Content uploaded by John Piletz
Author content
All content in this area was uploaded by John Piletz
Content may be subject to copyright.
Terms and Conditions for Use of PDF
The provision of PDFs for authors' personal use is subject to the following Terms & Conditions:
The PDF provided is protected by copyright. All rights not specifically granted in these Terms & Conditions are expressly
reserved. Printing and storage is for scholarly research and educational and personal use. Any copyright or other notices
or disclaimers must not be removed, obscured or modified. The PDF may not be posted on an open-access website
(including personal and university sites).
The PDF may be used as follows:
• to make copies of the article for your own personal use, including for your own classroom teaching use (this includes
posting on a closed website for exclusive use by course students);
• to make copies and distribute copies (including through e-mail) of the article to research colleagues, for the personal use
by such colleagues (but not commercially or systematically, e.g. via an e-mail list or list serve);
• to present the article at a meeting or conference and to distribute copies of such paper or article to the delegates
attending the meeting;
• to include the article in full or in part in a thesis or disser tation (provided that this is not to be published commercially).
This material is the copyright of the original publisher.
Unauthorised copying and distribution is prohibited.
2007, Vol. 21, No. 11 (pp. 885-900)
ISSN: 1172-7047
Leading Article
Agmatine and its Spectrum of Activity
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
CNS Drugs 2007; 21 (11): 885-900
L
EADING
A
RTICLE
1172-7047/07/0011-0885/$44.95/0
2007 Adis Data Information BV. All rights reserved.
Agmatine
Metabolic Pathway and Spectrum of Activity in Brain
Angelos Halaris and John Plietz
Department of Psychiatry and Behavioral Neurosciences, Loyola University Stritch School of
Medicine, Loyola University Chicago, Maywood, Illinois, USA
Agmatine is an endogenous neuromodulator that, based on animal studies, has
Abstract the potential for new drug development. As an endogenous aminoguanidine
compound (1-amino-4-guanidinobutane), it is structurally unique compared with
other monoamines. Agmatine was long thought to be synthesised only in lower
life forms, until its biosynthetic pathway (decarboxylation of arginine) was
described in the mammalian brain in 1994. Human arginine decarboxylase has
been cloned and shown to have 48% identity to ornithine decarboxylase. In
neurons of the brain and spinal cord, agmatine is packaged into synaptic vesicles
and released upon neuronal depolarisation. Other evidence of a neuromodulation
role for agmatine is the presence of a specific cellular uptake mechanism and a
specific metabolic enzyme (agmatinase; which forms putrescine).
Initially, agmatine was conceptualised as an endogenous clonidine-displacing
substance of imidazoline receptors; however, it has now been established to have
affinity for several transmembrane receptors, such as α2-adrenergic, imidazoline
I1 and glutamatergic NMDA receptors. In addition to activity at these receptors,
agmatine irreversibly inhibits neuronal nitric oxide synthase and downregulates
inducible nitric oxide synthase.
Endogenous agmatine is induced in response to stress and/or inflammation.
Stressful conditions that induce agmatine include hypoxic-ischaemia and cold-
restraint stress of ulcerogenic proportion. Induction of agmatine in the brain seems
to occur in astrocytes, although neurons also synthesise agmatine. The effects
of injected agmatine in animals include anticonvulsant-, antineurotoxic- and
antidepressant-like actions. Intraperitoneal or intracerebroventricular injections of
agmatine rapidly elicit antidepressant-like behavioural changes in the rodent
forced swim test and tail suspension test. Intraperitoneal injections of agmatine
into rats and mice also elicit acute anxiolytic-like behavioural changes in the
elevated plus-maze stress test. In an animal model of acute stress disorder,
intraperitoneal agmatine injections diminish contextual fear learning. Further-
more, intraperitoneal injections of agmatine reduce alcohol and opioid depen-
dence by diminishing behaviour in a rat conditioned place preference paradigm.
Based on these findings, agmatine appears to be an endogenous neuromodu-
lator of mental stress. The possible roles and/or beneficial effects of agmatine in
stress-related disorders, such as depression, anxiety and post-traumatic stress
disorder, merit further investigation.
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
886 Halaris & Plietz
1. Historical Background to Agmatine encodes a 460-amino-acid protein that shows 48%
identity to mammalian ornithine decarboxylase
(ODC), but has no ODC enzymatic activity.[3] In cell
Two main branches of arginine metabolism have culture transfection studies, this human copy DNA
been outlined in nearly every biochemical textbook (cDNA) has been shown to yield the expected enzy-
– one yielding citrulline plus nitric oxide (NO), and matic activity and corresponding increase in cellular
the other yielding ornithine and polyamines (figure agmatine concentrations that authenticate this
1). However, a third branch of arginine metabolism, cloned protein as a mammalian ADC.[6] Further-
which produces agmatine (decarboxylated argi- more, the protein has been studied in neurons and
nine), also exists (figure 1). This third branch has astrocytes (in vivo and in vitro) by a diversity of
only recently received attention from the biomedical methods (enzymatic activity, immunoblotting, mes-
community.[1,2] Being a guanidino compound (1- senger RNA [mRNA], RNA interference), with
amino-4-guanidinobutane), agmatine has properties clear findings.[6] Hence, little doubt remains that
that uniquely distinguish it from other monoamines ADC is a mammalian enzyme that produces agma-
because of the strong basic character of its guanidine tine.
group. Because of this property, agmatine is unusual Agmatine has been shown to be packaged into
in being protonated under physiological conditions. synaptic vesicles in the brain and spinal cord,[7] and,
Agmatine is present in low (pM–nM) concentrations upon depolarisation-induced release,[7,8] it acts on
in many organs, with enrichments in certain brain transmembrane receptors.[4] The proposed effects of
regions. It is synthesised by arginine decarboxylase agmatine in the brain have included anticonvul-
(ADC; EC 4.1.1.19),[3] a mammalian enzyme at- sant,[9,10] antineurotoxic[11-13] and antidepressant ac-
tached to the outer membrane of mitochondria.[4]
tions.[14,15]
It should be noted that the agmatine branch of the
metabolic pathway of arginine has been controver- Agmatine came to the attention of most of the
sial, with its existence in mammals even being ques- biomedical community when its presence in mam-
tioned as late as 2004.[5] The matter has become malian brain tissue was discovered in 1994.[1] It was
clearer with the cloning[3] and subsequent character- initially described as a natural clonidine-displacing
isation[6] of human ADC. The cloned human ADC substance (CDS) for imidazoline receptors.[1] This
HOOC
NH
NH
NH2
NH
L-Arginine
Arginine
decarboxylase
CO2
H2N
NH NH2
NH
Arginase
Urea
NOS
NO
Agmatinase
Urea
Putrescine
ODC
Polyamines Guanido-butanoic acid
Diamine oxidase:
predominates in the periphery
2
Ornithine Citrulline
Agmatine
Fig. 1. Arginine pathways. NO = nitric oxide; NOS = nitric oxide synthases; ODC = ornithine decarboxylase.
