Summary

Introduction.  Where does NA act to alter pain behavior?

Synthesis. What is the biochemical path, including the enzymes and precursors leading to NA?

Storage and Release.

Metabolism.  Once released NA has 2 possible faiths: reuptake or degradation

Adrenergic receptors.  What are the receptors on which NA can act to produce an effect?

I- Antinociceptive at the spinal level

-        Adding NA is antinociceptive

-        The story gets complicated because NA both inhibits and facilitates spinal nociceptive reflexes

-        Spinal NA increases with acute and chronic nociceptive stimuli.

-        NA tonically induces antinociception

-        NA spinal denervation produces hyper and hypoalgesia

-        NA spinal denervation retards the appearance of autotomy and neuropathic pain

-        NA phasically induces antinociception

-        Effect of removing NA at the spinal level on pain behavior

-        Removing NA produces hypersensitivity of the noradrenergic receptors

-        NA spinal denervation blocks serotoninergic induced antinociception

II- Anatomy

-        Direct bulbo-spinal pathways: Tracing, stimulation and lesions

-        Indirect bulbo-spinal pathways

III- Transgenic mice

-        DBH -/-

-        Alpha 2A point mutation

IV- Chemicals interacting with NA

-        5-HT

-        Opioids

-        Substance P

-        Adenosine

-        Nitrous oxide

-        Sex hormones

Introduction

This is WORKING DOCUMENT, therefore we ask our readers to forgive us for it incompleteness.  We will be regularly updating it.

   One of our main interests is to study the role of noradrenaline (NA) in pain behavior.  Because of its widespread distribution in the CNS it should not be forgotten that NA serves many other functions such as motor responses at the spinal level.  Also, supraspinal NA alters processes related to aspects of pain other than nociception such as learning, memory, attention and anxiety.

NA is generally reported to alter pain behavior by its action on spinal alpha2 adrenoreceptor.  There is evidence, however, that NA acting through alpha 2 receptors has antinociceptive effects by acting both at spinal and supraspinal sites including in the locus coeruleus [30, 77] .

Noradrenaline (NA)

Synthesis

   NA is synthesized by a metabolic pathway common to all three catecholamines: Dopamine (DA), Noradrenaline (NA) or norepinephrine (NE), and Adrenaline or epinephrine.

   NA is synthesized in the CNS noradrenergic cell groups, the peripheral sympathetic neurons and the adrenal medullae.  The common precursor is the amino acid tyrosine, which comes from the diet.  Tyrosine is converted to L-DOPA by the enzyme tyrosine hydroxylase (TH), which can be blocked by drugs such as alpha-methyl-p-tyrosine (AMPT).  L-DOPA is converted to dopamine by aromatic L-amino acid decarboxylase (AADC).  In DA neurons the pathway stops here.  Both these enzymes, i.e. TH and AADC, are cytosolic (i.e. in the cytoplasm) and dopamine is taken up into vesicles by an uptake mechanism in the vesicular membrane.  In NA neurons the vesicles contain the enzyme dopamine-beta-hydroxylase (DBH), which converts dopamine to NA.  Drugs like FLA-57 can inhibit DBH.  DBH is too large to cross the cell membrane but as it is inside vesicles, it is released alongside NA and can be found in the extracellular space. 

TH is the rate-limiting enzyme in the catecholamine synthesis.  The enzymatic activity of TH is blocked when NA or adrenaline binds to it[1].  Phosphorilation of TH by cAMP-dependant protein kinase A (PKA) causes a liberation of the bound catecholamine and increase enzyme activity at physiological pH [83] .  TH can also be phosphorilated by Ca²⁺/calmodulin-dependant kinase (Cam-K II), and by extracellular signal-regulated protein kinases (ERKs).  Compare to the 2 other kinases, however, phosphorilation by ERKs causes only a small change in TH.  Protein kinase C (PKC) also plays an important role since it activates ERKs in response to neurotransmitter or hormonal stimulation.

   The most popular TH inhibitor is AMPT[2] (trade name Demser).  It competes with the TH binding site and it acts both centrally and in the periphery.  Its major side effect is sedation in most patients.

   Benserazide, 3-hydroxybenzyl-hydrazine (NSD 1015), carbidopa, and alpha-methyl-DOPA (Aldomet) inhibit L-amino acid decarboxylase (AADC).

Storage & Release

   NA is stored in storage vesicles.  The transporter that brings NA, dopamine and serotonin in synaptic vesicles is probably one and the same for all neurons, but not for chromaffin cells[3].  Reserpine[4] blocks the uptake by synaptic vesicles of catecholamines and 5-HT.

