High homocysteine serum levels in young male schizophrenia and bipolar patients and in an animal model

Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191 www.elsevier.com/locate/pnpbp

Review article

High homocysteine serum levels in young male schizophrenia and bipolar patients and in an animal model Joseph Levine a, Ben-Ami Sela b, Yamima Osher a, R.H. Belmaker a,* a Ben Gurion University, Beersheva, Israel The Institute of Chemical Pathology, Sheba Medical Center, Tel-Hashomer, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel

b

Accepted 17 June 2005 Available online 22 August 2005

Abstract Elevated plasma homocysteine has been found to be a risk factor for Alzheimer’s disease as well as cerebral vascular disease, suggesting that some risk factors can accelerate or increase the severity of several CNS disease processes. We screened plasma total homocysteine levels of 193 schizophrenic patients vs. 762 controls for plasma homocysteine levels. The effect of schizophrenia was marked ( p < 0.0001) and mean homocysteine level was 16.3 T 12 (S.D.) AM in schizophrenic patients vs. 10.6 T 3.6 (S.D.) AM in healthy controls. The increase was almost entirely in young male schizophrenic patients. It seemed important to determine if this finding is already present in newly admitted schizophrenic patients. Serum homocysteine levels were studied in 184 consecutively admitted schizophrenic patients and 305 control subjects. Homocysteine levels were markedly increased in this population of newly admitted schizophrenic patients, especially in young males. However, no difference was found for CSF homocysteine levels between schizophrenia patients and controls. We also examined homocysteine levels in 41 euthymic outpatients with bipolar disorder. Functional deterioration in patients was rated as Fpresent_ or Fabsent_ by consensus of two treating clinicians. Young male bipolar patients were found to have higher homocysteine levels than controls. Among the male subjects, bipolar patients showing deterioration had homocysteine levels which were significantly higher than other patients. We attempted to develop a model of homocysteine neurotoxicity in mice. Mice were fed homocysteine in water at a dose of 200 mg/kg per mouse per day. Independent samples of animals were studied at 2 to 6 months with behavioral tests including apomorphine-induced stereotypy and spatial learning and memory in the Morris Water Maze. Homocysteine levels were elevated up to 800% at months 5 and 6 by this procedure. No homocysteine-induced defects were found in any behavioral test until month 5 when mild but statistically significant abnormalities in the Morris Water Maze were detected. D 2005 Published by Elsevier Inc. Keywords: Bipolar disorder; Cerebrospinal fluid; Cognitive impairment; Folate; Functional deterioration; Homocysteine; Mice; Schizophrenia; Vitamin B12

Contents 1. 2. 3. 4. 5. 6.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . High plasma homocysteine levels in young male schizophrenic inpatients . . . . High homocysteine levels in young male newly admitted schizophrenic patients CSF homocysteine levels are not elevated in schizophrenia . . . . . . . . . . . Homocysteine levels in euthymic bipolar patients . . . . . . . . . . . . . . . . Effects of chronic homocysteine administration on cognitive functions in mice . 6.1. Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Morris Water Maze (MWM) . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

1182 1182 1184 1185 1185 1186 1186 1187

Abbreviations: BBB, blood – brain barrier; EDTA, ethylevediamine tetra-acetic acid; HPLC, high pressure liquid chromatography; MTHFR, methylenetetrahydrofolate reductase; MWM, Morris Water Maze. * Corresponding author. Tel.: +972 8 6401602; fax: +972 8 6401621. E-mail address: [email protected] (R.H. Belmaker). 0278-5846/$ - see front matter D 2005 Published by Elsevier Inc. doi:10.1016/j.pnpbp.2005.06.029

1182

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

6.3. Apomorphine-induced stereotypy 6.4. Rotarod test . . . . . . . . . . . 6.5. Procedure . . . . . . . . . . . . 6.6. Homocysteine plasma levels . . . 6.7. MWM behavioral test . . . . . . 6.8. Modified MWM behavioral test . 6.9. Apomorphine behavioral test. . . 6.10. Rotarod behavioral test. . . . . . 6.11. Discussion of behavioral results . 7. Conclusions . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

1. Introduction Elevated homocysteine levels in plasma are considered a risk factor for cardiovascular and also cerebrovascular disease (Nygard et al., 1995). Recently, elevated plasma homocysteine has also been found to be a risk factor for Alzheimer’s disease as well (Seshadri et al., 2002), suggesting that some risk factors can accelerate or increase the severity of several CNS disease processes. Similarly APOE-4, originally found to be a risk factor for Alzheimer’s disease, is apparently also a risk factor for vascular dementia and a severity enhancer of other CNS degenerative disorders. Kruman et al. (2000) reported that homocysteine elicits a DNA damage response in neurons that promotes apoptosis and hypersensitivity to excitoxicity. An oral methionine load has classically been reported to exacerbate schizophrenia and is of course converted to homocysteine (Pollin et al., 1961). More than 30 years ago, Antun et al. (1971) following Pollin et al. (1961) reported that 10– 20 g lmethionine orally may exacerbate psychosis in chronic schizophrenic patients (Cohen et al., 1974). Methionine loading is known to increase plasma homocysteine levels. Early case reports from the 1970s reported schizophrenia like symptomatology in individuals with homocystinuria that may be ameliorated by folate treatment (Freeman et al., 1975). These reports raised hope that genetic aberrations in enzymes associated with the metabolism of homocysteine may be relevant for schizophrenia. However Berger et al. (1977) reported no significant difference in platelet methylene reductase activity between chronic schizophrenics, and either hospitalized or nonhospitalized age-matched control subjects. Elliott et al. (1978) found no change in N5,10-methylenetetrahydrofolate reductase (MTHFR) activity in autopsied brain parts of chronic schizophrenics and controls and suggested that, although it is possible that some subgroup of schizophrenics may be characterized by abnormal MTHFR activity, there does not appear to be a general association between the two. Godfrey et al. (1990) reported that folate treatment, known to reduce plasma hyperhomocysteinemia, improves symptoms of schizophrenia. Regland et al. (1994) reported a case of a young schizophrenic patient with a significantly

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

1187 1187 1187 1187 1187 1188 1188 1188 1188 1189 1190

increased serum level of homocysteine. She improved repeatedly on cobalamin (B12) injections and deteriorated in periods without treatment. Methylenetetrahydrofolate reductase (MTHFR) activity in cultured skin fibroblasts of this patient was reduced to a magnitude that is found among people with heterozygous deficiency. A defect in MTHFR activity causes a deficiency in methylenetetrahydrofolate, with a consequent reduction of the remethylation of homocysteine to methionine. The gene for MTHFR is polymorphic in the human population. In its homozygous form, a C677T mutation occurs in more than 5% of the population and produces a thermolabile variant which reduces the overall enzyme activity to less than 30% of normal (Regland et al., 1995). Several small studies and case reports suggest that homozygosity for the T677 allele of the MTHFR gene – encoding for the thermolabile enzyme associated with hyperhomocysteinemia – may be associated with increased occurrence of schizophrenia. Regland et al. (1997) studied patients with schizophrenia-like psychosis. Seven of 11 patients, six males and one female, were homozygous for thermolabile MTHFR, suggesting that homozygosity for thermolabile MTHFR is a risk factor for schizophrenia-like psychosis and that this risk may be reduced by folate supplementation. Susser et al. (1998) conducted a pilot study comparing homocysteine levels of schizophrenic patients and normal controls with and without low folate levels, reporting pilot data that are compatible with the hypothesis that a folatesensitive defect in homocysteine metabolism contributes to some cases of schizophrenia.

