Schizophrenia Is Associated With an Aberrant Immune Response to Epstein–Barr Virus

Faith Dickerson; Lorraine Jones-Brando; Glen Ford; Giulio Genovese; Cassie Stallings; Andrea Origoni; Colm O'Dushlaine; Emily Katsafanas; Kevin Sweeney; Sunil Khushalani; Robert Yolken


Schizophr Bull. 2019;45(5):1112-1119. 

In This Article


We found that individuals with schizophrenia have increased levels of IgG antibodies to EBV virions as compared to a control group. The differences were independent of demographic variables that are known to affect EBV exposure such as age, sex, race, and socioeconomic status as measured by maternal education. We also measured antibodies to specific viral proteins. These studies indicated that the individuals with schizophrenia had an aberrant immune response to EBV in that there were elevated levels of antibodies to EBV VCA but not to EBNA-1. It is likely that most of the participants in our study have undergone primary EBV infection at some time in the past as evidenced by the presence of low levels of antibody to EBV early antigen (EA). Generally, individuals with past infections to EBV have increased levels of IgG class antibodies to both VCA and EBNA-1 antigens.[13] The levels of EBV antibodies in the study populations are relatively stable over extended periods of time since the levels did not change significantly in the individuals for whom follow-up samples were available. This finding suggests that the aberrant immune response is unlikely to be related to timing of sampling but is relatively stable within an individual over time; additional longitudinal studies should be done to assess the stability of the markers over time. Antibodies to other herpesviruses including HSV-1, HSV-2, CMV, VZV, and HHV-6 were not significantly increased in the same population of individuals with schizophrenia. This finding indicates that the increased levels of antibodies to EBV virions found in the individuals are not reflective of a heightened immune response to human herpesvirus but is specific to EBV within this group of infectious agents.

As noted above, healthy individuals generally develop and maintain approximately equal levels of antibodies to both EBV VCA and EBNA proteins following the resolution of acute infection. The finding of increased levels of VCA but not EBNA-1 antibodies in individuals with schizophrenia is thus reflective of an aberrant immune response to EBV infection. Aberrant responses to EBV have been described in a range of EBV-associated disorders including autoimmune diseases such as multiple sclerosis[29] and systemic lupus as well as neoplastic conditions such as nasopharyngeal carcinoma.[30]

Increased levels of antibodies to EBV virion and VCA proteins were not associated with cognitive functioning or symptom scores but were associated with an increased prevalence of the deficit syndrome. EBV virion antibodies were also significantly increased in individuals who were receiving the mood-stabilizing medication valproate. This finding, which should be verified in larger samples, is of interest in light of the immunomodulatory effects of this medication which are likely related to its effects on histone deacetylation.[31] Valproate has also been used in the treatment of EBV-associated tumors based on its ability to modulate EBV gene expression.[31] Increased levels of EBV antibodies were also marginally associated with an increased prevalence of cigarette smoking.

The pattern of reactivity to EBV proteins in the study cohort was further analyzed by performing western blot analyses directed at IgG antibodies directed at 12 EBV proteins. These analyses confirmed significant reactivity to VCA proteins and other EBV proteins as well as the lack of reactivity to EBNA-1. Of interest is the significant increased reactivity to p27 in individuals with schizophrenia as compared to controls. The material provided by the manufacturer of the quantitative immunoblot system that we employed states that the viral strain used to produce the blot strips is P3HR1. Although we have not yet confirmed the identity of this 27kDa protein present on the blot strips, it is likely that this p27 represents the truncated form of the EBNA leader protein that Garibal et al[32] reported to be produced by P3HR1 EBV strain but not the predominant, wild-type strains, suggesting that individuals with schizophrenia might have been differentially exposed to an EBV variant with a mutation similar to this strain.

