Skip to content

CD3×5T4 In Vitro Activation and Cytotoxicity QSP

Fixed-effect in vitro model linking CD3×5T4 immune-synapse formation to T-cell activation, differentiation, and delayed tumor-cell killing.

An In Vitro Quantitative Systems Pharmacology Platform for Characterizing CD3-Bispecific Antibody-Mediated T-Cell Activation and Tumor Cell Cytotoxicity ↗

The AAPS Journal

Exploration notes · What are you exploring?

Notes stay on this browser.

Explore this model

Choose a starting point to view its result. Adjust key model inputs when you want to explore a different outcome.

MDA-MB-468 untreated, E:T 4:1

Null-control culture with 16,000 tumor cells and 64,000 resting T cells, matching the referenced assay setup.

Observation window

0–72 hour

Starting configuration

Study readouts

Immune-synapse trimers per T cell· trimers/T cellActivated or effector T cells· %Effector T cells· cells/wellTumor-cell cytotoxicity· %Viable and apoptotic tumor cells· cells/wellExploratory
Study purpose and interpretation limits

What this study explores

  • Does the no-drug culture preserve zero trimer, activation, effector differentiation, and drug-attributable cytotoxicity?

Interpretation limits

  • This is an in vitro null control, not a clinical no-treatment prediction.
  • Natural T-cell loss and tumor proliferation remain active.
  • Zero drug-attributable killing is evaluated against the parallel untreated tumor state.

No starting example is available for this configuration.

Edit inputs

Binding

DuoBody-CD3x5T4 value; the high-affinity bsIgG1 variant is 16 nM. Values between the endpoints are exploratory and values outside them are outside this runtime contract.

nM

Suggested: 341.5–1,025. You can explore beyond this range.

683 → —
Estimated population effect

MDA-MB-468 cell-line group 3 value.

molecule/cell

Suggested: 1.065–3.195. You can explore beyond this range.

2.13 → —

Upper rate of trimer-dependent resting-to-activated T-cell conversion in the fixed-effect model.

1/hr

Suggested: 0.0053–0.0159. You can explore beyond this range.

0.0106 → —

Aggregate trimer load across effector T cells producing half-maximal per-effector killing drive.

molecule

Suggested: 3,549–1.06e+4. You can explore beyond this range.

7098 → —

MDA-MB-468 cell-line group 4 and no-donor/donor group 2 value.

1/(hour*cell)

Suggested: 6.50e-6–1.95e-5. You can explore beyond this range.

0.000013 → —
Simulation time
hour
hour
72 → 168

Time between reported samples.

hour

1 change · Variation 1

Biology

Links per-cell CD3×5T4 trimer formation to resting-to-activated-to-effector T-cell dynamics and three-stage apoptotic tumor-cell loss across drug affinity, concentration, antigen density, and effector-to-target ratio.

A CD3×5T4 bispecific antibody bridges a resting T cell and a 5T4-positive tumor cell. Per-cell trimers activate and expand T cells, which differentiate into effectors. Effector cells push proliferating tumor cells through three apoptotic transit stages, producing activation and cytotoxicity readouts.

