Download Hypothesized Deficiency of Guanine

REVIEW ARTICLE
Hypothesized Deficiency of Guanine-Based Purines May
Contribute to Abnormalities of Neurodevelopment,
Neuromodulation, and Neurotransmission in
Lesch-Nyhan Syndrome
Stephen I. Deutsch, MD, PhD,*† Katrice D. Long, BS,* Richard B. Rosse, MD,*†
John Mastropaolo, PhD,*† and Judy Eller*
Abstract: The Lesch-Nyhan syndrome is a devastating sex-linked
recessive disorder resulting from almost complete deficiency of the
activity of hypoxanthine phosphoribosyltransferase (HPRT). The
enzyme deficiency results in an inability to synthesize the nucleotides
guanosine monophosphate and inosine monophosphate from the purine bases guanine and hypoxanthine, respectively, via the ‘‘salvage’’
pathway and an accelerated biosynthesis of these purines via the de
novo pathway. The syndrome is characterized by neurologic manifestations, including the very dramatic symptom of compulsive selfmutilation. The neurologic manifestations may result, at least in part,
from a mixture of neurodevelopmental (eg, a failure to ‘‘arborize’’dopaminergic synaptic terminals) and neurotransmitter (eg, disruption
of GABA and glutamate receptor–mediated neurotransmission) consequences. HPRT deficiency results in elevated extracellular levels of
hypoxanthine, which can bind to the benzodiazepine agonist recognition site on the GABAA receptor complex, and the possibility of
diminished levels of guanine-based purines in discrete ‘‘pools’’
involved in synaptic transmission. In addition to their critical roles in
metabolism, gene replication and expression, and signal transduction,
guanine-based purines may be important regulators of the synaptic
availability of L-glutamate. Guanine-based purines may also have important trophic functions in the CNS. The investigation of the LeschNyhan syndrome may serve to clarify these and other important neurotransmitter, neuromodulatory, and neurotrophic roles that guaninebased purines play in the central nervous system, especially the
developing brain. A widespread and general deficiency of guaninebased purines would lead to impaired transduction of a variety of
signals that depend on GTP-protein-coupled second messenger
systems. This is less likely in view of a prominent localized pathologic effect of HPRT deficiency on presynaptic dopaminergic
projections to the striatum. A possible more circumscribed effect of
a deficiency of guanine-based purines could be interference with
From the *Mental Health Service Line, VISN5, Department of Veterans
Affairs Medical Center, Washington, DC; and †Department of Psychiatry,
Georgetown University School of Medicine, Washington, DC.
This work was supported by the Department of Veterans Affairs VISN 5
Mental Illness, Research, Education and Clinical Center (MIRECC; Alan
S. Bellack, PhD, Director).
Reprints: Stephen I. Deutsch, MD, PhD, Director, Mental Health Service Line,
VISN5, Department of Veterans Affairs Medical Center, 50 Irving Street,
NW, Washington, DC 20422 (e-mail: [email protected]).
Copyright Ó 2005 by Lippincott Williams & Wilkins
28
modulation of glutamatergic neurotransmission. Guanosine has been
shown to be an important modulator of glutamatergic neurotransmission, promoting glial reuptake of L-glutamate. A deficiency of
guanosine could lead to dysregulated glutamatergic neurotransmission, including possible excitotoxic damage. Unfortunately, although
the biochemical lesion has been known for quite some time (ie, HPRT
deficiency), therapeutically beneficial interventions for these affected
children and adults have not yet emerged based on this elucidation.
Conceivably, guanosine or its analogues and excitatory amino acid
receptor antagonists could participate in the pharmacotherapy of this
devastating disorder.
Key Words: Lesch-Nyhan syndrome, hypoxanthine phosphoribosyltransferase, guanosine, hypoxanthine, GABA, glutamate
(Clin Neuropharmacol 2005;28:28–37)
LESCH-NYHAN SYNDROME
Disrupted Purine Metabolism
In addition to their essential roles in metabolism,
mitochondrial electron transport, and gene replication and
expression, nucleosides and nucleotides, particularly those
with purine bases (eg, adenosine, inosine, and guanosine), are
important modulators of neurotransmission, regulating the
synaptic availability of neurotransmitters, participating in
second messenger cascades, and acting as direct agonists of
cell surface receptors, among other functions. Chemically,
nucleosides are purine and pyrimidine bases covalently linked
via N-glycosidic bonds in the C1 carbon position to b-D-ribose
or b-D-deoxyribose, a 5-carbon ring sugar moiety. Nucleotides
are phosphorylated in the 5# carbon position of the pentose
sugar. In general, the literature on the involvement of
nucleosides and nucleotides in synaptic neurotransmission is
growing exponentially; conceivably, some of the recent data
pertaining to guanosine (and inosine) in this area is relevant to
our understanding of CNS manifestations of the Lesch-Nyhan
syndrome and neurodegenerative disorders.
The Lesch-Nyhan syndrome, a rare X-linked recessive
disorder, highlighted the importance of ‘‘salvaging’’ the purine
bases hypoxanthine and guanine; the disorder results from an
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
almost complete deficiency of the enzyme hypoxanthine
phosphoribosyltransferase (HPRT). This enzyme catalyzes the
reutilization of hypoxanthine and guanine in the energetically
favorable synthesis of the nucleotides inosine monophosphate
(IMP) and guanosine monophosphate (GMP), respectively;
when HPRT is absent or deficient, accelerated de novo purine
biosynthesis is stimulated as a result of the accumulation of 5phosphoribosyl-1-pyrophosphate (PRPP), a substrate for both
HPRT and the de novo pathway, and/or reduced end-product
inhibition.2,28
The neurologic manifestations of the Lesch-Nyhan
syndrome include spasticity, mental retardation, choreoathetosis, aggression, and the very dramatic compulsive selfmutilation. CNS symptoms of the disorder are manifested
within the first year of life and include irritability, feeding
difficulties, delayed motor development, and extrapyramidal
symptoms characterized by chorea and fine athetotic movements of the extremities. Moreover, although accurate estimates
of intelligence are difficult to make because of motor and
speech problems, the majority of patients are moderately to
severely mentally retarded. Self-injurious behavior, especially
self-biting of the mouth, tongue, and fingers, is a dramatic and
almost invariant symptom with a variable age of onset from 6
months to 16 years. The pathophysiological involvement of
purines and their metabolism in the mediation of these
symptoms has been sought but, nonetheless, remains both
uncertain and elusive. Normally, HPRT activity is highest in the
brain, especially the basal ganglia, and the soluble cytoplasmic
enzyme constitutes up to 0.05% of the brain’s total soluble
protein.29 Thus, the salvage of purine bases and their rapid and
energetically efficient reutilization for the synthesis of purine
nucleotides appears to be critical in the brain. Guanine-based
purines are involved in synaptic transmission; for example, the
exchange of GDP and GTP and the hydrolysis of GTP while
bound to heterotrimeric G proteins anchored to the membrane
are critical steps in the regulation of the initiation and duration
of several second messenger cascades.26
Recent work on the role that guanosine plays in the
modulation of glutamatergic neurotransmission prompted our
interest in the Lesch-Nyhan syndrome.9,10 Conceivably, diminished reutilization of free guanine bases secondary to
absent or negligent HPRT activity and relatively high guanase
activity in brain could lead to deficient pools of guanosine
associated with glutamatergic synapses in Lesch-Nyhan
syndrome. Guanosine promotes astrocytic reuptake of Lglutamate, an important excitatory neurotransmitter in the
brain. Thus, guanosine may be an important physiological
mechanism for dampening or terminating the synaptic actions
of L-glutamate. This hypothesis is interesting but difficult to
test because of the many small and rapidly turning over
metabolic and neurotransmitter pools of guanine-based purines. Nonetheless, if a guanosine deficiency were to exist in
Lesch-Nyhan syndrome, the administration of guanosine itself
or functional analogues that could promote glutamate reuptake
and/or excitatory amino acid receptor antagonists might be
useful pharmacotherapeutic agents to consider in the treatment
of this disorder.
