David Gyll, med. stud. Läkarprogrammet Uppsala Universitet Institutionen för neurovetenskap Sektionen för psykiatri Akademiska sjukhuset, Uppsala Handledare: Leif Lindström Examinator: Lars Hillered

___________________________________________________________________________

Significantly lower levels of hypoxanthine and

creatinine in the CSF of patients with schizophrenia

Abstract

Levels of hypoxanthine, xanthine, uric acid (UA), homovanillic acid (HVA), 5-hydroxyindoleacetic acid

(5-HIAA) and creatinine were measured in the cerebrospinal fluid (CSF) of 49 patients with

schizophrenia and 21 healthy controls. Twenty-six of the patients were first admissions who had

never received antipsychotic treatment. We found significantly lower levels of hypoxanthine in the

CSF of the patients (p=0.02) – the mean±SD was 3.09±0.57 µmol/l for controls and 2.72±0.69 µmol/l

for patients with schizophrenia. Female schizophrenics had significantly lower levels of creatinine in

their CSF than controls (p=0.01) –the mean±SD was 3.04±0.56 µmol/l for female controls and

2.67±0.74 µmol/l for females with schizophrenia. There were also numerous positive correlations

between the substances in the CSF. But these positive correlations between metabolites in the CSF

need to be interpreted with caution as it could be caused by a similar grade of dilution.

The finding of lowered levels of hypoxanthine in schizophrenia is interesting in the light of the purine

hypothesis of schizophrenia – which proposes that schizophrenia, at least in part, may be caused by

an imbalance in adenosine signaling. But also because of the recent interest the xanthine-oxidase

inhibitor, and hypoxanthine isomer, allopurinol has gained as an adjuvant treatment in

schizophrenia.

Sammanfattning

Koncentrationen av hypoxantin, xantin, urat (UA), homovanillinsyra (HVA), 5-hydroxyindolacetat (5-

HIAA) och kreatinin i cerebrospinalvätskan (CSF) uppmättes hos 49 patienter med schizofreni och hos

21 friska kontroller. Tjugosex av patienterna var nyinsjuknade och hade aldrig behandlats med

antipsykotika. Vi fann signifikant lägre nivåer av hypoxantin i ryggmärgsvätskan hos patienterna

(p=0.02) – medelvärdet±SD var 3.09±0.57 µmol/l hos kontrollgruppen och 2.72±0.69 µmol/l hos

patienterna med schizofreni. Kvinnor med schizofreni hade signifikant lägre nivåer av kreatinin i sin

ryggmärgsvätska jämfört med kvinnor i kontrollgruppen (p=0.01). Medelvärdet±SD var 3.04±0.56

µmol/l hos kvinnor I kontrollgruppen och 2.67±0.74 µmol/l hos kvinnorna med shizofreni. Vi fann

också flera signifikanta positiva korrelationer mellan ämnen i ryggmärgsvätskan. Dessa fynd måste

dock tolkas med stor försiktighet då det kan orsakas endast av en gemensam spädningsnivå.

Att patienter med schizofreni har lägre nivåer av hypoxantin isin ryggmärgsvätska är intressant med

tanke på purinhypotesen – som föreslår att störningar i adenosinsignaleringen åtminstone till viss del

kan ligga bakom schizofreni. Nivåer av hypoxantin är också värda att studera då allopurinol, en

hypoxantin-isomer och xantin oxidas-hämmare, har prövats och visat effekt i behandlingen av

schizofreni.

1

Significantly lower levels of

hypoxanthine and creatinine in the CSF

of patients with schizophrenia

Introduction

Schizophrenia is one of the leading causes of morbidity in the world. According to the WHO

schizophrenia is the 8th leading cause of Disability-Adjusted life years in people 15-44 years old.(1) It

causes a lot of suffering for both patients and relatives – illustrated by the fact that active

schizophrenia was ranked the health state with the highest disability weight in a recent Lancet

article.(2) Despite many decades of research and hundreds of thousands of publications the

pathophysiology of schizophrenia remains unclear and findings have often been discrete. These

findings point towards a complex aetiology with hereditary and environmental factors interacting in

promoting the disorder.(3)