SEROTONIN (5-HT)

   When sufficiently stimulated, the vesicles migrate to the synaptic area and NA is released. As you probably already know, NA binds to the adrenergic receptors (below).  In the CNS it produces antinociception but also many other effects in the periphery such as thermogenesis, piloerection, etc. 

Metabolism

NA metabolism involves two uptake mechanisms:

Uptake 1: After stimulating the adrenergic receptors, 85-90% of the NA is taken back up into the nerve terminal and stored in vesicles or metabolized by monoamine oxidase (specifically, MAO-A) in the mitochondria.  The NA transporter is located exclusively on noradrenergic neurons and terminals [60] and has similar affinities for dopamine and norepinephrine [80, 112] . 

Uptake 2: Some of the NA diffuses away from the receptors and is transported by extra-neuronal cells by uptake 2 and metabolized by catechol-O-methyl-transferase (COMT).  COMT plays a much smaller role in catecholamine dynamics than MAO.  COMT exists in both a soluble and a membrane-bound form. The soluble form of COMT is found in organs and it does not have as high of an affinity for catecholamines as the membrane-bound form.

Tidbits: cocaine, amphetamines, methylphenidate (Ritalin for ADD), nomifensine, and tricyclic antidepressants block Uptake 1 of NA.  Progesterone increases MAO and estrogen inhibits MAO.

   That reuptake inhibitors such as desipramine have an antinociceptive effect is supported by the demonstration of an increased pain threshold in mice lacking the NA transporter [11] .  The change in threshold is modest, however, and limited to the tail flick test being absent in the hot-plate test.  This would suggest that the antinociceptive effect of increased synaptic NA occurs at the spinal level.

Adrenergic Receptors

Alpha 1a, 1b, 1d, 2a, 2a, 2c, 2d  and Beta 1 and 2.

Alpha2 receptors subtypes are believed to mediate the analgesic[5] response of NA.  Alpha2 receptor stimulation blocks adenylyl cyclase through Gi/Go proteins[6], suppresses voltage gated Ca channels, and activates inwardly rectifying K channels [57] .

The effect of NA on alpha1 receptors is mediated through increased intracellular Ca++.

Adrenoreceptor agonists

The effect of NA on alpha2 and beta-receptors are coupled to adenylyl cyclase.

Adrenoreceptor antagonists[9]

Aminergic reuptake inhibitors

Enzyme inhibitors

Neurotoxins targeting aminergic neurons

NA coexists with other neurotransmitters

In the same synaptic vescicles, NA coexists with enkephalin (cats), vasopressin (rats), neuropeptide Y (NPY, rats and humans).  NA and NPY inhibit each other’s release.

ANTINOCICEPTIVE EFFECTS OF NA AT THE SPINAL LEVEL

To show that NA at the level of the spinal cord is a neurotransmitter of the endogenous pain inhibitory system investigators have:

-            Added NA and measured the effect on nociceptive responses

-            Constant release of NA is responsible for ongoing inhibition of spinal nociceptive neurons (Tonic antinociception).

-            Determine that NA increases after a nociceptive stimulus (Phasic antinociception)

-            Removal of NA from the spinal cord alters nociceptive responses

-             

Adding NA at the spinal level produces antinociception trough alpha-2 receptors

Spinal iontophoresis of NA or intrathecal administration of a non-selective alpha agonists inhibits stimulus induced depolarization of nociceptive neurons [34, 82] [17].  Because alpha-1 agonists such as phenylephrine are without analgesic effects, the alpha-2 receptors were concluded to mediate the antinociceptive effect of NA.

Alpha-1 receptors might be involved indirectly by reducing the excitatory activity of substantia gelatinosa (SG) on spinal projection neurons.  To do this, alpha 1 receptor stimulation increase the release GABA from SG interneurons, which activates inhibitory GABAa receptors, located on SG neurons [5, 6] .

The story gets complicated because NA both inhibits and facilitates spinal nociceptive reflexes

Initially, Wiesenfeld-Hallin determined the effect of i.t. NA on the hamstring flexion reflex to subcutaneous electrical shocks was examined in unanaesthetized, decerebrate, spinalized rats.  Low doses of NA depressed and high doses facilitated the reflex.  From this she suggests that the primary effect of NA in the dorsal horn is inhibitory while in the ventral horn it is excitatory.  Furthermore, dorsal horn neurons would be more sensitive to NA than those in the ventral horn [108] .

