GRO g Mouse

What are the major advances in genetically engineered mouse models for cancer research?

Genetically engineered mouse models have evolved significantly to better represent human cancers. Modern models can now specifically control the timing and location of mutations, even within single cells, allowing researchers to more accurately model sporadic human cancers. These advanced models permit the study of tumor development and its interaction with surrounding stroma as both evolve naturally .

These models have substantially advanced our understanding of:

• Cancer initiation mechanisms

• Immune system roles in tumorigenesis

• Tumor angiogenesis processes

• Invasion and metastasis pathways

• Molecular diversity patterns in cancers

Recent technological developments have enabled in vivo imaging that can track both primary and metastatic tumor development from much earlier stages than previously possible .

How do mouse models bridge the gap between human observation and experimental validation?

While human tumor studies have yielded insights into molecular changes in cancers, more rigorous testing through experimental manipulation is necessary to distinguish between causative changes (potentially targetable) and secondary changes. Mouse models provide an experimentally tractable mammalian system to test hypotheses generated from human tumor studies and can identify novel mechanisms to be confirmed in human tumors .

The power of these models lies in their ability to develop tumors de novo in the context of a normal immune system while coevolving with surrounding stroma. Furthermore, over a century of genetic research in mice has provided powerful strategies for assessing complex genetics of disease susceptibility, therapeutic responses, and associated toxicities that vary within human populations .

What is the Single Mouse Experimental Design and when should it be considered?

The Single Mouse Experimental Design is an alternative approach to traditional preclinical in vivo testing that enables greater inclusion of genetic diversity in cancer studies. In this design:

• Each mouse has a different patient-derived xenograft

• Endpoints are tumor regression and Event-Free Survival (EFS)

• No control (untreated) tumor is used

• One mouse per treatment group is utilized

This approach is particularly valuable when:

• Resources are finite but representation of genetic/epigenetic diversity is essential

• There is a need to identify biomarkers of response that can inform clinical trials

• Testing a large panel of models from a specific cancer type to identify sensitivity patterns

For example, conventional designs with 10 mice per treatment and control group would allow testing of just one model, while the single mouse design potentially allows inclusion of 20 models, dramatically increasing genetic heterogeneity representation .

How does the Single Mouse Design compare to conventional experimental designs in terms of statistical validity?

Conventional experimental designs determine group size according to the variance in tumor growth rates within a group and the statistical endpoint being applied. This approach necessitates large numbers of animals, especially when seeking to detect relatively small drug effects. With finite resources, this restricts the number of models that can represent each cancer type (typically 6-8 models per disease in programs like the Pediatric Preclinical Testing Program) .

A retrospective study indicated that using one mouse per treatment group was adequate to identify both active and inactive agents. In a prospective validation using PLX038A (a PEGylated SN-38 prodrug), the single mouse design successfully identified xenograft models sensitive to this agent, and these correlated with sensitivity to irinotecan, validating the design's ability to identify agents with the same mechanism of action .

The table below demonstrates the predictive power of the single mouse approach:

How can researchers identify potential annotation errors in mouse model genomic data?

Incorrectly annotated sequence data are becoming more commonplace as databases increasingly rely on automated techniques for annotation. Machine learning approaches designed to automatically predict protein functional classes can be employed to identify potential gene annotation errors .

In a study of 211 previously annotated mouse protein kinases, 201 of the GO annotations returned by AmiGO appeared inconsistent with the UniProt functions assigned to their human counterparts. In contrast, 97% of the predicted annotations generated using a machine learning approach were consistent with the UniProt annotations of the human counterparts and with available annotations in the Mouse Kinome database .

The analysis revealed a striking discordance in the distribution of Ser/Thr, Tyr, and dual specificity kinases in mouse versus human based on AmiGO annotations, while UniProt annotations showed similar distributions between species:

When annotations had evidence codes other than "RCA" (inferred from reviewed computational analysis), the machine learning classifier trained on human protein kinase data correctly labeled 85% of mouse protein kinases relative to the AmiGO reference .

