Aberrant Tyrosine Transport Across the Cell Membrane in Patients With Schizophrenia

Background  There is evidence that patients with schizophrenia exhibit abnormalities, not only in the brain but also in peripheral organs. An abnormal cell membrane composition has been suggested to be a common denominator, supported by findings of alterations in membrane phospholipid levels. In a previous study, the transport of amino acids across the plasma membrane was investigated with fibroblasts from patients with schizophrenia and controls. An isolated decrease in the maximal transport capacity (V_max) of tyrosine was observed in the cells from patients. In this context, tyrosine transport across the fibroblast membrane was investigated in patients with schizophrenia and healthy control subjects.

Methods  Skin fibroblasts were obtained from 36 patients with schizophrenia (15 first episode and 21 chronic) and 10 healthy controls. Tyrosine transport across the cell membrane was studied in cultivated fibroblasts. The V_max and the affinity of the tyrosine binding sites (K_m) were determined.

Results  Significantly lower V_max (F_1,41 = 12.80; P = .001; effect size = 1.36) and K_m (F_1,41 = 24.85; P<.001; effect size = 1.00) were observed in fibroblasts from the patients. The findings were present in both neuroleptic-naive patients with their first episode and patients with chronic schizophrenia.

Conclusions  The lower V_max and K_m are compatible with a cell membrane disturbance and support the view of schizophrenia as a systemic disorder. The decreased V_max and K_m observed in cells from schizophrenic patients probably reflect a genetic trait, as the changes were transmitted through several cell generations of cultured fibroblast.

THE SIGNIFICANCE of neurobiological changes in schizophrenia has become obvious through the past 2 decades of research. There is now evidence of abnormalities throughout the body involving not only the central nervous system but also peripheral organs. In studies of patients with schizophrenia, neuromuscular abnormalities have been shown, including histologic and electrophysiologic changes in muscle fibers.^1-4 Changes in the fatty acid composition of the cell membrane have been demonstrated,⁵ as has disturbed membrane phospholipid metabolism caused by enhanced activity of phospholipase A₂.^6,7 A disturbed phospholipid metabolism has been found not only in the periphery but also in the brain.^8,9 Moreover, reduced vasodilator responses to niacin and histamine as well as altered immunologic functions have also been observed.^10,11 In all, these findings suggest that schizophrenia is a systemic syndrome and not related only to brain function.

In a previous study,¹² the transport of amino acids across the plasma membrane was investigated by means of fibroblasts from patients with schizophrenia and control subjects. An isolated decrease in the maximal transport capacity (V_max) of tyrosine was observed in the cells from patients. Subsequently, a series of in vitro and in vivo studies showed evidence of aberrant tyrosine transport across the fibroblast membrane¹³ as well as the blood-brain barrier.^14,15 The decreased tyrosine transport into the cells from the patients could not be related to the function of the L-system, the major transport system for neutral amino acids, or any other known amino acid transport system. The aberrant tyrosine transport was suggested to be the result of a general dysfunction in plasma cell membrane in patients with schizophrenia.¹²

The aim of the present study was to investigate tyrosine transport kinetics across the cell membrane in a larger sample of patients with schizophrenia compared with control subjects by means of the fibroblast technique.¹² The patients in this study had previously undergone a series of investigations including clinical characterization, assessment of neurologic signs, muscle biopsy, and macroelectromyography.^4,16 Thus, a second objective was to investigate possible relationships between tyrosine kinetics and a family history of psychosis, neuromuscular and psychomotor findings, and clinical characteristics of the same patients.

Sixty-six consecutively recruited patients with psychotic symptoms were screened on admission for psychotic symptoms to a psychiatric clinic in Stockholm, Sweden. Forty-eight patients met the inclusion criteria, and 36 of these agreed to enter the study. Informed consent was obtained from all patients and control subjects, and the institutional ethics committee approved the protocol. For inclusion in the study, the patients had to meet the DSM-III-R criteria for schizophrenia and be between 18 and 45 years of age. The diagnosis of schizophrenia was independently made by 2 clinicians (L.F. and G.E.). The former, an experienced specialist in general psychiatry, made a clinical psychiatric investigation including the Positive and Negative Syndrome Scale¹⁷ and then applied the diagnostic criteria according to the DSM-III-R.¹⁸ The latter, a psychologist specializing in clinical psychology and trained in the computerized version of the Structured Clinical Interview for DSM-III-R,¹⁹ used it as a diagnostic tool. The two clinicians agreed in their diagnoses in 34 (94%) of 36 cases. In the 2 cases of disagreement, consensus was reached after consultation.