Sakitama and colleagues confirmed this dual effect of NA with various adrenergic agonists and antagonists.  They first confirmed that low doses of NA inhibited the flexor reflex, while high doses facilitated it.  In rats pretreated with the selective alpha 2-antagonist yohimbine the effect of NA shifted from inhibition to facilitation.  Intravenous administration of prazosin, a selective alpha 1-antagonist, dose-dependently antagonized the facilitation of the group II flexor reflex induced by NA in rats pretreated with yohimbine-HCl.  The selective alpha 1-agonist methoxamine and the alpha 2-agonist clonidine facilitated and inhibited the group II flexor reflex, respectively.  The effects of clonidine and methoxamine were almost the same as those of NA at low and high doses respectively.  These results suggest that NA facilitates and inhibits the flexor reflex via alpha 1- and alpha 2- receptors, respectively [90] .

NA phasically induces antinociception

Nociceptive stimuli will increase spinal release of NA after a short stimulus [100] or after a prolong stimulus such as the formalin test [73] or a chronic peripheral mononeuropathy [91] .

In confirmation when yohimbine (alpha2 adrenergic antagonist) or methysergide, (5-HT antagonist) are administered i.t. before or after starting a formalin test, the formalin induce pain behavior is notably increased.

In another series of the experiment, the tissue of the spinal dorsal horn of tuntreated rats and post-formalin stimulated rats were sampled and analyses for levels of monoamine and one of their metabolite.  The HPLC analysis showed that formalin injection induced significant increases in NA, MHPG, serotonin, and 5-HIAA concentrations in both the ipsi- and contralateral dorsal horns [73] .

This increase in NA supports the current view that a nociceptive stimulus triggers the release in the spinal dorsal horn of antinociceptive neurotransmitters such as NA and 5-HT as part of the endogenous pain inhibitory response.  A number of results both in our lab and in other labs shed doubt on how the changes in NA and its consequence after an acute nociceptive stimulus help us understand the role and dysfunction of the endogenous nociceptive system in chronic pain.  For instance, rats in which the spinal noradrenergic innervation was removed show a hypoalgesia on the second phase of the formalin test together with a decreased in stimulus evoked Fos expression in spinal laminae V-VI [63] .  This result could suggest that spinal NA is normally hyperalgesic.

NA tonically induces antinociception

Even in the absence of any nociceptive stimulus, NA would actively inhibit nociceptive neurons.

Accordingly depleting (85%) the spinal NA using the neurotoxin 6-OHDA[18] induces hyperalgesia [82]

Accordingly, i.t. injection of an alpha-2 noradrenergic antagonists produces a dose-dependent decrease in nociceptive threshold (hyperalgesia) [51, 89] .  The potency and duration of the hyperalgesia correlates with the relative potency of the antagonists for the alpha-2 noradrenergic receptor: yohimbine > phentolamine > WB-4101 > prazosin.  From these results, Sagen and Proudfit concluded that endogenous NA, which is tonically released from bulbospinal axon terminals, may interact preferentially with noradrenergic receptors of the alpha-2 type to affect nociception [89] .  The ready is reminded that these experiments are short term (a few hours to a few days) and therefore one should be cautious in applying the results to chronic pain.

Noradrenergic spinal denervation produces hyper and hypoalgesia

In contrast to pharmacological studies, lesion studies permit to assess the consequences of lack over long periods.  Regrettably, most studies were done over a few days only, making the correlation with clinical chronic pain limited at best.  Furthermore, the results of these studies are often contradictory with each other.

For instance, Basbaum and colleagues showed that 10 days after NA denervation of the spinal cord, rats show a modest hyperalgesia with the tail flick but not with the hot-plate [63] .  A previous study obtained the opposite result i.e. an hyperalgesia on the hot plate and no effect on the tail-flick [23] .  The only difference between these two studies is that the first used anti-DBH-Sap[19] to remove NA while the other used 6-OHDA.

Even more perplexing is the results obtain after lumbar i.t. 6- OHDA in mice to selectively remove spinal NA.  At day 3 post-lesion, hyperalgesia was found in the hot-plate test, while response latency in the tail-flick test were unchanged. Furthermore hypoalgesia was determined in the formalin test.  At day 14, however, there were no more any statistically significant differences from controls in any of the tests [23] .

The authors confirm that the NA was complete by showing that uptake of 3H-NA into synaptosomes from the lumbar spinal cord was reduced by 95%.  The uptake of 14C-5-hydroxytryptamine (14C-5-HT), on the other hand was unchanged.  Synaptosomal uptake of 3H-NA and 14C-5-HT in the brain is not altered [23] .

NA (and 5HT[20]) spinal denervation retards the appearance of autotomy[21]

To determine the role of NA on denervation-induced autotomy, male rats underwent unilateral ligation and transection of the sciatic and saphenous nerves 2, 7 or 14 days after being injected i.t. with 6-OHDA or vehicle.  The development of autotomy was then monitored. Cervicothoracic (C5-T1) and lumbosacral (L1-S1) NA, dopamine and 5-HT) spinal cord levels were analyzed by HPLC. 6-OHDA treatment (20 micrograms/10 microliters) produced a rapid (from day 2) and significant (90-95%) fall in NE content only at L1-S1.  DA levels remained essentially unchanged.  No differences in monoamine levels were detected among groups injected with vehicle. The main effect of spinal noradrenergic denervation was a significant delay in the onset of autotomy in the rats injected before neurectomy [24] .