Symptoms in psychotic disorders (including e.g. bipolar disorder and schizophrenia) can be classified

into 5 categories: I) Positive symptoms – delusions and hallucinations II) Negative symptoms –

encompassing lack of drive and motivation III) Cognitive impairment – loss of executive functions and

memory difficulties IV) Manic symptoms V) Depressive symptoms. Schizophrenia is characterised by

symptoms from the first three of these groups and differentiates itself from the other psychotic

disorders through less affective dysregulation.(4) According to numerous models schizophrenia could

be considered as a systemic disease which majorly affects the brain. A number of neurotransmitters

have been found altered in the schizophrenic brain and both dopaminergic, serotonergic,

glutamatergic, cholinergic and GABAergic models have been used to explain the diverse symptoms

seen in schizophrenia.(5) Dopamine is the main target for current treatment in schizophrenia and all

antipsychotics acts by blocking dopamine receptors. (4) Though relatively effective against the

positive symptoms in schizophrenia they are ineffective against cognitive deficits and negative

symptoms.(6) This is in line with increasing evidence that negative symptoms may be caused by a

hypodopaminergic state in the prefrontal cortex(7) contrary to the hyperdopaminergic state in the

basal ganglia which is thought to cause the positive symptoms.(8)

Purines Purines have multiple biologically fundamental roles – they are a crucial component of cells DNA and

RNA, they supply the cell with energy in the form of ATP and GTP. Beside this purine receptors are

2

evolutionary old and widely distributed in the human body – especially the nervous system.(9) The

diverse role of purine signalling can be understood by the fact that ATP functions as a cotransmitter

in all nerves in the central and the peripheral nervous system.(10) Purinergic receptors are classified

into two types based on what substance acts as the main effector – P1 receptors (with adenosine as

main effector) and P2 receptors (which reacts mainly on ATP).(11) Adenosine has an important role

as neuromodulator in the release of multiple other neurotransmitters. For example does the

adenosine receptor A1 seem to inhibit the release of all classical neurotransmitters(12) and the A2A-

receptor interacts with the dopamine D2-receptor which is also the main target of antipsychotic

medications.(13) Nucleotide signalling is also present early in embryonic neurodevelopment and has

numerous effects on neurogenesis(14), plastic remodelling and regeneration later in life.(9)

A purinergic hypothesis of schizophrenia was first laid forward by Lara et al. in 2000. According to

this theory an imbalance in the purinergic system would result in reduced adenosinergic activity

which could explain for the imbalance in both dopaminergic and glutamatergic neurotransmission. It

could also explain the neuro-immunological findings in schizophrenia.(15) (5) The hypothesis is mainly

built upon the fact that allopurinol as adjuvant therapy significantly improved symptoms in patients

with poorly responsive schizophrenia.(16) Allopurinol – which is otherwise mostly used in the

treatment of gout and hyperuricaemia – is a hypoxanthine isomer and acts as a xanthine oxidase

inhibitor in the body thus reducing formation of uric acid.(17) According to Brunstein et al. the first

psychiatric symptom which was treated with allopurinol was aggressive behaviour in dementia.(16)

The idea came from the fact that Lesch-Nyhan disease, a disorder of purine metabolism, is

characterised by mental retardation, self-mutilation and aggressive behaviour.(18)

Attached to this study is a mapping of the purine metabolism from the impressive Kyoto

Encyclopaedia of Genes and Genomes.(19) Metabolites analysed in this study is marked on the map

in red, adenosine in green and xanthine-oxidase, the therapeutic goal of allopurinol, in blue. The

picture also illustrates the complexness of metabolomics research and all the possible genes that

could play a role in shifting a transmitter balance.

