Bat-associated rabies virus in Skunks.

Rabies was undetected in terrestrial wildlife of northern Arizona until 2001, when rabies was diagnosed in 19 rabid skunks in Flagstaff. Laboratory analyses showed causative rabies viruses associated with bats, which indicated cross-species transmission of unprecedented magnitude. Public health infrastructure must be maintained to address emerging zoonotic diseases.

In North America, >90% of cases of rabies in animals occur in wildlife (); several mammalian taxa harbor characteristic rabies virus variants (RABVV). In Arizona, skunks (Mephitis mephitis) and gray foxes (Urocyon cinereoargenteus) maintain independent rabies enzootic cycles, and in indigenous bats, rabies has been diagnosed in 14 of 28 species (Arizona Department of Health Services, unpub. data). Although skunks live throughout Arizona, until 2001, rabid skunks had been found only in the southeastern quadrant of the state.

In the United States, bat RABVV are a source of infection for humans and other mammals (–). Typically, interspecies infection produces a single fatal spillover event; secondary transmission has rarely been observed. Antigenic typing of rabid carnivores in Arizona from 1996 through 2000 identified bat RABVV in 1 domestic dog and 2 gray foxes. This report describes the largest documented rabies epizootic among terrestrial mammals infected with bat RABVV, with perpetuated animal-to-animal transmission. Coincident with the zoonotic disease significance, this report provides contemporary insight into pathogen evolution ().

The Study

In January 2001, a homeowner contacted Flagstaff Animal Control about a dead skunk. Although no human had been exposed to the skunk, tissues were submitted to the Arizona State Health Laboratory, where rabies was diagnosed. This skunk was the first rabid terrestrial wild carnivore reported from the area. The Texas Department of State Health Services subsequently identified an RABVV associated with bats in tissues sent for antigenic characterization. From January through April, 14 more skunks, dead or exhibiting abnormal behavior, were found throughout a large residential subdivision within 4 km of the initial case. All were infected with the same bat RABVV. From April through July, 4 more skunks infected with bat RABVV were identified ≈9 km west of the initial focus (Figure 1). Control measures included prohibiting relocation of nuisance skunks, comprehensive public education, pet rabies vaccine clinics, and a 90-day emergency quarantine requiring pets to be leashed or confined and vaccinated (Figure 1). Additionally, 217 urban skunks were vaccinated and marked with ear tags during a 6-month phased program of trap, vaccinate, and release.

Figure 1

Temporal and geographic distribution of rabies outbreak in Flagstaff, Arizona. A) Timeline and control measures. TVR: trap, vaccinate, release program. B) Geographic location of rabid skunks (dark gray dots = subclade 1, light gray dots = subclade 2).

In Flagstaff and the surrounding county, during the decade before this epizootic, 2 rabid bats, on average, were reported each year. During the epizootic, 218 animals were submitted for rabies testing (Table). Rabies was confirmed in 19 (13%) of 145 tested skunks and 2 (9%) of 22 tested bats. Although most (18 [95%]) of the rabid skunks were identified and reported by lay citizens, no contact between these skunks and humans or domestic animals was reported.

Table

Animals from Flagstaff submitted for rabies diagnosis, January–July, 2001

Animal	Scientific name	No. submitted	No. rabid
Skunk	Mephitis mephitis	145	19
Bat	(Multiple spp.)	22	2
Domestic cat	Felis domesticus	12	0
Gray fox	Urocyon cinereoargenteus	9	0
Domestic dog	Canis familiaris	9	0
Squirrel	Species unknown	8	0
Coyote	Canis latrans	4	0
Raccoon	Procyon lotor	2	0
Porcupine	Erethizon dorsatum	2	0
Prairie dog	Cynomys ludovicianus	2	0
Badger	Taxidea taxus	1	0
Opossum	Didelphis virginiana	1	0
Bobcat	Lynx rufus	1	0
Total		218	21

Local baseline population estimates were not available to indicate whether skunk demography affected disease attributes. Synchronous with this outbreak, independent epizootic activity caused by well-established skunk RABVV was documented in southern Arizona, which suggests that regional skunk epizootiologic dynamics were similarly affected. Skunks' seasonal behavior may have contributed to transmission events. This epizootic was initially recognized when a dead skunk appeared in a snow-covered backyard, during a season when skunks are in communal dens. Given an incubation period of 2 months, most transmission would have occurred between late autumn (when skunks are in their dens) and late winter (when they are mating).The Flagstaff epizootic peak coincided with nationwide seasonal trends of rabid skunks (). Enhanced postepizootic surveillance in Flagstaff did not detect additional rabid terrestrial mammals for the next 3 years. However, in 2004, a total of 5 skunks found in the initially affected east Flagstaff neighborhood and 1 fox 28 km south of Flagstaff were infected with the same bat RABVV ().