Antagonizing alpha2-adrenoreceptors unveils neuropathic pain in rats in which it is latent without affecting the hyperalgesia in those in which it is established

Atipamezole, an alpha2- adrenoceptor antagonist, produced both mechanical and cold allodynia in those rats, which had not developed clear neuropathic symptoms after nerve ligation.  The same doses (50 microg i.t. or 1 mg/kg s.c.) did not increase the severity of symptoms in rats, which had developed them.  The opioid receptor antagonist naloxone (20 microg i.t. or 1 mg/kg s.c.) had no effect on the neuropathic symptoms [110] .

Removing NA produces hypersensitivity of the noradrenergic receptors

Fourteen days after spinal NA denervation (i.p. DSP-4), a potentiation of the NA effect upon pain sensitivity was observed in mice. Both an increase in the magnitude and duration of the antinociceptive responses to i.t. NA were recorded [78] .  Upon biochemical analysis of spinal cords, it was found that DSP4-treated mice had a 80% depletion of NA, whereas DA and 5-HT were unaffected.  Radioligand binding of [3H]clonidine in membranes prepared from spinal cord, showed no differences in density of alpha 2-adrenoceptors, but the affinity had been increased, probably explaining the supersensitivity [78] . 

In confirmation, rats previously treated with NA neurotoxins (systemic DSP4[22], neonatal i.p. 6-OHDA, or i.t. 6-OHDA) have a stronger antinociceptive response to alpha-2 adrenergic agonists [3, 79, 109] .

II- ANATOMY

All spinal NA[23] is of supraspinal origin [62]

Direct bulbo-spinal pathways: The A5, A6 (locus coeruleus/subceruleus), and A7 neuronal groups provide all the NA innervation of the spinal cord

In Sprague-Dawley rats, Harlan, Bantin and Kingman, or Wistar rats, A5, A6 and A7 all send direct noradrenergic projections to the dorsal horn [14-16, 54, 102] .  A5 noradrenergic neurons spinal projections travel in the ipsilateral DLF, terminate mostly in the intermediate gray of the cervical cord, the IML of the thoracic cord, and in all layers of the lumbar cord [16] .  A6 projections travel in the VLF [14, 48, 68] and terminates in the dorsal and ventral horn as well as in the IML [58] . 

Spinal projection originate from the caudal LC [107] while the forebrain projection originates in the compact cells [61] .  The projection of A5 on the IML would not be noradrenergic [58] .  Noradrenergic projections from A7 travels through the DLF and terminates in laminae I to IV of the dorsal horn [15] .

Stimulation studies provide evidence that projections from A5, A6, A7 are antinociceptive

Electrical stimulation or glutamate injection in the A5 area produces antinociception (tail flick) which is partially reverse by i.t. naloxone and phentolamine (alpha receptor antagonist) or yohimbine (alpha 2 antagonist) but not by alpha 1 or beta antagonists [13, 64] .

Electrical or glutamate stimulation of the LC (A6) also produces antinociception (Segal and Sandberg, 1997) and inhibits nociceptive responses of neurons in lamina IV and V (Hodge et al., 1981).  There are, however, differences according to the rat specie.  While the antinociception in Harlan rats is readily reversed by i.t. yohimbine or phentolamine.  In Sasco rats, in contrast, these antagonists do not alter the antinociception produced by LC stimulation.  Furthermore, the alpha 2-antagonist, idazoxan, does not alter the antinociceptive effect of LC stimulation in either group of rats.  Thus, electrical stimulation of NA neurons in the LC that innervate the spinal cord dorsal horn (Harlan rats) produces antinociception, but stimulation of LC noradrenergic neurons that project to the ventral horn (Sasco rats) does not produce antinociception [106] . 

After ipsilateral LC/SC[24] lesions, (DSP-4) there is a significant increase in inflammation-induced spinal Fos expression, especially in the ipsilateral superficial dorsal horn (ref?).

Electrical and chemical stimulation of A7 neurons produces and inhibition of dorsal horn nociceptive neurons [39, 117] and a behavioral antinociception that can be reduced by i.t. yohimbine [64, 114] .