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
Agmatine and its Spectrum of Activity 887
was of pharmacological interest because at that time In this review, special emphasis is placed on the
potential relevance of agmatine to psychiatry. We
there were no known endogenous transmitters for review studies in rats, mice and rhesus monkeys
imidazoline receptors to explain the novel actions of showing that peripherally injected agmatine crosses
clonidine (such endogenous transmitters are now the blood-brain barrier,[28] although with some limi-
known to exist, e.g. imidazoleacetic acid-ribotide). tation,[28] and that this leads to anti-stress and other
In the search for an endogenous neurotransmitter for related behaviours. For instance, low-dose peripher-
imidazoline receptors, agmatine thus emerged as the al or central injections of agmatine rapidly elicit
prime candidate. However, its relevance as a trans- antidepressant-like behavioural changes in the ro-
mitter for imidazoline receptors was soon question- dent forced swim test (FST) and tail suspension test
ed.[16,17] Skepticism firstly arose from the inability of (TST).[29-31] Intraperitoneal injections of agmatine
agmatine to mimic (or block) the antihypertensive into rats and mice also elicit acute anxiolytic-like
effects of clonidine when applied to brainstem nu- behavioural changes in the elevated plus-maze
clei where imidazoline receptors predominate.[18] (EPM) stress test.[32] In an animal model of acute
Furthermore, it was found (with variability but gen- stress disorder (ASD), intraperitoneal injections of
eral consensus) that agmatine has only modest (µM) agmatine also diminish contextual fear learning.[33]
affinity for imidazoline I2 receptors.[19] Although Scheduled intraperitoneal injections of agmatine al-
this skepticism initially hurt the imidazoline recep- so lower alcohol and opioid dependence by dimin-
tor field, it spawned new interpretations and better ishing what appears as ‘psychological’ dependence
methodologies that, in the ensuing years, led to more assessed by a conditioned place preference (CPP)
discoveries about agmatine. paradigm with rats.[34-36] All these findings support
the emerging hypothesis that agmatine is a neuro-
It is now recognised that agmatine acts on many modulator or neurotransmitter with properties that
receptors in addition to imidazoline receptors (al- relieve mental stress. Stress-related disorders (de-
though its I1 receptor activity has undergone a recent pression, anxiety, post-traumatic stress disorder,
renewal of interest[20-22]). The best studied receptors etc.) are highly prevalent in the modern world[37] and
on which agmatine acts are α2-adrenergic, I1 and new treatments are needed. We have proposed that
glutamatergic NMDA receptors.[23,24] At a physio- agmatine (or agents that enhance its availability)
logically relevant concentration, agmatine is capable could be used alone or as adjunctive pharmacother-
of irreversibly inhibiting the activity of neuronal apeutic agent(s) for these disorders.[14]
nitric oxide synthase (nNOS) by 50%.[25,26] The
initial findings of low endogenous agmatine concen- 2. Agmatine Pathway
trations have also been buttressed by findings of
higher concentrations of agmatine following various Agmatine had long been known to exist in bacte-
stressors and/or brain traumas.[20] For example, after ria and plants as a metabolic intermediate in the
hypoxic-ischaemia or cold-restraint stress of ulcer- biosynthesis of polyamines.[38] For many years, it
ogenic proportion,[20] endogenous agmatine can be was believed that ADC did not exist in higher orga-
found at >10µM in the brain. The induction of nisms and, therefore, ODC provided the only en-
agmatine synthesis by various stressors (described zyme for mammals to synthesise polyamines. A
in greater detail in section 4) has also been proposed surprise came in 1994 when the enzymatic decar-
to involve astrocytes,[27] although neurons also syn- boxylation of arginine was discovered to also occur
thesise agmatine.[6] Thus, as the cellular actions and in bovine brain to form agmatine (figure 1).[1,6] The
regulators of endogenous agmatine have been un- human form of ADC has now been cloned and
covered, they have revealed a much more dynamic characterised.[3,6] In rat brain, the highest levels of
system than when agmatine was first described as a ADC activity, immunoreactive protein and mRNA
CDS. exist in the hypothalamus, followed by the cortex,
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
888 Halaris & Plietz
Table I. Status of agmatine as a neurotransmitter or neuromodulator
Properties defining a neurotransmitter Status of agmatine References
Synthesised in neurons Arginine decarboxylase exists in neurons and glia. Agmatine 3,4,6
has been localised immunocytochemically to perikarya,
dendrites and axon terminals
Released presynaptically Released by calcium-dependent depolarisation. In at least 7,43
one situation, agmatine is co-packaged in synaptosomes
with glutamate and vasopressin
Mimics actions of nerve stimulation Not known
Drugs available to mimic or block its specific actions Possibly the imidazoline I1 receptor antagonists 21,22
Acts on receptors NMDA-R, α2-adrenoceptor, imidazoline I1, nicotinic, 4,23,24,44-48
serotonin 5-HT3 and voltage-gated receptors
Exists in specific population of neurons Distributed in various neurons, i.e. in the hypothalamus and 8,49
hippocampus
Taken up into synaptosomes Specific uptake process into cells 39,50,51
Inactivation by specific mechanism Agmatinase (ureohydrolase) 40,41,52
Other neuronal actions Inhibits neuronal nitric oxide synthase (nNOS) and inducible 25,26,53-56
nitric oxide synthase (iNOS). Also inhibits release of
noradrenaline (norepinephrine) and glutamate
with the least amounts in the locus coeruleus and should not be overlooked. Likewise, the receptors
medulla.[6] There is also now evidence for a specific and enzymes that interact with agmatine are mani-
transporter for agmatine to allow its entry into mam- fold and appear to vary across brain regions. Thus,
malian cells.[39] the agmatine pathway seems to impact on a variety
of systems.
Agmatine is catabolised to form putrescine by a
human ureohydrolase known as agmatinase, which Solid theoretical reasons exist for agmatine to
has also been cloned.[40,41] Putrescine is a protypical play a neuroprotective role. Based firstly on the
polyamine. In the liver and other peripheral tissues, observation that agmatine inhibits and downregu-
agmatine is alternatively oxidised by diamine oxi- lates inducible NOS (iNOS),[57] it has been suggest-
dase. It is then converted by aldehyde dehydroxy- ed that agmatine serves as a feed-forward inhibitor
lase to form guanido-butanoic acid, which is readily of iNOS induction.[53,58] A modified form of agma-
excreted from the body (figure 1). tine, agmatine-aldehyde, is also formed at least in
Two brain-enriched enzymes, ADC and agma- the kidney and is also a potent inhibitor of iNOS;[57]
tinase, are the molecular engines that drive the new- its role in the brain has yet to be determined. Since
ly-named ‘agmatine pathway’ of polyamine biosyn- high levels of NO are pro-inflammatory, suppres-
thesis. The agmatine pathway and the ODC pathway sion of iNOS should be anti-inflammatory. Juxta-
thus represent alternative pathways to form putres- posed with this are other, already mentioned, studies
cine and other polyamines in the brain (figure 1). that indicate that agmatine functions as a neurotrans-
mitter (table I).[4] In its apparent neurotransmitter
Fundamental intrigue arises from ADC directly capacity, agmatine acts on α2-adrenergic, I1 and
competing with nitric oxide synthases (NOSs) and NMDA receptors.[23,24] Furthermore, extracellular-
arginases. Yet, agmatine is far more than a diversion ly-applied agmatine has been shown to reduce gluta-
of substrate from NOSs and arginases, and far more mate release during seizures in rat brain.[54] If one
than a metabolic intermediate in polyamine synthe- considers these studies in concert, it is small wonder
sis.[42] Given the value placed on neurotransmitters, that agmatine has been found to be neuroprotec-
most of the excitement has been generated by the tive,[11,13] since all of these effects have been linked
fact that agmatine is released from synaptic vesicles to neuroprotection. The agmatine pathway not only
(table I).[7] Nevertheless, in terms of brain trauma seems to promote neuronal health via metabolism
and inflammation, the astrocytic source of agmatine
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
Agmatine and its Spectrum of Activity 889
of agmatine to polyamines, but also regulates neuro- and, thereafter, agmatine remains stable at –80°C for
long-term storage.[60] There are several good meth-
transmission and inflammation (probably via regu- ods for detecting agmatine by HPLC.[60-62]
lation of glutamatergic transmission, reduction
of glutamate release and inhibition of nNOS and A final obstacle has been the spectrum of sites/
iNOS). To our knowledge, these multi-faceted and receptors with which agmatine interacts, and partic-
unique aspects of the agmatine pathway have barely ularly the misperceptions this has caused. One ex-
ample is in the behavioural studies with agmatine.
been explored for their possible therapeutic poten- Given that agmatine blocks quite a number of chan-
tial and drug discovery. nels and receptors at high enough concentrations,
Despite progress and potential, research into there has been debate about whether the effects of
agmatine has been fraught with a number of obsta- agmatine in behavioural studies are specific effects.
cles that have spawned considerable debate. We Some have even wondered if the behavioural effects
wish to address three obstacles that have been cited observed with agmatine injections are not so global
commonly when debating the research findings with (i.e. affecting general ambulation, coordination or
agmatine. The first obstacle has been a lack of analgesia) as to allow no clear interpretation of the
suitable pharmacological tools. For example, no data. This suggestion that the behavioural effects of
specific inactivators (either of ADC or agmatinase, agmatine are global and, therefore, uninteresting,
or of agmatine uptake) or antagonists of agmatine has soundly been rebuffed by numerous controlled
are available to assist research into agmatine. Only studies using specific antagonists.[29-32,63,64] Conse-
very recently has it been reported[6] that the small quently, there appears to be reasonable evidence of
interfering RNA (siRNA) approach is useful in receptor specificity to the behavioural effects of
downregulating ADC. The elucidation of agma- agmatine (more description is provided in section
tinergic transmission has thus largely progressed by 5). With these three obstacles in mind, we now turn
studying either the effects of agmatine treatments, to the concept of agmatinergic transmission.
i.e. after agmatine injection, or by coadministration
with various types of antagonists that block other 3. Agmatinergic Transmission
receptors linked to agmatine. We agree that the lack
of suitable tools has been a serious limitation, partic- In the original report describing brain agmatine,
ularly when trying to determine the actual physio- Li and coworkers[1] noted that agmatine binds to two
logical role of agmatine. classes of receptors, the α2-adrenergic and imidazo-
A second obstacle has been the short half-life of line receptors. They further found that the concen-
agmatine in most, but not all, tissues.[59] Because of tration of agmatine in brain was comparable to that
this, many groups have found it difficult to detect of other neurotransmitters.[1] These findings raised
endogenous agmatine by conventional methods, i.e. the intriguing question of whether agmatine was a
high-performance liquid chromatography (HPLC). neurotransmitter.
