For several decades, inbred mice have been used to understand the genetic underpinnings of physiology and disease. Because inbred mice have identical genetic backgrounds, manipulating a specific gene can lead to clean conclusions on its role in biological processes—or so the thinking goes. Likewise, because identical inbred lines can be maintained across institutions, results from one lab should be easily reproduced in the next lab. But what happens when researchers want to move beyond mice, and translate their findings to humans? To get at this point, Abraham Palmer and colleagues from University of California (San Diego, CA) and University of Chicago (Chicago, IL) asked an even more fundamental question: what happens when researchers try to translate findings from one mouse line to another?

Using a clever experimental design, Abraham's group created 30 lines of mutant mice with null mutations for Cacna1c or Tcf7l2, genes with known disease associations and previously established behavioral/physiological effects in mice (Neuron 91, 1253–1259; 2016). C57BL/6J males heterozygous for either Cacna1c or Tcf7l2 null mutation were crossed with wild-type females from 30 different inbred strains, generating distinct lines of F1 progeny with siblings differing only at the targeted allele. Because their modified mouse lines had the same gene mutated (either Cacna1c or Tcf7l2), but also had different genetic backgrounds, the research team could directly test how well the genotype-phenotype relationships generalized across mouse lines.

For both Cacna1c and Tcf7l2-null lines, Abraham's group found strong interactions between phenotype and genetic background, and for some phenotypes the conclusions drawn from mutant mice of one line were the exact opposite of those found in another line. The results have significant implications for basic and preclinical research, which relies heavily on genetically modified mice. While Abraham's work doesn't invalidate the clear contributions that transgenic mice have made to biology and medicine, it is a tale of caution to researchers wishing to extrapolate data from rodents to humans: the results might not even apply to the mice next door.