The hypothesis developed in this paper focuses primarily on the pathophysiologic role of a deficiency of guanineq 2005 Lippincott Williams & Wilkins
Lesch-Nyhan Syndrome and Guanine-Based Purines
based purines in the Lesch-Nyhan syndrome. The neurobiology of adenine-based purines, especially adenosine, has been
more thoroughly studied.26 However, mathematical modeling of the effects of HPRT efficiency on purine metabolism
suggest that depletion of guanine-based purines may be a
more significant consequence than depletion of adenine-based
purines.
In Vivo and In Vitro Studies of
Neurodevelopmental Abnormalities of
Dopaminergic Systems
The earliest studies of the CNS manifestations implicated abnormalities of dopamine, especially in the basal
ganglia, which may explain some of the movement abnormalities.2,13,15 Specifically, the amount of dopamine and its
principal metabolite homovanillic acid, and the activities of
dopa decarboxylase and tyrosine hydroxylase, key enzymes
involved in dopamine biosynthesis, were significantly reduced
in striatum in 3 autopsied brains from patients with less than
1% of the normal activity of HPRT in striatal tissue and 1% to
2% of the normal HPRT activity in thalamus and cortex.15
There was also a significant decrease in the concentration of
dopamine and homovanillic acid in the nucleus accumbens,
whereas normal concentrations of dopamine were observed in
the substantia nigra. These neurochemical data suggested that
the density of ‘‘arborization’’ of the dopamine axon terminals
in striatum and limbic structures, whose cell bodies originate
in the midbrain, is reduced in Lesch-Nyhan syndrome.15
Dopamine is concentrated within axon terminals in striatal and
limbic structures.
An additional postmortem study of two brains from
patients with Lesch-Nyhan syndrome was consistent with a
reduction of the dopamine content in the caudate nucleus.20
Moreover, the density of cells that were immunocytochemically stained for the presence of the D2 dopamine receptor
was dramatically increased in the two putamens from the
patients with Lesch-Nyhan syndrome, relative to the mean
value of 5 control patients without any significant neuropathological findings. Also, the putamenal cells from the patients
with Lesch-Nyhan syndrome that were stained were more
densely labeled than the stained cells in the controls. Although
not as dramatic as the increased density of putamenal cells
stained for the D2 dopamine receptor, relative to the controls,
there were also small increases in the number of cells
immunoreactive for the D2 dopamine receptor in the caudate
nucleus and D1 dopamine receptor in the putamen and caudate
nucleus of the two patients with Lesch-Nyhan syndrome.
Consistent with the presence of pathology of the basal ganglia
in Lesch-Nyhan syndrome, neurons in the caudate nucleus and
putamen from the two patients with Lesch-Nyhan syndrome
were immunoreactive for methionine-enkephalin, an endogenous opioid that is involved in self-mutilation and nociception, whereas no methionine-enkephalin immunoreactivity
could be demonstrated in the caudate nucleus and putamen
from the controls. Finally, the density of medium-sized spiny
neurons was increased in the putamen of the two patients with
Lesch-Nyhan syndrome, relative to the controls. Interestingly,
the authors note that these findings are similar to the PET
findings of patients with Parkinson’s disease, who show an
29
Deutsch et al
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
increased density of the D2 dopamine receptor and content of
methionine-enkephalin in putamen. The authors ascribed the
increased density of D2 dopamine receptors in Lesch-Nyhan
syndrome to decreased activity of presynaptic dopaminergic
neurons.20 Also, the increased density of medium-sized spiny
neurons could reflect the maldevelopment of dopamine
terminals. The increased content of methionine-enkephalin
may have more to do with the motor abnormalities than selfmutiliation and decreased nociception. There were no differences in the immunoreactivity for methionine-enkephalin
along the pain pathways between the patients with LeschNyhan syndrome and the five controls. Of course, the postmortem findings must be regarded as tentative and provocative in view of the small sample sizes in the postmortem
studies.15,20
Blinded review of the magnetic resonance imaging scans
of a series of 7 patients with Lesch-Nyhan syndrome (age
range 22 to 35 years), including volumetric analysis of the
basal ganglia, revealed evidence of both reduced basal ganglia
volume, especially in the caudate nucleus, and total cerebral
volume, compared with 7 healthy age- and sex-matched controls.12 The patients with Lesch-Nyhan syndrome showed
greater variability in their total brain volume, whereas little
variability was observed in the controls. The patients had the
classic form of the illness with levels of HPRT activity below
1.6% in erythrocyte lysates and fibroblast cultures. On blinded
review, 3 patients were reported to have small and 2 patients
were reported to have probably small caudates; additionally,
the putamens of 2 patients were read as possibly small.
Consistent with atrophy of the caudate nucleus, the bicaudate
distances of the patients were larger. Quantitative volumetric
analysis showed that the volume of the caudate nucleus was
significantly reduced by 34%, and total brain volume was
reduced significantly by 17% in the patients. Clearly, the
HPRT deficiency is associated with disruption of normal brain
development, and the pathologic process in the basal ganglia is
likely to account for the dystonia and movement abnormalities
in Lesch-Nyhan syndrome.
Sequential analysis of CSF levels of homovanillic acid
(HVA), the principal dopamine metabolite, was performed in
4 patients with classic Lesch-Nyhan syndrome (erythrocyte
HPRT activity ,1%), whose ages were 1.5 to 17 years at the
time of study entry, over a 5-year period.22 With the exception
of 1 CSF sample, the CSF HVA levels of the patients were
lower than the mean HVA value of the age-matched controls.
Further, 10 of the 19 samples assayed for HVA were below the
control range, and 2 of them were more than 2 standard
deviations below the mean of the age-matched controls.
Sequential analysis of CSF levels of 5-hydroxyindoleacetic
acid (5-HIAA), the principal serotonin metabolite, was within
the control range. The diminished CSF levels of HVA are
consistent with diminished turnover of dopamine or its
accelerated egress out of the CSF compartment. However,
in view of the normal levels of CSF 5-HIAA and the fact that
HVA and 5-HIAA probably share a common mechanism for
transport out of the CSF, the data are most consistent with
diminished dopamine turnover in Lesch-Nyhan syndrome.
The dopamine that is released from terminals in the caudate
and putamen contribute significantly to the content of HVA in
30
CSF. In view of this, lowered CSF HVA levels in Lesch-Nyhan
syndrome implicate the basal ganglia in the emergence of
at least some of the CNS symptoms (eg, choreoathetosis).
Perhaps a diminished release of dopamine from terminals
within basal ganglia leads to a compensatory up-regulation in
the sensitivity of dopamine receptors.22
The density of the presynaptic dopamine transporter can
be measured in humans in vivo with the positron emission tomography (PET) ligand b-carbomethoxy-3b-4-fluorophenyltropane (CFT; WIN-35,428). This ligand has been used to
study the degeneration of dopaminergic nerve terminals,
particularly in the caudate and putamen. Six adult patients (age
range 19–35 years) with the classic presentation of the LeschNyhan syndrome were shown to have dramatic reductions
(.50%) in the density of dopamine transporters in the caudate
and putamen compared with 3 adult patients with RETT syndrome and with 10 age-matched normal volunteers, using
WIN-35,428 and 3 different approaches to measuring the
density of the dopamine transporter.30 Furthermore, volumetric measurements using MRI were performed on 5 of the
patients with Lesch-Nyhan syndrome and revealed a significant
30% reduction in caudate volume; even after correction for
a possible confounding effect of reduced caudate volume, the
significant reduction in the density of presynaptic dopamine
transporters in the caudate persisted (in fact, the magnitude of
the reduction was larger). The diminished binding of WIN35,428 in Lesch-Nyhan syndrome is consistent with a loss of
dopaminergic nerve terminals or down-regulation of the dopamine transporter in the basal ganglia in this disorder.
