• affective disorders;
  • antibodies;
  • Borna disease virus;
  • indirect immunofluorescence;
  • schizophrenia


  1. Top of page
  2. Abstract

Aim:  Data suggesting a pathogenetic role for Borna disease virus (BDV) in neuropsychiatric diseases are still inconclusive and it is unknown whether humans become persistently infected or clear the virus infection. The aim of the present study was therefore to investigate long-term BDV-specific antibody responses in psychiatric patients in order to gain new insights into human BDV infection and its pathogenicity.

Methods:  BDV-specific antibody titers and associations with clinical conditions were studied retrospectively in 94 seropositive patients with schizophrenia (n = 46), affective disorders (n = 19) and other psychiatric disorders (n = 29) who had been repeatedly tested for the presence of BDV-specific antibodies on indirect immunofluorescence assay between 1985 and 2006. Long-term titer dynamics were studied in 46 patients followed up for a period of >36 months.

Results:  A total of 25 of these 46 patients (54.3%) had persistent seropositivity, whereas seroreversion from positive to negative was observed in 21 (45.7%). Patients in the early course of schizophrenia had lower antibody titers compared to patients in the advanced course (P = 0.017), while a higher proportion of patients in the early course had titer increases (P < 0.05). There were no significant differences in antibody titers between patient subgroups with clinically stable and acute psychiatric disorders.

Conclusion:  Persistent seropositivity in a subgroup of psychiatric patients in the long-term analysis suggests chronic BDV infection in humans.

BORNA DISEASE VIRUS (BDV) is a single-stranded, non-cytolytic RNA virus known to cause chronic aseptic meningoencephalomyelitis with behavioral disturbances in a wide range of animal species.1,2

Serological studies reporting increased seropositivity rates in psychiatric patients and their relatives,3–5 and experimental studies showing that infection with the virus produces structural and functional disturbances in the brain resembling those in affective or schizophrenia psychoses6,7 sparked the hypothesis that BDV may play a role in the pathogenesis of human psychiatric disorders.8

The detection of BDV-RNA in peripheral blood of psychiatric patients and their relatives using ultrasensitive reverse transcription–polymerase chain reaction9–11 further supported this hypothesis, although sequence similarities between human BDV strains and laboratory strains highlighted the possibility of false-positive results due to contaminations,12,13 and the importance of rigorous safety conditions. A triple enzyme-linked immunosorbent assay (ELISA) developed by Bode et al. to detect BDV-specific circulating immune complexes (CIC), free antigen and antibodies in plasma14,15 provided additional important insights and is considered the currently most promising technique in BDV diagnostics.10

Indirect immunofluorescence assay (IFA) is a widely used method for the detection of BDV-specific antibodies.10,15–17 Detecting only fixed BDV antigen accumulating in the nucleus of infected cells, however, it may be less sensitive and specific than newer techniques recognizing native antigen.10 Concerns about the specificity of the test were supported by the finding that reactive antibodies in psychiatric patients were of low avidity,18 a typical feature of cross-reacting antibodies. The establishment of a peptide array-based screening test by Billich et al.19 allowed the further characterization of these antibodies. It demonstrated the presence of small amounts of BDV-reactive antibodies in human sera that specifically recognized various epitopes of BDV proteins. Most importantly, the purified epitope-specific antibodies were shown to bind to BDV antigen with high avidity when assayed on conventional IFA, supporting the view that the presence of BDV-reactive antibodies in human sera reflects an infection with BDV.19

While BDV is known to establish persistent central nervous system infections in most animal species, which may be clinically inapparent, lead to phasic disorders in behavior, sensitivity and motility or progressive, ultimately fatal meningoencephalomyelitis,1,2 data regarding a possible pathogenicity of human BDV infection are still inconclusive and it is unknown whether humans become persistently infected or clear the virus infection.

Hoping to gain additional insights into human BDV infection and its pathogenicity, we studied long-term dynamics of anti-BDV antibody responses in psychiatric patients and associations of humoral responses with clinical conditions.


  1. Top of page
  2. Abstract

After approval for several BDV studies by the local ethics committee had been obtained, blood sampling for BDV-specific antibody analyses was implemented in routine laboratory examinations of patients admitted to the Department of Psychiatry II of the University of Ulm from 1985 to 2006 (sample collection: K. Bechter, University of Ulm).