Lesions studies

A5.  Bilateral electrolytic lesions of the A5 nuclei produce a marked and long lasting antinociception as assessed by both the tail-flick and hot-plate tests.  Unilateral A5 lesions also produces a long-lasting elevation in hot-plate latency, but the elevation of tail-flick latency is smaller in magnitude and is only observed one day following the lesion. These finding are consistent with previous studies which have shown that blockade of the NA input to the NRM or microinjection of a NA antagonists in NRM produces antinociception.  These data indicate that neurons located in the A5 nucleus may be the origin of an NA projection to both the spinal cord and the NRM.

The elevation in tail-flick latency observed following A5 lesions is significantly attenuated by the i.t. injection of either the NA antagonist phentolamine or the serotonergic antagonist methysergide.  However, these monoaminergic antagonists do not significantly alter the elevation in hot plate latency.  Similarly, previous studies have shown that the elevation in tail-flick, but not hot-plate latency, produced by the microinjection of NA antagonists in the NRM is attenuated by the i.t. phentolamine or methysergide [88] .

When comparing the results of lesions and stimulations studies of A5, both of which produce antinociception, we are left without any explanation on the mechanisms.

A6  Bilateral LC/SC lesions, by microinjecting DSP-4, lead to an increase in inflammation-induced spinal Fos expression, especially in the ipsilateral superficial dorsal horn [104] .  See also [18, 29, 74, 106] .

- A7 (no one as done it to my knowledge)

OF NOTE.  Despite this triple noradrenergic innervation of the dorsal horn, this does not mean that A5, A6, and A7 all have the same effect on nociceptive responses.  For instance electrical stimulation of the region of A6 selectively inhibits spinal nociceptive transmission when dorsal horn neurons are excited by noxious and non-noxious stimuli.  In contrast, stimulation of A7 produces non- selective inhibition of both nociceptive and non-nociceptive responses of dorsal horn neurons of the spinal cord [117] .

OF NOTE.  Idazoxan, an alpha 2-adrenoceptor antagonist, blocks the spinal inhibition by administration of NA but not from electrical stimulation of the dorsolateral pons in 26 cats anaesthetized with sodium pentobarbitone.  The results suggest that alpha 2-adrenoceptors do not mediate inhibition of spinal nociceptive transmission from electrical stimulation of the locus coeruleus and the nucleus Kolliker- Fuse [117] .

OF NOTE.  While all of the 5-HT projections in the spinal white matter (DLF and VLF) are unmyelinated [9] , TH fibers are mostly unmyelinated in the dorsal DLF and myelinated in the ventral DLF and VLF.

OF NOTE: In Sasco Sprague-Dawley rats, each of these pontine noradrenergic cell group projects to different areas of the spinal cord in a dorso-ventral direction [17] :

A5 to the IML (pre-ganglionic) and lamina IV and VII of the spinal cord

A6 terminate mostly in the ventral horn (Sasco, female Srague-Dawley).

A7 projects to the superficial dorsal horn.

Indirect bulbo-spinal pathways: Aminergic brainstem groups are also involved in modulation of nociceptive and autonomic areas of the spinal cord through indirect projections to monoaminergic brainstem areas.

The A1 noradrenergic group has a direct NON-noradrenergic projections to the superficial dorsal horn [101, 102] .  The A1 area responds to noxious stimulation (see refs in discussion of [102] , p 91).  The neurotransmitter through which stimulation of the A1 area stimulation induced antinociception in the spinal cord would not be an opioid or acetylcholine (see refs in discussion of [102] , p 92).

A1, through its projections to A5 and C1, A1 modulates nociceptive and autonomic functions respectively [102] .  The projections to A5 would be non-adrenergic [101, 102] and reciprocal [101] .  Electrical stimulation of the caudal part (but not the rostral part) of A1 inhibits the excitation of lamina IV and V WDR neurons by peripheral C fiber stimulation (Morton et al. 1983), and the tail flick [43] .  Electrical lesions of the caudal but not the rostral part of the lateral reticular nucleus reduce tonic descending inhibition of the lamina IV and V WDR nociceptive neurons (Hall et al, 1982).  Glutamate injection in A1 produces antinociception [44] .

Rostral part of A1 projects to the IML but not to the dorsal horn (Blessing et al, 1981; McKellar and Lowey, 1982).  A1 would also act indirectly on the IML through its projections to C1 (adjacent adrenergic cell group).  C1 in the rat projects only to the IML [85, 86] .

Pharmacologic studies have shown that spinal serotoninergic receptors are involved in the antinociception from stimulation of the A1 area [26, 43] , but not from A5 [13, 64] .  The neurotransmitter involved might be NA since blocking noradrenergic receptors in the nu. raphe magnus produces antinociception [31, 32, 87] .

Bilateral destruction of the nuclei reticularis gigantocellularis (NGC) with a soma- selective excitotoxic neurotoxin, ibotenic acid, leads to an attenuation of hyperalgesia and a reduction of inflammation-induced spinal Fos expression [104] .

