In 1993, an outbreak of a mysterious respiratory illness took the lives of a number of otherwise healthy young people in the Four Corners area of the southwestern United States. Within a matter of months following this outbreak, the coordinated efforts o f scientists and health officials from a variety of local, state, and federal agencies resulted in an explanation: Sin Nombre virus, a strain of hantavirus, was identified and its spread associated with the deer mouse.

The deer mouse is the main carrier of the Sin Nombre strain of the hantavirus. Photo courtesy of Shattil and Rozinski.

Remarkable detective work led to the rapid discovery of Sin Nombre in the Southwest five years ago. Today, research into the potential spread of Sin Nombre continues at the Desert Research Institute (DRI), University of Nevada, Reno (UNR), and The Scripp s Research Institute.

Dr. Kenneth McGwire of DRI's Biological Sciences Center is part of an Emerging Virus Research Group funded by the National Institute of Allergy and Infectious Diseases (NIAID) and the National Aeronautics and Space Administration (NASA). The group is in the second year of a four year project investigating new viruses with a specific focus on Sin Nombre. Infection and transmission mechanisms are being explored in laboratory experiments. Trapping of deer mice in the field is also underway in order to lea rn more about population densities and how these are affected by the environment.

McGwire's role-which is funded by NASA--is to combine laboratory and field information with knowledge gained from satellite imagery to model, or mathematically predict, the spread of the virus on a much larger scale. In addition to McGwire, participants in the Emerging Virus Research Group include: Stephen St. Jeor, a microbiologist at UNR; Michael Buchmeir and Dennis Burton from Scripps; and John Boone, a postdoctoral researcher at UNR. Boone is helping to relate the field studies to McGwire's model ing effort.

McGwire is specifically focused on the spatial dynamics of Sin Nombre in the Great Basin. That is, how does the virus move through the mice populations in the region? Why is this important to humans? To understand, we need to learn more about Sin Nombr e.

In the West, the innocent looking deer mouse is the main host for this potentially deadly virus. Deer mice are often found where humans live and work, in homes and barns for instance. Transmission of Sin Nombre to humans can occur when mice feces or urin e--along with the virus--become suspended in the air as fine particles and inhaled. Although the chance of this occurring in most circumstances is not very high, the risk would be expected to increase along with growth in the population of infected deer mice.

Sagebrush is a common habitat for the deer mouse.

Scientists investigating the Four Corners outbreak speculated that heavy rains in 1993 led to an increase in the deer mouse food supply and, subsequently, the deer mouse population. McGwire and his colleagues are exploring the degree to which Sin Nombre infection rates among mice are related to the "connectivity" of mice from different areas. In other words, how easy is it for mice infected with Sin Nombre to pass the virus to mice in another location?

With a Ph.D. in Geography from the University of California, Santa Barbara, McGwire is an expert in the use of remote sensing and geographic information systems (GIS) to explore a broad range of environmental issues. Remote sensing uses images from space to identify features-such as vegetation-on the Earth's surface. Geographic information systems use maps to combine information in order to identify patterns across space. It is the application of these skills combined creatively with statistical analys is and modeling techniques that McGwire is using to better understand the spatial dynamics of Sin Nombre.

This map identifies the Walker River basin which is serving as a test area. Ultimately, the spread of Sin Nombre through the Great Basin will be analyzed.

According to McGwire, "One of the real driving points of the interdisciplinary nature of the work is the concept of integrating across scales. It is just mind boggling to try to relate things from the scale of the genetic sequence of a virus to the dist ribution of the virus among different habitat types across a landscape. We have people who understand genetics, ecological analysis in the field, and spatial analysis. No one of us alone could understand the whole picture. The hope is that by teaming t ogether people with this range of expertise we will have a unique opportunity to understand things from top to bottom."

Other studies are also underway in Texas and New Mexico. McGwire points out that the "New Mexico work involves large enclosure areas to control mice populations and monitor disease within these enclosures. The real contrast between the work in New Mexic o and what we're doing is that we are trying, as much as possible, to place no physical barriers within the study area. In these open systems, we are using indirect means like spatial statistics to identify such factors as rates of transmission." The test area used by McGwire and his colleagues is the Walker River basin which emerges from the alpine environment at the top of the Sierra and flows all the way down to salt desert. "We went into this work," according to McGwire, "with the idea that t he NASA-funded part of the project could look at habitat-specific parameters. We had the idea that this information might help us understand the population variability of the host species, the deer mouse; and we had the basic belief that the rate of tran smission and prevalence of the disease would be related to how densely packed the deer mice are in a specific area."