In our experience, a sizeable decomposition of The case that agmatine is indeed a neurotransmit-
agmatine occurs beyond 20 minutes from collecting ter is now fairly convincing (table I). It has been
blood or whole brain on ice. This decomposition established that agmatine is: (i) synthesised by a
even occurred in some samples that had been frozen unique enzyme, ADC;[3] (ii) stored in perikarya of a
rapidly after dissection (at –80°C). To overcome this specific population of central neurons;[49] (iii) taken-
problem, our laboratory instituted the immediate up into synaptosomes;[50] (iv) found in small vesi-
extraction of agmatine from blood and tissues (with- cles of axon terminals,[8] where it is co-localised
in 15 minutes of collection).[60] The extraction with classical neurotransmitters such as glutamate
method we currently use involves removal of pro- and arginine vasopressin;[43] (v) released from
teins into ice-cold ethanol.[61] Other extraction pro- synaptosomes by calcium-dependent depolarisa-
cedures (i.e. with trifluoracetic acid) also work well tion;[7] (vi) subject to a specific uptake process;[39]
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
890 Halaris & Plietz
(vii) inactivated by a specific enzyme known as exerts inhibitory actions on certain voltage-gated
agmatinase;[52] (viii) a high-affinity ligand at cell- channel receptors.[46]
surface receptors (i.e. NMDA, α2-adrenergic and 3.2 Imidazoline Receptors
imidazoline receptors);[44] and (ix) a modulator of
receptors in peripheral[65,66] and central[45] nerves. It Agmatine has high affinity for imidazoline recep-
is not yet known if agmatine has its own receptor in tors,[44] yet the classical antihypertensive responses
the classical sense, and thus it may only be active in of imidazoline receptors seem largely unaffected by
the face of ongoing synaptic activity, which is why agmatine.[69] This has created an obvious dilemma,
the term ‘neuromodulator’ is used to describe it. but the ability of agmatine to diminish opioid depen-
Furthermore, it is not yet known if its actions are dence appears to be mediated by the I1 receptor.[21,36]
primarily presynaptic or postsynaptic, or both. The evidence that agmatine acts through the I1 re-
ADC is 2- to 3-fold enriched in the hypothalamus ceptor comes from a tissue culture model system of
over other brain regions.[6] Hypothalamic agmatine opioid tolerance, assessed by cyclic adenosine
is co-released with vasopressin and oxytocin.[43] monophosphate (cAMP) overshoot responses, using
Some studies have also used a specific antibody co-transfection of IRAS (the I1 receptor clone) and
against agmatine to study its regional distribution the µ-opioid receptor.[22,44]
and, in so doing, found that glial, endocrine and
endothelial cells in the brain contain agmatine.[67,68] 3.3 α2-Adrenergic Receptors
ADC is also widely and unevenly distributed in Agmatine also has moderate (µM) affinity for
various cell types, including astrocytes.[6,49] These α2-adrenergic receptors.[44] An example of the ac-
findings have led to the conceptualisation of an tion of agmatine on α2-adrenergic receptors is in
‘agmatinergic transmission system’.[8] Herein, we terms of morphine-induced CPP.[70] Non-effective
will use this term not only to refer to the agmatine low doses of morphine (0.1, 0.05 and 0.01 mg/kg)
pathway (figure 1) but also to subsume receptors on per se are only able to induce CPP in mice pretreat-
neurons (and maybe on glia) that mediate the func- ed with agmatine 1, 5 and 10 mg/kg, respective-
tion(s) of this unique molecule. ly. Concomitant administration of brimonidine
The main receptors of agmatine are listed below. (UK-14304) [0.5 mg/kg], a highly selective α2-adre-
When reviewing this list, the reader is advised to noceptor agonist, with per se non-effective dose of
bear in mind that coincidence detection by multiple morphine (0.5 mg/kg), and also its combination with
receptors is a theoretical possibility, i.e. agmatine non-effective dose of agmatine (1 mg/kg) plus mor-
could simultaneously activate different receptors on phine (0.05 mg/kg), also produces significant CPP.
a single neuron (with an integrated output). It cannot Agmatine or brimonidine (0.05, 0.5 mg/kg), alone or
be ruled out that interactions with multiple receptors in combination, had no effects on CPP. The selec-
could explain at least some of the actions of agma- tive α2-adrenoceptor antagonists, yohimbine (0.005
tine. mg/kg) and RX-821002 (0.1, 0.5 mg/kg), block the
CPP induced by concomitant administration of
3.1 Ligand-Gated Channels agmatine (5 mg/kg) and morphine (0.05 mg/kg).
Together, these results indicate that pretreatment of
Agmatine is known to antagonise certain ligand- mice with agmatine enhances the rewarding proper-
gated calcium channels[46] (the channels themselves ties of morphine via a mechanism that involves
are ligand-gated by other ligands). These receptors α2-adrenergic receptors.[70]
include cholinergic nicotinic, serotonergic 5-HT3
and NMDA receptor channels.[23,47,48] Within the 3.4 Catecholamine Release
NMDA-glutamatergic complex, the blockade by
agmatine occurs at a site that is distinct from the Agmatine acts as a secretagogue, facilitating the
polyamine modulatory site.[46] In addition, agmatine release of catecholamines from adrenal chromaffin
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
Agmatine and its Spectrum of Activity 891
cells,[1] insulin from pancreatic β cells,[71] and cumulation of brain polyamines, while long-term
leutenising hormone-releasing hormone (LHRH) inhibition of polyamine synthesis depletes brain
from the hypothalamus in vitro and in vivo.[72] Con- polyamines and can result in altered emotional reac-
versely, agmatine inhibits the release of nora- tivity to stressors.[76] Furthermore, the brain PSR, in
drenaline (norepinephrine) from presynaptic nerve contrast to the peripheral PSR, can be blocked by
terminals,[73] and lowers extracellular levels of glu- long-term, but not short-term, treatment with lithi-
tamate during pentylenetetrazole-induced seizures um, an effective antimanic agent.[77] In addition,
in rat brain.[54] Thus, agmatine modulates the release a maladaptive PSR has been implicated in affec-
of other neurotransmitters, which produces an indi- tive disorders, based on the study by Elgun and
rect effect on neurotransmission. Kumbasar.[78] They showed that the enzymatic con-
version of arginine to ornithine is markedly elevated
3.5 Nitric Oxide Synthase in the serum of depressed patients, which is consis-
tent with elevated PSR and similar to our report of
The modulatory effect of agmatine on NO syn- elevated plasma agmatine levels in cases of depres-
thesis has also been studied by a number of laborato- sion.[79] The induction of agmatine synthesis and/or
ries.[25,26,53,55] Because the earliest reports showed enhanced agmatine release may, therefore, be
that the inhibitory activity of agmatine was low viewed as another arm of the PSR.
(high inhibition constant [Ki] values for NOS), this Is there an ‘agmatine stress response’ analogous
finding initially received little attention.[53] Howev- to the PSR? Recent studies have shown that mam-
er, in 2001, a careful kinetic study was published by
Demady et al.[26] showing that agmatine is actually malian ADC can be down- and upregulated.[27,80]
an irreversible inactivator of nNOS. It is now real- For instance, in rat astrocytes and a macrophage cell
ised that agmatine in concentrations in the low µMline (RAW 264.7), the activity of ADC has been
range irreversibly inactivates nNOS [26] and lowers reported to be downregulated by treatment with the
the level of iNOS.[27] Inactivation of nNOS occurs at anti-inflammatory cytokine interleukin-10, and up-
about the same concentration as agmatine is found regulated with transforming growth factor-β and
in brain (low µM range).[55,74] Agmatine thus lipopolysaccharide.[27,80] These effects on ADC ac-
downregulates this enzyme. tivity mediate changes in agmatine levels, and are
dose-dependent and reversible.[27,80]
4. Endogenous Agmatine: What The first in vivo evidence of mammalian ADC
Regulatory Mechanisms Underlie its regulation was provided by Gilad and coworkers,[81]
Beneficial Effects? who showed that ADC and ODC are upregulated
following global ischaemia in adult rat brain. We
If endogenous agmatine naturally plays a have extended the findings of Gilad et al.[81] to a
neuromodulatory role, then its concentration should paradigm of hypoxic-ischaemia in newborn rat
be tightly regulated. Transient alterations are al- pups, and found that from 5 minutes to 3 hours after
ready known to occur in polyamine metabolism in hypoxic-ischaemia the brains of the rat pups under-
response to a variety of stressors. This is the poly- go a 2- to 3-fold rise in endogenous brain agmatine
amine stress response (PSR).[75] The PSR occurs in levels.[82] Since the hypoxic-ischaemia paradigm did
many tissues (e.g. brain and liver). It occurs in not immediately cause histopathological changes in
response to physical, emotional and hormonal stres- either cerebral hemisphere, and yet the agmatine
sors, with a magnitude related to the intensity of elevations were seen in both hemispheres, we con-
stress.[76] The PSR is thought to be critical for cell cluded that agmatine is upregulated in response to
survival because polyamines are essential for cell brain stress rather than as a result of tissue damage.
growth. Activation of ODC and the consequential Additional evidence for the regulation of ADC is
rise in putrescine levels are two classic markers of
the PSR.[75] Severe continuous stress leads to ac- found in the opioid literature.[83] Following 3 days of
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
892 Halaris & Plietz
morphine exposure, the endogenous levels of agma- of rats leads to a marked rise in cortical and plasma
agmatine levels. Cold-restraint is a stress model of
tine and ADC activity were significantly reduced in ulcerogenic proportion. Interestingly, we did not
rat brain.[84] Fairbanks and coworkers[83] probed this find a rise in cortical agmatine levels with less
phenomenon with a selective anti-agmatine anti- severe experimental conditions of 2- or 4-hour re-
body (IgG) injected into the spinal cord to study straint at room temperature. The fact that room
antinociceptive tolerance. Their suggestion was that temperature restraint stress was insufficient to raise
endogenous agmatinergic tone is downregulated by brain agmatine levels may indicate that mild stres-
opioids, and that sequestration of agmatine by anti- sors are not able to cause a rise in agmatine levels.
agmatine IgG enhanced opioid tolerance and, per- We have conceptualised our findings from cold-
haps, other forms of neuronal plasticity induced by restraint stress to be aligned with the pendulum
morphine. hypothesis of agmatine regulation.
