BRM5013 Chap.8 Cancer Models and Genome Engineering
Cancer Models and Genome Engineering
Cancer models reveal different dependencies depending on scale, architecture, immune context and systemic exposure. This chapter compares patient-derived cells, organoids, xenografts and engineered animal models, then uses CRISPR perturbation, reciprocal correction and rescue to separate association, clonal effects and causal mechanisms.
What this chapter covers
- 01
Patient-derived cells and organoids
- 02
Xenografts and engineered animals
- 03
Scale versus physiological context
- 04
Guide targeting and nuclease logic
- 05
Off-target and clonal effects
- 06
Reciprocal editing and rescue
Worked example · free
Test a candidate dependency
- +1Use independent guides and confirm target loss.
- +1Measure baseline fitness and a relevant phenotype.
- +1Restore a guide-resistant version of the target.
- +1Repeat in another relevant model and comparison cell.
Key terms
- Patient-derived model
- An experimental system established from patient material to preserve relevant disease features.
- Guide RNA
- An RNA sequence that directs a CRISPR-associated system to a selected genomic target.
- Reciprocal edit
- A complementary genetic change used to test the direction and specificity of a phenotype.
Cancer Models and Genome Engineering FAQ
Which cancer model is best?
There is no best model in the abstract. Cell models favour scale and direct perturbation, organoids add architecture, and animal models add systemic context. Choose the least complex system that contains the mechanism, then name the biological layer still missing.
How does rescue strengthen a CRISPR result?
If restoring the target reverses the phenotype while an appropriate control does not, the result becomes more specific to that target. Rescue expression must still be verified and kept within a biologically reasonable range.
Can editing a variant establish clinical pathogenicity?
Editing can demonstrate a functional effect in a defined model, especially with reciprocal correction and independent replication. Clinical pathogenicity also depends on human evidence, penetrance and broader interpretation, so the experimental result should not be stretched beyond its scope. Verify the intended change and its molecular consequence before interpreting phenotype.
Compare independently derived clones and matched workflow controls so clonal variation or editing stress does not become the hidden explanation. Reciprocal correction, introduction of the variant and a complementary perturbation can provide convergent causal evidence. Even then, the finding belongs to the tested cell state, exposure and endpoint.
Human segregation, population frequency, clinical presentation and other evidence determine how that functional result contributes to pathogenicity assessment. A mechanistic model can strengthen interpretation without representing the full variability of disease in patients.
Assessment move
Revise cancer systems as a matrix of scale, microenvironment, immune context, time and best use. For genome-engineering questions, move from guide design to editing confirmation, phenotype, rescue and cross-model validation. End with the context that the chosen system does not contain. Build columns for established lines, patient-derived cultures or organoids and in vivo models.
Under each, record what can be controlled, what disease context is retained, what may be selected during establishment and which inference remains unsafe. Do not rank the models globally. Instead, take a question about target mechanism, drug penetration, stromal interaction or systemic response and decide which biological layer must be present.
Practise explaining why adding complexity can improve relevance while also introducing variability and confounding. For editing, draw a verification path that begins before phenotype measurement. State the intended genomic change, confirm it at the relevant locus, check the molecular consequence and compare independently derived clones with a matched workflow control.
Then ask whether an independent reagent or perturbation reproduces the result. A rescue is most persuasive when it restores the disrupted element without introducing a new difference that could explain the outcome. Keep an off-target check and clonal variation as competing explanations until replication addresses them. Work through response scenarios by separating exposure, target engagement and phenotype.
If an organoid appears resistant, consider penetration, composition, viability and growth before assigning a pathway mechanism. If an edited culture changes proliferation, confirm that the change is not a general consequence of editing stress. Practise writing conclusions at three strengths: association, causal contribution and sufficient mechanism.
Select the strongest one justified by perturbation and rescue rather than by how dramatic the image appears. For cross-model validation, explain what a second system contributes. Repeating the same assay in another line can test generality; moving to a system with immune or stromal context can test a missing mechanism.
Finish revision by taking one result and drawing its evidence boundary around model, genotype, exposure, endpoint and time. Anything outside that boundary becomes a specific next experiment, not an unsupported extension to all patients or all tumours.
Working through Cancer Models and Genome Engineering in BRM5013? Sia is AskSia’s AI Biology tutor — ask any BRM5013 Cancer Models and Genome Engineering question and get a clear, step-by-step explanation grounded in how BRM5013 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