Two-hit model and dopamine Lara et al.(20) present a two-hit model of schizophrenia which is consistent with the widely accepted

neurodevelopmental model. They postulate that increased levels of adenosine pre- and perinatally

have neurotoxic effects and can lead to defective neurodevelopment due to aberrant glutamergic

transmission. The increased concentrations of adenosine can also cause a loss of adenosine A1-

receptors which leads to the hyper-dopaminergic state in young adults - due to loss of inhibition.(11)

This view is supported by an experiment where mice 3-14 days old treated with an A1 adenosine

3

receptor agonist caused a reduction in A1 receptor expression. The treatment also caused white

matter reduction, ventriculomegaly, axon volume reduction and neuronal loss(21) - a picture in some

ways similar to findings in patients with schizophrenia.(22) The idea that A1-receptor expression

contribute to schizophrenia receives some support in a recent genetic study which link certain A1

receptor gene (ADORA1) polymorphisms to the disorder.(23)

Adenosine, glutamate and inflammation Apart from dopaminergic receptors adenosine also modulates glutamergic signalling in the brain

which according to Boison et al.(15) could give an explanation to the altered glutamergic transmission

seen in schizophrenia. They propose that an altered adenosinergic signalling could inhibit NMDA-

receptors (the main excitatory receptor of glutamate) in hippocampal regions which is believed to

play a role in the cognitive symptoms seen in schizophrenia. This hypothesis gains some support from

a recent study showing that transgenic mice with overexpression of adenosine kinase (the key

enzyme of adenosine clearance) – leading to lower levels of adenosine – cause attentional

impairments linked with schizophrenia.(24)

That there is an altered inflammatory response in schizophrenia is indicated by the strong negative

correlation between schizophrenia and rheumatoid arthritis.(25) In light of the increasing interest in

inflammations role in neuropsychiatric diseases purine signalling is also of interest since adenosine is

a potent regulator of neutrophil and monocyte function as well as a vasodilator and anti-

inflammatory agent.(26)

Previous findings Hypoxanthine and xanthine are end products in the catabolism of adenosine. The concentrations of

these metabolites in cerebrospinal fluid (CSF) have been shown to correlate with a number of

psychiatric and neurologic diseases. One study found significantly increased levels of xanthine in

patients with dementia of Alzheimers type(27) and both hypoxanthine and xanthine have also shown

correlation to certain psychiatric variables in patients with major depressive disorder.(28)

Concentrations of purine metabolites in plasma have been studied by Yao et al.(29) which have

shown significantly higher levels of xanthosine and lower levels of guanine in patients compared to

healthy controls. It is however difficult to draw conclusions about the concentrations of these

metabolites in the brain since plasma levels can be significantly lower than CSF-levels.(30) While Lara

et al.(5) emphasizes purines regulatory and neurodevelopmental role Yao et al.(31) focus on the

purine metabolisms role in the antioxidant defence system. Levels of purines have also been

measured in urine without finding any significant differences between patients and controls.(32)

4

Aim of this study The object of this study is to explore if there are differences in the concentrations of purine

metabolites in the CSF of patients with schizophrenia and if these concentrations correlate with

concentrations of monoamine metabolites. We expect to find similar correlations as in the above

mentioned plasma studies.

Method

This study was carried out on data collected among schizophrenic patients at Uppsala University

Hospital between 1980 and 1984.

Cerebrospinal fluid was collected by lumbar puncture and the levels of homovanillic acid (HVA), 5-

hydroxyindoleacetic acid (5-HIAA), xanthin, hypoxanthin, creatinine and uric acid (UA) were

measured in 49 patients with schizophrenia diagnosed using DSM-III. Of these 26 were first

admissions who had never received neuroleptic treatment and the remaining were patients in

relapse and had been off antipsychotic medication for at least two weeks. The control group consist

of 21 healthy volunteers without history of psychiatric disorder or family history of schizophrenia and

took no regular medication (except for oral contraceptives in seven females).

The lumbar puncture was performed in a standardized way between 8 and 9 a.m. with the person in

a sitting position. 18 ml of CSF was taken, immediately centrifuged at 1.000 g for 10 minutes at

ambient temperature, decanted, well mixed and thereafter divided into 1 ml-aliquots. The sample

was placed in -20°C within 20 minutes and in -80°C within 3 hours until assayed. There were no

restrictions of physical activity or food intake before the procedure. The control subjects were,

contrary to the patients, not hospitalized the night before the lumbar puncture.