Satriano and coworkers[57] have proposed a ‘pen-
dulum hypothesis’ concerning regulation of mam- 5. Possible Therapeutic Effects
malian ADC and agmatine in times of trauma. Their of Agmatine
hypothesis is based on evidence for ADC upregula- The therapeutic potential of agmatine may fall
tion and by the inhibitory actions of agmatine on into several realms, as outlined in sections 5.1–5.5.
iNOS (also mentioned in section 3.5). Their concept
is that the capacity to up- and, presumably, down- 5.1 Neuroprotection
regulate ADC allows for a pendulum (not over-
shooting or undershooting) of actions. This hypothe- In animal studies, agmatine, administered
sis starts with ADC being at a low basal state in intrathecally or systemically, reduces neuronal loss
healthy tissue at the time of trauma. This allows for produced by excitotoxins[23] or ischaemia.[82,85,86]
unabated iNOS activation immediately after trauma For instance, we performed a rat pup study[82] in
because of the low level of agmatine, thus leading to which we showed that intraperitoneally injected
a rapid surge in NO. The elevation in NO promotes agmatine (50 or 100 mg/kg) suppresses the inflam-
vasodilation and activating of defences, but the chal- matory rise in brain nitrite(s) [a measure of NO
lenge for the tissue is to make the rise in NO levels], an effect that was detected up to 6 hours
moderate enough and short-lived enough to avoid after hypoxic ischaemia. This resulted in a 50%
neurotoxicity. Sustained high levels of NO are neu- reduction in brain damage assessed 22 days later by
rotoxic because of calcium influxes that occur in brain weights and histological damage scores.[82]
cells, as NMDA channels open during this state. As Theoretically, this beneficial effect of agmatine
the pendulum hypothesis further states, a control for could relate to the ability of agmatine to inhibit
this system would be accomplished as ADC/agma- iNOS or to block NMDA receptors and/or voltage-
tine production becomes upregulated several hours dependent calcium channels. Blockade of either of
after the start of trauma (part of the PSR). This rise these signalling proteins would likely be neuropro-
in agmatine would act to inhibit iNOS and lower tective. Thus, agmatine might have potential for
glutamate release, and also block NMDA receptors. treating brain traumas of various kinds.
To our knowledge, Satriano and coworkers[57] 5.2 Pain Control and Prevention of
have only studied the agmatine pendulum concept in Morphine Tolerance
peripheral tissues. However, the same concept
should apply to the brain, assuming brain ADC is Intrathecally administered agmatine has been
regulated by stress and/or trauma. One method of shown to relieve experimentally-induced pain.[87,88]
assessing this concept in the brain might be a study For instance, a study by Aricioglu et al.[89] assessed
of restraint stress (rats held in a restrainer).[20] Using acute and mononeuropathic pain after chronic con-
this model, we have shown that 4-hour cold-restraint striction injury (CCI) in rats. CCI was created by
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
Agmatine and its Spectrum of Activity 893
four loose ligations around the right sciatic nerve. morphine plus agmatine reduced the development of
tolerance to the analgesic effect of morphine.
The analgesic threshold in rats was evaluated by
thermal hyperalgesia/allodynia at 4°C. The evalua- 5.3 Prevention of Opioid Addiction
tions were made preoperatively, on postoperative
day 15 and after drug administration. Agmatine (10, One study has evaluated the effect of agmatine on
20, 40, 80 and 100 mg/kg) was administered in- ‘psychological’ dependence induced by morphine in
traperitoneally for 5 days, beginning on postopera- the CPP assay in rats.[36] Agmatine (0.75–20 mg/kg,
tive day 15. Agmatine significantly reduced the subcutaneously) coadministered with morphine dur-
hyperalgesia in all doses applied. Intraperitoneal ing conditioning sessions completely abolished the
agmatine (10, 20, 40, 80, and 100 mg/kg studied acquisition of morphine-induced CPP in rats. Agma-
over 10–90 minutes) also increased the nociceptive tine administered on the test day also inhibited the
threshold in the tail-immersion test in a dose-depen- expression of place preference. After 30 days of
dent manner, but had no effect in the hot-plate test. extinction of CPP, agmatine, administered at 2.5 and
This effect of agmatine in the tail-immersion test 40 mg/kg, inhibited the priming effect of morphine
at 0.5 mg/kg on the place preference.
was blocked by both yohimbine (1 mg/kg) and Unfortunately, these interesting findings in rats
idazoxan (0.5 mg/kg), suggesting an α2-adrenergic- have been complicated by the fact that in a CPP
mediated mechanism. When agmatine was adminis- study of mice the effects of agmatine were in the
tered intracerebroventricularly (25–200 µg/10µLopposite direction, i.e. agmatine was reported to
studied over 10–90 minutes) it increased the noci- potentiate morphine-induced CPP.[70] The reason for
ceptive threshold in the hot plate but not in the tail- this is unknown; however, one might speculate that
immersion test.[89] These authors concluded that rats and mice differ in the way that opiates are
agmatine can modulate both acute and chronic pain. ‘psychologically preferred’, but more work is
Other studies have emphasised agmatine being needed to confirm this. Since place preference is a
more potent in relieving neuropathic pain than acute complex behaviour, these findings merit further in-
pain. Fairbanks and colleagues[87] reported that vestigation.
agmatine, exogenously administered to rodents, de-
creases hyperalgesia accompanying inflammation, 5.4 Attenuation of Drug
normalises the mechanical hypersensitivity (allody- Withdrawal Symptoms
nia/hyperalgesia) produced by chemical or mechani- Aricioglu-Kartal and Uzbay[92] have also shown
cal nerve injury, and reduces autotomy-like beha- that 72 hours after subcutaneous implantation of
viour and lesion size after excitotoxic spinal cord pellets of morphine in rats, a single injection of
injury, and agmatine produces these effects in the agmatine (20–40 mg/kg intraperitoneally) lead to
absence of antinociceptive effects in acute pain attenuation of all signs of naloxone-precipitated ab-
tests. stinence syndrome. This effect included decreased
Agmatine administered for pain relief also ap- jumping, wet dog shakes, writhing, defecation, pto-
pears to block the development of tolerance to mor- sis and teeth chattering. Li et al.[91] reported similar
phine.[90] A report by Li et al.[91] showed that agma- results with agmatine (2.5–40.0 mg/kg subcutane-
tine injected into mice at 0.125–2.5 mg/kg intraperi- ously) in mice.
toneally prevented the development of tolerance to Uzbay et al.[34] showed similar attenuation of
morphine in a dose-dependent manner. Pretreatment ethanol withdrawal symptoms in rats after 21 days
of the mice with morphine induced a >3-fold in- on a liquid alcohol diet (7.2% v/v) by a single dose
crease in the analgesic ED50 of an acute dose of of agmatine (20–160 mg/kg intraperitoneally) ad-
morphine, compared with the ED50 in mice pretreat- ministered 30 minutes before withdrawal. This in-
ed with saline. However, pretreatment of mice with cluded attenuation of stereotypical behaviour, wet
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
894 Halaris & Plietz
5.5.2 Antidepressant and/or Anxiolytic Effects
dog shakes and tremors. Control behavioural studies Zomkowski et al.[29] were the first to propose that
indicated that agmatine alone had no effect on gen- agmatine might be used as an antidepressant agent.
eral locomotion or coordination (for instance, on an In 2002, they reported that intraperitoneal agmatine
Accelerod performance test[34]). produces antidepressant-like effects in two stress
Thus, the emerging picture is that agmatine atten- models in mice – the FST and the TST. These
uates in a dose-dependent manner all drug with- effects were observed at low to moderate doses of
drawal symptoms. No obvious behavioural side ef- agmatine (0.01–50 mg/kg intraperitoneally) without
fects have been observed in rats and mice. accompanying changes in mouse ambulation in the
open field. They found that intracerebroventricular
5.5 Relief of Stress-Related Disorders injections of agmatine (1–100 nmol/lateral ventricle
site) also produced antidepressant-like effects in the
FST.