PET data obtained with [18F]fluorodopa, an analogue of
dopa, can provide in vivo information about the activity of
dopa decarboxylase and the storage of dopamine in presynaptic nerve terminals. With this compound, the presynaptic
uptake of [18F]fluorodopa and its conversion to [18F]fluorodopamine and storage in presynaptic dopaminergic terminals was
studied in 12 male patients with Lesch-Nyhan syndrome (age
range 10–20 years) and 15 controls (9 male and 6 female; age
range 12–23 years).8 In all of the patients, the activity of
HPRT in erythrocytes or fibroblasts was less than 1% of control values; further, the patients manifested both aggression
and self-injury. The patients were wheelchair-bound and required the use of full or partial restraints 100% of the time.
The patients and controls were compared on the amount of
[18F]fluorodopa activity in 4 dopaminergic regions of interest:
putamen, caudate nucleus, frontal cortex, and ventral tegmental complex. The ventral tegmental complex contained
the dopaminergic cell bodies that project to the striatum
and frontal cortex. The results showed that the patients with
Lesch-Nyhan syndrome had highly significant reductions of
[18F]fluorodopa activity in all 4 dopaminergic regions of
interest; the magnitude of the reductions ranged from 31% in
the putamen to 57% in the ventral tegmental complex. There
were also interesting associations that emerged between the
severity of aggression against others and [18F]fluorodopa activity in the putamen and ventral tegmental complex.8 Reduced
[18F]fluorodopa activity in the basal ganglia discriminated perfectly between patients and healthy controls. Moreover, the
reduced activity in the ventral tegmental complex and frontal
cortex suggests that the HPRT deficiency affects both the cell
q 2005 Lippincott Williams & Wilkins
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
bodies and terminals of the mesolimbic, mesocortical, and
nigrostriatal dopaminergic projections. However, these latter
conclusions must be viewed tentatively because of the low
levels of [18F]fluorodopa activity in the frontal cortex and
ventral tegmental complex and the small size of the ventral
tegmental complex; thus, measurements in these areas are less
reliable. Additionally, noradrenergic terminals contribute to
[18F]fluorodopa activity in the frontal cortex; dopa is the
precursor of both dopamine and norepinephrine. Thus, the
potential noise resulting from the contributions of noradrenergic projections must be considered. In any event, the data are
consistent with an effect of HPRT deficiency on dopamine
systems in brain.8
The HPRT-deficient knockout mouse, an animal model
of the Lesch-Nyhan syndrome, has been used to study the reduced density of dopaminergic fibers extending to the
caudoputamen and its relation to the enzyme deficiency.3 The
density of the neuritic arborization of dopaminergic neurons
cultured from the cerebral hemispheres of 14-day-old HPRTdeficient knockout mice and appropriate controls was studied
in vitro.3 The proliferation of glial cells was suppressed in
these cultures, and the dopaminergic neurons were identified
immunochemically with rabbit polyclonal antibodies to
tyrosine hydroxylase (TH), the enzyme catalyzing the ratelimiting step in catecholamine biosynthesis. Adrenergic neurons were distinguished from dopaminergic neurons by their
additional immunochemical staining with antibody to dopamine b-hydroxylase. A stereologic methodology was employed using light microscopy to measure the total fiber length
of individual dopaminergic neurons (ie, those stained only
with the polyclonal antibodies to TH). The ‘‘dendrites’’ of 100
HPRT-deficient and 100 control neurons were measured at
different time points. On the eighth day in culture, although
there were no significant differences in the total number of
dopaminergic neurons, significant differences emerged with
respect to the average fiber length per dopaminergic neuron:
the neuritic length of the HPRT-deficient cells was 32% shorter
(P , 0.025).3 Thus, the morphologic consequence of the
enzyme deficiency is an apparent defective development of
dopaminergic neurons characterized by decelerated dendritic
outgrowth; they are able to survive but cannot make an
enriched network of synaptic connections.
In contrast, in an in vitro study examining the trophic
effects of glial cell line–derived neurotrophic factor (GDNF)
on the survivability, morphologic development, and function
of dopaminergic cells in primary cultures derived from wildtype and HPRT-deficient mouse embryonic midbrain (day
16.5), the 2 cell lines showed no significant differences in their
survivability or morphology in the presence or absence of
GDNF.24 Dopaminergic cells were identified in these cultures
immunocytochemically with mouse monoclonal antibody to
TH. In the absence of GDNF, a survival-promoting factor for
dopamine neurons, the number of TH-positive wild-type and
HPRT-deficient cells decreased rapidly; approximately 50% of
the cells were lost by the fifth day in culture, continuing to
approximately 10% to 15% of the original number by the 10th
day in culture. GDNF stabilized the survival of both
embryonic wild-type and HPRT-deficient dopamine cells with
no difference observed between the cell lines in their
q 2005 Lippincott Williams & Wilkins
Lesch-Nyhan Syndrome and Guanine-Based Purines
sensitivity to GDNF’s cytoprotective effect. Similarly, from
a morphologic perspective, the wild-type and HPRT-deficient
cells were equally responsive to the ability of GDNF to
stimulate the extension of neuritic processes. Thus, there was
no differential effect of GDNF on the survivability and fiber
arborization of the wild-type and HPRT-deficient cells. Also,
in the absence of GDNF, in this study, the HPRT-deficient
TH-positive embryonic mesencephalic cells did not show any
differences in terms of their survivability or morphologic
differentiation, compared with the wild-type controls. However, although differences did not emerge between the wildtype and HPRT-deficient cells in terms of survivability and
morphology, under the influence of GDNF, the cultured wildtype cells showed statistically significant increases in both
dopamine uptake and dopamine content.24 Thus, important
functions of the neuritic extensions of the wild-type embryonic
mesencephalic dopaminergic neurons, especially dopamine
uptake and storage, were more responsive to the trophic
influence of GDNF. The authors concluded that in view of
GDNF’s inability to restore the functioning of dopaminergic
neurons in the HPRT-deficient condition, this specific neurotrophic factor may have little, if any, therapeutic value in
Lesch-Nyhan syndrome.24
Using a quantitative morphologic approach to the
assessment of differentiation in cell culture, N2aTG, the
HPRT-deficient nondopaminergic neuroblastoma cell line,
differentiated more over a period of 2 days than control HPRTpositive N2A cells.7 Measures of differentiation included the
percentage of cells with neurites, the average neurite length,
and the average number of neurites per cell. The cell lines also
differed with respect to their proliferation: the HPRT-deficient
cells proliferated less. The density of the cultured cells was
a factor that influenced proliferation. Thus, the in vitro data
show a clear link between HPRT activity and proliferation and
morphologic development of neurons, which may be relevant
to the disruption of the brain’s microarchitecture in LeschNyhan syndrome.7
When grown in vitro, HPRT-deficient cell lines derived
from rat (B103-4C) and mouse (N2aTG) neuroblastoma
showed abnormalities of proliferation and differentiation compared with their control lines; these cells serve as models of the
development of nondopaminergic neurons.6,7 These HPRTdeficient cells are more ‘‘adhesive’’ and have altered intracellular concentrations of nucleotides. The mouse N2aTG
HPRT-deficient cell line was derived from spinal cord, and the
rat B103-4C HPRT-deficient cell line was derived from brain.
