Modelling complex, multi-gene scenarios in mice has historically been challenging and risky. New genetic engineering techniques, such as CRISPR, are fast removing those obstacles.Credit: Roger Harris/Science Photo Library/Getty Images
Mice have been the backbone of biomedical research for more than a century. More than 90 percent of our genes have direct counterparts in mice and 85 percent of their protein-coding regions are identical to ours. Biologists long have leveraged those similarities to better understand human disease and elucidate potential mechanisms. Manipulating the mouse genome has historically required genetic modification of mouse embryonic stem (ES) cells, which are then injected into a developing mouse blastocyst—an early precursor of the embryo. The tissues in the resulting chimera contain a mix of normal and modified cells. If the modified cells are present in the germline tissues that produce sperm and ova, the mice can be interbred to produce a fully genetically engineered animal.
“ES cells and gene targeting technologies developed in the late 80s were probably the key innovation that entrenched the mouse in all of our lives,” says Bill Buaas, Associate Director of Genetic Engineering Technologies at The Jackson Laboratory (JAX).
Yet, as sophisticated as mouse models have become, there are limitations. One can only perform such manipulations in a small subset of mouse strains from which viable ES cells can be routinely generated. Most ES cell work is done in a strain called 129, but, Buaas says, “It doesn’t really stand out as a good strain for studying diabetes or other conditions.”
As a consequence, these modified mice are typically subjected to additional rounds of breeding with a popular laboratory mouse strain known as C57BL/6J. “JAX has more than 11,000 different models in its collection, and the vast majority are on that one genetic background,” says Rob Taft, Senior Product Manager at JAX.
C57BL/6-based strains have become powerful tools for understanding gene function, but they too have shortcomings. Because the strains do not reflect the genetic variation in human population, they are not well suited to the study of complex metabolic, neurological, and autoimmune diseases, leaving researchers with limited options.
CRISPR and the new model pipeline
Genome editing using CRISPR can overcome some of those limitations. The process entails injecting a fertilized ovum with a DNA-cutting enzyme known as Cas9 and a short guide RNA targeting that enzyme to a specific site in the genome. This process can disrupt the coding sequence of a gene, essentially knocking it out. However, one can add a strand of donor DNA that is subsequently integrated at the cut site, allowing researchers to introduce new genomic sequences or precisely targeted modifications to existing genes.
This procedure can be more efficient than ES cell-based approaches, and it enables modifications into non-C57BL/6J strains. “This is a great opportunity to essentially ‘reboot’ the mouse” says Buaas. “We can make it a much better platform for biomedical science.”
Using CRISPR alongside existing knockout, knockin, or knockdown methods, researchers can now introduce modifications at multiple genomic sites simultaneously in virtually any laboratory mouse strain.
JAX embraced this technology several years ago. “We can now accommodate the vast majority of our projects using CRISPR” Taft says. That move enabled the development of a high-throughput pipeline for new, sophisticated mouse models.
JAX is part of the International Knockout Mouse Project which aspires to generate a comprehensive library of models with genomic deletions encompassing every protein-coding gene. “We’re doing about ten of their projects every Tuesday like clockwork,” Buaas says, an unimaginable task without such a pipeline.
JAX is also using this technology as part of its custom model generation service, giving academic and industry scientists access to the expertise they need to design and generate a model in the mouse strain most appropriate for their research. “We go from a concept to a sequence-validated mouse,” says Buaas. “We basically run a turnkey operation.”
JAX is extending these services to include gene and protein expression, which help reduce risk by further characterizing a model. Paired with JAX’s other In vivo services, researchers can go from concept to data without worrying about the challenges of designing, generating and managing a new mouse model.
JAX researchers are engineering an improved strain of the well-known non-obese diabetic mouse, allowing researchers to better understand diabetes.Credit: dra_schwartz/Getty Images
A model for complex human disease
Increasingly, researchers are seeking models to study complex diseases that arise from the interplay of numerous genes. For example, researchers at JAX have been developing better mouse models of type 1 diabetes. This condition arises from an autoimmune response to a patient’s pancreatic islets, disrupting the body’s ability to regulate blood sugar levels through insulin signaling. The non-obese diabetic (NOD) mouse strain, first developed through a selective breeding procedure in 1980, contains numerous genetic mutations that produce an autoimmune response closely resembling what is observed in humans.
The genetic features that contribute to this disease phenotype are not fully understood, and the NOD mouse also fails to replicate certain features of human type 1 diabetes. Genetic modification could address both problems, allowing researchers to observe the impact of manipulating individual genes on disease symptoms or to introduce additional variants that produce a higher-fidelity model.
Such manipulations have proven challenging with ES cell-based approach where modifications must be introduced individually into C57BL/6J mice, which are then bred with NOD animals. “That could require somewhere between 5 and 10 generations,” Taft says.
In contrast, CRISPR enables direct engineering of the NOD genome. At least 50 genetic variants have been linked to diabetes in this model and JAX researchers are now dissecting their contributions by examining the physiological consequences of disrupting these genes with CRISPR. In two recent studies1,2, JAX researchers demonstrated how a pair of gene variants contribute to abnormalities in the generation and selection of properly functioning B and T cells in the NOD mouse’s immune system.
“We’ve also seen people looking into how to use antibody drugs to suppress some of the autoimmune features of the NOD,” says Buaas. His team is also using CRISPR to understand the mysteries of Alzheimer’s disease as part of the multi-institutional Model-AD consortium. For this effort, JAX scientists are introducing selected mutations identified in human patients into a special mouse strain that is genetically predisposed to experience Alzheimer’s-like neurodegeneration.
The validity of mice as a surrogate for studying human disease has been questioned in recent years. Taft believes these perceptions are a consequence of the simplistic models that were feasible in the early days of genome manipulation. With today’s tools, much greater sophistication and reproducibility are now possible. “We think there's much more that could be done to make the mouse a better model system,” Taft says. “But it requires thinking differently about how we’re using the mouse, and generating models that more closely recapitulate the genetic diversity you see in the human population.”