Theory of cargo and membrane trafficking

Lionel Foret, Lutz Brusch, Frank JülicherarXiv 2014 · DOI · arXiv

TL;DR

This paper presents a theoretical framework that explains how the macroscopic distribution of cargo within a cell emerges from the stochastic, microscopic interactions of individual vesicles (endosomes). By applying kinetic equations typically used in physics to biological imaging data, the authors demonstrate that we can infer hidden cellular parameters—like vesicle fusion rates—using only a few snapshots of fixed cell samples.

What Problem Does This Paper Address?

Every eukaryotic cell is a logistical powerhouse, constantly taking in nutrients and signaling molecules (cargo) from its environment through a process called endocytosis. Once inside, this cargo is distributed across a dynamic network of hundreds of "early endosomes" (EEs)—small membrane-bound sacs that constantly fuse together and split apart.

The central mystery is one of emergent properties: how does a system of hundreds of independent, noisy, and seemingly random vesicles manage to efficiently concentrate cargo for delivery to specific destinations? Biologists observe that cargo distribution often follows a "power law" (meaning a few vesicles carry a massive amount of cargo while most carry very little), but linking this broad pattern back to the specific molecular "rules" of vesicle interaction has historically been difficult.

Methods

The authors tackle this by treating the endosomal network as a statistical system. They identify an exhaustive list of six basic microscopic processes that can change cargo distribution:

  1. Biogenesis: Creation of new endosomes.
  2. Heterotypic Fusion: An endosome absorbing a small uptake vesicle.
  3. Homotypic Fusion: Two endosomes merging into one.
  4. Homotypic Fission: One endosome splitting into two similar ones.
  5. Heterotypic Fission: An endosome budding off a tiny recycling vesicle.
  6. Conversion/Maturation: An endosome exiting the "early" population to become a "late" endosome.

They translate these rules into a kinetic governing equation (similar to the Smoluchowski coagulation equation used in physics). To validate the theory, they use two approaches:

  • Stochastic Simulations: Using the Gillespie algorithm to model the random "walk" of individual vesicles.
  • Analytical Modeling: Deriving a simplified "entry-fusion-exit" model to see if it can predict the cargo distributions seen in real-world experiments.

Datasets

The theory was tested against experimental data from two primary sources:

  • In Vivo: Fluorescence microscopy images of mouse liver sections, where researchers tracked early endosomes in hepatocytes.
  • In Vitro: HeLa cell cultures internalizing labeled Low-Density Lipoprotein (LDL) cargo. These experiments provided time-course data, measuring cargo amounts at intervals from 3 to 60 minutes.

Key Insights and Ideas

  • Self-Organization from Noise: The "power law" distribution of cargo isn't explicitly programmed into the cell; it is a self-organized property that emerges naturally from repeated rounds of fusion and fission.
  • The "Entry-Fusion-Exit" Model: For certain types of cargo like LDL, a surprisingly simple model—where vesicles are born, fuse at a constant rate, and eventually "mature" out of the system—is enough to perfectly describe the complex data.
  • Data Collapse (Self-Similarity): One of the most striking "aha" moments is the discovery of self-similarity. When the authors rescaled cargo distributions from different time points, the curves "collapsed" onto a single universal curve ($x^{-3/2} \exp(-x)$). This proves that the underlying physical mechanism remains constant even as the network grows.
  • Geometric Sorting: The paper highlights a clever physical trick: when two spherical vesicles fuse, their volume increases faster than their surface area. This creates "excess" membrane, which the cell uses to bud off recycling tubules. This "geometric sorting" allows the cell to concentrate degradative cargo in the bulk volume while recycling receptors from the surface.

Key Results

  • Predictive Accuracy: The mathematical model (Equation 3) accurately predicted the growth and saturation of the total number of endosomes over time.
  • Parameter Inference: By fitting the model to experimental data, the authors were able to estimate microscopic kinetic rates (like the fusion rate constant $K$) that are nearly impossible to measure directly under a microscope.
  • Robustness: The theory held true across different concentrations of LDL influx, showing that the endosomal network's organizational logic is robust to changes in external "traffic" levels.

Limitations and Future Work

The primary limitation acknowledged is that the current model treats cargo like LDL as "passive"—it goes along for the ride but doesn't change the behavior of the vesicles. However, many signaling molecules (like EGF) are "active" and can actually recruit proteins to speed up or slow down fusion.

Future directions include: * Active Cargo Modeling: Adjusting the rate functions to account for cargo that "talks back" to the cellular machinery. * Crosstalk: Using the theory to understand how different signaling pathways interact within the same endosomal network. * In Vivo Diagnostics: Since the model works with fixed samples (snapshots), it could potentially be used to diagnose transport defects in tissues where live-cell imaging is impossible.

Summary updated

Abstract

Endocytosis underlies many cellular functions including signaling and nutrient uptake. The endocytosed cargo gets redistributed across a dynamic network of endosomes undergoing fusion and fission. Here, a theoretical approach is reviewed which can explain how the microscopic properties of endosome interactions cause the emergent macroscopic properties of cargo trafficking in the endosomal network. Predictions by the theory have been tested experimentally and include the inference of dependencies and parameter values of the microscopic processes. This theory could also be used to infer mechanisms of signal-trafficking crosstalk. It is applicable to in vivo systems since fixed samples at few time points suffice as input data.
BibTeX
@misc{foret2014theory,
  author = {Lionel Foret and Lutz Brusch and Frank Jülicher},
  title = {{Theory of cargo and membrane trafficking}},
  year = {2014},
  eprint = {1412.1399},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1412.1399},
}
Figure