrone.

CONCLUSION

Our results demonstrate that the increased 5-HT amount on the central nervous system is not per se capable to induce antinociception. The l-tryptophan interacts synergistically with dipyrone and paracetamol both orally and by i.p. route. This effect is dependent on the biotransformation of l-tryptophan to 5-HT and involves kynurenic acid and melatonin receptors.

Antinociceptive interactions between anandamide and endomorphin-1 at the spinal level

1. Although it is well known that the combined administration of synthetic or plant-originated opioids with cannabinoids (CB) results in synergistic antinociception, the effects of combined administration of endogenous ligands acting at micro-opioid and CB receptors are not known. The aim of the present study was to determine the interaction between anandamide (AEA; a CB(1) receptor agonist) and endomorphin-1 (EM-1; a micro-opioid receptor agonist) after intrathecal administration. 2. Nociception was assessed by the paw-withdrawal test after carrageenan-induced inflammation in male Wistar rats. 3. Endomorphin-1 (16.4 pmol to 16.4 nmol) and AEA (4.3-288 nmol) alone dose-dependently decreased carrageenan-induced thermal hyperalgesia, although the highest dose of AEA also exhibited pain-inducing potential. The potency of AEA was approximately 59-fold lower than that of EM-1 (35% effective dose (ED(35)) 194.4 vs 3.3 nmol, respectively). Coadministration of these ligands revealed that combinations of 16.4 pmol EM-1 plus 28.8 or 86.5 nmol AEA were more effective than either drug alone, but other combinations were no more effective than the administration of EM-1 itself. Therefore, coadministration of AEA did not significantly shift the dose-response curve to EM-1. 4. The results of the present study indicate that the coadministration of AEA and EM-1 results in potentiated antihyperalgesia only for a combination of specific doses. Because AEA activates other receptor types (e.g. TRPV1) in addition to CB(1) receptors, the results of the present suggest that, after the coadministration of EM-1 and AEA, complex interactions ensue that may lead to different outcomes compared with those seen following the injection of exogenous ligands.

The role of TRPV1 receptors in the antinociceptive effect of anandamide at spinal level

While it is well known that the endogenous cannabinoid receptor ligand anandamide also activates the transient receptor potential vanilloid1 (TRPV1) receptors, there has been no in vivo study indicating the role of the TRPV1 receptors in the antinociceptive effect of anandamide at spinal level. The goal of this study was to determine the effect of inhibition of TRPV1 receptors by capsazepine on the antinociceptive potency of anandamide after intrathecal administration. Anandamide alone (1, 30 or 100 microg) dose-dependently decreased carrageenan-induced thermal hyperalgesia, however, the highest dose caused temporary excitation and vocalization, suggesting the pain-inducing potential of anandamide. Capsazepine (10 or 20 microg) by itself did not change the pain sensitivity markedly, but the lower dose increased it, and the higher dose decreased the antinociceptive effect of 30 microg anandamide. Furthermore, both doses of capsazepine decreased the efficacy of the largest dose of anandamide. These results show that TRPV1 receptor activation plays a substantial role in the antinociceptive effects of anandamide at spinal level. The effect of the inhibition on TRPV1 receptors depended on the dose applied. We presume that coactivation of the cannabinoid and TRPV1 receptors by anandamide provides elevated antinociception through the release of antinociceptive endogenous ligands at spinal level.

Antinociceptive interactions of triple and quadruple combinations of endogenous ligands at the spinal level

A very interesting and rapidly developing field of pain research is related to the roles of different endogenous ligands. This study determined the antinociceptive interactions of triple and quadruple combinations of different endogenous ligands (endomorphin-1, adenosine, agmatine and kynurenic acid) on carrageenan-induced inflammatory pain model at the spinal level. Intrathecal infusion (60 min) of these drugs alone, in double, triple or quadruple combinations, was followed by a 60-min observation period. During the infusion, antihyperalgesic effect of 0.3 microg/min endomorphin-1 was higher in the triple combinations than those in the double combinations. After cessation of drug administration, only the combination of 0.3 microg/min endomorphin-1, 1 microg/min agmatine, and 0.3 microg/min adenosine was more effective than the double combinations. In quadruple combinations, the antinociceptive effects of both 0.1 and 0.3 microg/min endomorphin-1 were significantly potentiated by the otherwise ineffective triple combination of adenosine, agmatine, and kynurenic acid. No side effects could be observed at these doses. These results demonstrate that triple and quadruple combinations of these endogenous ligands caused more effective antihyperalgesia compared with double combinations. Accordingly, the doses of these substances could be further reduced, thus, reinforcing the view that complex activation and/or inhibition of different systems can be sufficiently effective in blocking nociception without adverse effects. Because all of these drugs had effects on various receptors and systems, the possible types of these interactions were discussed.