2. High plasma homocysteine levels in young male schizophrenic inpatients We screened plasma homocysteine levels in schizophrenic patients compared with a control population. One hundred and ninety-three chronic consenting schizophrenic patients (DSM-IV) (APPI, 1994), males and females, age 18– 70, with no history of neurologic or cardiovascular disorders or drug and alcohol abuse participated in the study. All were drug-treated and were in a variety of clinical

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

1183

Table 1 Homocysteine levels (AM) in young schizophrenia hospitalized patients vs. controls (mean T S.D.) Age

18 – 30 30 – 40 40 – 50

Schizophrenia

Controls

Male

Female

Male

Female

15.0 T 9.1 (n = 45)* 19.1 T15.5 (n = 53)* 20.1 T13.1 (n = 31)*

10.5 T 3.0 (n = 7) 10.8 T 4.7 (n = 8) 10.8 T 5.1 (n = 11)

8.0 T 3.1 (n = 87) 9.9 T 3.3 (n = 244) 12.6 T 4.1 (n = 65)

8.9 T 3.1 (n = 77) 9.7 T 3.0 (n = 75) 11.3 T 2.7 (n = 46)

Table modified from Levine et al. (2002). * Statistical difference from the corresponding control group (ANOVA, p < 0.05).

settings including acute inpatient units, chronic inpatient units, hostel care and outpatient care. Controls were between the ages of 18 and 70, males and females and were randomly selected from plasma homocysteine determinations performed at the Sheba Medical Center Institute of Clinical Pathology as part of employee health screening, run in the laboratory over the same time period as the schizophrenic patient samples. After written consent and fasting, AM blood samples were obtained in ice-cooled EDTA tubes. Plasma was separated by centrifugation at 5 -C and stored in 20 -C. Total homocysteine levels were measured by HPLC technology with fluorescence detection following labeling of homocysteine with monobromobimane as a modification of the method of Araki and Sako (1987). Results are presented in Table 1. One-way ANCOVA with age and sex as covariants was performed. The effect of schizophrenia was marked ( F = 135.7, df = 1951, p < 0.0001) and mean homocysteine level was 16.4 T 11.8 (S.D.) in schizophrenic patients vs. 10.6 T 3.6 (S.D.) in healthy controls (covariance adjusted means 16.3 vs. 10.6 in controls). The increase was entirely in male schizophrenic patients up until age 50. This was the first comprehensive study of homocysteine levels in schizophrenia to include a range of age groups. Virgos et al. (1999) found plasma homocysteine no different in schizophrenic inpatients (n = 210) compared with controls (n = 218). However, there was an average age of almost 60 for both schizophrenic patients and controls, and in this age group our data also do not find any difference. Increased homocysteine levels in young male schizophrenic patients could be related to the pathophysiology of aspects of this illness. For instance, it is known that onset of schizophrenia is younger in males than in females and the illness more often has a chronic deteriorating course in young males. Homocysteine has been shown to be neuro-

Table 2 Percentage of low plasma folate (< 3.7 ng/mL) among male schizophrenia patients (age < 51 years) Age in years

n

%

18 – 30 31 – 40 41 – 50

49 44 28

86 73 70

Table modified from Stahl et al. (2005).

toxic (Lipton et al., 1997) and it has been shown that stress can open the blood – brain barrier to some neurotoxic substances (Friedman et al., 1996). It is possible that the stress of acute psychosis allows high homocysteine levels to enter the brain and cause neurodegeneration, clinical deterioration and chronicity. This hypothesis does not make any assumption as to the origin of high homocysteine in young male schizophrenics. It could be caused by smoking, lack of exercise or by poor diet. Yet via a biochemical mechanism it could negatively affect the course of illness. In Alzheimer’s disease, biochemical and genetic studies clearly point to h-amyloid and its metabolism as pathogenetic factors. Yet plasma homocysteine levels could explain 16% of attributable risk in Alzheimer’s (Seshadri et al., 2002), demonstrating the powerfully multifactorial nature of most common diseases. In an attempt to determine the causes of elevated homocysteine in schizophrenia, we assessed an enlarged group of the above schizophrenic patients on a variety of variables. The patient group previously described by Levine et al. (2002) included 258 schizophrenic patients (201 males, 57 females) from the south of Israel. Fasting blood samples were obtained for determination of total plasma homocysteine was as previously described (Araki and Sako, 1987; Levine et al., 2002). A nutritional analysis of the average weekly menu served at the hospital, at the time blood levels were drawn, was performed. Articles of food and their quantities were taken from the Hospital Nutritional Records. Analysis of essential nutrients was done using a nutritional analysis computer program provided by the Israeli Ministry of Health showing no shortage in essential nutrients including that of folate and B12. Almost 75% of schizophrenic male patients showed low serum folate levels and 15% displayed low serum vitamin B12 levels. Similar results were obtained for schizophrenic female subjects where 65% had low plasma folate levels and 12% showed low plasma B12 levels (data not shown). Table 2 shows the percentage of male patients with low plasma folate. We were unable to find a clinical subgroup of schizophrenic patients that was defined by increased plasma homocysteine. A reasonable hypothesis to explain increased homocysteine was a medication effect, since many atypical neuroleptics cause metabolic abnormalities, for example. Use of olanzapine, clozapine or risperidone showed no trend

1184

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

for association with increased homocysteine, however. Anticonvulsants are often co-administered in schizophrenia, to treat violence for example, and valproate and carbamazepine are reported in the literature to have anti-folate properties that could increase homocysteine (Desouza et al., 2002). No such effect was found in our sample.

3. High homocysteine levels in young male newly admitted schizophrenic patients Our previous study (Levine et al., 2002) finding of markedly elevated plasma homocysteine levels in young male schizophrenia patients was done in a sample of patients that included chronically hospitalized, recently hospitalized patients and hostel patients, but entirely from an institutionalized setting. Therefore, it seemed possible that poor hospital nutrition may have contributed to the elevated homocysteine levels. It therefore seemed important to study plasma homocysteine in newly admitted patients. We therefore studied (Applebaum et al., 2004) plasma homocysteine levels in 184 newly admitted schizophrenia patients versus 305 control subjects. Newly admitted patients to the Beersheva Mental Health Center with DSM-IV schizophrenia had blood drawn fasting in the morning hours in a clot tube and transferred to the lab within less than 1 h at room temperature. In the laboratory, tubes were put on ice, centrifuged at 5 -C, and then returned to ice until an aliquot was taken for serum homocysteine, which was then frozen at 20 -C. An experienced clinician went over the patient’s chart within the next month and excluded samples from patients with a diagnosis other than schizophrenia, with physical illnesses including neurological, liver, kidney endocrinological (including diabetes) or cardiovascular diseases, or with alcohol or drug abuse in the previous 6 months. Aliquots of the remaining samples were then transferred for analysis of serum homocysteine levels as previously described (Araki and Sako, 1987). The comparison group consisted of individuals whose plasma homocysteine was drawn as part of an employee health program and the samples were run simultaneously with the schizophrenia patients’ samples. This control group was separate and new from the control group previously reported (Levine et al., 2002). Patients from Levine et al.’s (2002) study were not included in the current study.