Assumptions & evidence

Assumptions & limitations

  • The authors did not deposit the pooled 48 activation and 1525 cytotoxicity observations.
  • The population-modeling environment project, culture-level random effects, and residual-model input rows were not deposited.
  • Population fixed effects and disclosed covariate-group values are implemented; stochastic between-culture variability is not.
  • Paper figures do not provide machine-readable coordinates, so curated scenarios are source-aligned mechanistic probes rather than strict figure reproductions.
  • This is an in vitro assay model and must not be interpreted as a clinical dose, safety, or efficacy model.
  • This is an in vitro null control, not a clinical no-treatment prediction.
  • Natural T-cell loss and tumor proliferation remain active.
  • Zero drug-attributable killing is evaluated against the parallel untreated tumor state.
  • The trajectory uses reported population fixed effects and is not an individual-culture prediction.
  • Raster-digitized Figure 2, 3, 6, and 7d comparators support private pointwise reproduction, but repeated figure renderings are non-independent and do not replace source arrays or culture-level donor records.
  • This assay concentration must not be converted into a clinical dose recommendation.
  • Compare this condition with the E:T 4:1 and 8:1 scenarios as an assay-design sensitivity.
  • The model's R_ET factor makes per-cell synapse kinetics depend dynamically on both T-cell and tumor-cell pools.
  • The result is not a clinical immune-cell infiltration prediction.
  • Interpret per-cell trimer and aggregate trimer load together.
  • This is a source-aligned assay condition, not a patient-level tumor microenvironment.
  • Raster-digitized Figure 5, 6, and 7d endpoint comparators support private pointwise reproduction, but they are repeated renderings of one assay observation rather than independent validations.
  • Compare this condition with the E:T 1:1, 4:1, and 8:1 scenarios.
  • The result is an in vitro fixed-effect assay reconstruction, not a patient immune-infiltration prediction.
  • The published concentration grid was recovered by figure alignment and remains explicitly provenance-labeled.
  • The no-donor category maps to donor group 2 as disclosed in Supplement Table S4.
  • This comparison changes several measured or estimated cell-line attributes and is not a single-parameter causal attribution.
  • No tumor-type clinical efficacy inference is supported.
  • The 16 nM 3D and 174 molecules/cell 2D KD values are paired measured endpoint values.
  • Affinity can alter both potency and hook-region behavior; do not assume monotonic benefit at every concentration.
  • This is candidate-design exploration in an in vitro model, not clinical optimization.
  • Supplement Table S2 confirms the condition family, but the requested figures do not expose its exact observed endpoint at 6.86 nM.
  • The no-donor category maps to donor group 2.
  • Treat the trajectory as source-aligned and source-limited, not pointwise paper-validated.
  • The cell-line bundle changes antigen, growth, activation threshold, and killing susceptibility together.
  • The no-donor fixed effect is donor group 2.
  • Do not infer clinical pancreatic-tumor efficacy.
  • The paper uses both EPLC-272H and EPLC272H spellings; this scenario uses the Table S1 spelling.
  • Do not attribute the response to antigen abundance alone.
  • The cell-line killing group is a fitted categorical covariate and is not derived from 5T4 abundance.
  • Do not infer clinical esophageal-tumor efficacy.

Parameter origins

  • Released executable model definition; exact publication locator not reported
  • Yuan 2026 Supplement Table S5
  • Yuan 2026 Table II population estimate
  • Yuan 2026 Table I measured construct endpoints
  • Yuan 2026 Supplement Tables S4-S5

Validation & references

The implementation has validation evidence, but unavailable source material limits full reproduction.

An In Vitro Quantitative Systems Pharmacology Platform for Characterizing CD3-Bispecific Antibody-Mediated T-Cell Activation and Tumor Cell Cytotoxicity ↗

Available reference curves can be added to the plot when their scenario, readout, time and units match.

