Developing an in vitro approach to study transmission of respiratory viruses.

Influenza is a highly contagious disease that is spread by the transmission of virus through the air from one infected person to a new host. Many different influenza viruses are present in animals and occasionally these cross over into humans and cause outbreaks or pandemics. Animal influenza viruses can only cause pandemics if they spread efficiently between humans by passing through the air. We do not currently understand why some animal viruses can do this and others don't. One way to study this host range barrier is to use animal models such as ferrets. Ferrets can be infected with many different strains of human and animal influenza viruses but only the viruses that are already adapted to humans seem to pass through the air from one infected ferret to another animal housed in the next cage. A large body of work using the ferret model is emerging and informing the field but the model is far from ideal. One particular problem is that only small numbers of animals are used, and this makes it difficult to measure small differences between the transmission efficiency of different viruses, even though these small differences might account for important epidemiological effects. The problem might be solved by using many more animals. However here we propose an alternative strategy which will reduce the numbers of animals used because it will replace many of them with cell culture systems. These cultures are made up of different types of human cells that are present in the target tissue of the upper respiratory tract. The cells will become infected if the virus successfully passes through the air over a certain distance and can initiate infection through a mucus layer such as the one that coats the human nose and throat. Importantly the system will allow us to measure the length of time over which one infected animal is contagious. This contagiousness period can have a very large impact on the spread of a virus through the community and is an important parameter for public health planners to know. Eventually we propose that our system could be used to measure this parameter in humans. Once we have validated and calibrated the system using a small number of infected ferrets, we will begin to compare the length of contagiousness of different strains of virus to understand more about the likelihood that animal viruses will cause human outbreaks. Later in the project we propose to completely replace even the small number of ferrets with cell cultures. These will be infected with virus and air passing over them will mimic the normal breathing of a human. This will allow us to measure some transmission parameters of other important human viruses that spread through the respiratory route such as mumps, influenza type B and parainfluenza viruses. Ultimately we hope that understanding transmission will help us to use interventions more appropriately to control the spread of respiratory viruses.

A key area of current influenza research attempts to understand the genetic basis of transmissibility. We do not know why the particular swine influenza virus that emerged in Mexico in 2008/9 caused a pandemic where as other swine influenza viruses, to which humans are exposed, did not. However the large number of experiments currently performed with ferrets use small animal numbers and crude readouts, and do not adequately measure difference in transmissibility between different viruses that might have large epidemiological impacts. For example, the length of contagiousness is not measured. Instead of using many more sentinel animals, we propose to develop methodology whereby sentinel animals are replaced by cultures of human airway epithelium that recapitulate the innate barriers that respiratory viruses must overcome before they initiate infection. The project will use the H1N1 2009 pandemic strain of influenza A virus as a model because we and others already have a body of results generated in ferrets to compare and validate the in vitro alternative. The idea is to allow infected donor ferrets to breathe into a chamber as if in a small room or confined space. We will ask whether the air breathed out will and initiate infection on a cell culture placed downstream. The days on which infected animals are contagious will be compared for different viruses to measure increases or decreases in transmission. Once methodology is established, we will move to replacing the infected donor animals with infected airway cultures to assess the environmental conditions that affect transmission of different respiratory viruses through the air. Experiments showing influenza virus transmission to be optimal at dry and cold conditions are often used to explain the winter seasonality of influenza. But parainfluenza virus outbreaks occur in summer, so we aim to assess the effect of humidity and temperature on transmission of PIV and other respiratory pathogens through the air.

The main impact of this project on the NC3Rs will be to reduce the numbers of ferrets used for influenza virus transmission studies.
At the moment our standard experiments to measure transmissibility of each influenza virus we test involves a minimum of 8 animals, which comprise 4 donor animals and 4 sentinels. However as we argue in the case for support, this crude assay lacks the abilty to differentiate between the efficiency of transmission of different viruses so we need to move to a scheme we recently devised in which donors are exposed to different groups of sentinels on different days. Even in its simplest form with two different days of exposure, we would need to use 12 animals per virus tested. If we moved to exposure on 4 different days that number would be 20 animals and of we used exposure through the entire period in which we detect virus shedding from the nose, ie 7 days after donor infection we would use 32 animals.
In the first part of our proposal we remove the need for the sentinel animals so each virus tested would require 4 animals instead of 32. If we tested 4 different viruses as proposed in the first year of work, we would then use 96 fewer ferrets to compare the length of contagiousness of each virus.
If others take up our methodology, the saving in animal numbers could be much larger.
In the second year of the project we propose to completely replace the use of animals even as donors. For our own work this would save an additional 4 animal per virus tested, cumulating to 16 fewer ferrets for the 4 proposed experiments in year 2 of this project.
More importantly if others who work with different human respiratory viruses take up our method, animals that are inappropriate models for mumps, PIV or RSV will no longer be used for this work the cumulative saving although difficult to quantify will surely be much larger.

In summary
year 1 reduction in animal numbers will be 96 ferrets
year 2 replacement of 112 ferrets
year 3 replacement of 16 ferrets