5.5.1 Relief of Contextual Fear Symptoms Aricioglu and Altunbas[30] have reported that an
There have also been reports from one research intraperitoneal injection of agmatine also produces
team[33,93,94] on the behavioural effects of agmatine antidepressant-like effects in the rat FST. Rats were
in a contextual fear paradigm in rats. The paradigm injected with various doses of agmatine (10–100
has face validity as a model of ASD in humans. The mg/kg) or imipramine (30 mg/kg), a known antide-
paradigm involves Pavlovian conditioning to a sin- pressant, according to the 2-day model of the classic
gle peripheral shock. In the paradigm, the cage in rat FST.[95,96] Results showed that all doses of agma-
which the rats are shocked becomes the conditioned tine were as effective as imipramine at decreasing
stimulus. The rat is returned to this cage 26 hours immobility time. The data did not display the dose
after the initial stimulus and defensive freezing is responsivity of the earlier mouse FST data of
assessed. In controlled experiments, the research Zomkowski et al.,[29] but it is likely that if lower
team showed that ascending agmatine doses did not doses of agmatine had been used (i.e. 0.01, 0.1 and
affect baseline motor activity or exert analgesic ef- 1.0 mg/kg intraperitoneally), dose responsivity
fects in the rats as response thresholds to the primary would have been observed in the rat FST. As in the
stimulus. However, intraperitoneal injections of study by Zomkowski et al.,[29] Aricioglu and Al-
agmatine (5 and 10 mg/kg) prior to contextual fear tunbas[30] observed no changes in ambulation at any
conditioning, or up to 5 hours afterwards, impaired dose of agmatine. The data are in keeping with the
the acquisition of contextual fear, as measured by known properties of antidepressants, or some anxi-
defensive freezing. Thus, there was an anti-contex- olytics, in both mouse and rat FSTs.
tual fear effect of agmatine not related to an artifact Agmatine has also been studied in an anxiolytic
of less pain from the shock or a general effect on screening paradigm[30,32] using Sprague-Dawley
motor activity. The research team implicitly sug- rats. In the EPM paradigm, agmatine 40 mg/kg
gested that “low agmatine concentrations in the hip- (intraperitoneally) was found to be equipotent to the
pocampus may permit the acquisition and/or consol- positive control (a benzodiazepine injection).[30]
idation of contextual fear stimuli”.[33,93,94] Unfortu- However, two higher doses of agmatine (80 and 100
nately, their speculation on this brain region has not mg/kg) were as ineffective in the EPM as saline
been accompanied by measurements of endogenous control injections. To explain this, one might re-
rat brain agmatine levels or by use of pharmacologi- member that agmatine at high doses can be toxic,
cal blocker(s) to determine by what mechanism(s) causing uraemia. This concern comes from investi-
agmatine may reduce defensive freezing response. gating the effects of varying doses of agmatine
The series of reports from this team should be repli- administered intravenously to a rhesus monkey.[28]
cated using other models of contextual fear and The monkey appeared lethargic and lost weight for a
measuring endogenous agmatine levels.[33,93,94] few days after receiving high doses of agmatine. It
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
Agmatine and its Spectrum of Activity 895
Clinical Effects in Major Depressive Disorder
was also found that plasma urea levels were high To date, our laboratory is the only site to collect
with doses of agmatine comparable to the 80–100 human data on agmatine. Based on a long-standing
mg/kg range administered to rats. Therefore, it may interest in the neurobiology of depressive illness, we
be speculated that only lower doses of agmatine will began by measuring endogenous plasma agmatine
reveal true positive effects in the EPM paradigm. levels in depressed patients before and after antide-
To elucidate the antidepressant-like mechanisms pressant treatments. In the first study,[79] assessment
of action of agmatine in mice, Zomkowski and of plasma agmatine levels in untreated depressed
colleagues[29] have performed a number of patients (n = 16) revealed an 84% statistically signif-
coadministration studies. First, they reported that icant elevation compared with healthy controls
agmatine significantly enhanced the anti-immobility (n = 8) matched to the patients for age and gender
effect of imipramine in the FST and TST. They then (p = 0.004).[79] In a follow-up study of comparable
reported that agmatine did not modify the design (n = 13), we noted a 24% statistically non-
behavioural consequences of dizocilpine (MK-801), significant elevation in agmatine levels pretreatment
a known NMDA receptor antagonist.[29] Zomkoski (Halaris A, unpublished observation). Because de-
et al.[29] also found that pretreatment with yohimbine pression is closely associated with stress, an eleva-
(a mixed α2-adrenoceptor antagonist and 5-HT1A tion in agmatine levels tends to support the hypothe-
receptor agonist) reversed the effect of agmatine on sis that agmatine may be induced as part of the stress
immobility in the FST. This suggested that the response. On the other hand, the lack of statistical
behavioural effect of agmatine might be mediated significance in our follow-up findings casts doubt on
the hypothesis. However, these two studies, even
by either α2-adrenergic or 5-HT1A pathways, be- when combined, might be under-powered as the
cause yohimbine binds to both these receptors in total number of subjects is still low (n = 29). There is
vivo. A subsequent report from the same group[63]
also the possibility that plasma agmatine levels do
clarified this matter, since two 5-HT1A receptor not accurately reflect brain agmatine levels; the lat-
antagonists (WAY 100635 and NAN-190) also ter may be robustly elevated, compared with plasma
blocked the behavioural effects of agmatine in the levels, in depression. A direct study of human brain
mouse FST.[63] A third report from the same group agmatine levels would be technically difficult be-
linked part of the antidepressant effect of agmatine cause it would have to be limited to the study of
in the mouse FST to δ- and µ-opioid receptors.[64] postmortem brain tissue. Because ornithine and
Similar antagonist studies have not been replicated agmatine degrade postmortem, the measurement of
using the rat FST model. agmatine might be problematic.
Of course, it may seem surprising that agmatine Our other clinical studies with antidepressants
is active in both the FST and the EPM.[30,32] Typical- have been somewhat clearer.[79] The effect of
ly, antidepressants are ineffective in the EPM, this venlafaxine on plasma agmatine levels in 14 de-
model being a screen for anxiolytics. Some anxi- pressed patients has been particularly robust
olytics have an effect in the FST, but these agents (p = 0.02) [Halaris A, unpublished observation]. The
also alter ambulation in an open field, which agma- level of agmatine fell in response to venlafaxine by
tine does not. Thus, agmatine seems unusual in 50% compared with pretreatment baseline after
being active in both the FST and EPM without 4 weeks of treatment (Halaris A, unpublished obser-
detriments in open-field ambulation. To date, no one vation). By comparison, bupropion lowered agma-
has compared agmatine to a stimulant (another pos- tine levels in 12 depressed patients by 20% com-
sible reason for a false positive in the FST), nor has pared with pretreatment baseline after 8 weeks of
anyone reported antagonist studies to investigate treatment (p = 0.17). A possible interpretation of
which receptors may be involved in the effects of these post-treatment results is that agmatine repre-
agmatine in the EPM. sents a homeostatic relief response to psychological
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
896 Halaris & Plietz
stress. The implication is that relief was achieved temic injection, the optimally active dose of agma-
because of successful antidepressant therapy and, tine in rodents has been reported to be reasonably
consequently, the levels of endogenous agmatine low (10 mg/kg intraperitoneally).[29] Ostensibly, not
declined.[14] Since plasma agmatine levels declined much urea should be formed from this low dose of
after patients responded to treatment, and all pa- agmatine; however, if the dose were to increase,
tients responded to venlafaxine, the stress associated intentionally or accidentally, or if intake were to be
with depressive illness should have been relieved. prolonged, uraemia may become a problem. Fur-
We are awaiting the completion of a third ongoing thermore, the rapid peripheral metabolism of agma-
study (of escitalopram) before a complete under- tine[28] could pose a problem for steady-state dosing.
standing of these findings is achieved. Since all known antidepressants and anxiolytics
must be taken for long periods of time, the same
would likely be required for agmatine. Long-term
6. Making Clinical Use of Agmatine administration of a short-lived compound such as
A pertinent question is whether agmatine should agmatine could lead to unforeseen situations where
be studied in clinical drug trials for its possible the levels of agmatine are incorrectly maintained.
