Report A Potentially Fatal Mix of Herpes in Zoos

Zoological Gardens Wuppertal, Hubertusallee 30,42117 Wuppertal-Sonnborn, GermanySummaryPathogens often have a limited host range, but some canopportunisticallyjumptonewspecies.Anthropogenicactiv-ities that mix reservoir species with novel, hence suscep-tible, species [1] can provide opportunities for pathogensto spread beyond their normal host range. Furthermore,rapidevolutioncanproducenewpathogensbymechanismssuch as genetic recombination [2]. Zoos unintentionallyprovide pathogens with a high diversity of species fromdifferent continents and habitats assembled within aconfinedspace.Institutionsalerttotheproblemofpathogenspread to unexpected hosts can monitor the emergence ofpathogens and take preventative measures [3]. However,asymptomatic infections can result in the causative patho-gens remaining undetected in their reservoir host. Further-more, pathogen spread to unexpected hosts may remainundiagnosed if the outcome of infection is limited, as inthe case of compromised fertility, or if more severeoutcomes are restricted to less charismatic species thatprompt only limited investigation. We illustrate this problemhere with a recombinant zebra herpesvirus infecting charis-matic species including zoo polar bears over at least fouryears. The virus may cause fatal encephalitis and infects atleast five mammalian orders, apparently without requiringdirect contact with infected animals [4–8].ResultsVirus IdentificationIn June 2010, two cohoused polar bears (Ursus maritimus),a threatened species, suffered epileptiform seizures at theZoological Gardens Wuppertal, Germany. The 20-year-oldfemale, Jerka, presented symptoms first and died 8 days afterthe onset of clinical signs (see Movie S1 available online).Necropsy indicated moderate to severe nonsuppurativeencephalitis and gliosis of unknown etiology as the cause ofdeath.Theperivascularlyaccentuatedinflammationconsistedof lymphocytes and plasma cells. The lesions were consistentwith virus infection, but inclusion bodies that are associatedwith some viral pathogens, including herpesviruses, were notidentified. The 16-year-old male, Lars, survived after medicalintervention, which included intravenous administration ofa hypertonic electrolyte solution and diazepam to preventseizures (Movie S1). It took several weeks for him to fullyrecover.Faced with many potential causative agents, we appliedPCR targeting eight plausible encephalitis pathogens (Experi-mental Procedures) and high-throughput DNA microarrays(ViroChip) to extracted DNA and RNA from Jerka’s brain [9].The ViroChip can test for the presence of several thousandknown DNA and RNA viral sequences in a single assay but isbiased toward human viruses and is less sensitive than othermethods such as PCR or deep sequencing [10, 11]. PCR isvery sensitive but prone to false negatives if the sequencetargeted by the primers does not match perfectly. The onlypathogen detected in Jerka was a virus related to equidherpesvirus1(EHV1)[12])identifiedbyquantitativePCR.Addi-tional PCR assays and the ViroChip did not detect othercandidate pathogens (Table 1). Western blot analysis usingantibodies against the EHV1 IR6 protein [13] supported thefindingof herpesvirus nucleic acids anddemonstrated proteinexpression exclusively in Jerka’s brain (Figure 1; Supple-mental Experimental Procedures).Virus PhylogeneticsEquid herpesvirus 9 (EHV9), a virus thought to have originatedin the plains zebra (Equus burchelli) and that is closely relatedto EHV1, has been known to cause fatal encephalitis in polarbears [14, 15]. EHV1 related to domestic horse strains hasalso been shown to cause fatal encephalitis in black bears(Ursus americanus)[8]. To determine the identity of the viralstrainfromJerka,wesequencedportionsofsixgenes,namelyUL49.5,484bp;UL45,667bp;gC(UL44),1,523bp;Pol(UL30),810bp; gB (UL27), 1,140 bp; and IR6, 342bp, and phylogenet-ically compared the obtained sequences to known EHV1,EHV9,andmoredistantlyrelatedEHV4sequences(alignmentsavailable in the Supplemental Data Sets). All amplicons weredirectly sequenced and cloned into standard plasmid vectors.Multiple individual clones per amplicon were also sequencedto confirm the sequences obtained from direct sequencing.Thesequencinganalysesprovidednoevidenceforcoinfectionwith more than one EHV strain.Jerka’s sequences clustered with horse (Equus caballus)EHV1 strains but were divergent (1%–3%) at the nucleotidelevel and formed a sister lineage for the UL49.5, gC (UL44),Pol (UL30), and IR6 genes (Figure 2A). The UL45 and gB(UL27) gene sequences and part of Pol (UL30) were nearlyidentical to a distinct EHV1 strain of plains zebras, designatedhere as zebra EHV1, which is clearly different from the strainthat caused fatalities in black bears (Figure 2A, UL45 tree)[6]. The remaining genes that we sequenced from Jerka havenot been determined for the zebra EHV1 strain. Many of thesubstitutions were nonsynonymous and changed amino acidsequenceswhencomparedtothereferencehorseEHV1strainAb4(TableS1).ZebraEHV1iswidespreadinzoozebrasandisknown to infect and cause encephalitis in four other orders ofmammals following experimental inoculation or after naturalinfection in zoos (Figure 2)[4–7].Surprisingly, the Pol (UL30) gene was a recombinantbetween EHV1 and EHV9, with the 5