Inference of weak-form partial differential equations describing migration and proliferation mechanisms in wound healing experiments on cancer cells

Patrick C. Kinnunen,Siddhartha Srivastava,Zhenlin Wang,Kenneth K.Y. Ho,Brock A. Humphries,Siyi Chen,Jennifer J. Linderman,Gary D. Luker,Kathryn E. Luker,Krishna Garikipati

2024-10-24

Abstract:Targeting signaling pathways that drive cancer cell migration or proliferation is a common therapeutic approach. A popular experimental technique, the scratch assay, measures the migration and proliferation-driven cell closure of a defect in a confluent cell monolayer. These assays do not measure dynamic effects. To improve analysis of scratch assays, we combine high-throughput scratch assays, video microscopy, and system identification to infer partial differential equation (PDE) models of cell migration and proliferation. We capture the evolution of cell density fields over time using live cell microscopy and automated image processing. We employ weak form-based system identification techniques for cell density dynamics modeled with first-order kinetics of advection-diffusion-reaction systems. We present a comparison of our methods to results obtained using traditional inference approaches on previously analyzed 1-dimensional scratch assay data. We demonstrate the application of this pipeline on high throughput 2-dimensional scratch assays and find that low levels of trametinib inhibit wound closure primarily by decreasing random cell migration by approximately 20%. Our integrated experimental and computational pipeline can be adapted for quantitatively inferring the effect of biological perturbations on cell migration and proliferation in various cell lines.

Cell Behavior

What problem does this paper attempt to address?

### Problems the paper attempts to solve

The paper aims to derive a weakly - formulated partial differential equation (PDE) that describes the mechanisms of cancer cell migration and proliferation by combining high - throughput scratch assays, video microscopy, and system identification techniques. Specifically, the paper attempts to solve the following problems:

1. **Enhancing the analytical ability of scratch assays**:
- Scratch assays are a commonly used experimental technique for measuring how defects in a cell monolayer close under the action of cell migration and proliferation. However, traditional scratch assays are unable to measure dynamic effects. The paper improves the analytical ability of scratch assay data by introducing weakly - formulated system identification techniques.

2. **Deriving a PDE model of cell - density dynamics**:
- The paper uses live - cell microscopy and automated image - processing techniques to capture the evolution of the cell - density field over time. Then, using weakly - formulated system identification techniques, a first - order kinetic PDE model that describes cell migration and proliferation is derived from these data.

3. **Validating the effectiveness of the method**:
- The paper compares the results of the proposed method with those of traditional inference methods and validates the effectiveness and accuracy of the new method in handling previously analyzed one - dimensional scratch assay data.

4. **Applying to high - throughput two - dimensional scratch assays**:
- The paper demonstrates the application of this method in high - throughput two - dimensional scratch assays and finds that low - dose trametinib inhibits wound closure mainly by reducing random cell migration by approximately 20%.

5. **Quantifying the impact of biological perturbations on cell migration and proliferation**:
- The paper proposes an integrated experimental and computational pipeline that can be adapted to quantitatively infer the specific impacts of biological perturbations on cell migration and proliferation in different cell lines.

### Main research contents

1. **Experimental design**:
- Scratch assays are carried out using MDA - MB - 231 breast cancer cells, and the change in the cell - density field over time is extracted by fluorescence microscopy and automated image - processing techniques.

2. **Data processing**:
- The influence of numerical noise on derivative calculations is reduced through smoothing in space and time.

3. **Model derivation**:
- A PDE model that describes cell migration and proliferation is derived from experimental data using variational system identification techniques. The model includes a diffusion term (random motion), a convection term (directed motion), and a reaction term (cell proliferation or death).

4. **Parameter optimization**:
- Model parameters are further optimized through PDE - constrained optimization techniques to better fit the experimental data.

### Results

1. **Variational system identification of one - dimensional scratch assay data**:
- Variational system identification was carried out on scratch assay data under different initial cell densities. The results show that the three - term model can fit the data well in most cases and the convection term is excluded.

2. **PDE - constrained optimization**:
- By performing PDE - constrained optimization on the three - term model, the fitting effect of the model is further improved. In particular, in the case of 18,000 cells, after expanding to a four - term model, the cost function is significantly reduced.

3. **Application to high - throughput two - dimensional scratch assays**:
- Applying this method to high - throughput two - dimensional scratch assays, it is found that low - dose trametinib inhibits wound closure mainly by reducing random cell migration.

### Conclusions

By combining high - throughput scratch assays, video microscopy, and variational system identification techniques, the paper successfully derives a PDE model that describes the mechanisms of cancer cell migration and proliferation. This method not only improves the analytical ability of scratch assay data but also can quantify the specific impacts of biological perturbations on cell behavior.