psychotropic profile in humans, more specifically its Thus, the long-term use of agmatine seems likely to
potential antidepressant and/or anxiolytic proper- lead to problems and challenges. For instance, one
ties. Since agmatine is a naturally occurring sub- study of acute agmatine injection (30 mg/kg intrave-
stance, it could alternatively be packaged and sold as nously) in a rhesus monkey has shown the com-
a health food supplement with an indication for pound to cause transient uraemia.[28]
these disorders. Agmatine is absorbed from the gas- Another clinical concern has been raised by Abe
trointestinal tract in a manner consistent with the et al.[56] Their study showed that agmatine at high
existence of a specific transporter.[51] In principle, concentrations (200–800µM) in the presence of ele-
we support both clinical and dietary trials with vated potassium levels, as occurs in certain disease
agmatine. states (>20mM), is neurotoxic to cerebellar granule
What shortcomings might there be for consuming neurons. This neurotoxic effect of agmatine at high
agmatine in large amounts? Let us consider the case concentrations was attributed to enhanced glutamate
of arginine, which is classified as a semi-essential or release.[56] This proposed mechanism could also ac-
conditionally essential amino acid, depending on count for U-shaped dose response curves reported in
diet, developmental stage and health status of the vivo for agmatine in the EPM in rats.[30] We also
individual. Since agmatine is a direct metabolite of found a U-shaped dose response curve for the ef-
arginine, it is reasonable to assume that endogenous fects of agmatine in the rat pup model of hypoxic
agmatine could also be conditional, meaning it ischaemia.[82] In that study we found agmatine was
would be influenced by diet and highly regulated, as neuroprotective only between 50–100 mg/kg in-
is the case with arginine.[51] Under certain patho- traperitoneally, but not at 150 mg/kg intraperitoneal-
physiological conditions, the de novo synthesis of ly.[82]
agmatine could also become insufficient to keep up Thus, while agmatine may eventually prove to be
with demand, as occurs with arginine. If endogenous useful for treating stress-related disorders, we be-
agmatine levels became exceedingly low, a proper lieve its dosage will be problematic because of the
dose of exogenous agmatine might be difficult to likely narrow therapeutic window this substance
titrate. In other words, the effect of taking agmatine may possess. As a result of these considerations, we
dietarily may vary under different human dietary or began searching for an agmatinase inhibitor.
disease states that are related to arginine availabili-
ty.[51] Besides this, the breakdown process of agma- Why would an agmatinase inhibitor be a better
tine is also complicated because of formation of a antidepressant and/or anxiolytic agent than agma-
toxic metabolite, urea. By means of studies of sys- tine itself? From a theoretical perspective, an ideal
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
Agmatine and its Spectrum of Activity 897
antidepressant or anxiolytic agent should target just State University by our collaborator (Dr Ken Lee),
namely 3-aminopropylguanidine and trans-4-
the brain region(s) affected by the stressor and/or aminocyclohexyl guanidine. By QSAR modelling
involved in the stress response that is intricately of the agmatinase inhibitory data, a best fit equation
related to the clinical condition, such as major de- for the percentage of remaining agmatinase activity
pression. Of course, to identify and target only cer- was arrived at:
tain brain region(s) affected by a given stressor, one 0.3225 D + 72.76 D1916 + 64.97 D1920 – 192.58
would need to know the stress response pathway and H21 – 253.09, r = 0.89.
a way to selectively target that pathway. We believe
the agmatine branch of the PSR is a candidate for In this computational chemistry approach, D rep-
such a pathway, and an agmatinase inhibitor would resents the calculated dipole moment of the com-
be an ideal way to target that pathway selectively. pound, D1916 and D1920 are the distances between
N19 and N16, and between N19 and N20, respec-
The PSR is induced only in discrete brain regions in tively, and H21 is the charge on the H21 in the
response to various forms of stress. It is localised to agmatine derivatives. This QSAR equation indicates
the hippocampus and cortex in response to hypox- the kind of structure an ideal agmatinase inhibitor
ia,[97] two key regions of antidepressant drug action. should have. The findings suggest that a selective
Since ADC is reportedly induced during stress as a agmatinase inhibitor should have the following
branch of the PSR, we hypothesise that an agent that properties relative to agmatine: overall similar size
blocks agmatinase should lead to a selective rise in to agmatine, smaller dipole moment, smaller dis-
agmatine levels in just those brain regions where the tance between N19 and N16, smaller distance be-
PSR occurs, because these would be the only re- tween N19 and N20, and greater charge on H21.[99]
gions where agmatine is induced. Other brain re- An example of such a molecule is piperazine
gions without a PSR would not be expected to show carboxamidine. To our knowledge, this novel com-
much of an increase in agmatine levels in response pound has never been studied as a pharmacological
to an agmatinase inhibitor. This concept cannot, agent. Our plan is to synthesise piperazine carbox-
however, account for the possibility that brain diam- amidine, as well as other related compounds, to test
ine oxidase (DAO) could circumvent the presumed for agmatinase inhibitory capacity. These com-
utility of an agmatinase inhibitor. Under normal pounds will hopefully show clinical utility similar to
situations, levels of DAO are so low as to account agmatine but without the uraemic liability men-
for only a minimal fraction of the brain metabolism tioned earlier.
of agmatine, but it too could be induced by stress.
Thus, the role of DAO in stress also requires study. 7. Conclusion
The idea of an agmatinase inhibitor at this point
is, of course, still highly theoretical. However, pur- Agmatine has been established as a multi-faceted
suing this line of reasoning, we embarked on identi- neuromodulater with a wide spectrum of activity,
fying a specific agmatinase inhibitor. We screened such as neuroprotection, analgesia, effects on opioid
14 candidate compounds in test-tube enzyme inhibi- withdrawal and, possibly, stress protection. In
tion assays and performed computational quantita- animal experiments agmatine has also displayed
tive structure activity relationship (QSAR) analy- antidepressant and anxiolytic properties. The only
ses.[98] Each of the 14 compounds was chosen for human data on the use of agmatine available in the
slight differences in the guanidine moiety found in literature derive from studies of depressive illness.
agmatine, and each compound was studied by ab These preliminary data show possible elevations in
initio Hartee-Fock descriptors of the carbon (C), plasma levels of agmatine during acute episodes of
nitrogen (N) and hydrogen (H) atoms in this moiety. illness and subsequent reductions in levels follow-
Two compounds that proved most selective for in- ing recovery associated with two different antide-
hibiting agmatinase were synthesised at Jackson pressant agents. Agmatine may have potential as a
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
898 Halaris & Plietz
16. Reis DJ, Piletz JE. The imidazoline receptor in control of blood
therapeutic agent if its efficacy and safety can be pressure by clonidine and allied drugs. Am J Physiol 1997;
demonstrated in clinical trials. Alternatively, an 273: R1569-71
17. Eglen RM, Hudson AL, Kendall DA, et al. ‘Seeing through a
agmatinase inhibitor could be developed for clinical glass darkly’: casting light on imidazoline ‘I’ sites. Trends
study. Pharmacol Sci 1998; 19: 381-90
18. Raasch W, Schafer U, Qadri F, et al. Agmatine, an endogenous
Acknowledgements ligand at imidazoline binding sites, does not antagonize the
clonidine-mediated blood pressure reaction. Br J Pharmacol
The authors have no commercial interests in agmatine or 2002; 135: 663-72
related products; however, John Piletz is principal co-inven- 19. Reis DJ, Li G, Regunathan S. Endogenous ligands of
tor on US patent 20005/0220707 entitled ‘Mammalian agma- imidazoline receptors: classic and immunoreactive clonidine-
displacing substance and agmatine. Ann N Y Acad Sci 1995;
tinase inhibitory substance’ (awarded November 2005). The 763: 295-313
patent currently has no royalties or licensee, but it could have 20. Aricioglu F, Regunathan S, Piletz J. Is agmatine an endogenous
commercial value in the future. No sources of funding were factor against stress? Ann N Y Acad Sci 2003; 1009: 127-32
used to assist in the preparation of this review. 21. Wu N, Su R, Xu B, et al. IRAS, a candidate for I(1)-imidazoline
receptor, mediates inhibitory effect of agmatine on cellular
morphine dependence. Biochem Pharmacol 2005; 70 (7):
References 1079-87
1. Li G, Regunathan S, Barrow CJ, et al. Agmatine: an endogenous 22. Wu N, Su RB, Liu Y, et al. Modulation of agmatine on calcium
clonidine-displacing substance in the brain. Science 1994; 263: signal in morphine-dependent CHO cells by activation of
966-9 IRAS, a candidate for imidazoline I1 receptor. Eur J
2. Satriano J. Arginine pathways and the inflammatory response: Pharmacol 2006; 548 (1-3): 21-8
interregulation of nitric oxide and polyamines: review article. 23. Olmos G, DeGregorio-Rocasolano N, Paz Regalado M, et al.