Table 3 presents the results for the newly admitted patients. Serum homocysteine was analyzed with three-way ANOVA by age, sex and diagnosis. There was a significant effect of diagnosis ( F = 16.04, df = 1473, p < 0.001) and sex ( F = 13.66, df = 1473, p < 0.001) but no significant effect of age. There was a significant interaction between sex and diagnosis ( F = 8.32, df = 1473, p < 0.005). The fact that young newly admitted schizophrenia patients have the same elevated blood homocysteine as institutionalized schizophrenic patients suggests that the increase is not due to institutional food. However, it is of course also possible that patients in the weeks or months prior to admission have poor nutrition even at home. In the previous sample of Levine et al. (2002) plasma vitamin* B12 and folic acid demonstrated an inverse correlation with plasma homocysteine but did not explain a majority of the variance (Stahl et al., 2005). Smoking may raise homocysteine by 1– 2 AM/L (O’Callaghan et al., 2002) but this is not a large enough effect to explain our findings. Recently Goff et al. (2004) reported no increase in homocysteine blood levels in a sample of 91 schizophrenic outpatients from an urban community in northeast USA, as did Muntjewerff et al. (2003) in the Netherlands. Goff et al. (2004) may have found normal homocysteine in his patients because they were well stabilized outpatients rather than the inpatients as studied by Levine et al. (2002) and Applebaum et al. (2004). Moreover, the bread in Goff’s area of the USA has been fortified with folate (Choumenkovitch et al., 2002) whereas the Israeli diet has been reported to be low in folate (Kaluski et al., 2003; Levy et al., 1975) and this effect may be exacerbated in schizophrenic patients. Cholesterol, for example, is associated with heart disease in some populations but not in others, because some populations consume a diet that is below threshold or above ceiling for variance in cholesterol to have significant effects. This does not contradict the key role that cholesterol can play as a target for treatment and prevention in heart disease. Homocysteine levels may vary between populations of schizophrenic patients depending on folate and B12 status. Such consumption may be dependent on the food supplied, on folate fortification strategy, on eating patterns and on motivation to eat and care for oneself. Elevated plasma homocysteine could contribute to the cognitive deterioration and cerebral atrophy perhaps most

Table 3 Homocysteine levels (AM) in newly admitted schizophrenia patients vs. controls (mean T S.D.) Age

Schizophrenia Male

18 – 30 31 – 40 41 – 50 51 – 70

18.2 T 16 13.7 T 5 14.1 T 6 12.8 T 4.3

Controls Female

(n = 36)* (n = 36)* (n = 36)* (n = 19)

11.1 T 2.9 11.2 T 3.2 10.4 T 4.0 11.7 T 2.8

(n = 18) (n = 15) (n = 16) (n = 8)

Table modified from Applebaum et al. (2004). * Statistical difference from the corresponding control group (ANOVA, p < 0.05).

Male

Female

8.3 T 2 (n = 29) 10.5 T 2 (n = 32) 11.6 T 2 (n = 35) 13.3 T 2.2(n = 55)

9.2 T 2 9.9 T 2.5 10.7 T 2.4 12.1 T 2.0

(n = 36 (n = 31) (n = 30) (n = 57)

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

marked in young male schizophrenia patients. Bleich et al. (2002, 2003) and Sachdev et al. (2002) showed a significant positive relationship between plasma homocysteine levels and brain atrophy. Estrogen on the other hand, may exert a protective effect in young high risk vulnerable females similarly to its suggested role in Alzheimer’s (Zandi et al., 2002).

4. CSF homocysteine levels are not elevated in schizophrenia Homocysteine is formed during methionine recycling in the cell but is rapidly extruded from the intracellular environment to extracellular compartments (Selhub, 1999). Such extracellular compartments include the plasma in the case of peripheral tissues and the CSF in the case of brain tissue. Since amino acids can not pass the intact blood – brain barrier (BBB) freely (Kruse et al., 1985), it was suggested that the main source of CSF homocysteine is homocysteine synthesized in brain glial cells and neurons (Serot et al., 2003). Data suggesting that homocysteine is toxic to endothelium of the vascular tissue (Stuhlinger et al., 2003) may suggest that hyperhomocysteinemia may induce impairment of the BBB. Thus the statement regarding the brain as the main source of CSF homocysteine may not be true in cases of grossly increased plasma homocysteine levels. CSF homocysteine was previously reported to be elevated in a variety of disorders including Alzheimer disease (Teunissen et al., 2002). We studied (Levine et al., 2005) homocysteine levels in CSF of two research groups of schizophrenia patients. CSF homocysteine levels were measured in 40 schizophrenia and 28 healthy control subjects. The samples were obtained from two separate studies. Study A included 20 physically healthy Irish schizophrenia subjects from St Patrick Hospital, Castlerea County Roscommon and the Regional Hospital in Galway, Ireland and 20 healthy controls from the sample described in Preble and Torrey (1985). Study B included 21 physically healthy schizophrenia subjects admitted to a research inpatient service following exacerbation of psychosis and eight healthy controls as described in Garver et al. (2003). All patients in Study B were neuroleptic free for at least 2 months. 0.5 ml of each CSF sample was concentrated by lyophilization in a Speed-Vac to about 0.05 ml. The exact concentration factor for each sample was used to calculate the original CSF homocysteine level. CSF samples were assayed using HPLC technology with fluorescence detection following labeling of homocysteine with monobromobimane as a modification of the method reported by Araki and Sako (1987). There was no significant difference in mean CSF homocysteine in Study A between schizophrenic patients (x T S.D. =0.038 T 0.05) and controls (0.012 T 0.04, p = 0.255)

1185

or in Study B (schizophrenic patients, 0.022 T 0.03 vs. controls, 0.018 T 0.03). We recently found glycogen synthase kinase-3h to be reduced in schizophrenia both in brain postmortem (Kozlovsky et al., 2000) and in CSF (Kozlovsky et al., 2004). Both amyloid and tau proteins, abnormal in brain in Alzheimer’s disorder (Grundke-Iqbal and Iqbal, 1999; Selkoe, 1993) are also markedly abnormal in CSF (Arai, 1996; Galasko, 2003). However, these findings were all of proteins and homocysteine is a relatively ephemeral amino acid. Elevated plasma homocysteine in schizophrenia may be caused by dietary insufficiency of folate and other peripheral environmental influences (Stahl et al., 2005) and cause CNS damage only when a damaged blood – brain barrier allows entry of elevated homocysteine into the brain. Thus plasma in the case of homocysteine in schizophrenia may be a more fruitful tissue of study than CSF.