Patients found to be seropositive were repeatedly but not systematically examined further for the presence of BDV-specific antibodies during hospital or outpatient treatment. IFA were carried out by a single, very experienced investigator (S. Herzog, University of Giessen). The initial test protocol16 was optimized during the study period. Antibody titers ≥5 were defined as positive. Based on a registry of BDV-seropositive patients, antibody titers were retrospectively extracted from medical records and assigned inpatient or outpatient status. Moreover, psychiatric diagnoses and the time of onset of psychiatric disease were noted. All patients with repeated titer examinations and a follow-up time of ≥4 months were included in the study.

Ninety-four anti-BDV-seropositive psychiatric patients (47 men, 47 women; mean age, 40.70 ± 14.56 years; median, 38 years; range, 19–84 years) were followed up for a mean period of 53.5 ± 47.3 months (median, 39 months; range, 4–154 months). The mean number of serum analyses was 4.2 ± 2.5 (median, 4; range, 2–19). Forty-six patients suffered from schizophrenia spectrum disorders (ICD-10: F2x.x), and 19 patients from affective spectrum disorders (F3x.x) with the diagnoses of major depressive episode (F32; n = 5), recurrent depressive disorder (F33; n = 12) or bipolar affective disorder (F31; n = 2). The patient group with other psychiatric diagnoses (OPD; n = 29) consisted of patients with alcohol-related disorders (n = 14), multiple substance abuse (n = 1), Alzheimer's disease (n = 2), vascular dementia (n = 1), organic delusional or affective disorders (n = 3), anxiety disorders (n = 5), somatoform disorders (n = 2) or post-traumatic stress disorders (n = 1).

Statistical analysis

The non-parametric Kruskal–Wallis test was used to compare medians of metric variables between the three patient groups for schizophrenia, affective disorders and OPD. Comparisons between two patient groups were done using Mann–Whitney U-test. For categorical data, Fisher's exact test was used. For comparison of antibody titers during hospital versus outpatient treatment, the paired Wilcoxon test was used. Spearman's rank correlation was used to test for an association between age and antibody titer values. P < 0.05 was considered statistically significant.


  1. Top of page
  2. Abstract

Anti-BDV-antibody titers and baseline patient characteristics

Baseline characteristics of patient groups according to ICD-10 diagnosis are summarized in Table 1.

Table 1.  Baseline patient characteristics, antibody titers and seroreversions
 (F2x.x) (n = 46) Median (1st/3rd quartile)(F3x.x) (n = 19) Median (1st/3rd quartile)OPD (n = 29) Median (1st/3rd quartile)
  • P are uncorrected for multiple testing. P = 0.014 (F2x.x vs F3x.x); P = 0.046 (F2x.x vs OPD); §P = 0.006 (F2x.x vs OPD).

  • Mean titers were calculated for each subject with negative titers included as titer value 0.

  • OPD, other psychiatric diagnoses.

Age at initial testing (years)32.0 (27.0/46.25),38.0 (33.0/54.0)42.0 (34.0/49.5)
Female: n (%)20 (43.5)12 (63.2)15 (51.7)
Serum titer4.0 (3.0/5.0)4.0 (2.0/5.0)3.0 (2.0/5.0)
Observation period (months)59.5 (24.9/116.5)§37.5 (18.5/58.0)12.0 (5.5/64)§
Mean antibody titers28.35 (9.85/87.08)44.0 (20.0/128.0)20.0 (8.75/84.38)
Seroreversion: n (%)16 (34.8)5 (26.3)10 (34.5)

Schizophrenia patients were younger than patients with affective disorders (P = 0.014) and OPD (P = 0.046) while observation periods were significantly shorter in OPD compared to schizophrenia patients (P = 0.006).

No differences between groups were found concerning sex distribution, numbers of serum titer analyses or antibody titer values.

Undulating antibody titers with increases or decreases of ≥2 titer steps were observed in 61 of 94 patients (65%). Seroreversion defined as seronegativity after previously documented seropositivity occurred in 31 out of 94 patients (33%) after a median period of 10 months. Proportions of patients showing seroreversion did not differ between patients with affective disorders, schizophrenia and OPD (Table 1).

Long-term anti-BDV humoral responses

Long-term analysis of BDV-specific humoral responses was done on the basis of data from 46 patients with the diagnoses of schizophrenia spectrum disorders (n = 27), OPD (n = 10) and mood disorders (n = 9) followed up for more than 36 months (range, 37.5–154 months; median, 83 months). The median number of titer analyses was five (range 4–19).

Twenty-five out of these 46 patients (54.3%) had persistent seropositivity, whereas seroreversion from positive to negative was observed in 21 patients (45.7%).

As shown in Fig. 1, mean positive antibody titers were significantly lower in patients with seroreversion (P < 0.001) compared to persistently seropositive patients.