A patient was classified as having a first episode if he or she met the diagnostic criteria for schizophrenia for the first time and as having chronic disease if he or she had had a diagnosis of schizophrenia for at least 2 years before the present admission. Thirteen patients (4 in their first episode and 9 with chronic disease) were recruited from the outpatient facilities and 23 (11 first episode and 12 chronic) from the inpatient wards. Only 2 patients had been hospitalized for more than 1 month before the present admission (6 months and 1 year). Exclusion criteria were any history of alcohol or other drug abuse, head injury, and neurologic or serious somatic disease. The median prodromal period was 9 months in the first-episode group (range, 0.5-48 months). The patients with a history of illness of less than 6 months were followed up to confirm the diagnosis.

Thirty-two patients (89%) were taking neuroleptic drugs. Three of the first-episode patients had never taken neuroleptics, and 1 chronic patient had been without medication for 3 months. Twenty-four patients (67%) had taken conventional neuroleptics and 8 (22%) had taken clozapine in monotherapy. Four patients (11%) were taking anticholinergic drugs. Additional medication for daytime and nighttime sedation (benzodiazepines) was allowed. The duration of neuroleptic medication (in months) and the daily dosage (in equivalents of chlorpromazine hydrochloride) were established for each patient²⁰ (Table 1).

Data regarding a family history of psychosis were collected by a psychiatrist (L.F.) in interviews with the patients and at least 2 of their first-degree relatives. Information was obtained about all first- and second-degree relatives of the 36 patients according to the family history method described by Andreasen and coworkers.²¹ Patients with at least 1 first- and/or 1 second-degree relative with a psychotic condition (schizophrenic, schizoaffective, schizophreniform, delusional, or brief reactive psychotic disorder) were classified as family history positive (n = 21). If no first- or second-degree relative with a psychotic disorder was reported, the patient was classified as family history negative (n = 9). Six patients had a family history of other psychiatric disorders, but they were not included in the comparative analyses because of the small number and the heterogeneity of the group. There were no significant differences between the family history–positive and family history–negative groups in mean age (32.1 and 28.2 years, respectively; t₂₉ = 1.51; P = .14) and sex (P = .43, Fisher exact test).

Psychiatric assessments of all included subjects were performed by an experienced specialist in general psychiatry (L.F.) using the Positive and Negative Syndrome Scale,¹⁷ the Global Scale of Adaptive Functioning,¹⁸ and the Extrapyramidal Symptom Rating Scale.²² The social, occupational, and housing situation was rated by means of a modified Strauss-Carpenter outcome scale²³ (Table 1).

Ten healthy control subjects from previous studies,^4,16 similar to the patients in age and sex, were included in the study (Table 1). The first 10 among 55 control subjects recruited for a series of studies^4,16 by advertising in the press and at the Stockholm University were included in the present study. They underwent a thorough psychiatric, anamnestic, somatic, and laboratory investigation to exclude those with a history of head injury; drug abuse; systemic disease; significant somatic, neurologic, or psychiatric symptoms; or a history of psychiatric illness in their families.

Skin biopsies were performed within a week of the clinical measurements (all these investigations were made by L.F.). At the time of the skin biopsy, 7 of the 15 first-episode patients were still in an acute phase of the illness, and all except 1 of the patients with chronic schizophrenia were in a stable phase.

The skin cell culture procedures were performed as described in detail by Hagenfeldt and coworkers.¹² Briefly, fibroblasts were cultured from skin biopsy specimens by means of Eagle minimum essential medium, containing fetal calf serum (10%, vol/vol), penicillin V (125 U/mL), streptomycin sulfate (125 µg/mL), and tylosin (an antimycotic agent) (6 mg/mL), in plastic tissue culture flasks in a humidified atmosphere of 5% carbon dioxide in air at 37°C. Stocks of the primary cultures of individual fibroblast cells were frozen in 10% dimethylsulfoxide in Dulbecco minimal essential medium and 10% fetal calf serum and stored in liquid nitrogen for later use. Fibroblasts were used experimentally between the 8th and 16th passages. The cells were frequently checked for Mycoplasma and bacterial contamination.

Transport experiments of tyrosine in fibroblasts from the 36 patients and 10 normal subjects were performed by means of the cluster-tray method for rapid measurement of tyrosine flux in adherent intact fibroblast cells as described earlier.¹² Fibroblast cells were seeded in a multiwell tray (2-cm diameter; Costar Europe Ltd, Costar, NY) and cultured for 5 days to confluence. The cells were washed and preincubated for 1 hour at 37°C with glucose (10 g/L) to deplete endogenous amino acid pools. The initial rate of tyrosine uptake in the cells was then measured during an incubation for 60 seconds at 37°C in phosphate-buffered saline, containing tyrosine labeled with carbon 14 and varying concentrations of unlabeled tyrosine. Tyrosine uptake was measured at 12 concentrations. The experiment was repeated after 3 to 6 weeks with the use of new fibroblasts from the same incubation to diminish the risk of artifacts caused by the condition of the cells.

