We retrospectively investigated 3 lions for the presence of RusV. The lions were identified in 2 zoos in northern and western Germany; they exhibited neurologic signs and nonsuppurative meningoencephalitis. Lion 1 died in 1980 in a zoo in Lower Saxony, whereas lions 2 and 3 were submitted for pathological examination in 1989 by a zoo in North Rhine-Westphalia. All 3 lions displayed a mild, multifocal, lymphohistiocytic meningoencephalitis and vasculitis (Figure 1, panel A) and occasional glial nodules. Inflammatory infiltrates were most prominent in the cerebral gray matter and less prominent in cerebral white matter, cerebellum, and meninges. The spinal cord was not available for analysis. We tested archived formalin-fixed, paraffin-embedded (FFPE) tissues for the presence of RusV RNA and antigen by quantitative reverse transcription PCR (qRT-PCR), in situ hybridization (ISH) and immunohistochemistry (IHC). We included FFPE tissues originating from 8 lions without nonsuppurative meningoencephalitis (lions 4–11) as controls (Appendix).

FFPE brain samples from lions 1–3 tested positive for RusV RNA by the broadly reactive qRT-PCR assay panRusV-2 (6). Cycle quantification (Cq) values were 29–38. We detected no RusV RNA in central nervous system (CNS) samples from any of the 8 control animals (Appendix Table 1).

We determined a partial host-genome RusV sequence 409 bp long for all 3 RusV-positive animals by Sanger sequencing of overlapping RT-PCR products (Appendix). The sequences shared 97.8% nucleotide identity; phylogenetic analysis revealed all 3 sequences to form a single clade together with the sequence from a domestic cat in Hannover, Lower Saxony, in 2017 (6). Of note, this subclade was more closely related to sequences from cats with staggering disease in Austria than to sequences from zoo animals, domestic cats, and wild rodents in northeastern Germany (Figure 2).

IHC investigation for the presence of RusV capsid antigen using monoclonal antibody 2H11B1 (6) revealed multifocal, cytoplasmic, granular reactions, predominantly in cerebral cortical perikarya and their axons, in few astrocytes as well as in Purkinje cells of all 3 PCR-positive lions (Figure 1, panel B). Likewise, we detected RusV-specific RNA using a newly designed ISH probe (Appendix) in the brains of lions 2 and 3, but not of lion 1. We found viral RNA as a cytoplasmic granular signal in cortical perikarya (Figure 1, panel C). We observed RusV-specific capsid antigen and RNA in cerebral cortical neurons adjacent to perivascular infiltrates and also in neurons in more distant areas not associated with inflammatory changes. Neither IHC nor ISH revealed positive signals in any of the examined peripheral organs of the 3 RusV-positive animals and RusV-negative lion 7 (Appendix Table 1) or in the CNS of control animals. IHC staining for dsRNA using the dsRNA antibodies K1 and J2 (11) provided positive results in the CNS of all tested animals (Appendix Table 1). Immunolabeling with anti-dsRNA antibody 9D5 (11) remained negative for all 3 RusV-positive animals, whereas the RusV-negative lions 7 and 9 tested positive (Appendix Table 1).

The results of this study strongly indicate RusV as the potential cause of fatal lymphohistiocytic meningoencephalitis in lions from Germany in the 1980s. The animals were reported to have had neurologic disorders characterized by fever, depression, ataxia, and prolapse of the tongue (7). These clinical and histopathological findings are similar to those described previously for RusV-infected zoo animals and domestic cats (1,4–6); they also resemble RuV-induced encephalitis in humans (12).

A partial colocalization of RusV antigen and RNA detection with histopathologic lesions has been observed previously (1,4,6). Although the pathogenesis of RusV infection has not been elucidated, a virally triggered immune response that remains present even after focal virus clearance may provide an explanation for this phenomenon (13). In addition, vasculitis caused by a type III hypersensitivity reaction should be considered.

The lack of viral antigen and RNA in organs other than the CNS of the infected lions is consistent with previous findings in RusV-infected zoo animals of other species, in which RusV RNA was predominantly detected in the CNS and only sporadically in other organs (1,4,5). These results indicate a strong neurotropism of RusV also in lions.

In this study, we consistently detected RusV RNA and antigen in the affected animals using 3 independent methods (qRT-PCR, ISH, and IHC). The lack of viral RNA detection by ISH in the brain of lion 1, positive by qRT-PCR, could be a result of lower sensitivity of ISH (14) or of a higher degree of RNA degradation in this >40-year-old sample. Furthermore, the crosslinking of proteins caused by formalin has been shown to influence the quality and accessibility of DNA or RNA in FFPE material (15).

Immunohistochemical investigation for dsRNA revealed positive results in the brains of all investigated lions, regardless of their RusV infection status. Although it is possible that the RusV-negative lions were infected by other neurotropic RNA viruses, this scenario appears unlikely because no CNS lesions were observed in control animals 4–11. Thus, this method appears unsuitable for reliable detection of viral dsRNA in the brains of lions and perhaps other animals.

In summary, our study reveals that RusV was present in northern and western Germany in the 1980s. Detecting RusV in lions indicates an even broader host range of RusV, encompassing a variety of different species (1,4–6) and suggests that other wild and captive felids may be susceptible to RusV infection. As described previously (6), fulfilling Henle-Koch’s postulates by experimental reproduction of the disease has not been possible because of lack of RusV isolates. Nevertheless, the association between RusV detection and disease demonstrated in this study, combined with previous studies on RusV infections in zoo animals and domestic cats, strongly suggests RusV as a causative agent of meningoencephalitis in lions.