Figure 1. Borna disease virus (BDV)-specific antibody titers in patients with seroreversion vs persistent seropositivity. §Only positive titers were included in calculation of mean Ab titers of individual subjects. Boxplots, anti-BDV antibody titers in patient subgroups. Upper and lower ends of boxes are at first and third quartile, length of boxes indicates inter-quartile range. The line within the box marks the median. Whiskers extend to the smallest and largest observations within 1.5-fold the inter-quartile range. °Outlier.

Download figure to PowerPoint

Patient groups with persistent seropositivity and seroreversion did not differ with respect to follow-up time and number of titer analyses. Among patients with persistent seropositivity, antibody titers remained stable in 15/25 patients (60%), while undulating titers with increases or decreases of ≥2 titer steps were observed in 10/25 patients (40%). Eleven of 21 patients (52.4%) who had undergone seroreversion, later had re-conversion.

Based on these findings, two subtypes of human BDV-specific antibody responses may be differentiated: (i) persistent seropositivity with comparatively high antibody (Ab) titers; and (ii) low and variable antibody titers with seroreversions and sometimes re-conversions.

Ab titer courses of three patients illustrative of different subtypes of Ab responses are shown in Fig. 2.


Figure 2. Subtypes of Borna disease virus (BDV)-specific Ab titer courses. Titer steps: 1:5 (step 1); 1:10 (2); 1:20 (3); 1:40 (4); 1:80 (5); 1:160 (6); 1:320 (7). (inline image) Patient 1: female, age at initial testing 38 years, ICD-10 diagnosis: F33. Titer course with relatively low titers and seroreversion/reconversion. (inline image) Patient 2: female, 37 years, F10.2. Persistent seropositivity with titer undulations. (inline image) Patient 3: female, 31 years, F33. Persistent seropositivity with relatively high and stable titers.

Download figure to PowerPoint

Anti-BDV-antibody titers versus age in schizophrenia

As shown in Fig. 3, serum antibody titers were significantly lower in schizophrenia patients aged ≤30 years compared to patients aged >30 years at the time of first titer analysis (P = 0.017). A significant positive correlation was found between age at first examination and first antibody titer values in schizophrenia patients (Spearman ρ = 0.27; P = 0.034). In contrast, such a correlation could not be demonstrated in the remaining patients (OPD and mood disorders, P = 0.39). Analyzing antibody titer dynamics in age-related subgroups of schizophrenia patients, titer increases of ≥2 titer steps were observed predominantly in patients aged 25–30 years (9/14, 64.3%) compared to a lower proportion in other age groups (11/32, 34.4%; P < 0.05).


Figure 3. Initial Borna disease virus (BDV)-specific serum antibody titers in schizophrenia patients aged ≤30 vs >30 years. Boxplots indicating initial anti-BDV antibody titers in patient subgroups. Upper and lower ends of boxes are at first and third quartile, length of boxes indicates inter-quartile range. The line within the box marks the median. Whiskers extend to the smallest and largest observations within 1.5-fold the inter-quartile range. °,#Outliers.

Download figure to PowerPoint

Antibody titers were also lower in schizophrenia patients in the early (<5 years) versus advanced (>5 years) course of schizophrenia (median 20, 1st/3rd quartile 15/20 vs median 40, 1st/3rd quartile 20/160; P = 0.024).

Anti-BDV antibody titers and clinical parameters

In 24 patients with affective or schizophrenia spectrum disorders, blood samples were collected first after admission to hospital for treatment of acute disease or exacerbations and again later during outpatient treatment in a clinically stable condition. Antibody titers in these patients were higher during inpatient treatment than during outpatient treatment but differences were not statistically significant (median 40, 1st/3rd quartile 18.75/130 vs median 20, 1st/3rd quartile 5/80; P = 0.17).


  1. Top of page
  2. Abstract

Increased seropositivity rates in psychiatric patients have been a key finding suggestive of a possible role of BDV in human psychiatric diseases.3–5 With the aim of contributing to a better understanding of human BDV infection and its pathogenicity, the present study provides long-term data on BDV-specific humoral responses from a relatively large sample of psychiatric patients. We found that approximately one half of anti-BDV-seropositive psychiatric patients remained seropositive for several years with comparatively high titers, while the other half displayed markedly lower antibody titers with incident seroreversions sometimes followed by re-conversions.

The most likely explanation for the observation of variable and frequently low BDV-specific Ab titers and seroreversions is the binding of antibodies to free BDV antigen in plasma of psychiatric patients resulting in the formation of CIC, demonstrated for the first time by Bode et al. in 2001.14,15 This milestone in BDV diagnostics helped resolve previous long-lasting controversies about low antibody titers and ‘disappearing’ antibodies.

