Elsevier

Biological Psychiatry

Volume 52, Issue 8, 15 October 2002, Pages 823-830
Biological Psychiatry

Original article
Correlations between peripheral polyunsaturated fatty acid content and in vivo membrane phospholipid metabolites

https://doi.org/10.1016/S0006-3223(02)01397-5Get rights and content

Abstract

Background

There is evidence for membrane abnormalities in schizophrenia. It is unclear whether the observed membrane deficits in peripheral cells parallel central membrane phospholipid metabolism. To address this question we examined the relations between red blood cell polyunsaturated fatty acids and brain phospholipid metabolites from different regions of interest in schizophrenia and healthy subjects.

Methods

Red blood cell membrane fatty acids were measured by capillary gas chromatography and in vivo brain phospholipid metabolite levels were measured using a multi-voxel 31P Magnetic Resonance Spectroscopy technique on 11 first-episode, neuroleptic-naïve schizophrenic subjects and 11 normal control subjects.

Results

Both the total polyunsaturated fatty acids and the individual 20:4(n-6) contents were significantly correlated with the freely-mobile phosphomonoester [PME(s-τc)] levels (r = .5643, p = .0062 and r = .6729, p = .0006, respectively). The 18:2(n-6) polyunsaturated fatty acids content correlated positively with freely-mobile phosphodiester [PDE(s-τc)] levels (r = .5573, p = .0071). The above correlations were present in the combined right and left prefrontal region of the brain, while other regions including the basal ganglia, occipital, inferior parietal, superior temporal and centrum semiovale yielded no significant correlations.

Conclusions

Our preliminary data support the association between the decreased red blood cell membrane phospholipid polyunsaturated fatty acids content and the decreased building blocks [PME(s-τc)] and breakdown products [PDE(s-τc)] of membrane phospholipids in the prefrontal region of first-episode, neuroleptic-naïve schizophrenic subjects.

Introduction

There is substantial evidence for peripheral membrane abnormalities in chronic schizophrenic patients Vaddadi et al 1989, Glen et al 1994, Yao et al 1994, Yao et al 1996, Yao and van Kammen 1994, Yao and van Kammen 1996, Peet et al 1996, and a relative paucity of such evidence in early schizophrenia (Reddy et al 1999). On the other hand, there is convincing evidence for abnormal brain phospholipid metabolism in first-episode and chronic schizophrenic patients Pettegrew et al 1991, Keshavan et al 2000. The relation between peripheral findings and brain membrane alterations is unknown. If confirmed, such a relationship provides a unique opportunity to simultaneously investigate central and peripheral biochemistry, and their relations to clinical measures. Relations between peripheral and central phospholipid metabolism, if found, will also have several important implications for schizophrenia research. Firstly, the findings to date in peripheral tissue will assume greater relevance for understanding aspects of pathophysiology of schizophrenia. This issue has been vigorously debated because of examples in the literature, where peripheral measures either failed to adequately reflect central pathophysiology or did not serve as reliable biological markers. Secondly, the findings will support the notion that membrane abnormalities are present in both neural and extra-neural tissues. However, the functional consequences of the membrane deficits may differ across the tissue types. There are several conditions such as Down’s syndrome, phenylketonuria, and various lipidoses (Scriver et al 1989), where the metabolic abnormalities are expressed in both neural and peripheral tissues, but the functional consequences are most profound in the central nervous system. This paradigm may also apply to schizo-phrenia. Thirdly, if peripheral membrane indices parallel central phospholipid metabolism, and perhaps also neuromorphometric findings, then there exists the possibility that alterations in peripheral membrane indices on longitudinal follow ups (repeated measures) can reflect central nervous system membrane function over the course of illness.

