Human Infections with Borna Disease Virus 1 (BoDV-1) Primarily Lead to Severe Encephalitis: Further Evidence from the Seroepidemiological BoSOT Study in an Endemic Region in Southern Germany

Abstract

More than 40 human cases of severe encephalitis caused by Borna disease virus 1 (BoDV-1) have been reported to German health authorities. In an endemic region in southern Germany, we conducted the seroepidemiological BoSOT study ("BoDV-1 after solid-organ transplantation") to assess whether there are undetected oligo- or asymptomatic courses of infection. A total of 216 healthy blood donors and 280 outpatients after solid organ transplantation were screened by a recombinant BoDV-1 ELISA followed by an indirect immunofluorescence assay (iIFA) as confirmatory test. For comparison, 288 serum and 258 cerebrospinal fluid (CSF) samples with a request for tick-borne encephalitis (TBE) diagnostics were analyzed for BoDV-1 infections. ELISA screening reactivity rates ranged from 3.5% to 18.6% depending on the cohort and the used ELISA antigen, but only one sample of a patient from the cohort with requested TBE diagnostics was confirmed to be positive for anti-BoDV-1-IgG by iIFA. In addition, the corresponding CSF sample of this patient with a three-week history of severe neurological disease tested positive for BoDV-1 RNA. Due to the iIFA results, all other results were interpreted as false-reactive in the ELISA screening. By linear serological epitope mapping, cross-reactions with human and bacterial proteins were identified as possible underlying mechanism for the false-reactive ELISA screening results. In conclusion, no oligo- or asymptomatic infections were detected in the studied cohorts. Serological tests based on a single recombinant BoDV-1 antigen should be interpreted with caution, and an iIFA should always be performed in addition.

Keywords: Borna disease virus 1 (BoDV-1); ELISA; diagnostics; encephalitis; endogenous Borna-like elements; epidemiology; indirect immunofluorescence assay (iIFA); linear epitope mapping; molecular mimicry; solid organ transplantation.

Figure 1

BoDV-1 screening ELISA. Three cohorts of healthy blood donors (n = 216), outpatients after solid organ transplantation (n = 280), and patients with requested TBE diagnostics (n = 288) were screened for anti-BoDV-1-IgG antibodies by an ELISA system using recombinant viral nucleocapsid (N), X, or phosphoprotein (P) as antigens. (A) shows the signal-to-cut-off ratios (S/CO) for individual samples and percentages of reactive samples for each cohort and antigen. Fractions of samples with IgG antibodies directed against none (negative), one (single positivity), two (double positivity), or three antigens (triple positivity) are depicted in (B). (C) shows the absolute counts of negative and reactive samples for each group and antigen. The variables “cohorts” and “test outcome” were statistically dependent for anti-BoDV-1-N-IgG (p = 0.0012) and anti-BoDV-1-P-IgG (p < 0.0001) based on chi-square tests. As post hoc test, the p-values of residuals were analyzed using Bonferroni correction. Residuals were significantly elevated for outpatients after solid organ transplantation (anti-BoDV-1-N-IgG, anti-BoDV-1-P-IgG) and significantly diminished for patients with the requested TBE diagnostics (anti-BoDV-1-P-IgG), as indicated by an asterisk.

Figure 2

BoDV-1 immunoblots of selected samples with reactive ELISA results. Immunoblots were performed using the same recombinant viral proteins as used for ELISA. Under (A), confirmations of antigen-directed IgG antibodies by immunoblots are shown for Np40 protein (His-tagged) on the left, for Xp10 protein (tagged with a 26kDa GST-tag) in the middle, and for Pp23 (His-tagged) on the right. The values given below represent ELISA screening results of the corresponding sample. The probability of an immunoblot band increased with higher S/CO values of the screening ELISA for all used antigens (B). Shown are the best-fit lines for logistic regression (0 = no immunoblot band, 1 = immunoblot band) with 95% confidence intervals. Vertical dotted lines represent the ELISA cut-off.

Figure 3

Heatmaps of serological epitope mapping for N (top row), X (middle row), and P proteins (bottom row). The 15-mer peptides are represented in the y-direction from N- (top) to C-terminus (bottom). Individual samples of different cohorts are represented in the x-direction. Values represent the base 2 logarithm-transformed intensity values after background subtraction. An arbitrary cut-off of 13.5 was set to minimize unspecific staining of negative controls samples. Values below the cut-off were set to 0.

Figure 4

Principal component analysis (PCA) of serological epitope mapping. All 15-mer peptides of N, X, and P proteins were included as individual variables. A PCA test was performed for each membrane separately. Dots represent individual samples of the cohorts of negative controls (green), BoDV-1 patients (red), transplant patients with positive ELISA (blue), healthy blood donors with positive ELISA (purple), and healthy blood donors with negative ELISA (turquoise). Patients with BoDV-1 infection were best separated by principal component 1 (PC1) and PC4 (N membrane), by PC1 and PC3 (X membrane) and by PC1 and PC6 (P membrane). Percentages in parentheses provide the contribution to total variance for each PC. The most important loadings of PC in order to separate patients with BoDV-1 infection are given below.