Interaction of endogenous ligands mediating antinociception

It is well known that a multitude of transmitters and receptors are involved in the nociceptive system, some of them increasing and others inhibiting the pain sensation both peripherally and centrally. These substances, which include neurotransmitters, hormones, etc., can modify the activity of nerves involved in the pain pathways. Furthermore, the organism itself can express very effective antinociception under different circumstances (e.g. stress), and, during such situations, the levels of various endogenous ligands change. A very exciting field of pain research relates to the roles of endogenous ligands. Most of them have been suggested to influence pain transmission, but only a few studies have been performed on the interactions of different endogenous ligands. This review focuses on the results of antinociceptive interactions after the co-administration of endogenous ligands. The data based on 55 situations reveal that the interactions between the endogenous ligands are very different, depending on the substances, the pain tests, the species of animals and the route of administrations. It is also revealed that only a few of the possible interactions between endogenous ligands have been investigated to date, in spite of the fact that the type of antinociceptive interaction between different endogenous ligands could hardly be predicted. The results indicate that the combination of endogenous ligands should not be omitted from the pain therapy arsenal. Attention will hopefully be drawn to the complex interdependence of endogenous ligands and their potential use in clinical practice.

Effects of baclofen on spinal reflexes and persistent inward currents in motoneurons of chronic spinal rats with spasticity

In the months after spinal cord injury, motoneurons develop large voltage-dependent persistent inward currents (PICs) that cause sustained reflexes and associated muscle spasms. These muscle spasms are triggered by any excitatory postsynaptic potential (EPSP) that is long enough to activate the PICs, which take > 100 ms to activate. The PICs are composed of a persistent sodium current (Na PIC) and a persistent calcium current (Ca PIC). Considering that Ca PICs have been shown in other neurons to be inhibited by baclofen, we tested whether part of the antispastic action of baclofen was to reduce the motoneuron PICs as opposed to EPSPs. The whole sacrocaudal spinal cord from acute spinal rats and spastic chronic spinal rats (with sacral spinal transection 2 mo previously) was studied in vitro. Ventral root reflexes were recorded in response to dorsal root stimulation. Intracellular recordings were made from motoneurons, and slow voltage ramps were used to measure PICs. Chronic spinal rats exhibited large monosynaptic and long-lasting polysynaptic ventral root reflexes, and motoneurons had associated large EPSPs and PICs. Baclofen inhibited these reflexes at very low doses with a 50% inhibition (EC50) of the mono- and polysynaptic reflexes at 0.26 +/- 0.07 and 0.25 +/- 0.09 (SD) microM, respectively. Baclofen inhibited the monosynaptic reflex in acute spinal rats at even lower doses (EC50 = 0.18 +/- 0.02 microM). In chronic (and acute) spinal rats, all reflexes and EPSPs were eliminated with 1 microM baclofen with little change in motoneuron properties (PICs, input resistance, etc), suggesting that baclofen's antispastic action is presynaptic to the motoneuron. Unexpectedly, in chronic spinal rats higher doses of baclofen (20-30 microM) significantly increased the total motoneuron PIC by 31.6 +/- 12.4%. However, the Ca PIC component (measured in TTX to block the Na PIC) was significantly reduced by baclofen. Thus baclofen increased the Na PIC and decreased the Ca PIC with a net increase in total PIC. By contrast, when a PIC was induced by 5-HT (10-30 microM) in motoneurons of acute spinal rats, baclofen (20-30 microM) significantly decreased the PIC by 38.8 +/- 25.8%, primarily due to a reduction in the Ca PIC (measured in TTX), which dominated the total PIC in these acute spinal neurons. In summary, baclofen does not exert its antispastic action postsynaptically at clinically achievable doses (< 1 microM), and at higher doses (10-30 microM), baclofen unexpectedly increases motoneuron excitability (Na PIC) in chronic spinal rats.