A2 (commissural NTS) would also be involved in antinociception.  Electrical stimulation or injection of glutamate but not morphine in A2 inhibits the tail flick [70] .

A5 is also involved in antinociception through its projection to nu. raphe magnus (NRM) but not through 5-HT receptors (I have to look more closely at this stuff).

The PAG act on adrenergic systems to produce antinociception through A7 and A5 [7] .  Bicuculline (BIC, 15 ng) microinjected into the ventrolateral PAG produced a consistent inhibition of the responses of nociceptive dorsal horn neurons. Local iontophoresis of the selective alpha2-adrenoceptor antagonists idazoxan or yohimbine but not the selective alpha1 antagonist benoxathian significantly reversed PAG-BIC-evoked inhibition. At low ejection currents, clonidine, an alpha2-adrenoceptor agonist, markedly reduced noxious heat-evoked responses but had no consistent action on the responses to iontophoresed excitatory amino acids [EAA; N-methyl-- aspartate (NMDA) or kainic acid]. At ejection currents higher than required to block descending inhibition, idazoxan potentiated responses to both heat and EAA iontophoresis. At higher ejection currents, EAA responses were inhibited by clonidine. This indicates that both presynaptic and postsynaptic alpha2 receptors are capable of inhibiting the recorded neurons. Activation of the alpha1 adrenoceptors by iontophoresis of methoxamine often led to a marked increase in the responses to kainic acid and, to a lesser extent, to NMDA iontophoresis or noxious heat. Together with previously reported work, the current experiments demonstrate that PAG neurons inhibit nociceptive dorsal horn neurons primarily through an indirect alpha2 adrenoceptor mechanism. In this same population of dorsal horn neurons, NA has a direct alpha1-mediated excitatory effect [12] .

The rostral ventral medulla (nu raphe magnus and gigantocellularis pars alpha) produces antinociception in part through a noradrenegic transmission.  The antinociception produced by injection of the cholinergic agonist carbachol in nu raphe magnus or the nu gigantocellularis pars is blocked by injection of tetracaine of cobalt chloride in the A7 area [72] .

III- Transgenic mice further our understanding of NA in pain

DBH -/-

DBH -/- mice have lost the gene coding for DBH.  They completely lack NA and adrenaline.  Of great interest, is that these mice are conditional mutants in that NA can be restored to the adrenergic terminals by administering a synthetic amino acid precursor of NA, L-threo-3,4-dihydrox-yphenylserine (DOPS).

DBH -/- mice are helping us to understand several of the key roles of NA in pain behavior.  Please see our recent article on DBH -/- mice at:

http://www.pnas.org/cgi/content/full/99/2/1029

Because DA is the endogenous precursor of NA, noradrenergic terminals release DA instead of NA in the DBH-/- mice.  As a weak agonist at the adrenergic receptors, DA may ameliorate potential phenotypes due to the absence of NA.

Alpha 2A point mutation

These mice have a normal nociceptive threshold in the tail-flick and hot-plate tests [56, 99] .  Depending on the assay, morphine has a decreased potency (intrathecal substance P) or normal potency (tail flick and hot plate) [56, 99] .  Dexmedetomidine and UK 14,304 have a net decrease potency on pain behavior in these mice [56, 99] .  Dexmedetomidine also has a decreased sedative and anesthetic effect [56] .

IV- Neurotransmitters interacting with NA at the spinal level

5HT

  Spinal NA appears an important tonic factor modulating the function of the descending 5-HT.  Spinal NA depletion in rats, via either systemic DSP4 or i.t. 6-OHDA, reverses and/or abolishes the analgesic effects of i.t. 5-HT, or the 5-HT agonists, 5-methoxy-N,N- dimethyltryptamine (5-MeODMT) and p-chloroamphetamine (PCA), in shock titration, hot plate and tail-flick tests in both rats and mice [3, 65, 67] .  In the tail-flick test the analgesia induced by 8-OH-DPAT was reversed to an hyperalgesia [2] .

Spinal 5-HT depletion, via intrathecal 5,7- dihydroxytryptamine (5,7-DHT), only attenuated 5-MeODMT-induced analgesia in the tail-flick test but potentiated the 5-MeODMT effect in the hot-plate test. Intrathecal 5,7-DHT treatment caused a drastic potentiation of NA-induced analgesia in the shock titration and tail- flick tests but not in the hot-plate test [3] , and combined NA + 5- MeODMT induced antinociception in the hot-plate and tail-flick tests [66] .

