Dry Lab

Models

Iterative Design-Build-Test-Learn cycles that transformed TRACER from concept to reality

Computational Framework

In Silico Model Development

Computational design guided every stage of TRACER development, from concept to implementation

Key Design Transition: From Dual to Single SynNotch

  • ▸Dual cascade introduced excessive complexity and amplified noise
  • ▸Quantified false positives from basal activity and trafficking kinetics
  • ▸Single SynNotch achieved same goal with cleaner threshold behavior

Initial Concept & Modelling

Built comprehensive in-silico framework to model SynNotch pathway and reduce trial-and-error

●Cell-based biosensing for CD147 overexpression detection
●Threshold-dependent SynNotch activation system
●In-silico modeling before wet-lab testing
●Explored design options and anticipated failure points

Leak Suppression & Promoters

Designed constitutive inhibitor to block unintended activation until threshold signaling

●Addressed basal 'leaky' activation in SynNotch systems
●Modeled 6 constitutive promoters: CAG, CMV, EF1α, U6, PGK, SFFV
●Predicted expression strengths before construct synthesis
●Optimized stability across mammalian systems

Reporter Selection

Humanized Gaussia luciferase (GLuc) selected for non-invasive monitoring

●Minimal immunogenicity with rapid renal clearance (10-20 min)
●Previously validated in animal models
●Predicted secretion rates via computational framework
●Matched detection thresholds to desired assay window

Quantitative Predictions

Parameter scanning identified optimal promoter ranges for fine-tuned behavior

●SFFV: 295.6 mM | EF1α: 197.1 mM | CMV: 172.5 mM
●CAG: 118.3 mM | PGK: 19.28 mM | U6: 13.9 mM
●Explored dose-response curves and activation thresholds
●Adjusted design parameters virtually before synthesis

Structural Optimization

AlphaFold modeling guided epitope tag placement without wet-lab iteration

●C-terminal 3×FLAG compromised TetR interaction
●N-terminal tag preserved binding affinity
●Final design: N-terminal 3×FLAG-inhibitor fusion
●Avoided costly validation cycles

Dry-Lab Refinement

Continued in-silico refinement during customs delays shortened experimental cycle

●All 6 inhibitor variants + SynNotch + reporter finalized
●Tested combinations during synthesis delays
●Clear parameters ready for wet-lab testing
●Effectively shortened experimental timeline

All modeling predictions were validated against literature benchmarks and refined iteratively

Mathematical Models

Differential equations governing TRACER's synthetic biology dynamics

STEP 1 OF 3

Qualitative SynNotch–Inhibitor Model

Minimal model capturing SynNotch receptor dynamics with transcription factor production and negative feedback inhibition