The total amount of HPRT-deficient N2aTG mouse cell
proliferation and the rate of proliferation of the rat HPRTdeficient B103-4C cells were reduced over 4 days in culture.
The morphologic complexity of both lines of HPRT-deficient
cells was described as more intricate. Over a period of 2 days
in culture, a higher percentage of the HPRT-deficient cells had
neurites, and the average neuritic length per cell was longer,
compared with their controls. Further, the HPRT-deficient
mouse cell line showed a higher average number of neurites
per cell. The somal area of the HPRT-deficient cells was also
greater. These in vitro data clearly support an effect of HPRT
deficiency on the proliferation and differentiation of neurons.
Also, the effect of HPRT deficiency on morphologic measures
31
Deutsch et al
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
of development may differ depending on whether the cell is
dopaminergic or nondopaminergic.6 When challenged to differentiate, the ‘‘dopaminergic’’ HPRT-deficient PC12 neuroblastoma cell line shows a reduced neurite outgrowth. In any
event, although the connection between HPRT deficiency and
neuronal development is not known, the failure of nerve cells
to salvage free purine basis has an effect on their proliferation
and differentiation.
There are some suggestive clinical data in support of
dopamine receptor supersensitivity in Lesch-Nyhan syndrome.11 Two patients (ages 20 months and 15 years) were
treated with fluphenazine. The 20-month-old patient was
reported to show improved motor ability and decreased selfbiting. The 15-year-old patient did not show any beneficial
effect while on active medication; however, the frequency and
severity of self-biting increased after the fluphenazine was
stopped. In the first instance, fluphenazine may have blocked
up-regulated dopamine receptors, whereas in the second case,
fluphenazine may have led to further up-regulation of dopamine receptor sensitivity. Self-biting, a consequence of the
additional up-regulation, may have been unmasked with fluphenazine’s discontinuation in the second case. Because
haloperidol, which is a more selective D2 dopamine receptor
antagonist than fluphenazine, was not shown to be effective in
Lesch-Nyhan syndrome, the authors suggested that D1 dopamine receptors may be up-regulated.11
The arborization of the dopaminergic terminals in the
striatum, which begins at about 3 months of age in the human,
may heighten the demand for purine nucleotides.2 This
heightened demand may explain the vulnerability of developing nigrostriatal dopamine terminals to HPRT deficiency
and the time of emergence of the first symptoms in LeschNyhan syndrome. Additionally, metabolic consequences of
HPRT-deficiency may include oxidative stress; nigrostriatal
dopamine neurons are particularly sensitive to the lethal and
toxic effects of oxidative stress.26,27 Thus, this sensitivity of
dopaminergic projections to oxidative stress may serve as the
link between HPRT-deficiency and striatal involvement in
Lesch-Nyhan syndrome.
Potential Pathophysiological Role of Purines,
Especially Guanine-Based Purines,
in Neurologic Manifestations
Altered purine metabolism has been studied in vitro in
cultures of neurons and astroglia prepared from the brains of
HPRT-deficient knockout mice and an HPRT-deficient
neuroma cell line.31 Collectively, these studies show clear
consequences of this biochemical lesion with respect to
intracellular concentrations of both purine and pyrimidine
nucleotides and the free purine and pyrimidine bases and their
concentrations in the culture media. However, these studies
may be only suggestive of possible consequences for small,
rapidly turning over pools of purine and pyrimidine bases and
their derivatives that may exist within discrete regions of
specific neurons and glial elements, where they may be
involved in processes such as synaptic transmission. In both
cultured neuroma and astroglia, the studies showed that
accelerated incorporation of radiolabeled adenine into nucleotides occurs as a consequence of the virtual absence of the
32
incorporation of radiolabeled hypoxanthine into nucleotides
in the HPRT-deficient condition. Further, the rate of de novo
purine synthesis, as reflected in the incorporation of [14C]formate into purines, is increased about 4- to almost 10-fold in
cultured neuroma and astroglia, respectively.31 These metabolic results were expected and probably are attributable to the
increased availability of PRPP. In cultured HPRT-deficient
neuroma and astroglia, relative to HPRT-positive control cells,
there was an increased rate of disappearance of prelabeled
adenine nucleotides from within the cells after a 24-hour
period of incubation, which was associated with an increased
concentration of radiolabeled purines appearing in the culture
medium. The increased concentration of labeled purines in the
culture medium of HPRT-deficient cells may result primarily
from an increased excretion of hypoxanthine by these cells.
Thus, because the hypoxanthine endogenously produced by
the HPRT-deficient cells is not reutilized, it is excreted. Elevated levels of hypoxanthine may influence GABAergic neurotransmission, serving as a low-affinity endogenous benzodiazepine-like ligand.1,23 In view of the high levels of guanase
activity in the brain, the in vitro studies suggest that HPRTdeficient cells may be relatively deficient in, or depleted of,
intracellular guanine and its derivatives; that is, exogenous
guanine cannot be salvaged, and what exists is a substrate for
guanase. Guanase catalyzes the conversion of guanine to
xanthine and ammonia. There is the possibility that as a result
of the failure to salvage guanine in Lesch-Nyhan syndrome
and the catabolic efficiency of guanase, neurotoxic accumulation of ammonia may occur.17,31
As noted, in neurons, the rate of the degradative
deamination of guanine to xanthine by guanase is significantly
greater than its anabolic incorporation into nucleotides by
HPRT; this is especially true in mature neurons.4 As a result
of the relatively lower availability of guanine because of the
high activity of guanase in brain, and the relatively higher
availability of hypoxanthine because of the absence of
xanthine oxidase activity, the HPRT-catalyzed incorporation
of hypoxanthine into nucleotides is about 3-fold greater than
that of guanine. Guanase activity in brain and its regulation of
guanine availability may be very important to synaptic
neurotransmission. The enzyme has a cytosolic localization,
and its activity is high and unevenly distributed in the brain.
The highest levels of guanase activity are found in the
thalamus and cerebral hemispheres. Interestingly, inhibition
of guanase activity was associated with decreased binding
of benzodiazepines to human brain membranes.4 Moreover,
under basal conditions, radioactivity incorporated into guanine
nucleotides is quickly transferred to purine degradation
products, consistent with the fast turnover rate of the guanine
nucleotide pool; the turnover rate of the guanine nucleotide
pool in brain is estimated to be at least twice that of the adenine
nucleotide pool.4 Thus, it can be expected that the metabolic
stress of HPRT deficiency in brain could lead to a depletion of
guanine-based purines in areas critical to normal synaptic
neurotransmission, including the area of glutamatergic
synapses.