Bode et al. also reported long-term follow-up data from psychiatric patients monitored longitudinally for the presence of anti-BDV antibodies, BDV antigen and CIC on triple ELISA.14,15 Antibody titers in these patients showed considerable undulations corresponding to varying amounts of plasma antigen and CIC. Thus, the formation of CIC offers an appealing explanation for frequently low and undulating Ab titers observed in the present study, although it cannot be excluded that titer changes have been partly caused by variations in IFA methodology.

The predominant finding of the present study, however, is persistent seropositivity observed over several years in a proportion of psychiatric patients. This suggests that in analogy to many animal species, BDV may also establish persistent infection in humans. This observation is in accordance with serologic long-term studies in horses,20 in which continuous seropositivity was reported in a subpopulation with comparatively high anti-BDV antibody titers. More importantly, the present results are in line with long-term observations by Bode and Ludwig demonstrating the persistence of BDV antigen, anti-BDV antibodies and CIC for several years in the blood of patients with chronic psychiatric disorders.15

Although infections of animals with BDV may be clinically inapparent or lead to overt disease, the pathogenicity of human BDV infection remains a critical and unresolved issue. BDV-specific antibody titers in the present study did not differ significantly between different groups of psychiatric patients, which was not indicative of an association between BDV and a specific psychiatric illness.

Nevertheless, subgroup analysis of schizophrenia patients indicated low and increasing antibody titers in the early course of schizophrenia. Notably, humoral antiviral responses in chronic human infections may gradually develop for many months after infection, and a gradual shift from a T-helper 1 (Th1) to Th2 immune response has been observed in animals progressing to chronic BDV infection.21 Thus, low and increasing Ab titers in patients with early schizophrenia may reflect titer dynamics early after BDV infection, compatible with a contribution of BDV to the pathogenesis of schizophrenia. But because we cannot exclude a comparable association between age and antibody titer values in healthy subjects, further investigations are needed to clarify the significance of this finding.

BDV may lead to neuronal dysfunction independent of immune mechanisms22 or as a result of virus-induced T-cell-mediated immunopathology.6 The failure to detect significant differences in Ab titers between patient groups with stable and acute psychiatric disorders appears to be in line with studies demonstrating that antibodies do not contribute to BDV immunopathogenesis.23,24 In contrast, immune complex formation may have been expected to result in lower Ab titers in acute versus stable diseases, considering previously reported correlations between the severity of symptoms and concentrations of antigenemia in affective disorders, with peak antigenemia during acute stages.14,15

The major shortcomings of the present study were its retrospective approach and modifications of the IFA protocol undertaken during the study period, limiting the comparability of titer values. These methodological limitations, however, appear to be of little relevance with regard to the key observation of persistent BDV-specific humoral responses for several years in a subgroup of psychiatric patients.

This finding, suggesting chronic BDV infection in humans, is the most important result of the present study.


  1. Top of page
  2. Abstract

We thank K. Bechter and S. Herzog for their contributions to the study. K. Bechter and S. Herzog did not agree with the co-authorship because they had the following concerns (comment by K. Bechter and S. Herzog): the analysis of anti-BDV antibody titers in sera of individual patients collected in the long term is generally very interesting and important, but indirect IFA had been optimized during this time, therefore it is not possible to compare all antibody titers from each patient. For this reason all sera of one patient should be tested in one assay. Also, the serological result <1:5 does not mean negative. The original data may give some preliminary view. But we have disagreements in several points: in the introduction and discussion there are citations of papers that were disproved, although the paper that cleared up the facts, however, is not cited (Dürrwald et al., Rev. Med. Virol. 2007; 17:181–203; note: this is cited in the final form). And the interpretation of this paper is based on publications of a method (ELISA) that is non-specific (Wolff et al., J. Clin. Virol. 2006; 36: 312–313; Herzog et al., Pferdeheilkunde 2008; 24: 766–774). Therefore we cannot agree with co-authorship. In conclusion, we only agree with being listed in the acknowledgement in case our comment is published at the same time.