5. Homocysteine levels in euthymic bipolar patients Accumulating evidence suggests that at least some portion of bipolar patients demonstrate functional deterioration in the course of their illness. The Chicago Follow-up Study reported that over a period of 8 years, more than half of the bipolar patients exhibited some degree of functional impairment and 10 – 15% of patients showed serious disabilities in multiple dimensions of functioning (Goldberg and Harrow, 1999). This is consistent with the literature on impaired aspects of neurocognitive functioning in bipolar patients in their euthymic state (Clark et al., 2002; Ferrier et al., 1999; MacQueen et al., 2001; van Gorp et al., 1998; Zubieta et al., 2001). There are also reports of reductions in the volume of prefrontal cortical structures of bipolar patients, reported to be involved in executive functions. Therefore we designed a study examining whether homocysteine levels are elevated in euthymic bipolar disorder compared with controls and also examining functional deterioration and homocysteine levels in these patients. The patient group consisted of 41 euthymic bipolar disorder patients (23 males, 18 females, age range of 18 – 68 years) with no physical illnesses including diabetes, cardiovascular, renal endocrinological and neurological disease. Comparison subjects (N = 305, 151 males and 154 females, age range of 18 –70 years) were randomly selected from among participants in a large employee health screening program. In the patient group, homocysteine levels were measured in serum from blood taken via venipuncture for measurements of lithium, valproate or carbamazepine levels. Deterioration was rated by consensus of two clinicians blind to homocysteine levels but familiar with the patients. Deterioration was defined as decrements in occupational, social, intellectual, or interpersonal functioning in the bipolar patients while in euthymic state and not explainable by normal life changes such as retirement.

1186

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

Table 4 Plasma homocysteine levels (AM) in young bipolar patients (age < 50 years) according to deterioration status and in comparison subjects (mean T S.D.) Age

18 – 39 40 – 49

Bipolar patients with deterioration

Bipolar patients with no deterioration

Comparison subjects

Males

Females

Males

Females

Males

Females

17.8 T 2.0 (n = 3) 15.0 T 6.2 (n = 4)

9.2 T 2.1 (n = 2) –

11.0 T 2.8 (n = 5) 10.5 T 1.5 (n = 4)

7.6 T 1.2 (n = 6) 8.0 T 1.4 (n = 2)

9.4 T 2.3 (n = 60) 11.5 T 1.9 (n = 35)

9.6 T 2.3 (n = 67) 10.7 T 2.4 (n = 30)

Table modified from Osher et al. (2004).

Homocysteine data were analyzed by two-way ANCOVA, with group (patients vs. controls) and sex as main factors, and age as a covariant. Significant main effects were found for age ( F(1,341) = 91.7, p < 0.001) and for sex ( F(1,341) = 24.4, p < 0.001) and a significant two-way interaction was found for group by sex ( F(1,341) = 14.1, p < 0.001). One-way (by group) ANCOVAs were then performed separately for each sex. For the male subjects, a main effect was found for group ( F(1,171) = 10.7, p = 0.001) and for age ( F(1,171) = 41.2, p < 0.001); young male bipolar patients had higher homocysteine levels than controls. Bipolar patients were then separated into those with and without functional deterioration, and two-way ANCOVA was performed, sex and deterioration status as independent variables and age as covariate. Results of young bipolar and comparison subjects are summarized in Table 4. Significant main effects were found for age ( F(1,399) = 122, p < 0.001) and for sex ( F(1,399) = 11.1, p = 0.001) and a significant interaction effect was found for group by sex ( F(2,399) = 3.2, p < 0.05). In order to further explore the interaction, separate oneway ANOVAs were conducted for each sex. Among the males, a significant main effect was found for group ( F(2,171) = 5.19, p = 0.006). Sheffe’s post-hoc comparisons of plasma homocysteine levels (AM) of male bipolar and comparison subjects according to deterioration status showed that bipolar deteriorated subjects versus control subjects had statistically significant higher homocysteine levels ( p = 0.007), whereas deteriorated versus non-deteriorated bipolar subjects showed a trend for higher homocysteine levels ( p < 0.08) in deteriorated bipolar patients. Since valproic acid may have anti-folate effects, we analyzed homocysteine levels of patients with and without this medication. In fact, those patients (N = 10) taking valproate had slightly non-significant lower homocysteine levels. These results suggest that bipolar patients who show functional deterioration have an elevation of serum homocysteine levels. Elevated homocysteine levels thus seem to be present only in the one third of bipolar patients with functional deterioration suggesting that homocysteine, a neurotoxin (Kruman et al., 2000), may be related to deterioration in bipolar disorder as was suggested for Alzheimer’s (Seshadri et al., 2002). We speculated above that negative symptoms and cognitive impairments in schizophrenia may lead to functional deterioration, neglect of diet and increase of

homocysteine levels and that these high homocysteine levels may lead to further cognitive impairment and negative symptoms because of homocysteine’s neurotoxic effects (Seshadri et al., 2002). Similarly in deteriorated bipolar euthymic patients the neglect of adequate diet and B vitamin consumption, decreased physical activity and increased smoking may lead to high homocysteine levels which due to their neurotoxic effects may induce more cognitive impairment, functional deterioration, and so on. Thus, plasma homocysteine elevation seems especially interesting since it is found both in schizophrenia and in that subgroup of bipolar patients exhibiting a key characteristic associated with schizophrenia-deterioration. This concept could have importance in the treatment or prevention of functional deterioration in patients, as plasma homocysteine levels can be markedly reduced by folic acid and vitamin B12 supplements. In this context, it should be noted that folic acid has previously been reported to reduce affective relapses in bipolar patients (Coppen et al., 1986).

6. Effects of chronic homocysteine administration on cognitive functions in mice The study of heuristic models for schizophrenia has several goals as cited by Lipska and Weinberger (2002) including testing the plausibility of new theories derived from emerging research data about the disorder, probing the explanatory power of new biological findings about the disorder and uncovering mechanisms of schizophrenia-like phenomena. Heuristic models for schizophrenia include models in which stereotypic behavior is induced (i.e., apomorphine-induced stereotypic behavior) and models showing cognitive deficits (i.e., Morris Water Maze (MWM) impaired performance model). In the present study we tried to develop a model of homocysteine neurotoxicity in mice, studying whether homocysteine administration to male mice may increase apomorphine-induced stereotypic behavior and/or lead to alterations in MWM performance suggesting cognitive deficit. 6.1. Animals One hundred and sixty male ICR mice, 1 month old weighing 25 – 30 g were maintained under controlled temper-

