In APOPTO-SENSE 2.0, we applied synthetic biology engineering principles to design a modular biosensing system for detecting chemotherapy-induced apoptosis in cancer cells. Following the Design-Build-Test-Learn (DBTL) cycle, we engineered a dual-cell platform using synthetic Notch (synNotch) receptors and orthogonal transcription factors to convert phosphatidylserine (PS) exposure—a key apoptosis marker—into quantifiable fluorescent signals. This addresses the need for rapid, personalized drug sensitivity testing in oncology.
Our engineering process emphasized modularity, orthogonality, and dose-dependency, drawing from established tools like synNotch (previously used in iGEM for antigen sensing) and Gal4-UAS systems. We iterated through multiple DBTL cycles to optimize specificity, linearity, and robustness, starting with a linear amplification MVP and planning expansions for multiplexed cell death detection. For project context, see Description; for protocols, see Experiments.
We designed APOPTO-SENSE 2.0 as a composable system with two core modules: a sensor for apoptosis detection and a reporter for signal amplification. Key principles included:
Initial Design: Fused Annexin V (extracellular domain for PS binding) to a Notch transmembrane core and Gal4-VP64 intracellular payload. This creates a "plug-and-play" receptor activated by PS exposure on apoptotic cells.
Rationale: Annexin V's high affinity (Kd ~1-10 nM) and calcium-dependency ensure specificity to apoptotic membranes. Gal4-VP64 provides orthogonal nuclear translocation upon cleavage.
Engineering Considerations: We selected a minimal Notch TMD to minimize steric hindrance and ensured the construct fits within PiggyBac vectors for stable integration.
Initial Design: Gal4-VP64 binds 5x UAS repeats upstream of a minimal promoter driving TagBFP expression. Constitutive PGK-mCherry serves as an internal reference.
Rationale: UAS multiplicity amplifies signal linearly (output proportional to input), avoiding bistable "all-or-nothing" responses from positive feedback. TagBFP/mCherry ratio normalizes for variables like cell number.
Engineering Considerations: Minimal promoter reduces basal expression (<5% of induced levels); spectral separation (TagBFP: 390/450 nm; mCherry: 587/610 nm) enables ratiometric quantification.
Design: Assemble synNotch and reporter plasmids via Gibson cloning, targeting HEK293T cells for engineering.
Build: Synthesized genes, amplified in E. coli, and transfected using PiggyBac for stable lines (see Experiments for protocols).
Test: Induced apoptosis in HL-60 cells with raphasatin (10 µM) and co-cultured with sensor cells. Measured TagBFP/mCherry ratio via fluorescence microscopy.
Learn: Low activation observed (ratio ~0.3 vs. expected >1.0). Hypothesized: Insufficient UAS repeats or suboptimal transfection efficiency. No off-target activation in controls (specificity confirmed).
Design: Increased UAS repeats from 5x to 9x for stronger amplification; optimized transfection reagent ratio (2:1 DNA:lipid).
Build: Re-assembled plasmids; verified via sequencing and gel electrophoresis.
Test: Repeated co-culture assays with serial dilutions of apoptotic HL-60 (10^4-10^6 cells). Quantified linearity (R² >0.85) and LOD (~5% apoptotic fraction).
Learn: Improved ratio to >1.5 in apoptotic samples, but variability in expression (CV ~20%). Identified issue: Transient transfection instability; switched to stable lines for consistency.
Design: Added EDTA controls to block Ca2+-dependent Annexin V binding, confirming PS specificity. Incorporated constitutive mCherry for normalization.
Build: Generated polyclonal stable HEK293T lines with dual selection (puromycin/blasticidin).
Test: Co-cultured with non-apoptotic cells or PS-masked apoptotic cells. Assessed crosstalk via flow cytometry (no activation in negatives).
Learn: Achieved high specificity (>95% signal reduction in controls) and precision (CV <10%). However, extended co-culture (>24h) led to sensor cell stress; optimized to 15h.
For full results, see Results.
Engineering success was validated through:
This iterative engineering process demonstrates how DBTL cycles transformed a conceptual design into a functional biosensor, advancing synthetic biology in oncology.