The CSF was analysed for hypoxanthine, xanthine, creatinine and UA by means of a reversed phase

high performance liquid chromatography (HPLC) using a UV-absorbance detector as described by

Niklasson et al.(33) Assays for HVA and 5-HIAA were performed according to Swahn et al.(34) using a

HPLC equipped with a multiple ion detector.

Statistical analyses The statistical analyses were carried out using R Commander, an open source graphical interface of

the R programming language for statistical computing. The two groups were compared in terms of

sex, age, length and weight. We also checked if there were differences between the first admission

patients and the relapse patients. Levels of metabolites were checked for non-normality using

Shapiro-Wilk’s test with a confidence level of p<0.1. The distribution was also visually examined

5

through histograms. Correlations between metabolite concentrations and physical characteristics

(height, weight and age) were tested using Spearman rank-order on all the data, both patients and

controls. Correlations between metabolites were checked using Pearson product moment or

Spearman rank-order as appropriate of complete pairwise observations both in the patient group

and in the control group. P-values of all metabolite correlations were adjusted using Holm’s method.

Levels of metabolites were compared using Welch t test or Wilcoxon rank sum test as appropriate.

Concentrations of metabolites were compared both between patients and controls and between

patients without previous treatment and patients who previously had received neuroleptic drugs. To

adjust for group differences in height, weight and sex an ANCOVA was carried out for each of the

metabolites.

Results

A finding of lower levels of HVA and no difference in levels of HIAA in patients with schizophrenia has

previously been reported from this patient material, why this is not discussed further here.(35)

There were no major differences between controls and schizophrenic patients in terms of age,

weight or height. There was a larger proportion males in the patient group. (Data on weight and

height were missing for 2 of the controls and 2 of the patients.)

Table 1 – Comparison of age, weight, height and sex of patients and controls.

Age (years) Weight (kg) Height (cm) Sex

Mean SD n Mean SD n Mean SD n Male Female

Controls 26.6 8.7 21 67.0 10.5 19 173 8.7 19 9 12

Schizophrenics 26.6 5.2 49 68.0 12.3 47 176 8.7 47 35 14

There were no major differences between first-episode patients and patients in relapse.

Table 2 – Comparison of age, weight, height and sex in first episode patients and patients in relapse.

Age (years) Weight (kg) Height (cm) Sex

Mean SD n Mean SD n Mean SD n Male Female

First episode 25.0 3.7 26 64.8 10.2 25 176 10.2 25 18 8

Relapse 28.5 6.0 23 71.7 13.6 22 176 7.0 22 17 6

Among the controls all metabolites except creatinine had normally distributed concentrations –

p>0.1 in Shapiro Wilks normality test. But as can be seen in the histogram below creatinine was also

6

quite normally distributed. In the patient group all metabolite concentrations except HVA and UA

were normally distributed according to Shapiro Wilks test. But as the histograms below indicate

these were also quite normally distributed. We conclude that the data fit the normality assumption

and that it can be compared using Welch two sample t-test.

Correlations between metabolites and physical characteristics (age, weight and height) were

calculated in both the patient group and the control group using Spearman rank-order between

pairwise complete values (n=18-21 in the control group; n=46-49 in the patient group). In the control

group there were significant correlations between height and urate (rs=-0.50; p=0.03). In the patient

group there were significant correlations between age and HVA, age and UA, height and HIAA, height

and UA, weight and HIAA, and weight and UA. (see table 3)

Table 3 – Correlation between physical characteristics and metabolites in the patient group. *=p<0.05, **p=<0.01

Correlation – Patients

Spearman rank-order

Age Height Weight

HIAA -0.02 -0.39** -0.45**

HVA -0.34* -0.24 -0.20

Hypoxanthine -0.07 -0.10 -0.21

Xanthine -0.10 -0.08 -0.25

UA 0.37** 0.30* 0.30*

Creatinine -0.10 0.26 0.02

Correlations between metabolites were calculated using Pearson product-moment with p-values

adjusted using Holm’s method for complete pairs (n=20-21 in the control group and n=48-49 in the

patient group). Among the controls there were significant positive correlations between HVA-

Figure 1 – Histograms of the metabolites whose concentrations were non-normal according to Shapiro-Wilks normality test, p>0.1.