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

ature and light controlled conditions (25 -C, 12-h light/12-h dark). Each cage contained 10 mice. All mice had one week in their cages for acclimatization. Food administered was standard mouse diet from Koffolk Ltd. (Petach-Tikva, Israel) along with drinking water ad libitum. The experimental group had homocysteine 2 mg/cm3 added to the drinking water, changed daily. 6.2. Morris Water Maze (MWM) A round tank, 100 cm in diameter and 40 cm deep, filled with water was put in a small room designed for the experiment. The water temperature was maintained at 25 -C. A platform submerged 1 cm under the water surface was placed in the center of one of four imaginary quadrants of the tank and maintained in the same position during all first four days of the trial (this phase is defined as the ‘‘Acquisition’’ phase). Several distal visual cues were placed on the walls surrounding the water tank. Each day the mice were put successively four times in the maze, each time from a different imaginary quadrant of the water tank. The escape latency to find the platform was measured in seconds as the period between putting each mouse in the tank until it found the platform. If the mouse could not find the platform in the first trial of the acquisition phase within 2 min, he was put on the platform for 30 sec, enabling him to learn the location of the platform. On day 5 the platform was taken out (Extinction phase) and the time each mouse spent in the quadrant where the platform used to be (training quadrant) was measured. On days 6 and 7 the position of the platform was changed to another quadrant and the escape latency was measured. Trials were terminated after 2 min. After each trial all mice were kept in room temperature. 6.3. Apomorphine-induced stereotypy Two mice, one from the study group and the other from the control group were put each into two identical chambers. Chambers were made of a covered cylinder glass, 14 cm in diameter and 18 cm high. Prior to starting the study, the mice were held in chambers for 10 min for the purpose of acclimatization. The mice were then taken out of the chambers, injected s.c. with 0.5 mg/kg apomorphine and were then put back into the chambers for another 10 min after which the time spent by the mice in climbing behavior (climbing behavior is defined as a behavior in which both two forward paws of the mice are put on the walls of the chamber) was measured for another 10 min. The experiment was recorded by a video camera and a stopwatch was used to measure the time spent climbing the walls.

1187

its long axis by three partitions, enabling the examination of four mice at a given trial. The rotarod was accelerated in a range 4 –40 rpm. Mice had to keep walking on the rotating cylinder in order to avoid falling. Mice were given a practice trial while the cylinder was rotating in 4 rpm until they demonstrated coordinated walking which enabled them to stay on the rotarod. The frequency of the rotation rod was gradually increased within 10 min up to 40 rpm. The time each mouse spent on the rotating cylinder until falling (starting after practicing) was defined as the ‘‘latency time’’. 6.5. Procedure Eighty mice were administered homocysteine in their drinking water (the study group). The average dose of homocysteine consumed by the mice in the homocysteine group was 200 mg/kg/day, since each mouse drank about 3 cm3 of water per day and the average weight of mice at baseline (after 1 week of acclimatization) was 30 g. The control group included 80 mice treated with the standard diet and drinking water only. MWM test was performed after 2, 3, 4, 5 and 6 months of treatment. Each month, 10 mice from the study group and 10 mice from the control group were tested with the MWM behavioral test. Apomorphine stereotypy was tested after 3, 5 and 6 months of treatment. For each time point 10 mice from the study group and 10 mice from the control group were tested. Rotarod test was conducted at 5 month of treatment. 20 mice were tested with the rotarod test (10 mice from the study group and 10 mice from the control group). Mice tested with the rotarod test had been tested before in the MWM test. At 1, 2, 3, 4, 5 or 6 months of treatment after testing in the above behavioral tests, mice were sacrificed by decapitation and carotid blood homocysteine levels were taken. All treatment and testing procedures were approved by the Animal Care Committee of Ben Gurion University of the Negev and were carried out in accordance with the NIH Guide for the Care and Use of Laboratory Animals. 6.6. Homocysteine plasma levels Homocysteine levels at 1, 2, 3, 5, 6 months starting from month 1 to 6 of treatment were elevated between 300% (after 1 month) and 800% (at 6 months) of the control group by this procedure. Mean homocysteine levels of the study and control groups at the end of the study were 49.6 T 9.6 AM and 6 T 1.5 AM, respectively. 6.7. MWM behavioral test

6.4. Rotarod test The rotarod test is used to study motor coordination and skill learning. The rotarod is a rotating cylinder. Its diameter is 3 cm and its long axis is about 30 cm. It is divided along

No homocysteine-induced alterations in the MWM behavioral test were found in months 2, 3, and 4 (data not shown). A mild but statistically significant difference was found in the MWM between the study and the control

1188

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

Time (sec)

A

MWM at 6 months of treatment (mean ± SEM). control

60 50 40 30 20 10 0

hcy

1

2

3

4

Trials

B

MWM at 6 months of treatment (modified paradigm)(mean ± SEM).

control

Time (sec)

70 60

hcy

50 40 30 20 10 0 1

2

3

4

1

2

3

4

Trials Fig. 1. (A) Control mice spent statistically more time than the homocysteine fed mice at the quadrant where the platform used to be at the extinction phase. Two-way ANOVA between study and placebo groups, F(1,18) = 22.8, p < 0.0002; between trials, F(3,54) = 3.14, p < 0.03; interaction between group and trials, F(3,54) = 11.1, p < 0.001. (B) Homocysteine fed mice spent statistically significant more time on day 6 (reversal phase) in finding the platform. Three-way ANOVA between study groups F(1,11) = 4.7, p < 0.04.

groups at months 5 and 6. The groups, receiving homocysteine for 5 and 6 months, compared with the appropriate control group spent statistically significant less time on day 5 in the quadrant where the platform used to be on previous days (extinction phase) (two-way ANOVA in the extinction phase: between study and placebo groups, F(1,18) = 22.84, p < 0.0002; between trials, F(3,54) = 3.14, p < 0.03; interaction between group and trials, F(3,54) = 11.1, p < 0.001; Fig. 1A). No statistically significant changes were found for the acquisition and reversal phases. 6.8. Modified MWM behavioral test The MWM experimental design was modified at month 6 so that all parameters were the same except that the number of days applied was 6 instead of 7; days 1 – 3 for the ‘‘Acquisition phase’’ (instead of 4 days in the previous experiments), day 4 for the ‘‘Extinction phase’’ and days 5 and 6 for the ‘‘Reversal phase’’. The reason for such modification was the finding of a ‘‘floor effect’’ for learning at day 3 of the acquisition phase, (mean time for mice to find the platform was less than 20 sec) so that the ‘‘Extinction phase’’ could be started at day 4. After changing the paradigm as depicted above, the study group compared with the appropriate control group showed no difference in the acquisition and extinction phases (Fig. 1A and B) whereas

homocysteine fed mice spent statistically significant more time on day 6 (reversal phase) finding the platform. Threeway ANOVA between study groups F(1,11) = 4.7, p < 0.04 (Fig. 1B). 6.9. Apomorphine behavioral test No homocysteine-induced alterations in the apomorphine behavioral test were found at months 3, 5 and 6 of treatment. Mean T S.E.M. climbing times for all experiments held was 329 T 42 sec for the homocysteine group whereas it was 404 T 63 sec for the control group. 6.10. Rotarod behavioral test No homocysteine-induced alterations in the rotarod behavioral test were found at 5 months of treatment (mean T S.E.M. for the study group receiving homocysteine was 42.4 T 11.2 sec whereas it was 39.54 T 8.4 sec for the control group). 6.11. Discussion of behavioral results Previous studies used much higher doses of homocysteine-administered i.p. (i.e., 1200 mg/kg) acutely, a design aimed at inducing epileptic seizures in mice Such studies

