October 1, 2012 (Vol. 32, No. 17)
Kate Marusina Ph.D.
Animal models contribute significantly to our understanding of molecular mechanisms underlying disease pathologies. However, few models predictably translate preclinical findings into what will happen in humans.
Investigational drugs are able to cure mice from many diseases, but continue to fail in clinical trials. This fact is largely attributed to poor model designs that do not sufficiently reflect the pathophysiology of disease in humans. In addition, tremendous diversity of human genetic background, co-medications, dosing, timing of treatment, and many other factors greatly influence the treatment outcome.
The new generation of animal models, described in this article, takes into consideration previous shortcomings. These models aim to reflect the human condition as closely as possible and to close the gap between translational research and the bedside.
Inbred lab mouse strains with fixed and highly reproducible genotypes are a powerful tool for genetic manipulation. Mouse embryonic stem cells are easily amenable to genetic modifications, and thousands of genetically modified mouse strains were developed in the last 20 years.
“And yet, the inbred lab strains proved to be a poor system for discovery of specific genes associated with a particular trait, such as obesity or high blood pressure,” comments Gary A. Churchill, Ph.D., professor and principal investigator, The Jackson Laboratory.
“This is a result of their limited genetic diversity and large ‘blind spots’ largely devoid of genetic variations. We cannot resolve trait association at the level of an individual gene.
“In contrast, human genome association studies map individual genes to traits with a high degree of accuracy, but this is not enough to make a conclusive disease diagnosis.
“Therefore, discovery of a genetic basis of a complex trait required a radically new genetic strategy.”
Jackson Labs is a key participant in the International Collaborative Cross (CC) project, with the goal to create new types of inbred strains based on eight parents selected from the existing laboratory and wild strains.
“CC mice demonstrate high levels of genetic diversity,” continues Dr. Churchill. “Jackson Labs used this opportunity to take CC ideas to the next level.”
The same progenitor lines served as the parent lines for the JAX Diversity Outbred (DO) Population. This unique mouse population is maintained by a carefully designed outbreeding strategy. While still not as diverse as humans, DO mice more accurately reflect human genetic architecture and may provide better insight into genetic mechanisms of human diseases.
“The DO animals proved to be an excellent tool for mapping trait-associated loci to a higher level of resolution,” says Dr. Churchill.
Using DO mice, his team mapped a cluster of genes conferring sensitivity to doxyrubicine, a common chemotherapy agent. Within the cluster, protective, susceptible and neutral alleles were identified. The Jackson lab is collaborating with the National Institute for Environmental Health and Safety to identify susceptibility genes for other environmental pollutants.
“We hope that in the future these results will support toxicology analysis of human therapeutics,” says Dr. Churchill.
Linking Genotype with Phenotype
“Genetics of the outbred mouse population seem to provide theoretical consistency with human population,” agrees Michael D. Hayward, Ph.D., group leader, Taconic.
“However, it would be difficult to use such a population as a model to study function of individual genes. Taconic fully recognizes the profound influence of the genetic background of inbred strains on the phenotype. But we use this fact to our advantage to design our phenotyping methodology.”
The genetic background of lab mice may vary even between populations of the same strain maintained at different locations. Such sub-populations accumulate minor polymorphisms, leading to “genetic drifts.” To control for the background diversity in their phenotyping experiments, Taconic produces a litter of heterozygous progeny of a genetically engineered mouse (knockout or transgenic) that are congenic to a characterized background strain.
The second cross of heterozygous littermates produces mice that are genetically identical except for the gene of interest. This approach is especially important to correlate for subtle phenotypical manifestations such as behavioral or psychiatric changes.
To tease out these differences, Taconic developed what essentially represents a high-throughput screening of animal phenotypes in vivo. Marketed as “PhenoTac,” the analysis platform is a panel of fully validated assays representing several therapeutic areas of interest (i.e., obesity, diabetes, inflammation, neurology).
“In contrast with traditional approaches that use a large number of mice to achieve statistical significance, we conduct PhenoTac assays in the same animals, which decreases the number of mice and associated breeding costs,” continues Dr. Hayward.
“To perform phenotypic assessment of each genetic modification on most of the major aspects of mouse physiology we need only 40 experimental animals and the same number of controls. This is 10 times less than what would be required otherwise.
“Validation of our sequential strategy confirmed that the outcomes of the assays conducted later in the sequence are not affected by assays conducted in the beginning.”
Using its breeding approach and PhenoTac analysis, Taconic uncovered numerous clinically relevant phenotypes in laboratory strains with targeted and spontaneously occurring mutations.
“Our next step is to harness the power of the platform to study novel therapeutic compounds,” says Dr. Hayward. “In collaboration with industry partners, we plan to characterize phenotypes of animals carrying selected human genes. Using this carefully designed strategy, genetically modified mice can provide information for drug development that is not practical using any other method.”