What are the implications of annotation errors and how can they be addressed?

Annotation errors can propagate across multiple databases through the widespread use of information derived from available annotations. For example, 136 rat protein kinases had annotations transferred from mouse protein kinases based on homology with erroneously annotated mouse protein kinases. Of these, 94 labeled as "Ser/Thr" kinases by UniProt had AmiGO annotations of "Tyr" or "dual specificity" kinase, and 42 labeled as "Tyr" kinases by UniProt had AmiGO annotations of "Ser/Thr" or "dual specificity" .

To address these issues, researchers should:

• Cross-validate annotations from multiple databases

• Consider the evidence codes associated with annotations

• Compare annotations with orthologous proteins from other species

• Utilize machine learning approaches to predict and verify functional annotations

• Report potential errors to database curators for correction

How can mouse models inform personalized medicine approaches?

Genetically engineered mouse models are increasingly being used in preclinical guidance for human clinical trials. These models can accelerate discovery by enabling cross-species comparisons that help identify parameters key to developing personalized medicine .

The extensive genetic research in mice provides powerful strategies for assessing:

• Complex genetics of disease susceptibility

• Therapeutic response variations

• Associated toxicities that are diverse among human populations

While current technologies have accelerated human epidemiological association studies, cross-species comparisons using mouse models can significantly accelerate discovery of personalized medicine parameters .

Several initiatives are underway to use de novo mouse cancer models in preclinical guidance for human clinical trials. Although challenging due to the complexity of human diseases and the difficulty of modeling them accurately in mice, early studies indicate that this approach offers great promise, opening up a new era of translational research using engineered mouse models .

What biomarkers correlate with sensitivity to specific treatments in mouse models?

Identifying biomarkers that correlate with treatment sensitivity is a key advantage of testing multiple mouse models that represent diverse genetic backgrounds. In a study of PLX038A (a PEGylated SN-38 prodrug), biomarkers that correlated with model sensitivity included:

• Wild type TP53

• Mutant TP53 but with a mutation in 53BP1 (indicating a defect in DNA damage response)

Similarly, preclinical testing in large panels of adult melanoma xenografts has shown that BRAF mutant models respond to BRAF inhibitors, whereas those with wild type BRAF are less sensitive .

Using a larger number of models representative of a histotype may improve the prediction of activity in clinical trials by:

• Identifying genetic characteristics that segregate with drug sensitivity

• Potentially revealing subclassifications within a disease type

• Allowing for more accurate prediction of clinical outcomes across diverse patient populations

What technologies are available for manipulating the mouse genome for cancer research?

With the complete sequence of the mouse genome available, along with technology to manipulate it and well-defined inbred strains, researchers now have impressive capabilities to engineer mice for testing hypotheses of tumorigenesis. Experiments can be undertaken to assess outcomes when gene function is:

• Lost

• Mutated

• Underexpressed

• Overexpressed

These manipulations can be performed in the appropriate cell types to model specific cancer types or mechanisms .

Recent advances include:

• Conditional gene expression systems that control timing and location of mutations

• Single-cell mutation technologies

• In vivo imaging techniques for tracking tumor development

• Systems for assessing complex genetics of disease susceptibility

How can researchers ensure the translatability of mouse model findings to human disease?

Despite the differences between mice and humans, several strategies can enhance the translatability of mouse model findings:

• Use models that accurately represent the genetic and molecular diversity of human tumors

• Study tumors as they develop de novo in the context of a normal immune system

• Account for tumor-stroma interactions during development

• Select models that reflect the specific genetic alterations found in human cancers

• Validate findings through cross-species comparisons

• Consider the limitations of mouse models when interpreting results

The most effective translational research using engineered mouse models involves:

• Testing hypotheses generated from human tumor studies

• Identifying novel mechanisms to be confirmed in human tumors

• Using cross-species comparisons to accelerate discovery

• Applying findings to guide clinical trial design and patient stratification strategies