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

served as a model for epilepsy but bear no clinical relevancy to schizophrenia. Only few studies applied a design aimed at reaching mild to moderate plasma homocysteine elevations in mice. Thus Zhou et al. (2001) induced hyperhomocysteinemia in apolipoprotein E-deficient mice. Hyperhomocysteinemia was induced by either homocysteine or methionine supplementation: low methionine (11 g/kg food), high methionine (33 g/kg food), low homocysteine (0.9 g/L drinking water), and high homocysteine (1.8 g/L drinking water). The treatment was administered up to 12 months demonstrating that homocysteine and methionine supplementation significantly raised plasma total homocysteine levels by 4- to 16-fold above those observed in mice fed a control diet. The homocysteine diet containing 1.8 g/L drinking water led to aortic root plaque, increase in the amount of collagen in plaques and atherosclerosis and plaque fibrosis. Homocysteine plasma levels after 3 months of high homocysteine treatment (1.8 g/L) increased to 35.6 T 8.9 Amol/L and after 12 months of treatment to 146.1 T16.7 Amol/L. Interestingly in our study, mice administered with homocysteine 2 g/L demonstrated an increase of homocysteine plasma levels to about 20 Amol/L after 3 months and to about 50 Amol/L after 6 months of treatment. Other studies aiming at inducing mild to moderate hyperhomocysteinemic states in mice used vitamin B deficiency diets. Thus Duan et al. (2002) using folate deficient diet administered for 3 months by i.p. injections of 20 mg/kg of folate as the only source of this vitamin. These authors report that such a design led to increased homocysteine plasma levels of up to 25 Amol/L at the end of the treatment period (reaching approximately the same levels demonstrated in our experiment). We chose to elevate homocysteine by direct administration of this compound rather than by folate restriction in order to distinguish between brain effect of hyperhomocysteinemia and folate deficiency. On the whole, the paradigm applied in this study failed to demonstrate changes in apomorphine-induced stereotypic behavior in male mice. The current study reports no homocysteine-induced impairments in the Morris Water Maze until months 5 and 6. At month 5 homocysteine plasma levels increased by about 600% compared with baseline to about 35 Amol/L, whereas at month 6 plasma homocysteine levels increased by about 800% of baseline to about 50 Amol/L. The study did demonstrate alterations in spatial learning and memory as exhibited by the MWM following 5 and 6 months of homocysteine treatment. We found that on 5 and 6 months mice administered homocysteine spent less time on day 5 of the experiment (‘‘Extinction’’ phase) in the quadrant where the platform used to be suggesting a more rapid extinction of spatial memory at trials 2 – 4 in the face of ‘‘no platform’’ compared with the control mice. Another possible explanation of such behavior is that the mice exhibit faster and better spatial learning of the new

1189

constellation where no platform was present is less likely since no difference as to the spatial learning was demonstrated in the ‘‘acquisition phase’’ of the experiment (days 1 –4). Using the modified MWM mice demonstrated on month 6 a reduced ability for spatial learning on the ‘‘reversal phase’’ a phase where the platform was put in a new quadrant of the maze. It may be suggested that the ‘‘extinction’’ and ‘‘reversal’’ phases are more sensitive indexes of spatial memory and learning in the face of changing environments. Interestingly such difficulty to shift between strategies or modes of operation in the face of new paradigm was also demonstrated in schizophrenic patients (Goldberg and Weinberger, 1994). Thus chronic moderate elevation of homocysteine plasma levels may induce mild cognitive impairments related to spatial learning and memory. Is our finding relevant for schizophrenia? Wood et al. recently suggested that spatial working memory ability may be a marker of risk-for-psychosis and Glahn et al. (2003) reported that spatial working memory impairment may serve as an endophenotype for schizophrenia.

7. Conclusions We found in two large separate samples of Southern Israeli newly admitted and hospitalized schizophrenia patients that homocysteine levels are higher in schizophrenia patients compared with healthy controls. The increase was found almost entirely in young male schizophrenic patients. Multiple regression analysis suggested that folate and B12 levels explained about one fourth of the variance leaving three-quarters yet to be explained. CSF homocysteine levels showed no difference between schizophrenia patients and controls. Homocysteine in CSF is present in much smaller concentrations than in plasma and may be less relevant for the metabolic effects. Bipolar disorder and schizophrenia share biological and phenomenological characteristics. We have found that euthymic bipolar patients compared with controls have higher homocysteine levels. The increase was mainly found in young male patients who showed functional deterioration. Although the sample of bipolar patients was rather small, the results suggest that plasma homocysteine levels may be associated with functional deterioration per se rather than with a specific diagnosis. One cannot simply extrapolate our findings to schizophrenia and bipolar disorder populations studied at other geographical areas. Goff et al. (2004) did not find elevated plasma homocysteine in schizophrenia. Plasma homocysteine levels are determined by an interplay of a host of environmental, life style and genetic factors which may differ between different geographical areas. High homocysteine levels may affect the course of schizophrenia and bipolar disorder but are not mandatory for the appearance of illness episodes.

1190

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191

The animal study exemplifies that chronic exposure to hyperhomocysteinemia may lead to mild cognitive impairments. Functional deterioration may follow cognitive impairment and the functional impairment reported above associated with high homocysteine levels may be mediated via homocysteine-induced cognitive impairment.

