We've updated our Privacy Policy to make it clearer how we use your personal data. We use cookies to provide you with a better experience. You can read our Cookie Policy here.

Advertisement

How Diverse Mouse Models Can Drive Drug Discovery Innovation

A person holding a white lab mouse with a gloved hand.
Credit: iStock
Listen with
Speechify
0:00
Register for free to listen to this article
Thank you. Listen to this article using the player above.

Want to listen to this article for FREE?

Complete the form below to unlock access to ALL audio articles.

Read time: 7 minutes

Testing drugs before they reach human trials typically involves the use of animal models. For decades, mice have been one of the most established preclinical models. However, these “inbred” strains are often unable to replicate human diseases, particularly for complex conditions like diabetes and cancer. This creates inconsistent and sometimes unreliable results. Another alternative – using lab-grown cell cultures – can also fail to reproduce complex diseases.


Instead, researchers from The Jackson Laboratory (JAX) suggest using a combination of genetically diverse mouse models plus both human and mouse cells to model diseases and predict responses to drugs. This stands to also reduce the use of mouse models while maximizing the scientific benefits — with the US Food and Drug Administration’s (FDA) Modernization Act 2.0 stoking the need to find more reliable solutions to animal testing.


Technology Networks spoke to JAX’s mammalian genetics scientific director, Prof. Nadia Rosenthal, to find out more about their new approach and how using genetically diverse mice that better reflect human diversity can help find better therapies for a variety of diseases.

Sarah Whelan, PhD (SW):

Could you explain a bit more about the FDA's recent decision to allow alternatives to animal testing through the Modernization Act 2.0?


Nadia Rosenthal (NR):

There is a growing sentiment in biomedical research that animal models too frequently fail to accurately predict the safety and efficacy of new treatments before they enter human clinical trials. The FDA’s recent decision to phase out animals from the preclinical screening process is a response to increasing concerns that discoveries in animal models often fail to lead to medical progress.


As a result, research in the mouse faces substantial criticism and questioning of its effectiveness as a model for human disease, but these concerns deserve closer scrutiny. To the betterment of standardization, the widespread use of single inbred mouse strains to assess functional outcomes in thousands of patients has narrowed the scope of our investigations. Inbred mouse models incorporating single gene deletions have been valuable for researching some rare monogenic disorders but fall short in modeling more complex diseases that are rarely attributable to a single gene’s function.


Increasing research into new alternatives for preclinical testing includes cell-based assays, organoids and even in silico predictions. Leveraging AI-powered analytics, these new platforms promise to deliver insights into patient responses to drug candidates to minimize animal testing and promote a more efficient pharmaceutical engine. The mere prospect of sidestepping Phase 1 safety trials, with the attendant savings in time and cost, is feeding an explosion of activity fueled by the pharmaceutical industry.


That said, a substantial degree of uncertainty is inherent in making the leap from testing compounds using animals, cells or algorithms to predicting the likely response of human patients. It is still unclear to what extent organoid models and microtissue systems will recapitulate the quantitative, experimentally malleable, predictive molecular and cellular aspects of dynamic disease states in vivo.


Precision medicine requires model systems that can accurately reflect the variable genotype–phenotype associations seen in patients and the temporal progression of individual pathological features. The failure to translate knowledge to medicine is partly due to failure to embrace the complexity of human disease in our models and reductionist approaches are unlikely to overcome this obstacle. Rather, we must develop models that embrace the genetic and environmental complexity of pathophysiological processes. Until this complexity is acknowledged, gaps will remain between researchers and clinicians, between model data and human patient data, and between proprietary data sets and those who can consolidate and analyze them. To make medicine better, those gaps need to be bridged.


These realizations have prompted researchers at JAX to improve the current practice of preclinical testing and narrow the uncertainty gap by developing new genetically diverse models, leading to more reliable outcomes in patients through re-envisioning our current models and their applications. 



SW:
How can genetically diverse mouse models be used to advance our understanding of complex human diseases and their treatment?