Primary EBV infection generally occurs in adolescence following viral transmission facilitated by oral contact. Infection in adolescents is often manifested by a syndrome of fever, pharyngitis, lymphadenopathy, and splenomegaly generally referred to as infectious mononucleosis. Most cases of infectious mononucleosis are self-limited and are followed by increases in antibodies to VCA and EBNA proteins. The timing of primary EBV infection is of interest in terms of the first manifestations of schizophrenia which also often occur in adolescence. It is of note in this regard that Khandaker et al found that previous exposure to EBV as measured by VCA antibodies was associated with subsequent psychotic experiences in adolescence.[33]

Our studies indicate that the source of EBV antigen in the test immunoassay is important in evaluating the association between EBV exposure and a schizophrenia diagnosis. It is thus of interest that DeWitte et al did not find an association between levels of antibodies to EBV and a schizophrenia diagnosis.[34] However, as they measured EBNA antibodies (DeWitte L, personal communication) as markers of EBV exposure, their findings are consistent with ours. The consistency of our results with other studies that have measured EBV antibodies in individuals with schizophrenia[35–37] is difficult to evaluate without additional information relating to the immune responses to defined EBV proteins.

The reasons for altered levels of antibodies to EBV proteins and specific EBV proteins are not known with certainty. Possible mechanisms for this association include ones relating to the virus and to the host response. In terms of the virus, a differential response to infection might be related to variation in the timing of primary EBV exposure, the genomic composition and pathogenicity of the infecting EBV, and possible re-exposure to differing strains of EBV.[38] Virological analysis of samples obtained from individuals in prospective studies will be required to address these possibilities. An altered immune response to EBV infection could also be based on host factors such as genetics or other environmental factors. In terms of genetic factors, we found an additive effect of increased EBV antibodies and increased genetic risk with a combined odds ratio of 8.86 for individuals with values ≥75th percentile of both measures. However, there was no statistically significant association interaction between the levels of EBV virion antibodies and the polygenic risk scores suggesting that the risks associated with genetic susceptibility and increased levels of EBV virion antibodies are independent. However larger sample sizes might be employed to detect low levels of interaction. Furthermore, additional genetic studies are required to define the genetic contribution to the aberrant response to EBV infection found in individuals with schizophrenia. The possibility use of combinations of immune and genetic markers for the diagnosis and management of schizophrenia is an important area for future study. Antigen-specific antibodies to additional herpesviruses and other viruses might also be examined with the goal of incorporating them into this approach.

In terms of other environmental factors, it is of note that we found an association between levels of antibodies to EBV virion proteins and cigarette smoking in individuals with schizophrenia. An interaction between EBV exposure and cigarette smoking has also been noted in other EBV-associated disorders such as multiple sclerosis, possibly based on the many immunomodulatory effects of cigarette smoking.[39] As for the above factors, the timing of the interaction between cigarette smoking and EBV exposure in terms of risk for schizophrenia should be addressed in prospective studies. The possible effects of lymphotropic viruses such as EBV and other environmental factors which can affect B-cell activity on the immunopathology of schizophrenia should be the subject of future investigations.

The neurobiological mechanisms by which increased levels of EBV virion antibodies might be associated with schizophrenia are not entirely known. One potential scenario is that individuals with aberrant EBV immune responses underwent previous replication of EBV within the central nervous system. This possibility is consistent with reports of psychosis in individuals undergoing EBV encephalitis and the finding of increased levels of EBV VCA antibodies in the CSF of some individuals with psychiatric disorders.[40–45] It is also possible that psychiatric symptoms are related to neuro-inflammatory effects on the brain including alterations in neurotransmitter receptor interactions, as has been found in a range of autoimmune disorders which affect the brain.[46]

In conclusion, this study indicates that many individuals with schizophrenia have an aberrant immune response to EBV. There are a number of therapeutic interventions available for the modulation of EBV infection including anti-viral medications and pharmacological compounds which can modulate the immune response.[47] An increased understanding of the role of EBV infection might thus lead to novel methods for the prevention and treatment of schizophrenia.