References Antun, F.T., Burnett, G.B., Cooper, A.J., Daly, R.J., Smythies, J.R., Zealley, A.K., 1971. The effects of l-methionine (without MAOI) in schizophrenia. J. Psychiatr. Res. 8 (2), 63 – 71. APPI, 1994. DSM-IV: Diagnostic and Statistical Manual of Psychiatric Disorders. APPI Press, Washington DC. Applebaum, J., Shimon, H., Sela, B.A., Levine, J., 2004. Plasma homocysteine levels in newly admitted schizophrenic patients. J. Psychiatr. Res. 4, 413 – 416. Arai, H., 1996. Biological markers for the clinical diagnosis of Alzheimer’s disease. Tohoku J. Exp. Med. 179 (2), 65 – 79. Araki, A., Sako, Y., 1987. Determination of free and total homocysteine in human plasma by high-performance liquid chromatography with fluorescence detection. J. Chromatogr. 422, 43 – 52. Berger, P.A., Elliott, G.R., Erdelyi, E., Watson, S.J., Wyatt, R.J., Barchas, J.D., 1977. Platelet methylene reductase activity in schizophrenia. Arch. Gen. Psychiatry 34 (7), 808 – 809. Bleich, S., Junemann, A., Von Ahsen, N., Lausen, B., Ritter, K., Beck, G., Naumann, G.O., Kornhuber, J., 2002. Homocysteine and risk of openangle glaucoma. J. Neural Transm. 109 (12), 1499 – 1504. Bleich, S., Bandelow, B., Javaheripour, K., Muller, A., Degner, D., Wilhelm, J., Havemann-Reinecke, U., Sperling, W., Ruther, E., Kornhuber, J., 2003. Hyperhomocysteinemia as a new risk factor for brain shrinkage in patients with alcoholism. Neurosci. Lett. 335 (3), 179 – 182. Choumenkovitch, S.F., Selhub, J., Wilson, P.W., Rader, J.I., Rosenberg, I.H., Jacques, P.F., 2002. Folic acid intake from fortification in United States exceeds predictions. J. Nutr. 132 (9), 2792 – 2798. Clark, L., Iversen, S.D., Goodwin, G.M., 2002. Sustained attention deficit in bipolar disorder. Br. J. Psychiatry 180, 313 – 319 (Show). Cohen, S.M., Nichols, A., Wyatt, R., Pollin, W., 1974. The administration of methionine to chronic schizophrenic patients: a review of ten studies. Biol. Psychiatry 8 (2), 209 – 225. Coppen, A., Chaudhry, S., Swade, C., 1986. Folic acid enhances lithium prophylaxis. J. Affect. Disord. 10 (1), 9 – 13 (Show: 5 10 20 50 100 200 500 Sort Author Journal Pub Date Text File). Desouza, C., Keebler, M., McNamara, D.B., Fonseca, V., 2002. Drugs affecting homocysteine metabolism: impact on cardiovascular risk. Drugs 62 (4), 605 – 616. Duan, W., Ladenheim, B., Cutler, R.G., Kruman, II, Cadet, J.L., Mattson, M.P., 2002. Dietary folate deficiency, elevated homocysteine levels endanger dopaminergic neurons in models of Parkinson’s disease. J. Neurochem. 80 (1), 101 – 110. Elliott, G.R., Sutherland, K., Erdelyi, E., Ciaranello, R.D., Barchas, J.D., Wyatt, R.J., 1978. N5,10-methylenetetrahydrofolate reductase activity in autopsied brain parts of chronic schizophrenics and controls and in vitro tryptoline formation. Biol. Psychiatry 13 (6), 695 – 708 (Write to the Help Desk NCBI | NLM | NIH Department of Health and Human Services Privacy Statement | Freedom of Information Act | Disclaimer). Ferrier, I.N., Stanton, B.R., Kelly, T.P., Scott, J., 1999. Neuropsychological function in euthymic patients with bipolar disorder. Br. J. Psychiatry 175, 246 – 251 (Show). Freeman, J.M., Finkelstein, J.D., Mudd, S.H., 1975. Folate-responsive homocystinuria and ‘‘schizophrenia’’. A defect in methylation due to deficient 5,10-methylenetetrahydrofolate reductase activity. N. Engl. J. Med. 292 (10), 491 – 496.

Friedman, A., Kaufer, D., Shemer, J., Hendler, I., Soreq, H., Tur-Kaspa, I., 1996. Pyridostigmine brain penetration under stress enhances neuronal excitability and induces early immediate transcriptional response. Nat. Med. 2 (12), 1382 – 1385. Galasko, D., 2003. Cerebrospinal fluid biomarkers in Alzheimer disease: a fractional improvement?. Arch. Neurol. 60 (9), 1195 – 1196. Garver, D.L., Tamas, R.L., Holcomb, J.A., 2003. Elevated interleukin-6 in the cerebrospinal fluid of a previously delineated schizophrenia subtype. Neuropsychopharmacology 28 (8), 1515 – 1520. Glahn, D.C., Therman, S., Manninen, M., Huttunen, M., Kaprio, J., Lonnqvist, J., Cannon, T.D., 2003. Spatial working memory as an endophenotype for schizophrenia. Biol. Psychiatry 53 (7), 624 – 626. Godfrey, P.S., Toone, B.K., Carney, M.W., Flynn, T.G., Bottiglieri, T., Laundy, M., Chanarin, I., Reynolds, E.H., 1990. Enhancement of recovery from psychiatric illness by methylfolate. Lancet 336 (8712), 392 – 395. Goff, D.C., Bottiglieri, T., Arning, E., Shih, V., Freudenreich, O., Evins, A.E., Henderson, D.C., Baer, L., Coyle, J., 2004. Folate, homocysteine, and negative symptoms in schizophrenia. Am. J. Psychiatry 161 (9), 1705 – 1708. Goldberg, J.F., Harrow, M., 1999. Poor outcome bipolar disorders. In: Goldberg II, J.F., Harrow, M. (Eds.), Bipolar Disorders, Clinical Course and Outcome. American Psychiatric Press, Washington DC, pp. 1 – 19. Goldberg, T.E., Weinberger, D.R., 1994. Schizophrenia, training paradigms, and the Wisconsin card sorting test redux. Schizophr. Res. 11 (3), 291 – 296. Grundke-Iqbal, I., Iqbal, K., 1999. Tau pathology generated by overexpression of tau. Am. J. Pathol. 155 (6), 1781 – 1785. Kaluski, D.N., Tulchinsky, T.H., Haviv, A., Averbuch, Y., Rachmiel, S., Berry, E.M., Leventhal, A., 2003. Addition of essential micronutrients to foods—implication for public health policy in Israel. Isr. Med. Assoc. J. 5 (4), 277 – 280. Kozlovsky, N., Belmaker, R.H., Agam, G., 2000. Low GSK-3beta immunoreactivity in postmortem frontal cortex of schizophrenic patients. Am. J. Psychiatry 157 (5), 831 – 833. Kozlovsky, N., Regenold, W.T., Levine, J., Rapoport, A., Belmaker, R.H., Agam, G., 2004. GSK-3beta in cerebrospinal fluid of schizophrenia patients. J. Neural Transm. 111 (8), 1093 – 1098 (Electronic publication 2004 Apr 27.). Kruman, II, Culmsee, C., Chan, S.L., Kruman, Y., Guo, Z., Penix, L., Mattson, M.P., 2000. Homocysteine elicits a DNA damage response in neurons that promotes apoptosis and hypersensitivity to excitotoxicity. J. Neurosci. 20 (18), 6920 – 6926. Kruse, T., Reiber, H., Neuhoff, V., 1985. Amino acid transport across the human blood – CSF barrier. An evaluation graph for amino acid concentrations in cerebrospinal fluid. J. Neurol. Sci. 70 (2), 129 – 138. Levine, J., Stahl, Z., Sela, B.A., Gavendo, S., Ruderman, V., Belmaker, R.H., 2002. Elevated homocysteine levels in young male patients with schizophrenia. Am. J. Psychiatry 159 (10), 1790 – 1792. Levine, J., Agam, G., Sela, B.A., Garver, D.L., Torrey, E.F., Belmaker, R.H., 2005. CSF homocysteine is not elevated in schizophrenia. J. Neural Transm. 112 (2), 297 – 302. Levy, S., Rachmilewitz, M., Grossowicz, N., Reshef, Y., Izak, G., 1975. Nutritional survey in an iron- and folate-deficient population. Am. J. Clin. Nutr. 28 (12), 1454 – 1457. Lipska, B.K., Weinberger, D.R., 2002. A neurodevelopmental model of schizophrenia: neonatal disconnection of the hippocampus. Neurotox. Res. 4 (5 – 6), 469 – 475. Lipton, S.A., Kim, W.K., Choi, Y.B., Kumar, S., D’Emilia, D.M., Rayudu, P.V., Arnelle, D.R., Stamler, J.S., 1997. Neurotoxicity associated with dual actions of homocysteine at the N-methyl-d-aspartate receptor. Proc. Natl. Acad. Sci. U. S. A. 94 (11), 5923 – 5928. MacQueen, G.M., Young, L.T., Galway, T.M., Joffe, R.T., 2001. Backward masking task performance in stable, euthymic out-patients with bipolar disorder. Psychol. Med. 31 (7), 1269 – 1277. Muntjewerff, J.W., van der Put, N., Eskes, T., Ellenbroek, B., Steegers, E., Blom, H., Zitman, F., 2003. Homocysteine metabolism and B-vitamins

