Remarkably little is known about how hantaviruses are transmitted in nature among their reservoir rodents. Vertical transmission (passage of virus from mother to pup before, during or immediately after birth) appears to be very unusual or nonexistent. All hantavirus reservoir rodents show an increase in hantavirus seroprevalence (fraction of animals infected) as they get older. For some rodent carriers, there is a markedly greater prevalence in older males (ref, ref). Yet the precise activities and conditions that favor transmission of the virus among reservoir rodents are almost completely unknown (ref).
Historically, only two methods have been available to researchers who wished to study hantavirus transmission in rodent carriers. The first, observation of rodents in the field and correlation with hantavirus seroprevalence, is limited by the inability to determine the nature of the rodents' interactions. It is also difficult or impossible to manipulate their behavior so as to test models provocatively. The second, experimental manipulation of rodents in cages in a laboratory, is limited by the inability to recreate the natural environment. A second issue with the use of laboratory rodent colonies is biosafety. Maintenance of many important hantaviruses in their rodent hosts in the laboratory is rated at Biosafety Level 4 (ref), since infected animals can transmit the virus to workers via aerosols within closed spaces. Currently only two laboratories in North America have BSL-4 containment, and those facilities are already heavily used.
Recently we have been developing a hybrid system that incorporates aspects of field observation and laboratory manipulations to try to overcome some of the limitations inherent in conventional field and laboratory-based investigations. To do so we started by creating a novel outdoor quarantine facility that allows (1) the safe maintenance of experimentally infected animals; (2) protection from escape, thermal extremes, and predation. In the quarantine facility, deer mice are housed outdoors in artificial "nest boxes" (burrows) that protect the animal from extreme hot and cold, as does a natural burrow. The nest box consists of a nested set of buried 5-gallon plastic buckets. The top bucket is filled with dirt for insulation, and the crawlspace between the top bucket and the lower bucket comprises the nest. A 1" diameter PVC pipe affords access to the nest.
Unlike real burrows, the nest boxes are easily accessed by the investigator and can be easily decontaminated. Each nest box is contained within a steel container: a 30-gallon trashcan, with lid largely replaced by steel cloth. The lid is secured to the trashcan with steel latches and hinges, including a locking latch. A total of 34 nest-boxes is held within a 20 x20 meter perimeter fence constructed of 5 ft x10 ft 24G steel plates. The plates are buried lengthwise 2 ft into the ground and thus extend 3 ft above the surface. The quarantine facility is capable of keeping up to 34 mice alive for prolonged periods (months) despite temperature extremes ranging as low as 0oF (-17.7oC).
The quarantine facility is the first facility that allows the safe maintenance of experimentally
infected deer mice outside of a BSL-4 containment laboratory. The facility is housed at the
Sevilleta National Wildlife Refuge in central New Mexico. The 400-square-mile SNWR is
owned by the US Fish and Wildlife Service for the purpose of scientific research, and is
off-limits to the public. The SNWR has a perimeter fence and is patrolled daily. The quarantine
facility, which is in the far interior of the SNWR, miles from any human habitation, is posted
with biohazard warning signs.
In addition to the quarantine facility, we are constructing 4 other 20x20 meter enclosures that allow introduced deer mice to interact with one another under careful experimental control to determine what interactions lead to transmission of virus from one animal to another. The introduced animals will be colony-bred deer mice and some will be experimentally inoculated with Sin Nombre virus before their release. The demographics and experimental conditions can be directly manipulated to increase or decrease specific types of interactions, and the animals' whereabouts within the enclosures can be monitored with radiotelemetry. The models we develop in this experiment will be used to cross-test models of hantavirus transmission in the large (4 hectare "very large mouse array") enclosures developed by Dr. Terry Yates.
We are grateful that this research is funded by the US National Institutes of Health (RO1 AI41692).