Depletion of guanine-containing compounds, especially
within discrete pools, may affect energy metabolism, second messenger systems, and modulation of glutamatergic
q 2005 Lippincott Williams & Wilkins
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
neurotransmission, among other consequences. These metabolic consequences of HPRT deficiency in the brain may
contribute to the neurologic manifestations of Lesch-Nyhan
syndrome. As noted, relative to other organs, the brain is more
dependent on the salvage pathway for the synthesis of purine
nucleotides than it is on de novo synthesis. Thus, it was of
interest that the cellular content of GTP and ATP did not differ
between HPRT-deficient and HPRT-positive cultured neurons;
these neurons were cultured from the HPRT-deficient knockout and wild-type mice, respectively. The content of UTP,
a pyrimidine nucleotide, in the HPRT-deficient cultured neurons was elevated. In contrast, the HPRT-deficient cultured
astroglial cells showed significant reductions of ATP and GTP
concentrations as well as elevated levels of UTP. Elevated
levels of pyrimidine nucleotides may be caused by the increased availability of PRPP, a limiting substrate for the
synthesis of purine and pyrimidine nucleotides. The data
suggest that astroglia may experience a greater impact of the
HPRT deficiency on their purine nucleotide pools than neurons.16,31 The astroglial deficit can be expected to impact
energy metabolism and neurotransmission. In summary, the
HPRT-deficient rat neuroma cell line, which serves as a model
system for HPRT-deficient neurons, revealed the following
alterations of purine metabolism: accumulation of PRPP and
accelerated de novo synthesis, excessive production and excretion of hypoxanthine, and increased turnover rate of adenine
nucleotides.16 The synthesis of pyrimidine nucleotides is also
enhanced, as reflected in increased cellular levels of UTP. In
addition to these findings, primary cultures of HPRT-deficient
astroglial cells, derived from the cerebral cortices of a strain of
newborn HPRT-deficient transgenic mice, showed lowered
cellular levels of ADP, ATP, and GTP. The lowered levels of
purine nucleotides in HPRT-deficient astroglial cells suggest
that, in cells that are heavily dependent on and deficient in the
salvage pathway for maintenance of their cellular pools of
purine nucleotides, accelerated de novo synthesis cannot keep
up with the cellular demands. The data suggest that elevated
extracellular levels of hypoxanthine and diminished intracellular pools of purine nucleotides within the brain contribute
to the neurologic manifestations of Lesch-Nyhan syndrome.
Further, the consequences of these metabolic derangements
may depend on and differ as a function of the specific stage of
neurodevelopment and neurogenesis. Also, in addition to
obvious effects on energy metabolism and the maintenance of
unequal ionic concentration gradients across membranes, the
decreased intracellular purine nucleotide content may disrupt
synaptic transmission as a result of a failure to replenish
rapidly turning over cellular pools in discrete synaptic areas.
Inosine and hypoxanthine were extracted, purified, and
identified as possible endogenous ligands existing within
bovine brain that are capable of inhibiting the competitive
binding of [3H]diazepam to the central benzodiazepine
binding site.1,23 The ability of resolved fractions of the bovine
brain extract to inhibit the binding of [3H]diazepam to both
a crude membrane preparation of rat cerebral cortex (receptor
binding assay) and rabbit antibody generated against diazepam
(radioimmunoassay) served as the assay procedures to purify
inosine and hypoxanthine from the extract and identify them as
possible endogenous ligands for the central benzodiazepine
q 2005 Lippincott Williams & Wilkins
Lesch-Nyhan Syndrome and Guanine-Based Purines
binding site. Although they bound with low affinity, inosine
and hypoxanthine exhibited specificity for the central benzodiazepine binding site, showing little ability to inhibit the
binding of [3H]diazepam to the peripheral benzodiazepine
binding site in kidney or liver, or the binding of radioligands to
opiate receptors ([3H]dihydromorphine), muscarinic acetylcholine receptors ([3H]quinuclidinyl benzilate), the GABAA
agonist recognition site ([3H]muscimol), and b-adrenergic
receptors ([3H]dihydroalprenolol). Moreover, the IC50 of guanosine (1.0 mM) to inhibit the specific binding of [3H]diazepam was almost identical to that of inosine and hypoxanthine (1.3 mM). Again, although the binding is of low
affinity, pathologic interference with the endogenous modulation of GABAergic neurotransmission may be a consideration in Lesch-Nyhan syndrome, especially with respect to
elevated concentrations of hypoxanthine. The purines may
also be physiologically relevant modulators of GABAergic
neurotransmission as well because, as noted by Skolnick et al23
and Asano and Spector,1 brain concentrations of inosine,
hypoxanthine, and adenosine rise after chemical or electrical
depolarization. Further, significant pharmacological effects of
benzodiazepines are observed under conditions when a relatively small percentage of central benzodiazepine binding sites
are occupied (eg, 10%–20%). For example, 50% of mice are
protected against pentylenetetrazole (PTZ)-induced seizures
when only about 15% of the central benzodiazepine binding
sites are occupied by diazepam.23 Thus, even at the estimated
normal physiological concentrations of inosine and hypoxanthine (reported to be between 20 and 60 mM), in spite of their
low affinity, they may modulate inhibitory tone in the brain.23
Using primary cultures of astrocytes prepared from the
cerebral hemispheres of 18- to 19-day-old rat fetuses, it was
shown that under basal conditions they spontaneously release
guanine-based purines; in fact, the amount of guanine-based
purines released over a 3-hour period was greater than that of
adenine-based purines.5 Moreover, the exposure of these
cultures to hypoxic and hypoglycemic conditions resulted in
a sustained severalfold increase in the release of guanine- and
adenine-based purines over basal values during and up to 90
minutes after the insult. Guanosine, in particular, showed
a progressive increase of its extracellular concentration after
the stress. Importantly, the increased release of guanine- and
adenine-based purines during and after the combined hypoxic
and hypoglycemic insults was not an artifact of diminished cell
viability.5 The release of guanine-based purines in this stroke
model is consistent with hypothesis that these compounds,
especially guanosine, may exert immediate modulatory effects
on synaptic transmission and more sustained trophic effects.
Conceivably, if discrete intracellular astrocytic pools of
guanine-based purines are depleted in Lesch-Nyhan syndrome, the ability to modulate synaptic transmission under
conditions of hypoxia and hypoglycemia would be compromised in this metabolic disorder.
The potential ability of exogenously administered
guanosine and inosine to provide an alternative source of
energy to ATP has been suggested as an explanatory hypothesis for their neuroprotective effects in the context of
oxidative stress and damage.14 For example, after exposure to
rotenone, an inhibitor of the mitochondrial respiratory chain,
33
Deutsch et al
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
and the induction of chemical hypoxia, inosine and guanosine
were shown to preserve the viability of astrocytes and neurons
in primary mixed cultures of spinal cord prepared from 13- to
14-day mouse embryos.14 Presumably, these primary cultures
represent ‘‘normal’’ cells, as opposed to transformed cultures
derived from tumor cells; thus, the neuroprotective effect of
the purine nucleosides to chemical hypoxia may be a normally
expected response. The spinal cord neurons were more
sensitive to the lethal toxic effects of rotenone (5 mM) than
the astrocytes; more than 90% of the neurons lost viability 260
minutes after the addition of rotenone. Inosine afforded dosedependent protection, delaying cell death and preserving
viability over a dosage range of 125 mM to 1 mM. From
a morphologic perspective, inosine was also able to delay
neuronal swelling and disintegration of dendrites in a dosedependent manner after the addition of rotenone to the cultures. The likely cause of neuronal swelling is diminished
production and decreased quantities of ATP as a result of
rotenone’s interference with mitochondrial respiration and
oxidative phosphorylation. The hydrolysis of ATP is essential
to the maintenance of unequal ionic concentration gradients
across cell membranes and preservation of normal osmolality
within the cell. As for inosine, the addition of guanosine (500 mM)
to the primary mixed cultures of mouse embryonic spinal cord
3 minutes before the addition of rotenone (5 mM) preserved
both total cell and neuronal cell viability. The protective effect
of inosine (500 mM) on neuronal cell viability decayed rapidly
when it was added to the cultures at increasing times after the
addition of rotenone; it was most potent when added 3 minutes
prior to and 5 minutes after the addition of rotenone, whereas
its neuroprotective effect was almost completely lost when it
was added 30 minutes after rotenone. Again, it was hypothesized that the ability of these purine nucleosides to maintain
cellular levels of ATP above a critical threshold under conditions of hypoxia serves as the mechanism for their
preservation of cell viability. Indeed, the addition of a purine
nucleoside phosphorylase inhibitor to the cultures, which
would interfere with a pathway for the participation of purine
nucleosides in the production of ATP under anaerobic conditions, attenuated their protective effect; this effect of the purine
nucleosides to preserve cell viability was most dramatic with
neurons. The data also suggested that neuronal protection by
purine nucleosides is either dependent on or enhanced by the
presence of glia.14
In brain, there are probably several discrete pools of
nucleosides and nucleotides subserving functions in both
metabolism and neurotransmission. Presumably, these pools
within the brain are perturbed in Lesch-Nyhan syndrome. As
already emphasized, relative to other organs, the brain is more
dependent on the salvage pathway catalyzed by HPRT to
maintain pools of nucleosides and nucleotides than the de
novo pathway for the biosynthesis of purines.2 The dependence on the salvage pathway may be especially pronounced in the basal ganglia. Unfortunately, various attempts
to replenish the presumptively depleted pools of nucleosides
and nucleotides in brain in patients with Lesch-Nyhan
syndrome with the exogenous dietary administration of
GMP, AMP, IMP, inosine, and adenine to treat CNS symptoms
have not met with success.2 Some of this failure could relate to
34
problems with intestinal absorption, poor transport across cell
membranes, uptake into peripheral pools (eg, within erythrocytes), metabolism within the periphery, and poor penetrability across the blood–brain barrier.2 However, there are data
supporting the existence of nucleoside transporters across
intestinal cells and the blood–brain barrier25; thus, poor
bioavailability of nucleosides may not account for their lack of
therapeutic efficacy.