SynTF Production
d[SynTF]dt=kprod−kL[SynTF]−k−L[LSynTF]−kact[SynTF]−γS[SynTF]\frac{d[\text{SynTF}]}{dt} = k_{\text{prod}} - k_{L}[\text{SynTF}] - k_{-L}[\text{LSynTF}] - k_{\text{act}}[\text{SynTF}] - \gamma_{S}[\text{SynTF}]
  • ▸SynTF produced at constant rate (k_prod)
  • ▸Can bind ligand to form LSynTF (k_L)
  • ▸Activates TF downstream (k_act)
  • ▸Degrades naturally (γ_S)
Ligand-Bound SynTF
d[LSynTF]dt=kL[SynTF]−k−L[LSynTF]−kactL[LSynTF]−γL[LSynTF]\frac{d[\text{LSynTF}]}{dt} = k_{L}[\text{SynTF}] - k_{-L}[\text{LSynTF}] - k_{\text{act}_{L}}[\text{LSynTF}] - \gamma_{L}[\text{LSynTF}]
  • ▸Forms from SynTF-ligand binding
  • ▸Activates TF with rate k_act_L
  • ▸Degrades naturally
Free Transcription Factor
KEY
d[TF]dt=kact[SynTF]+kactL[LSynTF]−kon[TF][I]+koff[TFI]−kdegTF[TF]−kbindDNA[TF][DNAfree]+kunbindDNA[DNAbound]\frac{d[\text{TF}]}{dt} = k_{\text{act}}[\text{SynTF}] + k_{\text{act}_{L}}[\text{LSynTF}] - k_{\text{on}}[\text{TF}][I] + k_{\text{off}}[\text{TF}_{I}] - k_{\text{deg}_{\text{TF}}}[\text{TF}] - k_{\text{bind}_{\text{DNA}}}[\text{TF}][\text{DNA}_{\text{free}}] + k_{\text{unbind}_{\text{DNA}}}[\text{DNA}_{\text{bound}}]
  • ▸Produced by activated SynTF/LSynTF
  • ▸Can bind inhibitor forming TF_I
  • ▸Can bind DNA to transcribe downstream genes
TF-Inhibitor Complex
d[TFI]dt=kon[TF][I]−koff[TFI]−kdegTFI[TFI]\frac{d[\text{TF}_{I}]}{dt} = k_{\text{on}}[\text{TF}][I] - k_{\text{off}}[\text{TF}_{I}] - k_{\text{deg}_{\text{TFI}}}[\text{TF}_{I}]
  • ▸Mass balance of sequestered TF
DNA Binding
d[DNAfree]dt=−kbindDNA[TF][DNAfree]+kunbindDNA[DNAbound]\frac{d[\text{DNA}_{\text{free}}]}{dt} = -k_{\text{bind}_{\text{DNA}}}[\text{TF}][\text{DNA}_{\text{free}}] + k_{\text{unbind}_{\text{DNA}}}[\text{DNA}_{\text{bound}}]
d[DNAbound]dt=kbindDNA[TF][DNAfree]−kunbindDNA[DNAbound]\frac{d[\text{DNA}_{\text{bound}}]}{dt} = k_{\text{bind}_{\text{DNA}}}[\text{TF}][\text{DNA}_{\text{free}}] - k_{\text{unbind}_{\text{DNA}}}[\text{DNA}_{\text{bound}}]
  • ▸TF binding to promoter sites
mRNA Transcription
d[mRNA1]dt=ktx[DNAbound]−kdegmRNA[mRNA1]\frac{d[\text{mRNA}_{1}]}{dt} = k_{\text{tx}}[\text{DNA}_{\text{bound}}] - k_{\text{deg}_{\text{mRNA}}}[\text{mRNA}_{1}]
d[mRNA2]dt=ktx2[mRNA1]−kdegmRNA2[mRNA2]\frac{d[\text{mRNA}_{2}]}{dt} = k_{\text{tx}_{2}}[\text{mRNA}_{1}] - k_{\text{deg}_{\text{mRNA}_{2}}}[\text{mRNA}_{2}]
  • ▸Cascade transcription proportional to DNA occupancy
Secondary TF & Inhibitor
d[TF2]dt=ktl2[mRNA2]−kdegTF2[TF2]\frac{d[\text{TF}_{2}]}{dt} = k_{\text{tl}_{2}}[\text{mRNA}_{2}] - k_{\text{deg}_{\text{TF}_{2}}}[\text{TF}_{2}]
d[I]dt=kprod−kon[TF][I]+koff[TFI]−kdegI[I]\frac{d[I]}{dt} = k_{\text{prod}} - k_{\text{on}}[\text{TF}][I] + k_{\text{off}}[\text{TF}_{I}] - k_{\text{deg}_{I}}[I]
  • ▸Translation of second TF
  • ▸Constitutive inhibitor production

Model Interpretation

  • ✓Captures feedback between SynTF, downstream TFs, and inhibitor
  • ✓DNA binding and mRNA production are included but simplified
  • ✓Ligand effects are not explicitly pulsed