NR:

Heterogeneity in patient presentation and disease progression is a well-recognized challenge in clinical trial design and endpoint selection. A growing appreciation for the profound influence of genetic variation on human disease risk and therapeutic outcomes has resulted in a re-evaluation of current genomic studies that have focused largely on populations of European ancestry, introducing racial bias in genomic discoveries and limiting their predictive power.


Mirroring some of the current shortcomings in human genomic research, most preclinical studies in mice explore a limited range of genetic variation present in their parent species. It is hardly surprising then that a similar lack of both genetic and environmental variation in laboratory animal experimentation can limit the breadth of preclinical findings and is likely a major contributing factor in cases where they fail to translate to humans.


The introduction of genetic diversity, along with dramatic increases in the precision of mouse–human comparative analyses, have significantly improved the utility of mice as models of human disease.


Where inbred mouse strains fall short in modeling more complex diseases that are rarely attributable to a single gene’s function, replacing the canonical “laboratory mouse” with genetically diverse (GeDi) reference populations has enabled more precise assessments of human genetic variation and has yielded more clinically relevant data in many common disease areas, including cardiovascular, dementia, diabetes and infectious diseases among others.


With the current emphasis on small molecules in drug development, introducing genetic diversity into preclinical study design can overcome the limitations of modeling toxicity responses in standard inbred mouse strains. Discoveries of heritable factors influencing drug addiction have met with limited success in human genome-wide association studies (GWAS) but are more readily identified in GeDi mouse panels, where the mechanistic basis of differential responses can be dissected.


The utility of introducing genetic diversity into the modeling pipeline can also enhance the preclinical testing of genome editing strategies, which heretofore have been mainly explored in mouse disease models using standard inbred backgrounds. Distributing rare Mendelian disorders across the entire genetic diversity landscape by crossing existing mouse models with GeDi mouse panels affords assessment of broad phenotypic ranges, which can lead to the identification of more clinically relevant phenotypes, the discovery of biomarkers and modifiers to serve as new drug targets. It’s an exciting time in medical genetic discovery.



SW:
Given the ethical considerations surrounding the use of animal models, can you discuss why you think they remain important? What are the strengths and weaknesses of in vivo and in vitro systems, such as human cell-based experiments, and how can they complement each other?

NR:

There is a growing awareness of the need to re-examine how we design live animal studies, balancing our ethical obligation to minimize the use of animals with the clinical relevance of results. The goal of animal testing is to enhance our understanding of the underlying genetic susceptibilities to disease in human patients using reduced animal numbers, while enabling the identification of specific targets for designing precise pharmaceutical interventions. At JAX, we are exploiting advances in continuous digital analysis of naturally behaving mice in different environments that approximate “free range” studies in the field and can be repurposed for improved preclinical animal studies to increase efficiency, scalability, translatability, objectivity and reusability of animal data. Such monitoring technologies remove subjectivity in current methods, are scalable for drug discovery, avoid the unnecessary stressor of contrived experimental procedures, optimize animal welfare through early detection of morbidities and enable interventions to generate data that can be reused for increased utility. In the end, it’s not the number of mice we analyze, but the quality of the model we develop that will best serve the 3R rule in animal research (replace, reduce, refine).


Most effectively employed as a complement rather than a replacement for animal testing, in vitro disease models derived from human pluripotent stem cells or tissue stem cells are finding increasing application in academia and in the biotechnology and pharmaceutical sectors. Although animal research has historically been an indispensable cornerstone of critical areas such as toxicology, pharmacogenetics and aging, new cell-based alternatives to traditional animal screens are being actively pursued in these and many other disciplines. To date, diseases studied in cellular screens have been largely restricted to cancer, but the field is moving towards large-scale surveys of therapeutic candidates in human cell and organoid panels, in hopes of mimicking the responses of patients to new drugs.