Dose-independent antinociceptive interaction of endogenous ligands at the spinal level

Adenosine, agmatine and kynurenic acid are endogenous ligands acting on different (e.g. adenosine, NMDA, alpha(2)-adrenergic and imidazoline) receptors with a potential role in nociception at the spinal level. Their antinociceptive effects have already been investigated as monotherapy, but only a few studies have reported on their effects on the potency of other drugs. The purpose of the present study was carried out to analyse their interactions during continuous intrathecal co-administration in a carrageenan-induced thermal hyperalgesia model in rats. A paw withdrawal test was used for nociceptive testing. The intrathecal infusion (60 min) of these three drugs was administered alone or in combinations (kynurenic acid+adenosine or agmatine; adenosine+agmatine), which was followed by an additional 60-min observation period. Kynurenic acid alone was ineffective, while adenosine and agmatine alone caused a slight increase in pain threshold. However, independently of the applied doses all of the combinations significantly (p<0.05) increased the paw withdrawal latencies on the inflamed side during and after the infusion, but were almost ineffective on the normal side. The adenosine+kynurenic acid combination was the most effective: namely, that it relieved thermal hyperalgesia in all the applied dose combinations. Treatment with the kynurenic acid-containing combinations also caused dose-dependent side-effects (motor impairment and excitation), despite the fact that monotherapy with kynurenic acid in the applied dose (0.1 microg/min) did not result in adverse effects.

Spastic long-lasting reflexes of the chronic spinal rat studied in vitro

Over the months following sacral spinal cord transection in adult rats, a pronounced spasticity syndrome emerges in the affected tail musculature, where long-lasting muscle spasms can be evoked by low-threshold afferent stimulation (termed long-lasting reflex). To develop an in vitro preparation to examine the neuronal mechanisms underlying spasticity, we removed the whole sacrocaudal spinal cord of these spastic chronic spinal rats (>1 mo after S(2) sacral spinal transection) and maintained it in artificial cerebral spinal fluid in a recording chamber. The ventral roots were mounted on monopolar recording electrodes in grease, and the reflex responses to dorsal root stimulation were recorded and compared with the reflexes seen in the awake chronic spinal rat. When the dorsal roots were stimulated with a single pulse, a long-lasting reflex occurred in the ventral roots, with identical characteristics to the long-lasting reflex in the awake spastic rat tail. The reflex response was low threshold (T), short latency, long duration ( approximately 2 s), and enhanced by repeated stimulation. Brief high-frequency stimulation trains (0.5 s, 100 Hz, 1.5 x T) evoked even longer duration responses (5-10 s), with repeated bursts of activity that were similar to the repeated muscle spasms evoked in awake rats with stimulation trains or manual skin stimulation. Stimulation of a given dorsal root evoked long-lasting reflexes in both the ipsilateral and contralateral ventral roots. Long-lasting reflexes did not occur in the sacrocaudal spinal cord of acute spinal rats (S(2) transection), which is similar to the areflexia seen in awake acute spinal rats. However, long-lasting reflexes could be made to occur in the acute spinal rat by altering K(+) (7 mM) or Mg(2+) (0 mM) concentrations, or by application of high doses of the neuromodulators norepinephrine (NE, >20 microM) or serotonin (5-HT, >20 microM). In chronic spinal rats, much lower doses of these neuromodulators (0.1 microM) enhanced the long-lasting reflexes, suggesting a denervation supersensitivity to 5-HT and NE following injury. Higher doses of NE or 5-HT produced a paradoxical inhibition of the long-lasting reflexes. The high dose inhibition by NE was mimicked by the alpha(2)-adrenergic receptor agonist clonidine but not the alpha(1)-adrenergic receptor agonist methoxamine. In summary, the sacral spinal in vitro preparation offers a new approach to the study of spinal cord injury and analysis of antispastic drugs.

