3D Computational Modeling of Defective Early Endosome Distribution in Human iPSC-Based Cardiomyopathy Models

Hafiza Nosheen Saleem, Nadezda Ignatyeva, Christiaan Stuut, Stefan Jakobs, Michael Habeck, Antje EbertCells 2024 · DOI

TL;DR

This paper introduces a novel 3D computational modeling approach to quantify the distribution of early endosomes within human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs). The method successfully identifies and quantifies defective endosome localization in iPSC-CMs modeling dilated cardiomyopathy (DCM) and demonstrates its ability to track the recovery of normal distribution upon therapeutic intervention.

What Problem Does This Paper Address?

The intricate network of endosomes plays a crucial role in intracellular cargo transport, with early endosomes being central to pathways like clathrin-mediated endocytosis (CME). However, the dynamic and complex nature of this system makes quantitative analysis challenging. In certain cardiac diseases, such as dilated cardiomyopathy (DCM), this endosomal trafficking is disrupted, leading to an abnormal accumulation of early endosomes at the plasma membrane. Existing methods struggle to precisely quantify these spatial distribution defects, hindering a deeper understanding of disease mechanisms and the development of targeted therapies. This paper aims to bridge this gap by developing a computational tool to accurately assess and characterize these endosome distribution anomalies.

Methods

The core of this work is a new computational modeling approach that uses 3D confocal microscopy images of iPSC-CMs. The method involves:

  • Image Preprocessing: Background noise in the confocal images is fitted with a Gaussian model to identify true fluorescence signals.
  • Cell Center Estimation: The center of each cell is estimated using a center-of-mass calculation on thresholded images.
  • Distance Calculation: Distances of all identified endosome vesicles (marked by EEA1) from the cell center are computed.
  • Spherical Modeling & Scaling: To account for variations in cell size, these distances are scaled using a "radius of gyration" derived from the cell's shape. This creates "scaled distance distributions."
  • Analysis of Distributions: The scaled distance distributions are analyzed to identify patterns. A "bi-modal" distribution, indicating two distinct populations of endosomes (e.g., at the plasma membrane and in the cytoplasm), is a key indicator of defects.
  • STED Microscopy Integration: The authors also demonstrate the utility of combining their modeling approach with super-resolution STED microscopy, which offers higher resolution and allows for more detailed analysis of individual vesicles.

Datasets

The study utilized iPSC-CMs derived from human subjects. Specifically, they used:

  • Wild-type (WT) iPSC-CMs: Serving as a control group.
  • Mutant (MUT) iPSC-CMs: These were engineered to carry specific mutations associated with dilated cardiomyopathy (DCM):
  • TPM1-L185F mutation (MUT1)
    • TnT-R141W mutation (MUT2)
  • Treated MUT iPSC-CMs: MUT iPSC-CMs treated with RhoA II, a peptide known to restore CME-dependent signaling.

The data consisted of 3D confocal and STED microscopy images of these cell populations, focusing on the localization of Early Endosome Antigen 1 (EEA1)-positive early endosomes. No new benchmark datasets were introduced; rather, existing cell models were used for evaluation.

Key Insights and Ideas

  • Bi-modal Distribution as a Biomarker: The most significant insight is the identification of a "bi-modal" scaled distance distribution as a robust indicator of defective early endosome localization in DCM iPSC-CMs. This pattern directly reflects the abnormal accumulation of endosomes at the cell periphery.
  • Computational Modeling for Disease Phenotyping: The developed 3D spherical modeling approach provides a quantitative and objective way to assess endosome distribution defects, moving beyond qualitative visual inspection.
  • Therapeutic Efficacy Assessment: The model successfully recapitulates the recovery of normal endosome distribution upon treatment with RhoA II, demonstrating its potential for evaluating the efficacy of therapeutic interventions.
  • STED Enhances Resolution: Combining the computational model with STED microscopy reveals finer details of vesicle distribution, highlighting the benefit of higher-resolution imaging for more accurate quantification.

Key Results

  • Defective Distribution in DCM Models: The 3D spherical modeling consistently identified a bi-modal distribution of early endosomes in iPSC-CMs carrying DCM mutations (MUT1 and MUT2) compared to their wild-type (WT) counterparts. This indicates a significant deviation from the homogeneous distribution seen in healthy cells.
  • Recovery with RhoA II Treatment: Treatment of MUT iPSC-CMs with RhoA II led to a reversion of the bi-modal distribution, showing a more homogeneous pattern, confirming the model's ability to track therapeutic effects.
  • Robustness of the Method: The modeling approach was shown to be robust against variations in cell size and sampling density, making it a reliable tool for analysis.
  • STED vs. Confocal: While both imaging techniques, when coupled with the model, showed comparable relative distance distributions, STED microscopy provided a substantially larger range of detected events, suggesting improved sensitivity and detail.

Limitations and Future Work

  • Cell Shape Variability: The authors acknowledge that the spherical approximation of cell shape is a limitation, as iPSC-CMs and especially mature cardiomyocytes can exhibit more complex, elongated, or irregular shapes. They suggest that adapting the model to an ellipsoid coordinate system or other more complex shapes could improve accuracy for mature cells.
  • Translational Application: While the platform shows promise for pre-clinical models, further integration of different imaging data types and multi-functional analysis algorithms is needed for direct translational application in clinical diagnostics.
  • Broader Disease Scope: The current study focuses on specific DCM mutations. Future work could involve applying this model to a wider range of DCM mutations and other cardiac diseases to build a more comprehensive understanding of endosomal dysfunction.
  • Integration with Machine Learning: The authors propose that this automated 3D computational modeling can serve as a foundation for integrative data platforms utilizing machine learning and deep learning tools for more advanced analysis and prediction.

Summary updated

Abstract

Intracellular cargo delivery via distinct transport routes relies on vesicle carriers. A key trafficking route distributes cargo taken up by clathrin-mediated endocytosis (CME) via early endosomes. The highly dynamic nature of the endosome network presents a challenge for its quantitative analysis, and theoretical modelling approaches can assist in elucidating the organization of the endosome trafficking system. Here, we introduce a new computational modelling approach for assessment of endosome distributions. We employed a model of induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) with inherited mutations causing dilated cardiomyopathy (DCM). In this model, vesicle distribution is defective due to impaired CME-dependent signaling, resulting in plasma membrane-localized early endosomes. We recapitulated this in iPSC-CMs carrying two different mutations, TPM1-L185F and TnT-R141W (MUT), using 3D confocal imaging as well as super-resolution STED microscopy. We computed scaled distance distributions of EEA1-positive vesicles based on a spherical approximation of the cell. Employing this approach, 3D spherical modelling identified a bi-modal segregation of early endosome populations in MUT iPSC-CMs, compared to WT controls. Moreover, spherical modelling confirmed reversion of the bi-modal vesicle localization in RhoA II-treated MUT iPSC-CMs. This reflects restored, homogeneous distribution of early endosomes within MUT iPSC-CMs following rescue of CME-dependent signaling via RhoA II-dependent RhoA activation. Overall, our approach enables assessment of early endosome distribution in cell-based disease models. This new method may provide further insight into the dynamics of endosome networks in different physiological scenarios.
BibTeX
@inproceedings{saleem2024d,
  author = {Hafiza Nosheen Saleem and Nadezda Ignatyeva and Christiaan Stuut and Stefan Jakobs and Michael Habeck and Antje Ebert},
  title = {{3D Computational Modeling of Defective Early Endosome Distribution in Human iPSC-Based Cardiomyopathy Models}},
  booktitle = {Cells},
  year = {2024},
  doi = {10.3390/cells13110923},
}

Loading related content...

Figure