Model details

33 parameters · 17 states · 17 equations · 18 derived outputs

Classification & API access
Therapeutic
DuoBody-CD3x5T4, bsIgG1-CD3x5T4
Modality
Bispecific antibody / T-cell engager
Target
CD3, 5T4
Disease / scope
In vitro solid-tumor cell-line assay (not disease-specific)
ParameterDefaultUnitDescription
Activated t cell doubling timeDT_Ta_hr12hrActivated t cell doubling time coefficient used by the released system of model equations.
Tumor cell doubling timeDT_tumor_hr40hrMDA-MB-468 default.
Trimer threshold for half-maximal activationEC50_activation_trimers_per_tcell2.13molecule/cellMDA-MB-468 cell-line group 3 value.
Generation count for half maximal differentiationEC50_differentiation_generations3dimensionlessGeneration count for half maximal differentiation at which the corresponding modeled effect reaches half of its maximum.
Aggregate trimer threshold for half-maximal killingEC50_killing_total_trimers7098moleculeAggregate trimer load across effector T cells producing half-maximal per-effector killing drive.
T cell activation hill coefficientGamma_activation0.623dimensionlessT cell activation hill coefficient coefficient used by the released system of model equations.
Tumor killing hill coefficientGamma_killing1.63dimensionlessTumor killing hill coefficient governing loss, degradation, or removal in the modeled biological process.
Activated t cell half-lifeHL_Ta_hr48hrActivated t cell half-life coefficient used by the released system of model equations.
Effector t cell half-lifeHL_Te_hr48hrEffector t cell half-life coefficient used by the released system of model equations.
Resting t cell half-lifeHL_Tr_hr168hrResting t cell half-life coefficient used by the released system of model equations.
CD3-arm 2D dissociation constant — high-affinity endpointKD_CD3_2D_high_molecules_per_cell11088molecules/cellCD3-arm 2D dissociation constant — high-affinity endpoint describing equilibrium binding affinity in the modeled interaction.
CD3-arm 2D dissociation constant — low-affinity endpointKD_CD3_2D_low_molecules_per_cell174molecules/cellCD3-arm 2D dissociation constant — low-affinity endpoint describing equilibrium binding affinity in the modeled interaction.
Measured high CD3-arm 3D dissociation endpointKD_CD3_3D_high_nM683nMMeasured high CD3-arm 3D dissociation endpoint describing equilibrium binding affinity in the modeled interaction.
Measured low CD3-arm 3D dissociation endpointKD_CD3_3D_low_nM16nMMeasured low CD3-arm 3D dissociation endpoint describing equilibrium binding affinity in the modeled interaction.
CD3-arm 3D dissociation constantKD_CD3_3D_nM683nMDuoBody-CD3x5T4 value; the high-affinity bsIgG1 variant is 16 nM. Values between the endpoints are exploratory and values outside them are outside this runtime contract.
5T4-arm 2D dissociation constantKD_TAA_2D_molecules_per_cell27molecules/cell5T4-arm 2D dissociation constant describing equilibrium binding affinity in the modeled interaction.
5T4-arm 3D dissociation constantKD_TAA_3D_nM3nM5T4-arm 3D dissociation constant describing equilibrium binding affinity in the modeled interaction.
Maximum T-cell activation rateMax_activation_per_hr0.01061/hrUpper rate of trimer-dependent resting-to-activated T-cell conversion in the fixed-effect model.
Maximum t cell differentiation rateMax_differentiation_per_hr0.2681/hrMaximum t cell differentiation rate coefficient used by the released system of model equations.
Maximum per-effector tumor killing rateMax_kill_per_effector_hr0.0000131/(hour*cell)MDA-MB-468 cell-line group 4 and no-donor/donor group 2 value.
Baseline CD3 abundancebase_CD3_molecules_per_tcell30000molecule/cellBaseline CD3 abundance coefficient used by the released system of model equations.
Baseline 5 T4 abundancebase_TAA_molecules_per_tumor31582molecule/cellMDA-MB-468 default from Supplement Table S1.
CD3 side complex internalization rateint_CD3_per_hr0.0661/hrCD3 side complex internalization rate coefficient used by the released system of model equations.
5 T4 side complex internalization rateint_TAA_per_hr0.0661/hr5 T4 side complex internalization rate coefficient used by the released system of model equations.
CD3 degradation ratekdeg_CD3_per_hr0.0661/hrCD3 degradation rate governing loss, degradation, or removal in the modeled biological process.
5 T4 degradation ratekdeg_TAA_per_hr0.0661/hr5 T4 degradation rate governing loss, degradation, or removal in the modeled biological process.
CD3 dissociation ratekoff_CD3_per_hr721/hrCD3 dissociation rate governing molecular dissociation in the binding model.
5 T4 dissociation ratekoff_TAA_per_hr3.7741/hr5 T4 dissociation rate governing molecular dissociation in the binding model.
Apoptosis transit ratektr_apoptosis_per_hr11/hrApoptosis transit rate coefficient used by the released system of model equations.
Molecule to medium concentration conversionmolecule_to_nM1.1e-11nM/moleculeMolecule to medium concentration conversion coefficient used by the released system of model equations.
Immune synapse areasynapse_area_um212.56um^2Immune synapse area coefficient used by the released system of model equations.
T cell radiustcell_radius_um3.5umT cell radius coefficient used by the released system of model equations.
Tumor cell radiustumor_radius_um10umTumor cell radius coefficient used by the released system of model equations.

Related models

Feedback: helpdesk@unibiointelligence.com.

Ubi Biologics