Icon Legend

This session is not in your schedule.

This session is in your schedule. Click again to remove it.

Bioinformatics, Computational and Systems Biology

Bioinformatics, Computational and Systems Biology - Poster Session C

Poster A6 - Incorporation of receptor recycling and asymmetric endosome distribution to improve performance of moving boundary model of epidermal growth factor receptor trafficking and signaling

Friday, October 25, 2024

10:00 AM – 11:00 AM EST

Location: Exhibit Hall E, F & G

Presenting Author(s)

Seth Boehringer (he/him/his)

Research Assistant
University of Virginia
Blacksburg, Virginia, United States

Co-Author(s)

Sung Hyun Lee

Graduate student
University of Virginia
Charlottesville, Virginia, United States

Primary Investigator(s)

ML

Matthew J. Lazzara

Professor
University of Virginia, United States

Introduction: Endocytic trafficking of receptors plays a key role in cell signaling by regulating the number of receptors available for activation at the plasma membrane. Endocytosed receptors sustain downstream signaling by interacting with diffusible cytosolic species and function as remote activators for plasma membrane-retained signalers (Sorkin,2023). Defects in receptor endocytosis frequently arise in cancer causing aberrant cell signaling (Huang,1997). Despite its clear importance, the impact of endocytosis on downstream signaling is not well understood. An unresolved question is whether natural variation in the number and position of endosomes among individual cells impacts heterogeneous pathway activation. To answer this, we built a novel reaction-diffusion model with moving endosomes based on the epidermal growth factor receptor (EGFR) – extracellular signal regulated kinase (ERK) pathway. A provisional version of the model predicted cells with more EGFR-containing endosomes have higher RAS/ERK activity by promoting a more homogeneous distribution of active Src-family kinases (SFKs) that connect endocytosed EGFR with RAS via reaction-diffusion processes. We anticipated the impact of the number of endosomes on RAS to also stem from reaction-diffusion processes involving GAB1-SHP2, a phosphatase required for complete ERK activation (Ahmed,2019). However, RAS activity was insensitive to changes in the GAB1-SHP2 diffusivity (DG1S2) in the provisional model. To quantify the distribution of cytosolic species, we expanded our model analysis pipeline to calculate their spatial variance. To improve model accuracy, we added rate processes and parameters describing the recycling of endocytosed EGFR to the plasma membrane and features allowing for asymmetric distribution of endosomes within the cell.

Materials and

Methods: The first version of the moving-boundary model was composed of 31 ordinary and 20 partial differential equations in MATLAB. The model simulates the activation of the RAS-ERK pathway by EGF binding to EGFR causing an assembly of cytosolic components at the plasma membrane. This model is novel in its consideration of endosomes as moving compartments within the cell. The reaction network is characterized by 72 parameters consisting of kinetic rate constants and diffusion coefficients for cytosolic species. We used the particleswarm global optimization method to minimize the cost function for the model parameter fits. To test the effects of varying numbers of endosomes on spatial distributions of cytosolic species’ activities, we created a MATLAB script that examines the activity of signaling species as a function of the spatial variance of cytosolic species. The extended model that includes EGFR recycling and degradation was created by adding 3 rate processes and parameters that describe a fraction of EGFR returning to the plasma membrane by diffusion and a fraction of EGFR degrading within the lysosome. EGFR recycling was implemented as a diffusive process rather than convective due to programmatic restrictions. The new parameters in the extended model were refitted. To model asymmetric endosome distribution about the cell center, we created an endosome coordinate library. This library stores (x,y) coordinates describing a specific position around the plasma membrane for each endosome. This is an updated version of the previous scripts, which evenly dispersed the endosomes around the membrane without specifying a position.

Results, Conclusions, and Discussions: Comparison of the original and extended models revealed that RAS activity marginally increased in the extended. Surprisingly, RAF activity, indicated by concentrations of membrane-bound RAF1-bound RAS, and total pERK molecules substantially increased revealing greater flux of the EGFR-ERK pathway. (Fig. 1A,B). This suggests that receptor recycling plays a role in ERK pathway activation. The extended model predicts that spatial variance of active SFKs is positively correlated with RAS activity yet negatively correlated to the number of endosomes (Fig. 1C,D). This conflicts with previous model predictions but aligns with immunofluorescence data that revealed an inverse relationship between variance in phosphorylated Src intensity and the number of endosomes. Local sensitivity analysis revealed that the addition of recycling parameters increased RAS sensitivity to DG1S2 perturbations (Fig. 1E), leading to the expected increase in ERK activity. Increased sensitivity of RAS to DG1S2 brings our model closer to biological accuracy, as GAB1-SHP2 binding to active receptors at the plasma membrane is a critical step for RAS activation (Montagner,2005). The new distribution function allows for changes to endosome position around the plasma membrane but is limited in that the endosomes cannot progress towards the cell center. Analysis of active SFK distribution revealed variations in SFK activity at the plasma membrane (Fig. 1F). Reflecting this attribute in our model may aid in lowering the cost of model fit and increase accuracy as immunofluorescence images of carcinoma cells previously displayed asymmetric endosome crowding (Myers,2023). The effect of endosome distribution on cytosolic signal activity is therefore a point of interest in our continued investigations.

Acknowledgements (Optional): The authors would like to acknowledge: Paul J. Myers (University of Virginia), Masoud Nickaeen (University of Virginia), Boris M. Slepchenko (University of Virginia), Leslie M. Loew (University of Virginia), Kevin S. Brown (Oregon State University) and Alexander Sorkin (University of Pittsburgh School of Medicine) for development and assistance with the moving boundary solver. Thank you to the SASCO Center for providing project funding made possible by the National Cancer Institute U54 grant -- NCI U54 CA274499. Thank you to the Systems Bioengineering REU led by Dr. Timothy Allen and Karen Sleezer and funded by (NSF) REU award 1950374.