While a correlation between red blood count (RBC) polyunsaturated fatty acids (PUFAs) and brain phospholipid metabolites may not indicate a direct causal relationship, there is a growing body of evidence suggesting that RBC membrane fatty acid changes parallel the brain membrane fatty acid changes. In an earlier study by Carlson et al (1986), fatty acid-enriched diet in rats resulted in a similar relative changes in phospholipid fatty acids of both neural and RBC membranes. Later, in fatty acid deficient juvenile rhesus monkeys fed with fish oil-rich diet, a parallel increase of 22:6(n-3) was also shown in brain and RBC membranes (Connor et al 1990). Following modification of maternal diet with enrichment in fish oil, parallel increases of 22:6(n-3) levels were found in brain and RBC membranes of the piglets (Arbuckle and Innis 1993) and suckling rats (Araya et al 1994). Similarly, breastfed infants had a greater proportion of 22:6(n-3) in their erythrocytes and brain cortex relative to those fed formula (Makrides et al 1994). Recently, Yeh et al (1998) have shown an increased accretion of 20:4(n-6) and 22:6(n-3) levels in brain and RBC membranes after infant rats fed with rat milk formulas supplemented with either 20:4(n-6) or 22:6(n-3). In two double-blind placebo-controlled pilot studies of eicosapentaenoic acid (EPA) in the treatment of schizophrenia, Peet et al (2001) demonstrated that patients taking EPA had significantly lower scores on the first-episode, neuroleptic-naïve schizophrenic (FENNS) rating scale. Moreover, there was a positive correlation between clinical improvement and rise in RBC arachidonic acid concentration (Peet et al 2002). Taken together, these findings provide somewhat indirect support for the notion that peripheral measures of RBC PUFAs could indeed be reflecting central PUFA metabolism. However, few previous studies have directly examined the central-peripheral correlation in phospholipid metabolism in humans.

Using in vivo phosphorus magnetic resonance spectroscopy (31P MRS), it has been reported that schizophrenic patients show alterations in membrane phospholipid metabolism in the prefrontal (Pettegrew et al 1991) and other regions (for review, see Keshavan et al 2000). Thus, the purpose of the present study is to test whether RBC levels of phospholipid polyunsaturated fatty acids are associated with in vivo 31P spectroscopy measures of brain phospholipid metabolites, particularly in the prefrontal region.

Section snippets

Subject selections and diagnoses

Eleven FENNS subjects (males, n = 6; females, n = 5; mean age, 26; and age range, 17–44) were recruited through Western Psychiatric Institute and Clinic (WPIC) inpatient and outpatient units, after informed consent. A structured clinical interview (SCID) and available clinical information was utilized to derive a “best estimate” DSM-IV diagnosis of patients with schizophrenia, schizoaffective, or schizophreniform disorder. Ammons Quick Test was performed to estimate IQ (range was 82–107). The

Group differences in RBC membrane fatty acids

In agreement with our previous findings in chronic schizophrenic patients (Yao et al 1994), the RBC levels of arachidonic acid, 20:4(n-6), were significantly lower in FENNS subjects than in normal controls (Table 1). In addition, levels of total polyunsaturated fatty acids were also lower in FENNS patients than in normal controls, although it is not statistically significant (p = 0.08). No significant differences were observed in either saturated or monounsaturated fatty acids between FENNS

Discussion

Combining the FENNS and control subjects in the correlation analysis, resulted in a significant association between the total PUFA content and PME(sc) levels; a finding that only was present in the prefrontal and not in the basal ganglia, superior temporal, inferior parietal, occipital or centrum semiovale regions. As for the association between the individual PUFA and 31P metabolites in the prefrontal, the arachidonic acid [20:4(n-6)] content correlated with the PME(sc) levels and PCr/Pi

Acknowledgements

This study was supported in part by grants from the National Institute of Mental Health MH58141 (JKY), MH45203 (MSK), and MH46614 (JWP), NARSAD Young Investigator Award (RDR), Office of Research and Development (Merit Review, JKY), Department of Veterans Affairs, and the Highland Drive VA Pittsburgh Healthcare System. The authors are grateful to C. Korbanic and L. McElhinny for their technical assistance. The authors also are grateful for the time domain fitting software package, which was

References (35)

Cited by (0)

View full text