7

creatinine, HVA-xanthine, hypoxanthine-xanthine and creatinine-xanthine. (see table 4) In the

patient group there were significant correlations between HIAA-HVA, HIAA-hypoxanthine, HIAA-

xanthine, HVA-hypoxanthine, HVA-creatinine, HVA-xanthine, hypoxanthine-creatinine,

hypoxanthine-xanthine and creatinine-xanthine. (see table 5)

Table 4 - Correlations between metabolites in the control group using Pearson R. P-values were adjusted using Holm’s method. *=p<.05; **=p<.01

Correlations controls HIAA HVA Hypoxanthine Xanthine UA Creatinine

HIAA 1.00 0.53 0.39 0.46 -0.09 0.28

HVA 0.53 1.00 0.27 0.70** -0.08 0.63**

Hypoxanthine 0.39 0.27 1.00 0.61** 0.15 0.41

Xanthine 0.46 0.70** 0.61** 1.00 0.04 0.88**

UA -0.09 -0.08 0.15 0.04 1.00 0.08

Creatinine 0.28 0.63** 0.41 0.88** 0.08 1.00

Table 5 - Correlations between metabolites in the patient group using Pearson R. P-values were adjusted using Holm’s method. *=p<.05; **=p<.01

Correlations Patients

(Pearson R) HIAA HVA Hypoxanthine Xanthine UA Creatinine

HIAA 1.00 0.50** 0.62** 0.50** -0.01 0.29

HVA 0.50** 1.00 0.59** 0.50** -0.27 0.43*

Hypoxanthine 0.62** 0.59** 1.00 0.75** 0.06 0.58**

Xanthine 0.50** 0.50** 0.75** 1.00 0.24 0.67**

UA -0.01 -0.27 0.06 0.24 1.00 0.30

Creatinine 0.29 0.43* 0.58** 0.67** 0.30 1.00

Using Welch two sample t-test we compared the concentrations of metabolites between patients

and controls, and between first-episode schizophrenic patients and previously medicated patients.

There were no significant differences in metabolite concentrations between previously medicated

and drug-naive patients.

There were no significant differences in levels of xantin, creatinine or urate between patients and

controls. (Differences in levels of HVA and HIAA in this material is reported elsewhere.(35)) Patients

with schizophrenia had significantly lower levels of hypoxanthine in their CSF (t=2.37, p=0.02); the

8

mean±SD was 3.09±0.57 for controls and 2.72±0.69 for patients with schizophrenia. The differences

were as large among the females as among the males.

To identify differences between the groups masked by differences in height, weight and sex an

ANCOVA was carried out for all the metabolites. The expression used was ‘metabolite ~ height + sex

+ type + weight’. Doing this the difference in concentration of hypoxanthine between patients and

controls remained signficant (p=0.03). The ANCOVA also showed significantly lower levels of

creatinine in the patient group compared to the control group when ruling out height, weight and

sex differences (p=0.03). The difference in concentration is only seen among the females.

Figure 2 – Box plot and plot of means comparing levels hypoxanthine between patients and controls, and between males and females. The whiskers in the plot of means indicate standard errors.

Figure 3 – Boxplots and plot of means comparing levels of creatinine patients and controls, females and males, and between female patients and female controls. Whiskers in the plot of means indicate standard errors.

9

Doing a Welch Two Sample test we saw that the female patients have significantly lower levels of

creatinine compared to female controls. (t = 2.76 p = 0.01); the mean±SD was 3.04±0.56 for controls

and 2.67±0.74 for patients with schizophrenia.

Discussion

Hypoxanthine Significantly lowered levels of hypoxanthine in CSF in patients with schizophrenia have not previously

been reported. Studies of plasma and urine have not found any differences between patients with

schizophrenia and controls.(29) (32) The cause of this finding is still a matter of speculation.

Hypoxanthine is known to be a marker of disturbed energy metabolism in the brain. This is due to

the fact that it reflects degradation of high energy phosphates (e.g. ATP)(36) and can be increased

more than 10-fold in case of ischemia.(37) In a study of patients suffering a number of neurological

diseases - multiple sclerosis, myelopathy, stroke, epilepsy and viral meningitis – one found 2-3 fold

increases of hypoxanthine. Stover et al.(36) interpreted this as a marker of glutamate-mediated

excitotoxicity as glutamate cause increased ATP-consumption.