GUANINE-BASED PURINES: MODULATORS OF
GLUTAMATERGIC NEUROTRANSMISSION AND
TROPHIC FUNCTIONS IN THE CNS
From a quantitative perspective, L-glutamate is one of
the most important neurotransmitters in mammalian brain,
mediating synaptic transmission at 30% to 50% of all synapses. Thus, regulation of its extracellular concentration is
crucial for normal physiological function; elevated levels of Lglutamate are excitotoxic and can lead to cell death. Astrocytes
have a sodium-dependent uptake system for L-glutamate that
may be particularly relevant at high (100 mM), as opposed to
low (1 mM), concentrations of L-glutamate.10 Thus, the astrocytic uptake of L-glutamate may be critical for maintaining
its extracellular concentrations below neurotoxic levels. By
use of primary astrocyte cultures prepared from the cortices of
1-day-old Wistar rats and rat brain cortical slices, guanosine
was shown to promote the sodium-dependent uptake of Lglutamate (100 mM) in a dose-dependent manner; uptake was
measured for 7 minutes.9 The minimum effective concentration of guanosine was 100 nM, which resulted in a 19%
increase over control values. The maximal stimulation of
uptake by guanosine was 63% over control values, and the
EC50 was 2.47 6 0.27 mM. Importantly, adenosine (100 mM)
affected neither basal uptake nor the stimulatory effect of
guanosine (1 mM). Theophylline (100 mM), a nonspecific
A1/A2 adenosine receptor antagonist, stimulated basal uptake
of L-glutamate but did not affect the stimulatory effect of
guanosine (1 mM). Finally, dipyridamole (40 mM), a nucleoside transport inhibitor, stimulated basal uptake, and this
stimulatory effect was additive with that of guanosine.9 The
data suggest that guanosine, a guanine-based purine, mediates
a relatively specific stimulatory effect on the astrocytic uptake
of L-glutamate because the effect was not mimicked by
adenosine, an adenine-based purine, nor blocked by theophylline. Thus, this effect is probably not mediated by adenosine
receptors. Further, the effect is an extracellular one, occurring
at the cell’s surface because it was not abolished by the
nucleoside transport inhibitor. In fact, inhibition of nucleoside
transport enhanced the stimulatory effect of guanosine (1 mM),
suggesting that more of this nucleoside was available extracellularly for this action, which is mediated at the cell’s
surface.9
This stimulatory effect of guanosine (100 and 300 mM)
on the uptake of L-glutamate (100 mM) by astrocytic cells in
culture could be mimicked by GMP (100 and 300 mM) and
GTP (100 and 300 mM). However, a significant additive effect
on uptake was not observed with the simultaneous addition
of all 3 guanine-based purines to the culture medium (ie,
q 2005 Lippincott Williams & Wilkins
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
guanosine, GMP, and GTP, 100 mM each), compared with the
addition of each compound alone. These data are consistent
with a saturating effect of guanine-based purines occurring at
or just above concentrations of 100 mM and the possibility that
1 of these 3 compounds is mediating the stimulatory effect on
uptake; the 3 compounds are metabolically interconvertible
with each other. Importantly, a poorly hydrolyzable analogue
of GTP was not able to stimulate the uptake of L-glutamate by
cultured astrocytes, nor could GMP (100 mM) after cultures
were pretreated with 100 mM of a,b-methyleneadenosine 5#diphosphate (AOPCP), the ecto-5#-nucleotidase inhibitor that
prevents the conversion of GMP to guanosine. Thus, guanosine appears to mediate the stimulatory effect of all the
guanine-based purines.10 Further, the stimulatory effect of
guanine-based purines on L-glutamate uptake was not observed with 100 mM concentrations of adenosine, AMP, ADP,
or ATP. Finally, guanosine (100 mM) did not affect the uptake
of GABA (1 mM) by cultured astrocytes over a period of 3, 10,
and 30 minutes. Therefore, guanosine mediates the stimulatory effect of guanine-based purines on the astrocytic uptake of
L-glutamate, which is a process that is relatively specific and
selective. This stimulatory effect of guanosine was present
after cells were treated with an inhibitor of nucleoside transporters, consistent with this action occurring after the binding
of guanosine to the cell surface. The astrocytic uptake of
L-glutamate is a mechanism for terminating its actions within
the synapse; thus, the facilitation of uptake by guanine-based
purines may be an important regulator of glutamatergic
neurotransmission, especially under excitotoxic conditions.
Guanine-based purines (including GTP and GMP)
decrease the sodium-independent binding of [3H]glutamate
(40 nM) to rat forebrain homogenates under conditions of
saturating G proteins with a bound nonhydrolyzable analogue
of GTP.19 The sodium-independent binding of [3H]glutamate
would most likely label extracellular receptors, whereas
binding measured in the presence of sodium ions would also
label transporter sites. Magnesium ions were also eliminated
from the incubations to measure the binding of [3H]glutamate
to cell-surface receptors because magnesium ions are involved
in the formation of the receptor/G-protein complex; thus, these
results are not likely to be confounded by mechanisms involving the participation of G-proteins. Measurable inhibition
of [3H]glutamate binding was observed with a concentration
of GMP as low as 0.1 mM, and maximal inhibition of about
30% to 40% was observed over a range of GMP concentrations from 1 mM through 1 mM. These data suggest that
guanine-based purines may inhibit the transduction of the
glutamate signal, independently of any G-protein ‘‘coupling’’
mechanisms.19
The intracerebroventricular infusion of quinolinic acid
(39.2 mM; 4 mL, 156.8 nmol), an excitatory amino acid
glutamatergic agonist, produces tonic-clonic seizures in 100%
of adult male Wistar rats.25 This dose of quinolinic acid was
the lowest dose associated with the induction of seizures in all
of the control animals. Using this chemical procedure for the
induction of seizures, guanosine and GMP (7.5 mg/kg) were
shown to have anticonvulsant effects, even after their
peripheral intraperitoneal administration 30 minutes before
the infusion of quinolinic acid.25 At an intraperitoneal dose of
q 2005 Lippincott Williams & Wilkins
Lesch-Nyhan Syndrome and Guanine-Based Purines
7.5 mg/kg of guanosine and GMP, about 50% of the animals
were protected and did not show any tonic-clonic seizure
activity over a 10-minute period after the infusion of quinolinic
acid. Importantly, the intraperitoneal administration of guanosine (7.5 mg/kg) or GMP (7.5 mg/kg) was associated with
a two- and three-fold increase in the levels of guanosine in the
CSF, respectively, whereas levels of GMP and adenosine were
unchanged. These data are consistent with the extracellular
conversion of guanine nucleotides to guanosine in brain and
the latter compound involved in the modulation of glutamatergic neurotransmission.25 The intracerebroventricular infusion of GMP prior to infusion of quinolinic acid provided
a dose-dependent protection against the induction of seizures.