When building experimental model systems, it is important to consider the likely outcome of genetic analysis of disease. Cellular disease modeling has been generally conducted in a retrospective fashion. Investigators begin with a well-defined (usually Mendelian) genetic syndrome with a clear patient presentation, introduce the causative variant into stem cells, differentiate the stem cells into a relevant cell type and obtain results that confirm what is known from clinical findings. The challenge going forward will be to analyze candidate genetic variants of uncertain significance, in a prospective fashion to functionalize disease variants and candidates whose pathogenic role is completely unclear. This is where the introduction of genetic diversity into our models has the greatest impact.



SW:
How can researchers use this new approach to combine the use of genetically diverse mouse models with cell-based assays?

NR:

To date, the extent to which organoid models and microtissue systems recapitulate disease states remains limited. The most pervasive human diseases are caused by the interplay of underlying heritable traits with a lifetime of environmental influences that are impossible to track in an individual patient, let alone replicate in a simplistic animal model. These same shortcomings in current model organism research design and practice are likely to extend to cell-based platforms.


Interrogating cell/organoid platforms from GeDi mouse panels helps to overcome these challenges. Extensive genetic, genomic and phenotypic information are already available for GeDi mouse populations. Introducing cell multiplexing (“villages in a dish”) into the screening process can establish face validity for cell-based platforms through iterative in vivo/in vitro multiomic profiling in genetically matched cells and animals. Readout from GeDi mouse cell and organoid platforms can then be aligned with genomic and medical records in human disease.


The use of human stem cell-derived systems in combination with panels of GeDi mouse stem cells further powers the identification of the optimal genetic backgrounds for study in vivo.  Combining rich datasets from in vivo analysis in GeDi mice with high throughput in vitro analysis of human and mouse cell-based data provides a means to establish ground truth for the clinical readout of specific genetic variants, uncovering genetic modifiers, new biomarkers for monitoring disease risk or progression and targets for early intervention.



SW:
Could you give us some examples of additional insights generated from the use of genetically diverse mouse models over standard ones?

NR:

An example from our own laboratory: COVID-19 has been associated with a wide range of outcomes in human infections, ranging from asymptomatic or subclinical disease to hyperinflammation, acute respiratory distress and death. Notably, candidate host genes affecting pathogenic disease outcomes have been primarily investigated using reverse genetic approaches in mice, evaluating the effects of ablation or humanization of specific genes, such as innate immunity genes. While these studies have yielded valuable information, they do not take into account the simultaneous contribution of genetic variants impacting multiple pathways that may influence infection outcomes.


In a previous study, we assessed the impact of host genetics on COVID-19 severity and immune responses by creating a panel of humanized GeDi mouse strains that accurately modeled the highly variable human response to infection with the ancestral SARS-CoV-2 strain. The resulting wide spectrum of survival, viral replication kinetics and immune profiles in these mice recapitulated the complexities of SARS-CoV-2 virus replication, dynamics and inflammatory profiles in patients, underscoring the importance of host genetics in interpreting or predicting individual responses to viral infection.


Since then, mutations in SARS-CoV-2 variants of concern (VOCs) have expanded the viral host range beyond primates to a limited range of other mammals including mice, affording the opportunity to exploit direct infection of mice without humanization to more accurately model the broad range of disease outcomes in patient populations.  SARS-CoV-2 VOC infections in GeDi mouse panels produced an equally broad range of viral burden, disease susceptibility and survival. Whereas most GeDi mouse strains were resistant to disease despite measurable lung viral titers, CAST/EiJ, a wild-derived strain, developed high lung viral burdens with severe pulmonary pathology without ectopic spread to the brain seen in other humanized models, and a dysregulated cytokine profile that resulted in morbidity and mortality.


These, and numerous other examples, have established GeDi mouse and cell panels as clinically valuable platforms to map genetic contributions to infectious disease severity. We are now refining the approach to evaluate therapeutic countermeasures against acute and long COVID-19, an avenue that can be readily and rapidly followed for other human viral pandemics in the future.