Viruses isolated from the rabid skunks exhibited monoclonal antibody patterns similar to RABVV associated with big brown (Eptesicus fuscus) and Myotis bats in the western United States (). These are among the most abundant bat species in Arizona and often roost in houses and outbuildings; however, no bat colonies were found in association with any of the rabid skunks. Restriction digests of PCR amplicons from the rabid skunks did not match patterns known for RABVV from North American terrestrial reservoirs (). Phylogenetic analysis of a 300-bp region of the N gene showed that the Flagstaff skunk RABVV was identical (100%) to Arizona bat RABVV (Table A1, Figure 2A), and differed by 22% from skunk and gray fox RABVV. A monophyletic clade (clade A) of 8/8 big brown, 5/14 Myotis, and 1/6 southern yellow (Lasiurus ega) bats shared >95% identity with Flagstaff skunk RABVV. An additional 44 samples, representing 11 bat species, differed by >8% from Flagstaff skunk RABVV.

Table A1

Representative rabies virus variants found in Arizona and included in phylogenetic analyses

CDC ID	State ID	Taxon name	Common name	Scientific name	Collection date	Collection Site
3629	504318	Ap1	Pallid bat	Antrozous pallidus	Jul 1997	Navajo
3626	498674	Ap2	Pallid bat	A. pallidus	Jul 1997	Maricopa
3628	500509	Ap3	Pallid bat	A. pallidus	Jul 1997	Wickenburg
3925	390721	Ap4	Pallid bat	A. pallidus	Sep 1995	Coconino
3927	735430	Ap5	Pallid bat	A. pallidus	Mar 1997	Tucson
4891	10793	Ef1	Big brown bat	Eptesicus fuscus	Sep 1999	Yuma
4862	947906	Ef2	Big brown bat	E. fuscus	Jun 1999	Davis AFB
4867	9630	Ef3	Big brown bat	E. fuscus	Aug 1999	Tucson
4850	15171	Ef4	Big brown bat	E. fuscus	Oct 1999	Tucson
4871	99026842	Ef5	Big brown bat	E..fuscus	Jun 1999	Coconino
4886	99046548	Ef6	Big brown bat	E. fuscus	Oct 1999	Yavapi
5450	1034778	Ef7	Big brown bat	E. fuscus	Jul 2001	Coconino
5442	1026934	Ef8	Big brown bat	E. fuscus	Jul 2001	Coconino
4074	98007821	Em1	Spotted bat	Euderma maculatum	Sep 1998	Maricopa
3057	396597	Ln1	Silver-haired bat	Lasionycteris noctivagans	Oct 1995	Maricopa
3285	432420	Le6	Southern yellow bat	Lasiurus ega	May 1996	Coconino
3046	264111	Le1	Southern yellow bat	L. ega	Aug 1993	Yuma
3870	814565	Le2	Southern yellow bat	L. ega	Oct 1997	Pima
3284	447760	Le3	Southern yellow bat	L. ega	Aug 1996	Yuma
3050	374106	Le4	Southern yellow bat	L. ega	Jun 1995	Yuma
3347	395408	Le5	Southern yellow bat	L. ega	Sep 1995	Yuma
5422	141014	Scsk1	Striped skunk	Mephitis mephitis	May 2001	Cochise
5423	142140	Scsk2	Striped skunk	M. mephitis	May 2001	Cochise
4995	1001810	Sk1	Striped skunk	M. mephitis	Jan 2001	Coconino/Flagstaff
4998	1004508	Sk2	Striped skunk	M. mephitis	Jan 2001	Coconino/Flagstaff
5079	1006928	Sk3	Striped skunk	M. mephitis	Jan 2001	Coconino/Flagstaff
5470	1006403	Sk4	Striped slunk	M. mephitis	Feb 2001	Coconino/Flagstaff
5074	1009087	Sk5	Striped skunk	M. mephitis	Feb 2001	Coconino/Flagstaff
5075	1010227	Sk6	Striped skunk	M. mephitis	Feb 2001	Coconino/Flagstaff
5076	1010229	Sk7	Striped skunk	M. mephitis	Feb 2001	Coconino/Flagstaff
5077	1011591	Sk8	Striped skunk	M. mephitis	Mar 2001	Coconino/Flagstaff
5081	1012600	Sk9	Striped skunk	M. mephitis	Mar 2001	Coconino/Flagstaff
