Antibodies to Heat Shock Proteins in Schizophrenic Patients: Implications for the Mechanism of the Disease

Heat shock proteins play an important role not only in neuronal protection but also in pathophysiological mechanisms of several degenerative and ischemic disorders of the central nervous system (CNS). Immunoreactivity against cellular chaperones has been described in CNS disorders such as multiple sclerosis, especially against 60 kilodaltons (kD) heat shock proteins (HSP60) and 70 kD heat shock proteins (HSP70), which are recognized as major antigens by the immune system after a viral or bacterial infection (1, 2). Since antibodies against HSP60 were detected in a subgroup of 20%–40% of medicated patients with schizophrenia (3, 4), we measured antibody titers against both HSP60 and HSP70 in patients with schizophrenia to evaluate the possible effects of treatment and course of disease.

METHOD

We investigated 30 patients with schizophrenia who had been free of neuroleptic medication for at least 4 weeks while they were in the acute state of illness. Fifteen of the patients were men, 15 were women, and their mean age was 32.9 years (SD=10). We investigated these patients a second time before they were dismissed from the hospital, while they were receiving antipsychotic medication (the mean duration of treatment was 2.4 months). Thirteen of the patients were suffering from their first episode of schizophrenia and had never been medicated with neuroleptics. We selected 31 healthy individuals as a comparison group. The comparison group was similar in age and sex to the patient group: 17 of the comparison subjects were men, 14 were women, and their mean age was 28.6 (SD=6.5).

Diagnoses were made by using the Structured Clinical Interview for DSM-III-R. Psychopathology was monitored by using the Positive and Negative Syndrome Scale score before treatment (first blood sampling) and at the second point of investigation, 2 days before discharge from the hospital. Individuals suffering from febrile, infectious, or autoimmune disorders as well as those with alcohol or substance abuse were excluded from the study. After complete description of the study to the subjects, written informed consent was obtained.

We measured antibody titers to human 60 kD and 70 kD heat shock protein by direct enzyme-linked immunosorbent assay, using recombinant human HSP60 and HSP70 (StressGen Biotechnologies Corp., Canada) as antigens and detecting human antibody classes IgA, IgG, and IgM by an enzyme-conjugated rabbit antihuman IgA, IgG, and IgM antibody (Dako Diagnostics, Denmark). Samples were measured in duplicate. Unspecific binding was tested for each sample and subtracted from the mean of the duplicates to exclude unspecific reactions. Highly positive sera were used to establish a standard curve. The threshold value for high anti-HSP60 and anti-HSP70 titers was defined as two standard deviations from the mean values of the comparison group. Inter- and intraassay variances were below 10%.

RESULTS

Seven (23%) of the 30 drug-free patients showed high antibody titers against HSP70. Six of the seven were experiencing their first episode of schizophrenia. Additionally, elevated anti-HSP60 antibody titers were found in another three (10%) of the unmedicated patients. No patient had both a high anti-HSP60 and a high anti-HSP70 titer.

On reinvestigation during antipsychotic treatment, the proportion of patients with anti-HSP70 antibodies decreased to 16% (five of 30) and the proportion of those with anti-HSP60 antibodies increased to 20% (six of 30). The six patients whose anti-HSP60 antibodies increased also showed an impairment in blood-brain barrier. As far as psychopathology was concerned, there was no marked relationship between antibody titers and Positive and Negative Syndrome Scale scores. There was only a tendency for unmedicated patients with elevated anti-HSP60 titers to have lower Positive and Negative Syndrome Scale positive scores (mean score=15.3, SD=0.6) than patients with normal titers (mean=21.4, SD=6.3) (statistical calculation was not recommended because of the small number of subjects).

In the healthy comparison group, only 3% (one of 31) had either anti-HSP60 or anti-HSP70 antibodies. Two-tailed Fisher’s exact test revealed that significantly more unmedicated patients had elevated antibody titers to HSP70 (p=0.01) and that more medicated patients had elevated antibody titers to HSP60 than did healthy individuals (p=0.05). Neither the decrease in the patients showing high anti-HSP70 antibody titers nor the increase in the patients showing high anti-HSP60 antibody titers was statistically significant.

Our results show parallels between schizophrenia and multiple sclerosis, where immune reactions to HSP60 and/or HSP70 have been shown repeatedly (5). The finding that six of seven never-medicated patients who were experiencing their first episode of schizophrenia had high anti-HSP70 titers suggests the presence of an immune process independent of antipsychotic medication. A strong overexpression of HSP70 in corticolimbic neurons has been described after administration of an N-methyl-d-aspartic acid (NMDA) receptor antagonist (6), and an endogenous NMDA antagonist has been discussed in pathogenesis of schizophrenia (7). Therefore, immunoreactivity against HSP70 could be crucially involved in a glutamatergic mechanism in schizophrenia. In the described model of NMDA antagonism, HSP70 overexpression was not detectable after treatment with haloperidol. This could indicate another parallel to the slight decrease in anti-HSP70 antibody titers in our patients over the course of treatment. A longer period of treatment before reinvestigation might lead to a more marked decrease.

DISCUSSION

Although immune reactivity to heat shock proteins is not specifically related to schizophrenia, cross-reactivity between these major antigens and brain-specific proteins may be an important mechanism in a subgroup of patients with schizophrenia. Such cross-reactivity has been shown for HSP60 and the activity-dependent neuro­­trophic factor, a neuroprotective peptide, by Brenneman and Gozes (8). The activity-dependent neurotrophic factor has been identified as protecting neuronal cells from toxicity of several agents like tetrodotoxin or NMDA. The strong sequence homology of the activity-dependent neurotrophic factor and HSP60 leads to neutralization of the activity-dependent neurotrophic factor by anti-HSP60 antibodies (8). An inhibition of protective factors by antibodies against heat shock protein might be another important mechanism in schizophrenia.

Further investigation is needed to elucidate the actual role of heat shock proteins in schizophrenia. Furthermore, the question of whether elevated antibody titers might indicate just an unspecific sign for an increased humoral immune responsiveness must be critically addressed.

The next step to more definitive investigation will be the measurement of antibody titers against heat shock proteins in the CSF of patients experiencing their first episode of schizophrenia compared with patients suffering from other psychiatric diseases or multiple sclerosis.

Presented in part at the 20th congress of the Arbeitsgemeinschaft Neuropsychopharmakologie und Pharmakopsychiatrie (AGNP), Nuremberg, Germany, Oct. 8–11, 1997. Received June 12, 1998; revision received Dec. 1, 1998; accepted Dec. 17, 1998. From the Psychiatric Hospital and the Institute for Immunology, University of Munich. Address reprint requests to Dr. Schwarz, Department of Neurochemistry, Psychiatric Hospital, Ludwig-Maximilian-University Munich, Nussbaumstr. 7, D-80336 Munich, Germany; [email protected] (e-mail). Supported by grant I 69/356-357 from the Volkswagen Foundation, Hanover, Germany.