In late 2021, an adult golden lion tamarin from a zoo in western Germany died (Appendix). On the basis of the symptoms and an initial diagnosis by the Hessian State Laboratory (Landeslabor Hessen, Giessen, Germany), we conducted further screening for LCMV. The Hessian State Laboratory and the zoo sent tissue samples from the golden lion tamarin and wild mice from the zoo (taken in 2009, 2021, and 2022) to the Friedrich-Loeffler-Institut (Greifswald–Insel Riems, Germany) for molecular and epidemiologic investigations. We extracted and screened nucleic acids for LCMV using conventional reverse transcription PCR (14). We detected LCMV RNA in the brain of the golden lion tamarin and in the kidneys of 55% of wild house mice (M. m. domesticus) from 2021 and 2022 (n = 53) but not in any of the house mice from 2009 (n = 82). On the basis of ≈340 nt sequences from the L segment of the virus (GenBank accession nos. OP938541–68), we selected 2 mice with the most dissimilar LCMV sequences and, together with brain tissue from the diseased golden lion tamarin, used them for high-throughput sequencing of complete genomes (GenBank accession nos. OP958777–82) (Appendix).

Figure

Figure. Phylogeny of the L protein encoding nucleotide sequences of lymphocytic choriomeningitis virus (LCMV) identified in Germany (blue arrows) and reference sequences, constructed by using Bayesian inference. Lunk virus from ...

We used the new complete coding-region sequences (L, GP, and NP) together with all published LCMV genomes to reconstruct phylogenetic trees using the general time reversible substitution model with invariable sites and gamma distribution (MrBayes 3.2.7, https://nbisweden.github.io/MrBayes) (15). LCMV sequences of the full coding regions of the L, NP, and GP proteins were almost identical between the golden lion tamarin and 1 of the mice, forming a monophyletic clade within LCMV lineage II (Figure; Appendix Figure). The sequences obtained from the other mouse fell within lineage I (Figure; Appendix Figure). According to the L segment sequences (≈340 nt) obtained from the remaining 27 LCMV-positive mice, both lineages were nearly equally represented in the zoo population (lineage I for 16 mice, lineage II for 11 mice).

We obtained sequences from the mitochondrial DNA d-loop of all LCMV-positive mice and several LCMV-negative mice from 2009 (n = 32), 2021 (n = 12), and 2022 (n = 41). All those sequences identified exclusively the house mouse subspecies M. m. domesticus (data not shown).

The high similarity between LCMV lineage II in a golden lion tamarin and a wild house mouse indicates that the virus was passed between wild and captive animals in the zoo. The large number of LCMV lineage I and II strains in the wild house mouse population at this site suggests either an outbreak after recent introduction from 2 different sources or long-term persistence in the local house mouse population but with very low prevalence in 2009.

Despite considerable effort by researchers to detect LCMV in Germany, the virus remains mostly elusive. Although the route through which LCMV entered the zoo is not known, this event most likely occurred after 2009. The virus may have been brought in through naturally occurring wild animals in the region (e.g., wild house mice) or, although unlikely, through infected zoo animals.

We provide evidence for LCMV lineage II in Germany within an area naturally occupied only by the M. m. domesticus subspecies of house mice. The occurrence of both LCMV lineages I and II in M. m. domesticus mice does not support the subspecies host specificity proposed by Fornůsková et al. (1). Further evaluation of LCMV association with house mouse subspecies in Germany and other parts of the world will help clarify potential expanded risk to animal and human health.