5080	1014008	Sk10	Striped skunk	M. mephitis	Mar 2001	Coconino/Flagstaff
5100	1015436	Sk11	Striped skunk	M. mephitis	Apr 2001	Coconino/Flagstaff
5101	1015687	Sk12	Striped skunk	M. mephitis	Apr 2001	Coconino/Flagstaff
5102	1015688	Sk13	Striped skunk	M. mephitis	Apr 2001	Coconino/Flagstaff
5103	1016511	Sk14	Striped skunk	M. mephitis	Apr 2001	Coconino/Flagstaff
5132	1016704	Sk15	Striped skunk	M. mephitis	Apr 2001	Coconino/Flagstaff
5133	1016707	Sk16	Striped skunk	M. mephitis	Apr 2001	Coconino/Flagstaff
5440	1023718	Sk17	Striped skunk	M. mephitis	May 2001	Coconino/Flagstaff
5441	1025813	Sk18	Striped skunk	M. mephitis	May 2001	Coconino/Flagstaff
5451	1036291	Sk19	Striped skunk	M. mephitis	Jul 2001	Coconino/Flagstaff
3858	721957	Mc1	California myotis	Myotis californicus	Dec 1996	Pima
3847	433224	Mc2	Western small-footed bat	M. ciliolabrum	May 1996	Holbrook
3848	445332	Ml1	Little brown bat	M. lucifugus	Aug 1996	Wickenburg
4882	99043079	Mu1	Myotis bat	Myotis sp.	Sep 1999	Bullhead City
3862	800694	Mu2	Myotis bat	Myotis sp.	Jul 1997	Pima
4887	99048912	Mu3	Myotis bat	Myotis sp.	Oct 1999	Pima
419	32455	Mu4	Myotis bat	Myotis sp.	Apr 1983	Mesa
3350	457788	Mu5	Myotis bat	Myotis sp.	Oct 1996	Maricopa
3351	390759	Mu6	Myotis bat	Myotis sp.	Sep 1995	Yuma
4881	99042806	Mu7	Myotis bat	Myotis sp.	Sep 1999	Yuma
4873	99038008	Mu8	Myotis bat	Myotis sp.	Aug 1999	Eager
4872	99028231	Mu9	Myotis bat	Myotis sp.	Jun 1999	Pima
3855	502675	Mv1	Cave myotis	M. velifer	Jul 1997	Maricopa
3852	489891	My1	Yuma bat	M. yumanensis	May 1997	San Carlos
3043	259807	Ph1	Western pipistrelle	Pipistrellus hesperus	Jul 1993	Maricopa
3860	747838	Ph2	Western pipistrelle	P. hesperus	Jun 1997	Pima
3044	259808	Ph3	Western pipistrelle	P. hesperus	Jul 1993	Maricopa
3924	455031	Ph4	Western pipistrelle	P. hesperus	Sep 1996	Navajo
2060	8543	Ph5	Western pipistrelle	P. hesperus	Sep 1981	Sedona
3863	805082	Ph6	Western pipistrelle	P. hesperus	Aug 1997	Pima
3345	466665	Ph7	Western pipistrelle	P. hesperus	Dec 1996	Maricopa
3859	735181	Ph8	Western pipistrelle	P. hesperus	Mar 1997	Oro Valley
3053	391577	Ph9	Western pipistrelle	P. hesperus	Sep 1995	Coconino
410	26202	Tb1	Mexican free-tailed bat	Tadarida brasiliensis	May 1985	Riveria
448	5646	Tb2	Mexican free-tailed bat	T. brasiliensis	Aug 1982	Thatcher
4860	943431	Tb3	Mexican free-tailed bat	T. brasiliensis	May 1999	Tucson
4857	821298	Tb4	Mexican free-tailed bat	T. brasiliensis	Dec 1997	Tucson
4889	99057833	Tb5	Mexican free-tailed bat	T. brasiliensis	Dec 1999	Phoenix
4863	848036	Tb6	Mexican free-tailed bat	T. brasiliensis	Jun 1999	Tucson
4868	9810071	Tb7	Mexican free-tailed bat	T. brasiliensis	Oct 1998	Phoenix
4864	949009	Tb8	Mexican free-tailed bat	T. brasiliensis	Jun 1999	Tucson
4866	949396	Tb9	Mexican free-tailed bat	T. brasiliensis	Jun 1999	Tucson
4865	949010	Tb10	Mexican free-tailed bat	T. brasiliensis	Jun 1999	Tucson
4847	10197	Tb11	Mexican free-tailed bat	T. brasiliensis	Sep 1999	Tucson
4884	99046042	Tb12	Mexican free-tailed bat	T. brasiliensis	Sep 1999	Winslow
3344	713793	Pt1	Townsend's big-eared bat	Plecotus townsendii	Oct 1996	Pima
3659	513855	Nm1	Big free-tailed bat	Nyctinomops macrotis	Oct 1997	Yavapai