  1. Top of page
  2. Abstract
  • 1
    Ikuta K, Ibrahim MS, Kobayashi T, Tomonaga K. Borna disease virus and infection in humans. Front. Biosci. 2002; 7: d470d495.
  • 2
    Ludwig H, Bode L, Gosztonyi G. Borna disease: A persistent virus infection of the central nervous system. Progr. Med. Virol. 1988; 35: 107151.
  • 3
    Fu ZF, Amsterdam JD, Kao M, Shankar V, Koprowski H, Dietzschold B. Detection of Borna-disease virus-reactive antibodies from patients with affective disorders by western immunoblot technique. J. Affect. Disord. 1993; 27: 6168.
  • 4
    Richt JA, Pfeuffer I, Christ M, Frese K, Bechter K, Herzog S. Borna disease virus infection in animals and humans. Emerg. Infect. Dis. 1997; 3: 343352.
  • 5
    Chen CH, Chiu YL, Wei FC et al. High seroprevalence of Borna virus infection in schizophrenic patients, family members and mental health workers in Taiwan. Mol. Psychiatry 1999; 4: 3338.
  • 6
    Stitz L, Bilzer T, Planz O. The immunopathogenesis of Borna disease virus infection. Front. Biosci. 2002; 7: 541555.
  • 7
    Hans A, Bajramovic JJ, Syan S et al. Persistent, noncytolytic infection of neurons by Borna disease virus interferes with ERK 1/2 signalling and abrogates BDNF-induced synaptogenesis. FASEB 2004; 18: 863865.
  • 8
    Planz O, Bechter K, Schwemmle M. Human Borna disease virus infection. In : CarboneK (ed.). Borna Disease Virus: Role in Neurobehavioural Disease. ASM Press, Washington, 2002; 179225.
  • 9
    Bode L, Zimmermann W, Ferszt R, Steinbach F, Ludwig H. Borna disease virus genome transcribed and expressed in psychiatric patients. Nat. Med. 1995; 1: 232236.
  • 10
    Thakur R, Sarma S, Sharma B. Role of Borna disease virus in neuropsychiatric illnesses: Are we inching closer? Indian J. Med. Microbiol. 2009; 27: 191201.
  • 11
    Nunes SO, Itano EN, Reiche EM et al. RNA from Borna disease virus in patients with schizophrenia, schizoaffective patients, and in their biological relatives. J. Clin. Lab. Anal. 2008; 22: 314320.
  • 12
    Schwemmle M. Borna disease virus infection in psychiatric patients: Are we on the right track? Lancet Infect. Dis. 2001; 1: 4652.
  • 13
    Dürrwald R, Kolodziejek J, Herzog S, Nowotny N. Meta-analysis of putative human bornavirus sequences fails to provide evidence implicating Borna disease virus in mental illness. Rev. Med. Virol. 2007; 17: 181203.
  • 14
    Bode L, Reckwald P, Severus E et al. Borna disease virus specific circulating immune complexes, antigenemia and free antibodies: The key marker triplet determining infection and prevailing in severe mood disorders. Mol. Psychiatry 2001; 6: 481491.
  • 15
    Bode L, Ludwig H. Borna disease virus infection, a human mental health risk. Clin. Microbiol. Rev. 2003; 16: 534545.
  • 16
    Rott R, Herzog S, Fleischer B et al. Detection of serum antibodies to Borna disease virus in patients with psychiatric disorders. Science 1985; 228 (4700): 755756.
  • 17
    Grabner A, Herzog S, Lange-Herbst H, Frese K. Die Intra-vitam-Diagnose der Bornaschen Krankheit bei Equiden. Pferdeheilkunde 2002; 18: 579586 (in German).
  • 18
    Allmang U, Hofer M, Herzog S, Bechter K, Staeheli P. Low avidity of human serum antibodies question their diagnostic value. Mol. Psychiatry 2001; 6: 329333.
  • 19
    Billich C, Sauder C, Frank R et al. High-avidity human serum antibodies recognizing linear epitopes of Borna disease virus proteins. Biol. Psychiatry 2002; 15: 979987.
  • 20
    Inoue Y, Yamaguchi K, Sawada T, Rivero JC, Hori Y. Demonstration of continuously seropositive population against Borna disease virus in Misaki feral horses, a Japanese strain: A four-year-follow-up study. J. Vet. Med. Sci. 2002; 445448.
  • 21
    Hatalski CG, Hickey WF, Lipkin I. Evolution of the immune response in the central nervous system following infection with Borna disease virus. J. Neuroimmunol. 1998; 137142.
  • 22
    Watanabe M, Lee BJ, Yamashita M et al. Borna disease virus induces acute fatal neurological disorders in neonatal gerbils without virus- and immune-mediated cell destructions. Virology 2003; 310: 245253.
  • 23
    Richt JA, Stitz L, Wekerle H, Rott R. Borna disease, a progressive meningoencephalomyelitis as a model for CD4+ T-cell mediated immunopathology in the brain. J. Exp. Med. 1989; 170: 10451050.
  • 24
    Briese T, Hornig M, Lipkin I. Borna immunopathogenesis in rodents: Models for human neurological diseases. J. Neurovirol. 1999; 5: 604612.