Date Published: August 1, 2019
Publisher: Public Library of Science
Author(s): Dennis Rubbenstroth, Kore Schlottau, Martin Schwemmle, Jürgen Rissland, Martin Beer, Andrew Mehle.
BoDV-1 (originally abbreviated BDV) was initially identified in a rather small area of Western Europe where domestic mammals such as horses or sheep were found to die of an encephalitis called “Borna disease.” BoDV-1 exhibits a remarkable neurotropism in most permissive species. Experimentally, it can infect a broad range of mammalian and even some avian species, in which it usually causes severe neurologic disorders resembling those observed in naturally infected horses and sheep. However, in certain models, rather subtle behavioral changes occurred in the absence of encephalitis [1–3]. Based on these intriguing observations and the broad experimental host range, it was soon proposed that the virus may also cause behavioral abnormalities in humans, and this not only in the endemic areas in Europe but perhaps globally [1, 4].
And so the race was on to demonstrate the presence of the virus in people with depression and other psychiatric or behavioral disorders. Rott and colleagues were the first to report bornavirus-reactive antibodies in sera from patients in Germany and the United States , and similar findings were soon published by other groups [1–3]. Given its potential clinical significance, bornavirus research flourished, rapidly leading to the molecular characterization of the virus and the analysis of its T lymphocyte–mediated immunopathogenesis in experimental rodent models [1–3]. There were some flies in the ointment, though: BoDV-1–reactive antibodies were also found in the blood of healthy donors worldwide who had no history of psychiatric disorders, and so were viral antigens, antigen–antibody complexes, and BoDV-1 RNA [1–3, 5]. Other studies, however, were unable to detect any signs of infection, regardless of whether they dealt with apparently healthy or sick people [1–3, 6]. Nevertheless, the overall conclusion was that BoDV-1 represented a true human pathogen that circulates in healthy humans worldwide and that occasionally causes psychiatric and behavioral disorders [1–3, 5].
The above phylogenetic analyses not only identified the suspected human BoDV-1 sequences and the few supposedly human isolates as laboratory contaminants, but also helped to better understand the epidemiology of Borna disease in animals and its remarkable restriction to a small endemic area. Clearly, there were distinct and nonoverlapping regional sequence clusters within the endemic area (Fig 1). This striking pattern suggested that BoDV-1 was not transmitted between horses and/or sheep but instead originated from an at that time unknown, less mobile and strongly territorial reservoir, such as a rodent or another small mammal . Screening of small mammals in endemic regions finally led to the detection of BoDV-1 in bicolored white-toothed shrews (Crocidura leucodon) . Infected bicolored white-toothed shrews apparently stay healthy and show no signs of neural inflammation, despite a broad range of infected tissues, viral shedding in urine, feces, and saliva, and probably lifelong virus persistence . These characteristics make them an ideal natural viral reservoir (Fig 1). In contrast, as mentioned above, in domestic mammals such as horses and sheep, BoDV-1 infection is strictly neurotropic and induces immune-mediated encephalitis [1–3]. It was conceivable, therefore, that domestic animals represent accidental dead-end hosts that do not contribute to the spread of the virus (Fig 1). By the same token, humans may likewise be spillover hosts. If so, infected humans should be found predominantly in the endemic area and show signs of encephalitis.
In 2015, variegated squirrel bornavirus 1 (VSBV-1), a close relative of BoDV-1, was discovered in healthy exotic squirrels kept as pets in European holdings . Surprisingly and sadly, at least four breeders and caretakers came down with lethal encephalitis apparently associated with infection by this novel virus [12, 13]. Thus, similar to BoVD-1, VSBV-1 does not seem to harm its reservoir host but might cause fatal Borna disease-like encephalitis upon transmission to other species. These alarming findings once again changed the view on the zoonotic potential of bornaviruses and prompted some clinicians to consider them in patients with severe encephalitis of unclear origin.
The recent cases demonstrate that not only VSBV-1 but also BoDV-1 is a zoonotic human pathogen associated with fatal encephalitis. Thus, bornavirus infections in humans cause a clinical picture that is strikingly different from that suggested three decades ago. Remarkably, all diagnostic tools for detecting BoDV-1 had been available already in the 1990s, with the sole exception of next-generation sequencing. True human BoDV-1 infections could have been discovered 20 years earlier, if scientists had studied brain samples from severe encephalitis cases occurring in the known endemic regions instead of focusing globally on psychiatric patient cohorts. The developments in this field thus teach us the important lesson that, to guard against blind alleys, it helps to have a good theoretical framework that fully integrates the available evidence, in this case including epidemiological and clinical data.