Figure 2

A) Phylogenetic tree of the 19 rabid skunk isolates and representative samples of known rabies virus variants (RABVV) from Arizona based on 300 bp of the nucleoprotein (N) gene (GenBank accession no. AY170226–304). B) Detailed analyses of clade including all 19 skunk isolates (clade B) based on 2221 bp of the N and glycoprotein (G) genes (GenBank accession no. AY170397–438). Phylogenetic analyses used PAUP* software (version 4.0b2, Sinauer Associates, Sunderland, MA, USA; 2000] using the neighbor-joining search algorithm (minimum evolution) with maximum likelihood to estimate Ti:Tv ratio and nucleotide base frequencies (HKY85 model). Numbers at tree nodes indicate nonparametric bootstrap proportions based on 1,000 replicates.

An analysis of clade A, which incorporates N and G genes, indicated that the Flagstaff skunk RABVV were more closely related to 2 bat RABVV (E. fuscus from Coconino County, M. velifer from Maricopa County) collected in 1999 and 1997 than to the 2 bat RABVV collected locally during the outbreak. In clade B, subclade 1 RABVV were collected from January through early April from the northeastern region of the outbreak, whereas subclade 2 RABVV were collected from early March through July from the southeastern and western regions of the outbreak (Figure 1). However, phylogenetic data do not support a wavelike spread from northeast to west because this would require nesting of subclade 1 within subclade 2. In contrast, both subclades exhibit independently derived mutations. East-to-west epizootic movement of RABVV within subclade 2 (sk16–19 form a monophyletic clade nested within subclade 2) during April is supported by the data and may be related to dispersal of infected skunks along river corridors or translocation by humans. One person reported trapping, moving, and releasing a skunk before the outbreak was known in the community. Alternatively, apparent shifts may be an artifact of intensified public awareness and reporting. Lack of sampling in the uninhabited forest between the eastern and western foci limits our ability to discriminate among these hypotheses.

Conclusions

This is the largest recorded cluster of bat RABVV infection in terrestrial mammals. Investigation of this novel outbreak showed evolution in action with the emergence of an RABVV that successfully adapted from Chiroptera to Carnivora. Previously documented clusters involving 3–4 to terrestrial mammals infected with a single insectivorous bat rabies virus variant did not corroborate sustained transmission (). Although >1 skunk may have been exposed to a single rabid bat, it is highly unlikely that each skunk was exposed to the same bat or that multiple bat-skunk exposures occurred. We could not ascertain the complete scope of this outbreak or whether it was the index event. Phylogenetic analyses support the evolution of 2 independent lineages, suggesting establishment for months or years. Additionally, virus isolation from salivary glands of 5 affected skunks and the reappearance of rabid skunks with the same RABVV in 2004 support the probability of independent transmission.

The recognition of this epizootic can be credited to a coordinated laboratory-based disease surveillance program to monitor sick and dead wildlife for potential zoonoses (plague, tularemia, rabies) even in situations lacking human or pet exposures. Comprehensive animal disease surveillance provides direct benefits to public health and animal health by promoting early recognition of risk and opportunities for disease control and prevention interventions.

Unpredictable health threats related to emerging zoonoses, especially those involving wildlife reservoirs, pose notable surveillance and control challenges (–). Recent bioterrorism initiatives emphasize integration of human and animal disease surveillance, and enhanced laboratory capacity, as essential functions in zoonosis detection (). Rabies surveillance and control programs serve as historic prototypes for effective, long-term, public health programs. Quintessential zoonotic disease programs require innovative and expanded capacities, commitments to public health and veterinary laboratory infrastructure, and appropriate interagency and interdisciplinary coordination and communication.