Consistent with the anticonvulsant effect of guanine-based
purines mediated by guanosine, this protective effect of
intracerebroventricularly administered GMP was significantly
reduced by prior intracerebroventricular infusion of AOPCP,
the inhibitor of the ecto-5#-nucleotidase that is responsible for
the conversion of GMP to guanosine. Further, AOPCP caused
a dramatic reduction in the concentration of guanosine in the
CSF after infusion of GMP. The data are also consistent with
a rapid extracellular conversion of GMP to guanosine in the
brain because, in the absence of AOPCP, the concentration of
guanosine was more than 17 times greater than the concentration of GMP five minutes after the intracerebroventricular
injection of 4 mL of a 240 mM solution (960 nmol) of GMP.
This same dose of GMP reduced the percentage of animals that
displayed tonic-clonic seizures after infusion of quinolinic acid
from 100% to about 20%, an anticonvulsant effect that was
attenuated by AOPCP. Although guanosine and GMP could
antagonize seizures precipitated by the intracerebroventricular
injection of quinolinic acid, they did not antagonize seizures
precipitated by the subcutaneous injection of picrotoxin (3.2
mg/kg).21 Picrotoxin is a noncompetitive GABAA receptor
antagonist; thus, the data are consistent with the specificity of
the effects of guanosine and GMP for the glutamatergic
system. These data are evidence of an important functional
role for guanosine as a modulator of glutamatergic neurotransmission in the intact animal; moreover, this anticonvulsant role of guanosine appears to be a direct one that is not
dependent on the generation of adenosine. Further, either
guanosine or a derivative may be a candidate compound(s) for
development as a medication with possible anticonvulsant and
neuroprotective effects after oral administration. The transport
of nucleosides across intestinal cells and the blood-brainbarrier has been characterized.25 Conceivably, diminished
neuromodulatory pools of guanine-based purines contribute in
some significant way to the pathophysiology of the LeschNyhan syndrome; unfortunately, replenishing these pools via
the exogenous administration of nucleosides may be insufficient to address the neurologic manifestations. These
manifestations result from early neurodevelopmental abnormalities, in addition to presumed alterations of synaptic
neurotransmission
The exogenous administration of guanosine and other
guanine-based purines, including their peripheral administration, has functional neurobehavioral consequences, which are
likely mediated by an effect on glutamatergic neurotransmission.18 Inhibitory avoidance in rats is a 1-trial aversive learning
35
Deutsch et al
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
procedure that is a useful paradigm to test effects of drugs on
learning and memory. In this paradigm, the effect on the
latency of rats for stepping off a platform is assessed 24 hours
after a training session during which they receive an electric
footshock immediately after stepping down from the platform.
Ordinarily, the latency to step-down is dramatically increased
24 hours after this training session, relative to the control
condition. Guanosine (2.0 mg/kg and 7.5 mg/kg, intraperitoneally) showed a significant dose-dependent ability to reduce
the latency to step-down from the platform when it was
administered 30 minutes before the training session 24 hours
earlier.18 This guanosine-induced impairment of inhibitory
avoidance performance was reversible: no deficit of treatment
with guanosine (7.5 mg/kg) was observed when rats showing
the guanosine-induced deficit were resubjected to the training
and testing procedures 1 week later. Further, whereas the
memory deficit associated with guanosine (7.5 mg/kg) was
observed when it was administered 30 minutes before training,
there were no significant deficits of inhibitory avoidance
performance when it was administered immediately after
training or 24 hours before testing. Thus, to affect the process
of memory consolidation, guanosine must be administered at
least a finite time before the aversive training procedure, and
it does not appear to affect memory retrieval for inhibitory
avoidance performance. Guanosine did not affect the nociceptive sensitivity to parameters of the electric footshock, nor
did it affect open-field behaviors when it was administered
before testing. Thus, there is a relatively selective behavioral
effect on a specific type of learning and memory involved in
the performance of this inhibitory avoidance behavior.
Guanosine’s ability to enhance the uptake of L-glutamate by
astrocytes suggests that the effect on inhibitory avoidance
performance may result from negative modulation or interference with glutamatergic neurotransmission. Glutamatergic mechanisms are very much implicated in learning and
memory, especially the induction of hippocampal long-term
potentiation.
In addition to their effects on neurotransmission, which
take place in a time frame of milliseconds to seconds, guaninebased purines and other purines also have important trophic
functions, affecting the development, structure, or maintenance of target cells in the brain.17 These trophic actions may
take hours to days to evolve fully. Guanosine (10 mM) and
GTP (10 mM) have been shown to stimulate the synthesis and
release of immunoreactive nerve growth factor from cultured
neonatal mouse astrocytes. Thus, some of the trophic actions
of purines may be indirect, occurring as a result of stimulating
the synthesis and release of trophic factors and/or enhancing
the effects of these specific trophic factors. High extracellular
levels of guanosine and GTP under conditions of hypoglycemia and hypoxia may secondarily result in increased
extracellular levels of adenosine and ATP, whose trophic
actions may be mediated, at least in part, by specific purinergic
receptors. Conceivably, guanine-based purines would increase
extracellular levels of adenine-based purines by interfering
with their uptake and metabolism as well as by stimulating
their release. Therefore, at least some of the trophic actions of
guanine-based purines may be mediated indirectly by adeninebased purines. For example, the ability of guanine-based
36
purines to stimulate proliferation of rat brain microglia in
a concentration-dependent manner appears to be mediated by
specific purinergic receptors that recognize adenine-based
purines. Guanosine and GTP can promote neuritic outgrowth
from cultured embryonic mouse hippocampal neurons.17
These ‘‘neuritogenic’’ effects may be synergistic with those
of nerve growth factor. Derivatives of hypoxanthine and
xanthine are being considered as potential therapeutic agents
because of their abilities to stimulate the synthesis and
secretion of a variety of growth factors from cultured mouse
astrocytes, including nerve growth factor, neurotrophin-3, and
a specific fibroblast growth factor. As noted, elevated extracellular levels of hypoxanthine are observed in the brains of
patients with Lesch-Nyhan syndrome. Importantly, with respect to a specific neurotrophic role for guanosine, elevated
levels may be maintained for up to a week after focal brain
injury. In any event, because of the metabolic consequences of
an HPRT deficiency and the possibility of diminished availability of guanine-based purines subserving trophic functions,
patients with Lesch-Nyhan syndrome may have problems with
both neurodevelopment and their ability to respond to a toxic
brain insult.
DISCUSSION
Guanine-based purines are undoubtedly involved in the
pathogenesis of the Lesch-Nyhan syndrome. Investigation of
this disorder and the neurobiological consequences of the HPRT
deficiency will almost certainly clarify our current understanding of the potential roles that guanine-based purines play in
neurodevelopment and as neuromodulators and neurotransmitters in addition to their more familiar roles in gene replication
and expression, metabolism, and second messenger cascades,
among other roles. Data support the possible involvement of
hypoxanthine and guanine-based purines, whose levels are
perturbed in Lesch-Nyhan syndrome, in neuorotransmission
mediated by GABA and L-glutamate, respectively. Additionally,
guanine-based purines may have important trophic functions in
the brain. HPRT-deficiency is associated with immaturity and
dysfunction of dopaminergic nerve terminals; however, the
mechanism of this neurodevelopmental abnormality of dopaminergic neurotransmission is unclear. Cell culture systems and
HPRT-deficient knockout mice provide good models for conducting many of these studies; additionally, behavioral models
have been used successfully to demonstrate the neuromodulatory role of guanine-based purines on glutamatergic systems.