Amino Acids 2004; 26: 321-9 Protection by imidazol(ine) drugs and agmatine of glutamate-
3. Zhu M, Iyo A, Piletz J, et al. Expression of human arginine induced neurotoxicity in cultured cerebellar granule cells
decarboxylase, the biosynthetic enzyme for agmatine. Biochim through blockade of NMDA receptor. Br J Pharmacol 1999;
Biophys Acta 2004; 1670: 156-64 127: 1317-26
4. Reis DJ, Regunathan S. Is agmatine a novel neurotransmitter in
brain? Trends Pharmacol Sci 2000; 21: 187-93 24. Yang XC, Reis DJ. Agmatine selectively blocks the N-methyl-
D-aspartate subclass of glutamate receptor channels in rat
5. Coleman CS, Hu G, Pegg AE. Putrescine biosynthesis in mam- hippocampal neurons. J Pharmacol Exp Ther 1999; 288: 544-9
malian tissues. Biochem J 2004; 379: 849-55
6. Iyo AH, Zhu MY, Ordway GA, et al. Expression of arginine 25. Galea E, Regunathan S, Eliopoulos V, et al. Inhibition of
decarboxylase in brain regions and neuronal cells. J Neu- mammalian nitric oxide synthases by agmatine, an endogenous
rochem 2006; 96: 1042-50 polyamine formed by decarboxylation of arginine. Biochem J
7. Goracke-Postle CJ, Nguyen HO, Stone LS, et al. Release of 1996; 316: 247-9
tritiated agmatine from spinal synaptosomes. Neuroreport 26. Demady DR, Jianmongkol S, Vuletich JL, et al. Agmatine
2006; 17: 13-7 enhances the NADPH oxidase activity of neuronal NO
8. Reis DJ, Yang XC, Milner TA. Agmatine containing axon synthase and leads to oxidative inactivation of the enzyme.
terminals in rat hippocampus form synapses on pyramidal Mol Pharmacol 2001; 59: 24-9
cells. Neurosci Lett 1998; 250: 185-8 27. Regunathan S, Piletz J. Regulation of inducible nitric oxide
9. Su R, Wei X, Zheng J, et al. Anticonvulsive effect of agmatine synthase and agmatine synthesis in macrophages and astro-
in mice. Pharmacol Biochem Behav 2004; 77: 345-9 cytes. Ann N Y Acad Sci 2003 Dec; 1009: 20-9
10. Riazi K, Honar H, Homayoun H, et al. The synergistic anticon- 28. Piletz J, May P, Wang G, et al. Agmatine crosses the blood-
vulsant effect of agmatine and morphine: possible role of alpha brain barrier. Ann N Y Acad Sci 2003; 1009: 64-74
2-adrenoceptors. Epilepsy Res 2005; 65: 33-40 29. Zomkowski AD, Hammes L, Lin J, et al. Agmatine produces
11. Zhu MY, Wang WP, Bissette G. Neuroprotective effects of antidepressant-like effects in two models of depression in
agmatine against cell damage caused by glucocorticoids in mice. Neuroreport 2002; 13: 387-91
cultured rat hippocampal neurons. Neuroscience 2006; 141 30. Aricioglu F, Altunbas H. Is agmatine an endogenous anxiolytic/
(4): 2019-27 antidepressant agent? Ann N Y Acad Sci 2003; 1009: 136-40
12. Gilad GM, Gilad VH, Finberg JP, et al. Neurochemical evidence 31. Li Y, Gong Z, Cao J, et al. Antidepressant-like effect of agma-
for agmatine modulation of 1-methyl-4-phenyl-1,2,3,6- tine and its possible mechanism. Eur J Pharmacol 2003; 469:
tetrahydropyridine (MPTP) neurotoxicity. Neurochem Res 81-8
2005; 30: 713-9 32. Lavinsky D, Arteni N, Netto C. Agmatine induces anxiolysis in
13. Wang WP, Iyo AH, Miguel-Hidalgo J, et al. Agmatine protects the elevated plus maze task in adult rats. Behav Brain Res
against cell damage induced by NMDA and glutamate in 2003; 141: 19-24
cultured hippocampal neurons. Brain Res 2006; 1084: 210-6 33. Stewart LS, McKay BE. Acquisition deficit and time-dependent
14. Halaris A, Piletz J. Relevance of imidazoline receptors and retrograde amnesia for contextual fear conditioning in agma-
agmatine to psychiatry: a decade of progress. Ann N Y Acad tine-treated rats. Behav Pharmacol 2000; 11: 93-7
Sci 2003; 1009: 1-20
15. Aricioglu F, Regunathan S. Agmatine attenuates stress- and 34. Uzbay IT, Yesilyurt O, Celik T, et al. Effects of agmatine on
lipopolysaccharide-induced fever in rats. Physiol Behav 2005; ethanol withdrawal syndrome in rats. Behav Brain Res 2000;
85: 370-5 107: 153-9
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
Agmatine and its Spectrum of Activity 899
35. Aricioglu F, Means A, Regunathan S. Effect of agmatine on the 55. Abe K, Abe Y, Saito H. Agmatine suppresses nitric oxide
development of morphine dependence in rats: potential role of production in microglia. Brain Res 2000; 872: 141-8
cAMP system. Eur J Pharmacol 2004; 504: 191-7 56. Abe K, Abe Y, Saito H. Agmatine induces glutamate release and
36. Wei X, Su R, Lu X, et al. Inhibition by agmatine on morphine- cell death in cultured rat cerebellar granule neurons. Brain Res
induced conditioned place preference in rats. Eur J Pharmacol 2003; 990: 165-71
2005; 515: 99-106 57. Satriano J, Schwartz D, Ishizuka S, et al. Suppression of induci-
37. Fullerton CS, Ursano RJ, Wang L. Acute stress disorder, post- ble nitric oxide generation by agmatine aldehyde: beneficial
traumatic stress disorder, and depression in disaster or rescue effects in sepsis. J Cell Physiol 2001; 188: 313-20
workers. Am J Psychiatry 2004; 161: 1370-6 58. Khoshnoodi MA, Motiei-Langroudi R, Tahsili-Fahadan P, et al.
38. Tabor CW, Tabor H. Polyamines. Annu Rev Biochem 1984; 53: Involvement of nitric oxide system in enhancement of mor-
749-90 phine-induced conditioned place preference by agmatine in
male mice. Neurosci Lett 2006; 399 (3): 234-9
39. Molderings G, Bruss M, Bonisch H, et al. Identification and
pharmacological characterization of a specific agmatine trans- 59. Roberts J, Grocholski B, Kitto K, et al. Pharmacodynamic and
port system in human tumor cell lines. Ann N Y Acad Sci pharmacokinetic studies of agmatine after spinal administra-
2003; 1009: 75-81 tion in the mouse. J Pharmacol Exp Ther 2005; 314: 1226-33
40. Mistry SK, Burwell TJ, Chambers RM, et al. Cloning of human 60. Feng Y, Halaris AE, Piletz JE. Determination of agmatine in
agmatinase: an alternate path for polyamine synthesis induced brain and plasma using high-performance liquid chromatogra-
in liver by hepatitis B virus. Am J Physiol Gastrointest Liver phy with fluorescence detection [published erratum appears
Physiol 2002; 282: G375-81 in J Chromatogr B Biomed Sci Appl 1997 Aug 15; 696 (1):
173]. J Chromatogr B Biomed Sci Appl 1997; 691 (2): 277-82
41. Iyer R, Kim H, Tsoa R, et al. Cloning and characterization of
human agmatinase. Mol Genet Metab 2002; 75: 209-18 61. Zhang W, Kaye D. Simultaneous determination of arginine and
seven metabolites in plasma by reversed-phase liquid chroma-
42. Morris S. Recent advances in arginine metabolism. Curr Opin tography with a time-controlled ortho-phthaldialdehyde pre-
Clin Nutr Metab Care 2004; 7: 45-51 column derivatization. Anal Biochem 2004; 326: 87-92
43. Gorbatyuk OS, Milner TA, Wang G, et al. Localization of 62. Zhao S, Wang B, Yuan H, et al. Determination of agmatine in
agmatine in vasopressin and oxytocin neurons of the rat hypo- biological samples by capillary electrophoresis with optical
thalamic paraventricular and supraoptic nuclei. Exp Neurol fiber light-emitting-diode-induced fluorescence detector.
2001; 171: 235-45 J Chromatogr A 2006; 1123: 138-41
44. Piletz JE, Chikkala DN, Ernsberger P. Comparison of the 63. Dias Elpo Zomkowski A, Oscar Rosa A, Lin J, et al. Evidence
properties of agmatine and endogenous clonidine-displacing for serotonin receptor subtypes involvement in agmatine an-
substance at imidazoline and alpha-2 adrenergic receptors. tidepressant like-effect in the mouse forced swimming test.
J Pharmacol Exp Ther 1995; 272: 581-7 Brain Res 2004; 1023: 253-63
45. Zheng J, Weng X, Gai X, et al. Mechanism underlying blockade
of voltage-gated calcium channels by agmatine in cultured rat 64. Zomkowski A, Santos A, Rodrigues A. Evidence for the in-
hippocampal neurons. Acta Pharmacol Sin 2004; 25: 281-5 volvement of the opioid system in the agmatine antidepressant-
like effect in the forced swimming test. Neurosci Lett 2005;
46. Askalany A, Yamakura T, Petrenko A, et al. Effect of agmatine 381: 279-83
on heteromeric N-methyl-d-aspartate receptor channels. Neu-
rosci Res 2005; 52: 387-92 65. Gonzalez C, Regunathan S, Reis DJ, et al. Agmatine, an endoge-
nous modulator of noradrenergic neurotransmission in the rat
47. Loring RH. Agmatine acts as an antagonist of neuronal nicotinic tail artery. Br J Pharmacol 1996; 119: 677-84
receptors. Br J Pharmacol 1990; 99: 207-11 66. Zhao D, Ren L. Non-adrenergic inhibition at prejunctional sites
48. Molderings GJ, Schmidt K, Bonisch H, et al. Inhibition of 5- by agmatine of purinergic vasoconstriction in rabbit saphenous
HT3 receptor function by imidazolines in mouse neuroblas- artery. Neuropharmacology 2005; 48: 597-606
toma cells: potential involvement of sigma 2 binding sites.