(NA mediated 5-HT release.  To be added)

NA spinal denervation blocks serotoninergic induces antinociception

Intrathecal administration of 6-OHDA abolished the antinociceptive effects of acute administration of 5-methoxy-N,N- dimethyltryptamine (5-MeODMT, 1 mg/kg, s.c.) in the hot plate, tail- flick and shock titration tests of nociception.  The antinociceptive effects of 5-MeODMT, abolished by the prior intrathecal 6-OHDA treatment, were restored by intrathecal administration (2 or 1 microgram) of NA, immediately prior to 5-MeODMT, in all three tests of nociception.  Biochemical analysis confirmed severe NA depletions (95 percent loss) in the lumbar and thoracic regions of the spinal and much lesser dopamine depletions (25-35 percent loss) [66] .

Opioids

   Substantial experimental evidence suggests that NA potentiates the antinociceptive effects of endogenous or exogenous opiates, most likely through alpha2 receptors, possibly through the alpha2A receptor, see Caron and colleagues [11] for data and other references.  Noradrenergic agonists, reuptake inhibitors and indirect-acting agonists such as amphetamine, have all been shown to synergize or potentiate the analgesic effect of opiates [28, 42, 50, 69, 75, 84] .  Lesion or inhibition of noradrenergic spinal afferents, in contrast, produces a state of acute hyperalgesia and reduced antinociceptive effects of opiates [8, 41, 89] .  For instance, depletion of spinal NA using DSP-4, attenuates morphine analgesia [118] .  Importantly, Also, neither brainstem nor spinal cord 5-HT is affected by DSP4 [118] .

To determine if the interaction of morphine and NA or 5-HT occurs at the spinal level, morphine was injected either, intrathecally (i.t.) or intraventricularly (i.c.v) in mice deficient in either NA, 5-HT or both[25].  Alternatively, morphine was injected i.c.v. and alpha2 or 5-HT1-2 antagonists were given i.t.  Results show that the antinociceptive effect of morphine depend on noradrenaline and 5-HT at the supraspinal level but not at the spinal level [4, 71] .  The antinociceptive action of morphine (s.c.) was attenuated only in animals with supraspinal depletion in 5-HT or NA [71] . 

In contrast, analgesia induced by the mu-agonist fentanyl, appears dependant on NA but not on 5-HT both at the spinal and supraspinal levels [19] .

There is also some evidence that NA induced antinociception at the level of the spinal cord is dependant on opiates.  This was demonstrated by rendering rats tolerant to spinal morphine using continuous i.t. infusion.  The antinociceptive effect of i.t. NA was found to be significantly attenuated in these opioid tolerant animals.  In contrast, no cross-tolerance is observed between morphine and 5-HT was observed [59] .

Hammond and Proudfit soot to determine the NA nucleus contributing to morphine antinociception.  They found that destruction of A7 but not A6 attenuates morphine antinociception.  Following lesions of both A7 and A6, nociceptive thresholds assessed by the tail flick and hot plate assays were not altered or by lesions of the PBV alone.  Those lesions which involved A6 altered NA content in the cortex, spinal cord and medial brain stem; however, no correlation could be demonstrated between the attenuation of morphine-induced analgesia and the changes in NA content of the brain regions examined [33] .

Data obtained in transgenic mice confirm the interaction in of NA with opiates.  First, a decreased potency of morphine was found in mice with non-functional alpha-2A adrenoreceptor [99] . 

Also, Caron and colleague [11] have studied a mouse lacking the NA transporter to study the antinociceptive effect of opioids.  The two most striking result of that study are that morphine and the warm water (33^oC) swim stress have a much greater antinociceptive effect on the tail-flick test in the knock out mice compare to the controls.  Surprisingly, these effects are not seen in the hot plate test, which lead the authors to conclude that the potentiating effect of NA on opiates (exogenous and endogenous) was occurring at the spinal level.

The conclusion that noradrenaline is necessary for opiate analgesia are challenged, by reports that the effect of morphine is unchanged when alpha-2 adrenergic receptors are blocked [97, 98] .  Also, noradrenergic denervation of the spinal cord was found to produce hypoalgesia and potentiation of morphine analgesia at longer term (14 days) [63] .  The interpretation of these results is further complicated by the previous finding that 7 to 12 days after removal of noradrenergic spinal innervation, nociceptive thresholds are unchanged while the antinociceptive effect of morphine is either reduced or unchanged depending on the nociceptive test [10, 93, 118] .

Substance P (SP) is blocked by NA

NA as well as opioids would block the release of SP and glutamate by primary nociceptive afferents.  NA also inhibits the hyperalgesic effects of SP by acting on spinal neurons (i.e. post-synaptically to the primary afferents) [20] .