The Antinociceptive Potencies and Interactions of Endogenous Ligands During Continuous Intrathecal Administration: Adenosine, Agmatine, and Endomorphin-1

Recently, a series of endogenous ligands related to inhibition of sensory transduction of noxious stimuli at the spinal level has been described, including endomorphins, agmatine, and adenosine, which act on different receptors; however, little data exist concerning their effect during continuous administration or their interactions. In this study, we investigated the antinociceptive properties of continuously administered (for 60 min) adenosine and agmatine on carrageenan-induced thermal hyperalgesia by means of a thermal paw withdrawal test in awake rats. The possible interaction between endomorphin-1 and adenosine or agmatine was also determined. Continuous administration of adenosine (0.3-3 microg/min) did not influence the paw withdrawal latencies of the normal or inflamed paws during the infusion but in larger doses it resulted in a significant increase in latencies after the cessation of the infusion. Agmatine (0.3-3 microg/min) dose-dependently decreased the hyperalgesia, but the largest dose caused a temporary excitation in 50% of animals. The continuous administration of adenosine or agmatine (3 micro g/min) potentiated and prolonged the antinociceptive effect of endomorphin-1 (1 microg/min). Our results revealed that adenosine and agmatine have a small antinociceptive efficacy during continuous intrathecal administration but that both potentiate the effect of endomorphin-1. These data suggest that the combination of these endogenous ligands might represent novel targets for the therapeutic modulation of pain; however, the systematic examination of side effects is essential.

IMPLICATIONS

Adenosine and agmatine have little antinociceptive efficacy during continuous intrathecal administration, as shown by the inflammatory pain test in rats, but both potentiate the effect of endomorphin-1. These data suggest that the combination of these endogenous ligands might represent novel targets for the therapeutic modulation of pain; however, the systematic examination of side effects is essential.

Persistent sodium and calcium currents cause plateau potentials in motoneurons of chronic spinal rats

After chronic spinal cord injury motoneurons exhibit large plateau potentials (sustained depolarizations triggered by brief inputs) that play a primary role in the development of muscle spasms and spasticity (Bennett et al. 2001a,b). The present study examined the voltage-gated persistent inward currents (PICs) underlying these plateaus. Adult rats were spinalized at the S2 sacral spinal level and after 2 mo, when spasticity developed, intracellular recordings were made from motoneurons below the injury. For recording, the whole sacrocaudal spinal cord was removed and maintained in vitro in normal artificial cerebral spinal fluid (nACSF), without application of neuromodulators. During a slow triangular voltage-clamp command (ramp) a PIC was activated with a threshold of -54.2 +/- 4.8 mV (similar to plateau threshold), with a peak current of 2.88 +/- 0.95 nA and produced a pronounced negative-slope region in the V-I relation. This PIC was in part mediated by Cav1.3 L-type calcium channels because it was low threshold and significantly reduced by 10 to 20 microM nimodipine or 400 microM Cd2+. The PIC that remained during a calcium channel blockade (in Cd2+) was completely and rapidly blocked by tetrodotoxin (TTX; 0.5 to 2 microM), and thus was a TTX-sensitive persistent sodium current. This persistent sodium current was activated rapidly about 7 mV below the spike threshold (spike threshold -46.1 +/- 4.5 mV), contributed approximately 1/2 of the initial peak of the total PIC, inactivated partly to contribute only approximately 1/3 of the sustained PIC (at 5 to 10 s), and deactivated rapidly with hyperpolarization (<50 ms). When TTX was added to the bath first, the nimodipine-sensitive persistent calcium current (L-type) was seen in isolation; it was slowly activated (>250 ms), had a low but variable threshold (either slightly above or below the spike threshold), contributed the other approximately 1/2 of the initial peak of the total PIC (before TTX), did not usually inactivate with time (contributed approximately two-thirds of the sustained PIC), and deactivated slowly with hyperpolarization to rest (in >300 ms). In summary, low-threshold persistent calcium (Cav1.3) and sodium currents spontaneously develop in motoneurons of chronic spinal rats and these enable large, rapidly activated plateaus that ultimately lead to spasticity.