Hypoxanthine could possibly have a role in protecting the brain against hypoxic events. This has been

shown through administration of hypoxanthine to rabbits subjected to hypoxia which significantly

improved their brain recovery.(38) Hypoxanthine may have this effect through inhibiting xanthine

oxidase that may play a role in forming oxygen reactive compounds in case of metabolic stress. A

deficiency in hypoxanthine may increase stress vulnerability and explain the therapeutic

improvements reported by Brunstein et al.(16) when complimenting treatment with allopurinol, a

hypoxanthine isomer. It would be bold to draw conclusions about levels of adenosine from our data.

Lower levels of hypoxanthine could either be a sign of less adenosine signalling and that it thus is less

adenosine that can be metabolized to hypoxanthine; or it could be a sign of less degradation and

thus higher concentrations of adenosine in the brain.

Creatinine Lowered concentrations of creatinine in the CSF of patients with schizophrenia have previously been

reported by Swahn et al.(39) They have also showed that levels of creatinine increased with

neuroleptic treatment. As opposed to this study they did not find any differences due to sex either in

the control group or in the patient group. We do not have any good explanation for the difference

seen, in this study, between males and females in the patient group. Creatinine may, just as

10

hypoxanthine, be an indicator of the brains energy metabolism as it is formed when cells catabolize

creatine-phosphate. CSF levels together with blood levels can be a measure of blood-brain barrier

permeability(40) – why blood samples taken at the time of the lumbar puncture would add an

interesting aspect to this finding.

Contrary to Yao et al.(29) this study showed no correlation between urate levels and levels of

hypoxanthine or xanthine. Both of these findings need to be interpreted with caution as they may

not properly represent levels of extracellular metabolite levels in the brain. (30) (41) It is also difficult

to know whether a deviant concentration of a substance is caused by an altered transmitter turnover

in the brain or an altered transport of the substance across the blood-brain barrier – which can be

either a cause or an effect of the disease.

The fact that both this study and Yao’s study only found positive correlations between metabolites

should lead to very careful interpretation of these findings as they can be solely a sign of degree of

dilution. It is to be expected that a fluid with higher concentration of one protein have higher

concentrations of other proteins as well. A negative correlation would be easier to interpret as a sign

of metabolic interaction.

Conclusion

There were lowered concentrations of hypoxanthine and creatinine in cerebrospinal fluid in patients

with schizophrenia compared to controls. There were also numerous positive correlations between

metabolite concentrations. Both these findings may be due to a more diluted CSF in patients with

schizophrenia. The finding of lowered levels of hypoxanthine in schizophrenia is interesting in the

light of the purine hypothesis of schizophrenia. But also because of the recent interest the xanthine-

oxidase inhibitor allopurinol has gained as an adjuvant treatment in schizophrenia.

Future research needs to validate these findings making use of the progress seen in metabolomic

biochemistry methods in the last 20 years. It would be interesting to study the metabolomic profile

of persons with high risk of schizophrenia to identify differences between those who do not develop

and those who do develop schizophrenia. The purinergic hypothesis of schizophrenia offers a new

approach in understanding the symptoms of schizophrenia – how it develops and possibilities for

new treatments. However – many parts of the theory still needs to be proved: for example could PET

with radioligand-studies explore if there is lower concentrations of A1 adenosine receptors in

patients with schizophrenia, and adenosine modulating medication needs to prove its place in

schizophrenia treatment.

11

Referenser 1. The World health report 2001: mental health: new understanding, new hope. Geneva:

World Health Organization; 2001.

2. Salomon JA, Vos T, Hogan DR, Gagnon M, Naghavi M, Mokdad A, et al. Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010. lancet. 2012 Dec 15;380(9859):2129–43.

3. Tandon R, Keshavan MS, Nasrallah HA. Schizophrenia, “Just the Facts”: What we know in 2008: Part 1: Overview. Schizophr Res. 2008 Mar;100(1–3):4–19.