Adenosine and NA interdependence

Through alpha2 receptors, NA synergizes with adenosine to produce antinociception [1, 94] .  To produce antinociception, adenosine would act through the A1 receptor and would induce NA release [27] .  The antinociceptive effect of NA would also depend on adenosine receptor stimulation since blocking adenosine receptors reduces the antinociceptive effect of NA, amytriptyline, or morphine but not of 5-HT [21, 113] .  In contrast, the hyperalgesic effect of blocking adenosine receptors[26] produces hyperalgesia through the alpha1 receptor [76] .  The latter conclusion is shaky since it is based on the observation that phenoxybenzamine, but not prazosin, blocked the hyperalgesic of theophylline.  The alpha1 antagonist might have produced a motor inhibition, which masked the hyperalgesia.

Nitrous oxide

Produces analgesia by increasing the release of NA in the spinal cord [116] .  This increase is mediate by nitrous oxide acting on supraspinal sites including the PAG.  The action of nitric oxide in the PAG is opioid dependent since it is block by local naloxone.

Sex hormones

See: Kritzer MF, Adler A, Marotta J, Smirlis T: Regionally selective effects of gonadectomy on cortical catecholamine innervation in adult male rats are most disruptive to afferents in prefrontal cortex. Cereb Cortex, 9: 507; 1999.

IV- The brainstem noradrenergic cells express sex hormone receptors?

A1 and A2 express estrogen receptors some of which double label for DBH [96] .

V. NA and serotonin act on different cell groups in the spinal cord

Noradrenergic spinal denervation (DSP-4) enhances noxious induced activity of superficial dorsal horn spino-parabrachial neurons, while serotoninergic denervation (5,7-DHT) enhances noxious induced activity of deep dorsal horn spino-parabrachial neurons [105] .

VI. noradrenergic cells and ventral medullae

Substance P (SP) neurons from the ventromedial medulla would induce antinociception by activating the A7 NA bulbospinal projection.

Stimulation of neurons located in the ventromedial medulla (VMM), including the nucleus raphe magnus (RMg), produces antinociception which appears to be mediated in part by activation of spinally- projecting noradrenergic neurons located in A7.  Retrograde tracing determined that numerous SP-immunoreactive cells in the RMg, gigantocellular reticular nucleus pars alpha and the paragigantocellular reticular nucleus project to A7 and are presumed responsible for the activation of spinal projecting noradrenergic pain inhibitory neurons [115] .

Noradrenergic excitation of raphe magnus neurons is pronociceptive.

Iontophoresis of NA in the raphe excites “on” cells [25] .  In the raphe, excitation of “on” cells facilitates spinal nociceptive reflexes [35] , therefore in the raphe, NA is pronociceptive.  The source of innervation of the raphe nuclei has not yet been reported.  The fact that i.t. noradrenergic antagonists block the antinociception produced by stimulation of the raphe, could be due to the occurrence of antidromic stimulation of A5 cells projecting to both the raphe and to the spinal cord (proposed by [63] ).  Such collateral projections have been shown for the PAG and spinal cord from A5 [53] .  The spinal noradrenergic antinociception obtained when stimulating the PAG is unlikely due to antidromic stimulation of collaterals since the antinociception produced by microinjection of glutamic acid [46] or morphine [22, 45, 111] in the PAG is reversible by i.t. alpha 2 antagonist.

X- Noradrenergic receptors do not change their expression but increase their affinity after noradrenergic depletion of the spinal cord.  (I have to complete this section)

XI. Decrease release of SP may contribute to the antinociceptive effects of NA

i.t. SP (5-20 micrograms) produces a dose-related antagonism of the effect of morphine, baclofen and NA, which persists for the entire time-course of the antinociceptive effect in each case [92] .

EXPERIMENTS TO DO

-        See if cells of the adrenergic cells from the pons send projections to both the spinal cord and the rostral ventral medullae.  

A5 neurons collateralize to the cord and PAG [53] .

-        See if adrenergic cells of the pons send projections to both the ventral and dorsal horns

-        If A5 noradrenergic cells are under GABA inhibition [55] , does bicuculine in A5 produces antinociception?  An is the noradrenergic input to A5 inhibiting GABA interneurons?

-        Is the noradrenergic input to raphe magnus on 5-HT cells or GABAergic cells, since lifting of the GABAergic inhibition in raphe magnus produces antincociception [36, 37] .

-        Check if the hypoalgesia that follows DBH-Sap i.t. is reversed by i.t. naloxone.  This is because removal of noradrenergic terminals might facilitate the release of opiates or produce denervation supersensivity of opiate receptors as well as adrenergic receptors.

-        See if cells of the locus project to both the spinal cord and the forebrain.

-        Study the long term (90 days) effect of noradrenergic denervation when re-innervation and placticity is completed [47] .

-        Because there is a regrowth (?) of noradrenergic terminals after their removal from the spinal cord