7

Reflections on the Workshop

Microbes—bacteria, Archaea, eukaryotes, and viruses—are everywhere, including in every animal model of disease, said Vincent Young in his reflections on the important lessons of the workshop. “Therefore, if you use an animal model of any human disease or physiological process, you have a microbioanimal model,” said Young. When considering the role microbes play, researchers may be forced to reevaluate what they understand about their model. In the end, he said, the microbiota may not matter, but doing the experiments to determine whether it does or does not is necessary.

Young discussed some of the issues raised by the speakers: scale, both in terms of cost but also with regard to which interactions are important to study in a given set of experiments. “How much are you going to focus on various aspects of the host, how much on the various aspects of the indigenous microbiota, and how much on pathogens?” asked Young. Complexity is another element both in terms of technical issues and with respect to what can be controlled in any set of experiments. Translatability to human conditions is yet another consideration, as are relevance and variation. How much variation should one allow in an experiment given that variability is part of what defines being human?

The choice of which animal model to use and how to evaluate it stems from the scientific question that drives a research project, said Young. In some cases, it will be important to define the exact microbial community, while at other times using “dirty mice” may be appropriate. “It all depends on what you are asking,” said Young. It is important, he added, “to control what you can and be ready, willing, and able to measure what you cannot.” Ultimately, he said, when there are more questions than answers, reach out to the rest of the community. “We talked about having things we can share—reagents, mice, methods, technologies, strains. Perhaps that is the right way to do science.”

Joseph Newsome, associate professor of pathology and clinical director of the Division of Laboratory Animal Resources at the University of Pittsburgh, reminded the workshop audience that the idea behind specific pathogen-free animal facilities and models arose in the 1950s and 1960s as a means of standardizing animal models in order to reduce variability and improve translation of results to humans. Today, researchers have developed a host of new models for investigating the microbiome, relying on the work of the field’s pioneers. At the

same time, many of the technologies currently used have been minimally changed over time. Newsome suggested that the research community at large, and the laboratory animal community specifically, begin to challenge dieticians, cage manufacturers, and animal vendors to examine the issues discussed at this workshop to improve the translatability of animal models.

Newsome, like Young, commented on the need for more collaboration given the complexity of microbiome research. “This probably requires us to think and ask for support in ways that do not currently exist,” said Newsome. “Can we streamline the interactions and sharing of data, create repositories, and get animals from one place to the other [faster]?” he asked.

Newsome mentioned the need for training and infrastructure development, and he referred to Betty Theriault’s suggestion that the nation might need centers of excellence to train investigators and share resources. He also referred to Vowles’s suggestion that the field would benefit from standards and regulations for gnotobiotic facilities. His final comment was the need to educate IACUCs on the specific requirements for working with gnotobiotic animals.