Hallmarks of Cancer Development
“In the near future oncology therapy will be undergoing a paradigm shift,” comments Murray Robinson, Ph.D., senior vp of research, Aveo Oncology.
“Physicians will have an ability to tailor therapy according to the tumor genotype and the gene-expression profile of each individual patient. Aveo is working on new animal models to support this emerging paradigm. Our approach is conceptually similar to the JAX Diversity Outbred mouse population. Only we create population-diverse tumors.”
A tumor originates when tissue cells acquire specific genetic changes. The progression of the tumor is dependent on its surrounding microenvironment. In the course of its growth the tumor’s genome spontaneously acquires additional mutations. This makes tumors in each patient subtly genetically different.
To reproduce this process, Aveo developed a unique in vivo mouse system that recapitulates the hallmarks of cancer: genetics, context, and variation. The tumor donor mouse is engineered to express desired oncogenes, such as Her2. Tumorigenesis is allowed to progress naturally, a process that includes the spontaneous acquisition of additional genetic alterations.
After the tumor has developed, it is removed and transferred into the anatomically appropriate locations (e.g., breast) of multiple mouse recipients. This strategy gives rise to a population consisting of hundreds of genetically different tumors originating from the same initial genetic model. Aveo performs a comprehensive molecular characterization of the parent and the progeny tumors.
“A Her2-driven tumor population would have a certain proportion of other intrinsic mutations previously associated with cancer. We can identify them because we know the starting genetic background. This would not be possible in classical xenograft models because of high background diversity,” continues Dr. Robinson.
“Our population-based model closely reflects what happen in humans. Now we have an opportunity to correlate activity of anticancer treatments with specific tumor genetics.”
Aveo’s leading clinical drug candidate, tivozanib, just completed Phase III trial as the first-line therapy for advanced renal carcinoma.
“We tested tivozanib in our population of Her2 tumors and identified two novel biomarkers of response,” adds Dr. Robinson. “Next, we looked for these biomarkers in the context of the clinical trial and found excellent correlation of the biomarker signature with the response to the drug.”
The company plans to employ the same approach to seek biomarkers of acquired resistance to oncology drugs and use this knowledge to develop rational combinations of therapies to overcome resistance.
Genetically engineered mice became a preferred animal model because, until very recently, they have been the only mammalian species where targeted genetic manipulations were possible, according to Kevin Gamber, Ph.D., product manager, Sigma Advanced Genetic Engineering (SAGE) Labs, Sigma-Aldrich.
But the rat remains the preferred species for neuroscience, cardiovascular, and toxicology research applications, among others, adds Dr. Gamber.
“Moreover, the anatomy, physiology, and social behavior of rat is closer to human than the mouse is, and we are seeing phenotypes modeling human disease in rat models that are not present in the equivalent mouse models,” he explains.
“The lack of rat knockouts has significantly hampered development of therapeutics.”
SAGE Labs reportedly opened new opportunities for drug development by creating rat knockout (KO) models using zinc finger nucleases (ZFNs). ZFNs recognize and cut specific DNA sequences and can be used on fertilized oocytes, opening the door to genetic manipulations in species other than mouse.
The introduced genetic change is hereditary and can be stably maintained by inbreeding. SAGE Labs produces a variety of rat KO lines covering a range from oncology to neurosciences to cardiovascular disease.
In collaboration with the Michael J. Fox Foundation, SAGE Labs designed several rat models lacking genes associated with Parkinson’s disease. The pathological hallmark of the disease is loss of dopamine neurons, a phenotype that so far has been difficult to reproduce in mouse models. In the absence of neurodegeneration, mouse models cannot be used to test novel neuroprotective agents.
“Two of our rat KO models show a progressive neurodegenerative phenotype,” continues Dr. Gamber. “Hind limb deficits occur at about five months of age and progress rapidly. More subtle motor deficits occur even earlier. The preliminary tissue analysis indicates significant loss of dopaminergic neurons.”
Highly developed social behavior in rats can be used to model diseases such as autism and Fragile X. SAGE Labs designed six KO rat models using genes linked to certain components of autism spectrum disorders. Disruption of the FMR1 gene, the primary cause of Fragile X syndrome, indeed results in curtailing of rats social play, a behavior that cannot be assessed in mouse.
In collaboration with Autism Speaks, SAGE Labs continues to design rat models for measuring neuronal signaling and its effect on social interactions.
“Genetically engineered rats complement mouse in many research areas, especially where translation of mouse to human has shown to be inadequate,” says Dr. Gamber.
“We will continue expanding our platform by providing comprehensive phenotyping capabilities to our genetically engineered rat models.”