Naunyn Schmiedebergs Arch Pharmacol 1996; 354: 245-52 67. Wang H, Regunathan S, Youngson C, et al. An antibody to
agmatine localizes the amine in bovine adrenal chromaffin
49. Otake K, Ruggiero DA, Regunathan S, et al. Regional localiza- cells. Neurosci Lett 1995; 183: 17-21
tion of agmatine in the rat brain: an immunocytochemical
study. Brain Res 1998; 787: 1-14 68. Regunathan S, Youngson C, Raasch W, et al. Imidazoline
receptors and agmatine in blood vessels: a novel system inhib-
50. Sastre M, Regunathan S, Reis DJ. Uptake of agmatine into rat iting vascular smooth muscle proliferation. J Pharmacol Exp
brain synaptosomes: possible role of cation channels. J Neu- Ther 1996; 276: 1272-82
rochem 1997; 69: 2421-6 69. Briaud S, Zhang BL, Sannajust F. Central actions of agmatine in
51. Molderings G, Heinen A, Menzel S, et al. Gastrointestinal conscious spontaneously hypertensive rats. Clin Exp Hyper-
uptake of agmatine: distribution in tissues and organs and tens 2005; 27: 619-27
pathophysiologic relevance. Ann N Y Acad Sci 2003; 1009:
44-51 70. Tahsili-Fahadan P, Yahyavi-Firouz-Abadi N, Khoshnoodi MA,
et al. Agmatine potentiates morphine-induced conditioned
52. Sastre M, Regunathan S, Galea E, et al. Agmatinase activity in place preference in mice: modulation by alpha(2)-adrenocep-
rat brain: a metabolic pathway for the degradation of agmatine. tors. Neuropsychopharmacology 2006; 31 (8): 1722-32
J Neurochem 1996; 67: 1761-5 71. Sener A, Lebrun P, Blachier F, et al. Stimulus-secretion coup-
53. Auguet M, Viossat I, Marin JG, et al. Selective inhibition of ling of arginine-induced insulin release: insulinotropic action
inducible nitric oxide synthase by agmatine. Jpn J Pharmacol of agmatine. Biochem Pharmacol 1989; 38: 327-30
1995; 69: 285-7
54. Feng Y, LeBlanc MH, Regunathan S. Agmatine reduces ex- 72. Kalra SP, Pearson E, Sahu A, et al. Agmatine, a novel hypotha-
tracellular glutamate during pentylenetetrazole-induced lamic amine, stimulates pituitary luteinizing hormone release
seizures in rat brain: a potential mechanism for the anticonvul- in vivo and hypothalamic luteinizing hormone-releasing hor-
sive effects. Neurosci Lett 2005; 390: 129-33 mone release in vitro. Neurosci Lett 1995; 194: 165-8
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
900 Halaris & Plietz
73. Molderings GJ, Gothert M. Inhibitory presynaptic imidazoline 88. Onal A, Delen Y, Ulker S, et al. Agmatine attenuates neuropath-
receptors on sympathetic nerves in the rabbit aorta differ from ic pain in rats: possible mediation of nitric oxide and noradren-
I1- and I2-imidazoline binding sites. Naunyn Schmiedebergs ergic activity in the brainstem and cerebellum. Life Sci 2003;
Arch Pharmacol 1995; 351: 507-16 73: 413-28
74. Regunathan S, Feinstein DL, Reis DJ. Anti-proliferative and 89. Aricioglu F, Korcegez E, Bozkurt A, et al. Effect of agmatine on
anti-inflammatory actions of imidazoline agents: are imidazo- acute and mononeuropathic pain. Ann N Y Acad Sci 2003;
line receptors involved? Ann N Y Acad Sci 1999; 881: 410-9 1009: 106-15
75. Gilad VH, Rabey JM, Kimiagar Y, et al. The polyamine stress 90. Kolesnikov Y, Jain S, Pasternak GW. Modulation of opioid
response: tissue-, endocrine-, and developmental-dependent analgesia by agmatine. Eur J Pharmacol 1996; 296: 17-22
regulation. Biochem Pharmacol 2001; 61: 207-13 91. Li J, Li X, Pei G, et al. Effects of agmatine on tolerance to and
76. Gilad GM, Gilad VH. Overview of the brain polyamine-stress- substance dependence on morphine in mice. Chung Kuo Yao
response: regulation, development, and modulation by lithium Li Hsueh Pao 1999; 20: 232-8
and role in cell survival. Cell Mol Neurobiol 2003; 23: 637-49
77. Gilad GM, Gilad VH. Brain polyamine stress response: recur- 92. Aricioglu-Kartal F, Uzbay IT. Inhibitory effect of agmatine on
rence after repetitive stressor and inhibition by lithium. naloxone-precipitated abstinence syndrome in morphine de-
J Neurochem 1996; 67: 1992-6 pendent rats. Life Sci 1997; 61: 1775-81
78. Elgun S, Kumbasar H. Increased serum arginase activity in 93. McKay B, Lado W, Martin L, et al. Learning and memory in
depressed patients. Prog Neuropsychopharmacol Biol Psychia- agmatine-treated rats. Pharmacol Biochem Behav 2002; 72:
try 2000; 24: 227-32 551-7
79. Halaris A, Zhu H, Feng Y, et al. Plasma agmatine and platelet 94. McKay B, Persinger M. Combined effects of complex magnetic
imidazoline receptors in depression. Ann N Y Acad Sci 1999; fields and agmatine for contextual fear learning deficits in rats.
881: 445-51 Life Sci 2003; 72: 2489-98
80. Sastre M, Galea E, Feinstein D, et al. Metabolism of agmatine in 95. Porsolt RD, Anton G, Blavet N, et al. Behavioural despair in
macrophages: modulation by lipopolysaccharide and inhibito- rats: a new model sensitive to antidepressant treatments. Eur J
ry cytokines. Biochem J 1998; 330: 1405-9 Pharmacol 1978; 47: 379-91
81. Gilad GM, Gilad VH, Rabey JM. Arginine and ornithine decar-
boxylation in rodent brain: coincidental changes during devel- 96. Porsolt RD, Deniel M, Jalfre M. Forced swimming in rats:
opment and after ischemia. Neurosci Lett 1996; 216: 33-6 hypothermia, immobility and the effects of imipramine. Eur J
Pharmacol 1979; 57: 431-6
82. Feng Y, Piletz JE, Leblanc MH. Agmatine suppresses nitric
oxide production and attenuates hypoxic-ischemic brain injury 97. Gilad GM, Gilad VH, Eliyayev Y, et al. Developmental regula-
in neonatal rats. Pediatr Res 2002; 52: 606-11 tion of the brain polyamine-stress-response. Int J Dev Neurosci
83. Fairbanks C, Kaminski L, Nguyen H, et al. Pre-treatment with 1998; 16: 271-8
antisera raised against agmatine sensitizes mice to plasticity- 98. Huang M, Regunathan S, Botta M, et al. Structure-activity
mediated events [abstract]. Soc Neurosci Abstr 2001; 27: 465 analysis of guanidine group in agmatine for brain agmatinase.
84. Aricioglu-Kartel F, Reis D, Regunathan S. Agmatine and mor- Ann N Y Acad Sci 2003; 1009: 52-63
phine tolerance/dependance: molecular mechanisms of inter- 99. Piletz J, Huang M, Lee K, inventors. Jackson State University,
actions [abstract]. Soc Neurosci Abstr 2001; 27: 685 assignee. Mammalian agmatinase inhibitory substance. US
85. Gilad GM, Salame K, Rabey JM, et al. Agmatine treatment is patent application 20050220707; 2004 Apr 5
neuroprotective in rodent brain injury models. Life Sci 1996;
58: 41-6
86. Gilad GM, Gilad VH. Accelerated functional recovery and Correspondence: Dr Angelos Halaris, Department of Psychi-
neuroprotection by agmatine after spinal cord ischemia in rats. atry and Behavioral Neurosciences, Loyola University
Neurosci Lett 2000; 296: 97-100 Medical Center, Maguire Center, Rm 1940, 2160 South First
87. Fairbanks CA, Schreiber KL, Brewer KL, et al. Agmatine Avenue, Maywood, IL 60153, USA.
reverses pain induced by inflammation, neuropathy, and spinal E-mail: ahalaris@lumc.edu
cord injury. Proc Natl Acad Sci U S A 2000; 97: 10584-9
2007 Adis Data Information BV. All rights reserved. CNS Drugs 2007; 21 (11)