J. Levine et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 29 (2005) 1181 – 1191 in schizophrenic patients: low plasma folate as a possible independent risk factor for schizophrenia. Psychiatry Res. 121 (1), 1 – 9. Nygard, O., Vollset, S.E., Refsum, H., Stensvold, I., Tverdal, A., Nordrehaug, J.E., Ueland, M., Kvale, G., 1995. Total plasma homocysteine and cardiovascular risk profile. The Hordaland Homocysteine Study. JAMA 274 (19), 1526 – 1533. O’Callaghan, P., Meleady, R., Fitzgerald, T., Graham, I., 2002. Smoking and plasma homocysteine. Eur. Heart J. 23 (20), 1580 – 1586. Osher, Y., Sela, B.A., Levine, J., Belmaker, R.H., 2004. Elevated homocysteine levels in euthymic bipolar disorder patients showing functional deterioration. Bipolar Disord. 6 (1), 82 – 86. Pollin, W., Cardon, P., Kety, S., 1961. Effect of amino acid feedings in schizophrenic patients treated with iproniazid. Science 133, 104 – 105. Preble, O.T., Torrey, E.F., 1985. Serum interferon in patients with psychosis. Am. J. Psychiatry 142 (10), 1184 – 1186. Regland, B., Johansson, B.V., Gottfries, C.G., 1994. Homocysteinemia and schizophrenia as a case of methylation deficiency. J. Neural Transm. 98 (2), 143 – 152. Regland, B., Johansson, B.V., Grenfeldt, B., Hjelmgren, L.T., Medhus, M., 1995. Homocysteinemia is a common feature of schizophrenia. J. Neural Transm. 100 (2), 165 – 169. Regland, B., Germgard, T., Gottfries, C.G., Grenfeldt, B., KochSchmidt, A.C., 1997. Homozygous thermolabile methylenetetrahydrofolate reductase in schizophrenia-like psychosis. J. Neural Transm. 104 (8 – 9), 931 – 941. Sachdev, P.S., Valenzuela, M., Wang, X.L., Looi, J.C., Brodaty, H., 2002. Relationship between plasma homocysteine levels and brain atrophy in healthy elderly individuals. Neurology 58 (10), 1539 – 1541. Selhub, J., 1999. Homocysteine metabolism. Annu. Rev. Nutr. 19, 217 – 246. Selkoe, D.J., 1993. Physiological production of the beta-amyloid protein and the mechanism of Alzheimer’s disease. Trends Neurosci. 16 (10), 403 – 409. Serot, J.M., Bene, M.C., Faure, G.C., 2003. CSF homocysteine, CSF folates and choroid plexus. Neurobiol. Aging 24 (4), 627 – 628 (discussion 629). Seshadri, S., Beiser, A., Selhub, J., Jacques, P.F., Rosenberg, I.H., D’Agostino, R.B., Wilson, P.W., Wolf, P.A., 2002. Plasma homocys-

1191

teine as a risk factor for dementia and Alzheimer’s disease. N. Engl. J. Med. 346 (7), 476 – 483. Stahl, Z., Belmaker, R.H., Friger, M., Levine, J., 2005. Nutritional and life style determinants of plasma homocysteine in schizophrenia patients. Eur. Neuropsychopharmacol. 15 (3), 291 – 295. Stuhlinger, M.C., Oka, R.K., Graf, E.E., Schmolzer, I., Upson, B.M., Kapoor, O., Szuba, A., Malinow, M.R., Wascher, T.C., Pachinger, O., Cooke, J.P., 2003. Endothelial dysfunction induced by hyperhomocyst(e)inemia: role of asymmetric dimethylarginine. Circulation 108 (8), 933 – 938 (Electronic publication 2003 Aug 11). Susser, E., Brown, A.S., Klonowski, E., Allen, R.H., Lindenbaum, J., 1998. Schizophrenia and impaired homocysteine metabolism: a possible association. Biol. Psychiatry 44 (2), 141 – 143. Teunissen, C.E., de Vente, J., Steinbusch, H.W., De Bruijn, C., 2002. Biochemical markers related to Alzheimer’s dementia in serum and cerebrospinal fluid. Neurobiol. Aging 23 (4), 485 – 508. van Gorp, W.G., Altshuler, L., Theberge, D.C., Wilkins, J., Dixon, W., 1998. Cognitive impairment in euthymic bipolar patients with and without prior alcohol dependence. A preliminary study. Arch. Gen. Psychiatry 55 (1), 41 – 46. Virgos, C., Martorell, L., Simo, J.M., Valero, J., Figuera, L., Joven, J., Labad, A., Vilella, E., 1999. Plasma homocysteine and the methylenetetrahydrofolate reductase C677T gene variant: lack of association with schizophrenia. NeuroReport 10 (10), 2035 – 2038. Zandi, P.P., Carlson, M.C., Plassman, B.L., Welsh-Bohmer, K.A., Mayer, L.S., Steffens, D.C., Breitner, J.C., 2002. Hormone replacement therapy and incidence of Alzheimer disease in older women: the Cache County Study. JAMA 288 (17), 2123 – 2129. Zhou, J., Moller, J., Danielsen, C.C., Bentzon, J., Ravin, H.B., Austin, R.C., Falk, E., 2001. Dietary supplementation with methionine and homocysteine promotes early atherosclerosis but not plaque rupture in ApoEdeficient mice. Arterioscler. Thromb. Vasc. Biol. 21 (9), 1470 – 1476. Zubieta, J.K., Huguelet, P., O’Neil, R.L., Giordani, B.J., 2001. Cognitive function in euthymic bipolar I disorder. Psychiatry Res. 102 (1), 9 – 20.