Salt desert and barren land represent a natural barrier to deer mice populations.

Vegetation maps were used originally to guide the field sampling. "We had to determine how to pick a limited number of points on the ground that represent the mouse habitat in an accurate way without missing any habitat types," according to McGwire. Usi ng satellite-derived information on habitat, trapping grids were established and mice were collected in specific locations, counted, and tested for the rate of infection.

Part of the challenge has been to identify patterns from wildlife information collected in the field. The whole idea is to pull together the laboratory and field information to answer certain questions. Is there any distinction between different habitat s and the deer mouse population? Can any of these patterns affect disease prevalence? McGwire adds, "The deer mouse is a generalist. It may be found practically everywhere, except it is much less common in areas of salt desert. We have found what appe ars to be a measurable effect on the transmission of the disease associated with an inter-fingering of salt desert and other habits within the Great Basin. There is a pattern which could be useful in understanding how the virus transmits through the regi on."

Dr. Ken McGwire analyzes satellite images of the Great Basin to figure out how deer mice are spreading hantavirus across the region.

Preliminary Results The Sin Nombre virus is present in the Walker River Basin. Mouse population numbers can vary significantly from year to year. Salt desert plays a role in determining the distribution of the deer mouse and hence Sin Nombre. There is low prevalence of the disease in low elevation, arid sites. Sin Nombre is more prevalent in male mice than in females. Sin Nombre is more prevalent in larger male mice than in smaller males. The infection rate within a habitat is not constant; sites close together in the Walker River basin show different rates of Sin Nombre infection. Centers of infection may span less than 1 kilometer. The density of mice in an area is not as strong a predictor of the rate of infection as the issue of salt desert versus other habitats. The salt desert barrier has been better defined using information derived from remote sensing and GIS. Prevalence varies over time with the observed incidence of infection dropping to zero in some sites.

The level of infection at various sites is being used to reconstruct spatial patterns. Other sites--called longitudinal sites--involve sampling over time in one location. McGwire states, "By combining information about how things change over space with information about how things change over time, we can begin to understand the transmission rate of the virus. This will ultimately be used in a simulation model. And if we have the process correct, we will be able to derive the rate of recurrence of the virus in a certain area." In other words, if the virus is not in an area now, the model can be used to predict how long before the virus comes back and once it's back, how long it will be there.

Sin Nombre is only one of many recently discovered causes of disease in humans. Since 1973, more than 30 new disease agents have been identified throughout the world ranging from HIV to hepatitis C. Some of these are also carried by animal or insect hos ts, increasing the need to understand the role of climate change and other environmental factors in the spread of disease. According to McGwire, "The use of spatial analysis in the study of the spread of disease is relatively young, and I think that ther e is an opportunity for more to be learned and for wider applications to be made."

The next challenge in the Sin Nombre study is going to be to extend the work using modeling, a mathematical way of taking what we know and projecting this knowledge to a larger area. McGwire asks, "Can we use the modeling approach to determine a reason able representation of how the virus transmits through a region as large as the Great Basin?" One of McGwire's advantages is the ability to integrate knowledge acquired in the field with model development. He is also considering the possibility of using the supercomputer located at the University of Nevada, Las Vegas to complete model simulations at the Great Basin scale.

At that point, McGwire says, "GIS may be used to combine the model results with demographic information. We will be able to look at whether a high prevalence of the virus in the deer mouse population will be likely in a particular area with defined demog raphic characteristics--a non-English speaking population, for example. In those areas, specific public education campaigns could be organized in the appropriate language to alert people to the health concern and the steps that can be taken to minimize t he risk of exposure."

In addition to modeling, what are the future directions of the Emerging Virus Research Group? McGwire hopes to explore the evolving nature of the virus over space and time. "If we had unlimited funds," he points out, "I would increase the number of peop le trapping mice in the field. It is an expensive and difficult proposition but necessary, particularly as we expand to the Great Basin scale. Also, in a recent proposal to the National Institutes of Health, we are seeking support for an international c ollaboration with researchers investigating hantavirus in Argentina." McGwire is clearly very interested in other health-related applications of spatial analysis. In his own words, "The real-world driver of health research is rewarding. There is a high degree of satisfaction in saying something definitive that might help people now."

Roger Kreidberg