The concentrations varied between 0.1 and 27 mg/dL (0.005 and 1.5 mmol/L) of tyrosine. The tyrosine uptake values obtained were used to calculate the kinetic parameters K_m (micromoles per liter) and V_max (micromoles per minute per milligram of protein) and the nonsaturable diffusion constant K_d (microliters per minute per milligram of protein) by means of the Lineweaver-Burke plot. The equation describing the Lineweaver-Burke plot is 1/V_o = (K_m/V_max[S]) + (1/V_max), where V_o is the initial transport velocity and [S] is the substrate concentration.

The parameter V_max reflects the maximal transport velocity at a saturating tyrosine concentration. The parameter K_m reflects the affinity of the binding sites for tyrosine and is numerically equal to the tyrosine concentration at which the transport velocity is equal to half of the V_max.

Results of finger-tapping tests,¹⁶ muscle biopsy investigations, and macroelectromyographic recordings⁴ previously performed on the present patients were used in the comparative analyses. The techniques used to assess the morphologic characteristics of the muscle biopsy specimens have been described in detail.⁴ The most frequent abnormality was atrophic muscle fibers, seen in 7 patients and 1 control subject (P = .05, Fisher exact test). Eight patients and none of the control subjects exhibited pathologic macroelectromyograms with increased amplitude and area of the motor unit (P = .03, Fisher exact test) but normal fiber density. (Increased fiber density is a sign of distal nerve degeneration.)

Sociodemographic and clinical data were summarized by means of standard descriptive statistics (means, SDs, medians, ranges, and frequencies). Provided the assumptions behind parametric methods (normal distributions, etc) were fulfilled, these methods were preferred. Thus, differences between patients and control subjects in tyrosine transport kinetic variables (K_d, V_max, and K_m) were analyzed with analysis of variance for repeated measures (group × experiment). Because of skewed distributions, relationships between clinical characteristics such as duration of illness, ratings of symptoms, and medication, and transport kinetic parameters were expressed as nonparametric Kendall rank order correlations. A Bonferroni correction was applied to adjust for the increased risk of type I errors at multiple comparisons. Relationships between discrete variables (eg, sex) with 2 or 3 categories and the transport kinetic variables were analyzed with Student t test and 1-way analysis of variance, respectively. Correlations between K_m and V_max values were calculated with Pearson product-moment correlation coefficient. An α level of 5% (2-tailed) was applied.

Significantly lower V_max (F_1,44 = 14.40; P = .001; effect size = 1.36) and K_m (F_1,44 = 29.68; P<.001; effect size = 1.00) values were obtained in the cell lines from patients with schizophrenia compared with healthy control subjects (Figure 1; Table 2). There was no significant difference in K_d, the nonsaturable diffusion constant, between the 2 groups (F_1,44 = 1.63; P = .21). No significant difference in tyrosine kinetics was found between first-episode patients and those with a chronic course. The V_max/K_m values of the 3 neuroleptic-naive, first-episode patients were 9.20:28.65, 10.40:16.05, and 11.3:19.80. All but 1 of these values were within the lower quartile of those of the controls (Figure 1). No significant correlation between V_max and K_m was found (r₄₄ = 0.20; P = .19).

A significant correlation was found between K_m and the item "accommodation" (τ₃₄ = 0.35; P = .02, corrected for 3 comparisons). Thus, better housing functions corresponded to a higher K_m. No significant correlations were found between the kinetic parameters of tyrosine transport (K_d, K_m, and V_max) and first episode or chronic state, Positive and Negative Syndrome Scale, Global Scale of Adaptive Functioning, occupational functioning, sex, prodromal period, and duration and dosage of neuroleptic medication. No relation to a family history of psychosis was found (data not shown). Thus, patients (n = 10) with values of V_max and K_m within the lower quartile of the value range of control subjects did not significantly differ from patients with higher values in any of the above-mentioned clinical parameters except in accommodation (see above).

No significant relationships were found between the tyrosine transport parameters (K_d, K_m, and V_max; Table 2) and the muscle biopsy, macroelectromyographic,⁴ and finger-tapping¹⁶ results (data not shown).

A reduced tyrosine transport capacity (V_max) and increased tyrosine affinity to the transport sites (lower K_m) compared with healthy controls was found in patients with schizophrenia. This is in accordance with previous findings in smaller patient series by 2 independent research groups.^12,24 The differences between patients and controls were highly significant with large effect sizes. First-episode patients as well as patients with a chronic course exhibited the changes (Figure 1).