Unfortunately, these studies have not yet led to more effective
pharmacological strategies for the treatment of Lesch-Nyhan
syndrome than the current empirical reliance on antipsychotic
medications, benzodiazepines, and antidepressants, which have,
at best, only modest efficacy.28 Specialized nonpharmacological
operant behavioral strategies may be valuable and essential
components of a multimodal therapeutic program for this
inherited disorder of purine metabolism. Importantly, a hypothesized deficiency of guanosine in discrete pools associated with
glutamatergic synapses might stimulate therapeutic strategies to
promote glutamate reuptake and/or antagonism of effects
mediated by glutamatergic receptors.
q 2005 Lippincott Williams & Wilkins
Clin Neuropharmacol Volume 28, Number 1, January – February 2005
ACKNOWLEDGMENT
The first author would like to thank Dr Rody P. Cox for
teaching him about the clinical and metabolic aspects of the
Lesch-Nyhan syndrome and serving as the ideal model of
a compassionate physician investigator.
REFERENCES
1. Asano T, Spector S. Identification of inosine and hypoxanthine as endogenous ligands for the brain benzodiazepine-binding sites. Proc Natl
Acad Sci USA. 1979;76:977–981.
2. Baumeister AA, Frye GD. The biochemical basis of the behavioral disorder
in the Lesch-Nyhan syndrome. Neurosci Biobehav Rev. 1985;9:169–178.
3. Boer P, Brosh S, Wasserman L, et al. Decelerated rate of dendrite
outgrowth from dopaminergic neurons in primary cultures from brains of
hypoxanthine phosphoribosyltransferase-deficient knockout mice. Neurosci Lett. 2001;303:45–48.
4. Brosh S, Sperling O, Dantziger E, et al. Metabolism of guanine and
guanine nucleotides in primary rat neuronal cultures. J Neurochem. 1992;
58:1485–1490.
5. Ciccarelli R, Iorio PD, Giuliani P, et al. Rat cultured astrocytes release
guanine-based purines in basal conditions and after hypoxia/hypoglycemia. Glia. 1999;25:93–98.
6. Connolly GP. Hypoxanthine-guanine phosphoribosyltransferase-deficiency produces aberrant neurite outgrowth of rodent neuroblastoma used to
model the neurological disorder Lesch Nyhan syndrome. Neurosci Lett.
2001;314:61–64.
7. Connolly GP, Duley JA, Stacey NC. Abnormal development of
hypoxanthine-guanine phosphoribosyltransferase-deficient CNS neuroblastoma. Brain Res. 2001;918:20–27.
8. Ernst M, Zametkin AJ, Matochik JA, et al. Presynaptic dopaminergic
deficits in Lesch-Nyhan disease. N Engl J Med. 1996;334:1568–1572.
9. Frizzo MES, Lara DR, Dahm KCS, et al. Activation of glutamate uptake by
guanosine in primary astrocyte cultures. Neuroreport. 2001;12:879–881.
10. Frizzo MES, Soares FAA, Dall’Onder LP, et al. Extracellular conversion
of guanine-based purines to guanosine specifically enhances astrocyte
glutamate uptake. Brain Res. 2003;972:84–89.
11. Goldstein M, Anderson LT, Reuben R, et al. Self-mutilation in Lesch-Nyhan
disease is caused by dopaminergic denervation. Lancet. 1985;1:338–339.
12. Harris JC, Lee RR, Jinnah HA, et al. Craniocerebral magnetic resonance
imaging measurement and findings in Lesch-Nyhan syndrome. Arch
Neurol. 1998;55:547–553.
13. Kopin IJ. Neurotransmitters and the Lesch-Nyhan syndrome. N Engl J
Med. 1981;305:1148–1150.
14. Litsky ML, Hohl CM, Lucas JH, et al. Inosine and guanosine preserve
neuronal and glial viability in mouse spinal cord cultures during chemical
hypoxia. Brain Res. 1999;821:426–432.
q 2005 Lippincott Williams & Wilkins
Lesch-Nyhan Syndrome and Guanine-Based Purines
15. Lloyd KG, Hornykiewicz O, Davidson L, et al. Biochemical evidence
of dysfunction of brain neurotransmitters in the Lesch-Nyhan syndrome.
N Engl J Med. 1981;305:1106–1111.
16. Pelled D, Sperling O, Zoref-Shani E. Abnormal purine and pyrimidine
nucleotide content in primary astroglia cultures from hypoxanthineguanine phosphoribosyltransferase-deficient transgenic mice. J Neurochem. 1999;72:1139–1145.
17. Rathbone MP, Middlemiss PJ, Gysbers JW, et al. Trophic effects of
purines in neurons and glial cells. Prog Neurobiol. 1999;59:663–690.
18. Roesler R, Vianna MRM, Lara DR, et al. Guanosine impairs inhibitory
avoidance performance in rats. Neuroreport. 2000;11:2537–2540.
19. Rubin MA, Medeiros AC, Rocha PCB, et al. Effect of guanine nucleotides
on [3H]glutamate binding and on adenylate cyclase activity in rat brain
membranes. Neurochem Res. 1997;22:181–187.
20. Saito Y, Ito M, Hanaoka S, et al. Dopamine receptor upregulation in
Lesch-Nyhan syndrome: a postmortem study. Neuropediatrics. 1999;30:
66–71.
21. Schmidt AP, Lara DR, Maraschin JDF, et al. Guanosine and GMP prevent
seizures induced by quinolinic acid in mice. Brain Res. 2000;864:
40–43.
22. Silverstein FS, Johnston MV, Hutchinson RJ, et al. Lesch-Nyhan
syndrome: CSF neurotransmitter abnormalities. Neurology. 1985;35:
907–911.
23. Skolnick P, Paul SM, Marangos PJ. Purines as endogenous ligands of the
benzodiazepine receptor. Fed Proc. 1980;39:3050–3055.
24. Smith DW, Friedmann T. Characterization of the dopamine defect in
primary cultures of dopaminergic neurons from hypoxanthine phosphoribosyltransferase knockout mice. Mol Ther. 2000;1:486–491.
25. Soares FA, Schmidt AP, Farina M, et al. Anticonvulsant effect of GMP
depends on its conversion to guanosine. Brain Res. 2004;1005:182–
186.
26. Visser JE, Bar PR, Jinnah HA. Lesch-Nyhan disease and the basal
ganglia. Brain Res Brain Res Rev. 2000;32:449–475.
27. Visser JE, Smith DW, Moy SS, et al. Oxidative stress and dopamine
deficiency in a genetic mouse model of Lesch-Nyhan disease. Dev Brain
Res. 2002;133:127–139.
28. Watts RWE, Spellacy E, Gibbs DA, et al. Clinical, post-mortem,
biochemical and therapeutic observations on the Lesch-Nyhan syndrome
with particular reference to the neurological manifestations. Q J Med
[New Ser]. 1982;51:43–78.
29. Wilson JM, Young AB, Kelley WN. Hypoxanthine-guanine phosphoribosyltransferase deficiency. The molecular basis of the clinical syndromes. N Engl J Med. 1983;309:900–910.
30. Wong DF, Harris JC, Naidu S, et al. Dopamine transporters are markedly
reduced in Lesch-Nyhan disease in vivo. Proc Natl Acad Sci USA.
1996;93:5539–5543.
31. Zoref-Shani E, Boer P, Brosh S, et al. Purine nucleotide metabolism in
cultured neurons and astroglia from HPRT-deficient knockout mice. Ital J
Biochem. 2001;50:9–13.
37