4. Van Os J, Kapur S. Schizophrenia. lancet. 22;374(9690):635–45.

5. Lara DR, Souza DO. Schizophrenia: a purinergic hypothesis. Med Hypotheses. 2000 Feb;54(2):157–66.

6. Pratt JA, Winchester C, Egerton A, Cochran SM, Morris BJ. Modelling prefrontal cortex deficits in schizophrenia: implications for treatment. Br J Pharmacol. 2008;153(S1):S465–S470.

7. Stone JM, Morrison PD, Pilowsky LS. Review: Glutamate and dopamine dysregulation in schizophrenia — a synthesis and selective review. J Psychopharmacol (Oxf). 2007 Jun 1;21(4):440–52.

8. Mehler-Wex C, Riederer P, Gerlach M. Dopaminergic dysbalance in distinct basal ganglia neurocircuits: Implications for the pathophysiology of parkinson’s disease, schizophrenia and attention deficit hyperactivity disorder. Neurotox Res. 2006 Sep 1;10(3-4):167–79.

9. Abbracchio MP, Burnstock G, Verkhratsky A, Zimmermann H. Purinergic signalling in the nervous system: an overview. Trends Neurosci. 2009 Jan;32(1):19–29.

10. Burnstock G. Purinergic signalling: past, present and future. Braz J Med Biol Res. 2009 Jan;42(1):3–8.

11. Lara DR, Dall’Igna OP, Ghisolfi ES, Brunstein MG. Involvement of adenosine in the neurobiology of schizophrenia and its therapeutic implications. Prog Neuropsychopharmacol Biol Psychiatry. 2006 Jun;30(4):617–29.

12. Dunwiddie TV, Masino SA. The Role and Regulation of Adenosine in the Central Nervous System. Annu Rev Neurosci. 2001;24(1):31–55.

13. Howes OD, Kapur S. The Dopamine Hypothesis of Schizophrenia: Version III—The Final Common Pathway. Schizophr Bull. 2009 May 1;35(3):549–62.

14. Nucleotide signaling in nervous system development - Springer. [cited 2013 May 6]; Available from: http://link.springer.com.ezproxy.its.uu.se/article/10.1007/s00424-006-0067-4/fulltext.html#Sec25

15. Boison D, Singer P, Shen H-Y, Feldon J, Yee BK. Adenosine hypothesis of schizophrenia – Opportunities for pharmacotherapy. neuropharmacology. 2012 Mar;62(3):1527–43.

16. Brunstein MG, Ghisolfi ES, Ramos FLP, Lara DR. A clinical trial of adjuvant allopurinol therapy for moderately refractory schizophrenia. J Clin Psychiatry. 2005 Feb;66(2):213–9.

12

17. Pacher P. Therapeutic Effects of Xanthine Oxidase Inhibitors: Renaissance Half a Century after the Discovery of Allopurinol. Pharmacol Rev. 2006 Mar 1;58(1):87–114.

18. Schretlen DJ, Ward J, Meyer SM, Yun J, Puig JG, Nyhan WL, et al. Behavioral aspects of Lesch–Nyhan disease and its variants. Dev Med Child Neurol. 2005;47(10):673–7.

19. Kanehisa M, Goto S, Kawashima S, Okuno Y, Hattori M. The KEGG resource for deciphering the genome. Nucleic Acids Res. 2004 Jan 1;32(suppl 1):D277–D280.

20. Rapoport JL, Giedd JN, Gogtay N. Neurodevelopmental model of schizophrenia: update 2012. Mol Psychiatry. 2012 Dec;17(12):1228–38.

21. Turner CP, Yan H, Schwartz M, Othman T, Rivkees SA. A1 adenosine receptor activation induces ventriculomegaly and white matter loss. neuroreport. 2002 Jul 2;13(9):1199–204.

22. Harrison PJ. The neuropathology of schizophrenia A critical review of the data and their interpretation. brain. 1999 Apr 1;122(4):593–624.

23. Gotoh L, Mitsuyasu H, Kobayashi Y, Oribe N, Takata A, Ninomiya H, et al. Association analysis of adenosine A1 receptor gene (ADORA1) polymorphisms with schizophrenia in a Japanese population. Psychiatr Genet. 2009 Dec;19(6):328–35.