The tyrosine transport shows Michaelis-Menten kinetics, and factors affecting its properties may be studied in vitro by means of the fibroblast technique. A low V_max means that the transport system has a lower capacity for tyrosine uptake, and a small K_m implies a high affinity between the transport molecules and tyrosine. Changes in V_max may be secondary to changes in K_m or vice versa. In the present study, however, V_max and K_m were not significantly correlated, indicating a more complex background. The in vitro data are not easily transferred to in vivo conditions, but they are in line with previous findings of disturbed tyrosine transport across the blood-brain barrier.²⁵

Cultivated skin fibroblasts provide an advantageous system for investigating tyrosine transport across the plasma membrane under controlled conditions. Furthermore, fibroblasts are useful for studies of systemic biochemical abnormalities with predominant consequences for the brain, as the fibroblasts resemble neurons in a number of aspects, such as cell-to-cell adhesion molecules, actions of growth factors, and membrane phospholipid-derived second messengers.²⁶

The present experiments were performed in cells that had grown for several generations in vitro. It is thus unlikely that the results could be affected by factors such as the medication being taken or the clinical condition of the patient at the time of the biopsy. Rather, the decreased V_max and K_m for tyrosine transport observed in the cells from patients might reflect a genetic trait determining membrane function transmitted through several cell generations. Although no relation to a family history of psychosis was found, this does not preclude a genetic contribution to the tyrosine kinetic parameters. The family history–sporadic distinction presumes the existence of predominantly environmental and predominantly genetic subgroups of patients with schizophrenia, which perhaps oversimplifies the issue.

One possible interpretation of the relative lack of relationships between tyrosine kinetics and clinical characteristics is that the former is genetically determined and the latter is partly environmentally determined. Another interpretation, not contradictory to the first, is that the disturbed tyrosine transport may reflect a generalized disturbance and, thus, the finding is not restricted to a subgroup of patients with certain characteristics. The correlation between K_m and the Strauss-Carpenter item accommodation may reflect the fact that both low K_m and impaired functioning are group characteristics of patients with schizophrenia. As no other relationships were found between the tyrosine kinetics and clinical characteristics, the relevance of this finding may be questioned. The findings of altered tyrosine transport kinetics and the lack of relations to the clinical characteristics may thus suggest that schizophrenia, although heterogeneous in its phenotype, may constitute an entity with genetically determined common underlying factors.

There are several systems for amino acid transport. The systems known to be involved in tyrosine transport are the L-system, the T-system, and the sodium-dependent A-system.¹² The L- and Atransport systems, but not the T-system, are present in human cultured fibroblasts.^27,28 Factors affecting the tyrosine transportation capacity of the L- and A-systems are (1) the intracellular tyrosine content,²⁷ (2) the extracellular concentration of the competing amino acids,¹² and (3) the presence of sodium.²⁷ In a series of in vitro experiments, the effects of these factors on tyrosine transport were investigated. No differences between the fibroblasts from patients and control subjects were found. Thus, the altered tyrosine transport kinetics V_max and K_m in patients with schizophrenia could not be explained in terms of known amino acid transport mechanisms.¹²

Several studies have demonstrated decreased levels of membrane phospholipids in erythrocytes,^29-31 platelets,^32,33 and fibroblasts.³⁴ In recent years, the phosphorus 31–magnetic resonance spectroscopy technique has enabled the study of phosphorus-containing compounds in the living brain. Independent investigators have reported increased levels of phosphodiesters in the frontal and temporal cortexes in patients with schizophrenia, both those who are drug naive^8,35 and those who are receiving medication.³⁶ These data were interpreted as indicative of increased membrane phospholipid breakdown thus taking place in the central nervous system as well as in peripheral organs. Since the state of all membranes, including neuronal, is dependent on their composition, even small changes in phospholipids can lead to a broad range of membrane dysfunctions in receptor binding,³⁷ electrophysiologic properties,³⁸ and probably also tyrosine transport.

A limitation of the present study was the small number of patients in the intragroup comparisons. The risk of false-negative results increases with small sample size, and multiple comparisons may lead to false-positive results. Thus, the relationship between K_m and accommodation must be interpreted with caution.

In conclusion, the main finding of the present study was the reduced V_max and K_m in patients with schizophrenia compared with controls, possibly representing a genetic trait. The increased tyrosine affinity, as indicated by the reduced K_m, and reduced tyrosine transport, as indicated by the low V_max, may be the result of changes in membrane fatty acid composition, but this needs to be studied further.

Accepted for publication March 23, 2001.

This study was supported by grants 732 and 8318 from the Swedish Medical Research Council, Stockholm.

Corresponding author and reprints: Lena Flyckt, MD, FoUU, Department of Psychiatry, Danderyds Hospital, S-18288 Danderyd, Sweden (e-mail: lena.flyckt@kids.ki.se).