24. Shen H-Y, Singer P, Lytle N, Wei CJ, Lan J-Q, Williams-Karnesky RL, et al. Adenosine augmentation ameliorates psychotic and cognitive endophenotypes of schizophrenia. J Clin Invest. 2012 Jul 2;122(7):2567–77.

25. Eaton WW, Hayward C, Ram R. Schizophrenia and rheumatoid arthritis: a review. Schizophr Res. 1992 Mar;6(3):181–92.

26. Haskó G, Cronstein B. Regulation of inflammation by adenosine. Front Immunol. 2013;4:85.

27. Degrell I, Niklasson F. Purine metabolites in the CSF in presenile and senile dementia of Alzheimer type, and in multi infarct dementia. Arch Gerontol Geriatr. 1988 Jun;7(2):173–8.

28. Agren H, Niklasson F, Hällgren R. Brain purinergic activity linked with depressive symptomatology: hypoxanthine and xanthine in CSF of patients with major depressive disorders. Psychiatry Res. 1983 Jul;9(3):179–89.

29. Yao JK, Dougherty GG, Reddy RD, Keshavan MS, Montrose DM, Matson WR, et al. Homeostatic Imbalance of Purine Catabolism in First-Episode Neuroleptic-Naïve Patients with Schizophrenia. Plos One. 2010 Mar 3;5(3):e9508.

30. Eells JT, Spector R. Purine and pyrimidine base and nucleoside concentrations in human cerebrospinal fluid and plasma. Neurochem Res. 1983 Nov 1;8(11):1451–7.

31. Yao JK, Keshavan MS. Antioxidants, Redox Signaling, and Pathophysiology in Schizophrenia: An Integrative View. Antioxidants Redox Signal. 2011 Oct;15(7):2011–35.

32. Bollard BM, Culpan RH, Marks N, McIlwain H, Shepherd M. Purines in the Urine of Normal and Schizophrenic Subjects. Br J Psychiatry. 1960 Oct 1;106(445):1250–72.

33. Niklasson F. Simultaneous liquid-chromatographic determination of hypoxanthine, xanthine, urate, and creatinine in cerebrospinal fluid, with direct injection. Clin Chem. 1983 Aug 1;29(8):1543–6.

13

34. Swahn C-G, Sandgärde B, Wiesel F-A, Sedvall G. Simultaneous determination of the three major monoamine metabolites in brain tissue and body fluids by a mass fragmentographic method. Psychopharmacology (Berl). 1976 Jan 1;48(2):147–52.

35. Lindström LH. Low HVA and normal 5HIAA CSF levels in drug-free schizophrenic patients compared to healthy volunteers: Correlations to symptomatology and family history. Psychiatry Res. 1985 Apr;14(4):265–73.

36. Stover JF, Lowitzsch K, Kempski OS. Cerebrospinal fluid hypoxanthine, xanthine and uric acid levels may reflect glutamate-mediated excitotoxicity in different neurological diseases. Neurosci Lett. 1997 Nov 28;238(1–2):25–8.

37. Hillered L, Kotwica Z, Ungerstedt U. Interstitial and cerebrospinal fluid levels of energy-related metabolites after middle cerebral artery occlusion in rats. Res Exp Med (Berl). 1991 Dec 1;191(1):219–25.

38. Mink R, Johnston J. The effect of infusing hypoxanthine or xanthine on hypoxic–ischemic brain injury in rabbits. Brain Res. 2007 Maj;1147:256–64.

39. Swahn C-G, Sedvall G. CSF creatinine in schizophrenia. Biol Psychiatry. 1988 Mar 15;23(6):586–94.

40. Niklasson F, Agren H. Brain energy metabolism and blood-brain barrier permeability in depressive patients: analyses of creatine, creatinine, urate, and albumin in CSF and blood. Biol Psychiatry. 1984 Aug;19(8):1183–206.

41. Degrell I, Nagy E. Concentration gradients for HVA, 5-HIAA, ascorbic acid, and uric acid in cerebrospinal fluid. Biol Psychiatry. 1990 Apr 15;27(8):891–6.