Saurabh S. Mogre

Cells appear extraordinarily complex, yet aspects of their organization may arise from a small number of influences acting within geometric constraints. Here, we extend and use cellPACK, an open‐source framework for rule‐based spatial modeling, to compare observed and simulated 3D distributions of peroxisomes and endosomes (puncta) in human induced pluripotent stem cells. Simulations conditioned on observed cell and nuclear geometries along with puncta counts show that geometric variation can generate apparent structure in organelle localization, highlighting the need to account for geometry when interpreting spatial statistics. We therefore employ an occupancy‐based metric that normalizes puncta density by available intracellular volume, revealing modest enrichment near the nuclear surface against an otherwise near‐uniform cytoplasmic distribution. We use these occupancy patterns to fit combinations of predefined spatial bias models and evaluate the resulting 3D packings against observed data. To support open and reproducible modeling while enabling a more intuitive visual exploration, we present cellPACK Studio , a browser‐based interface that allows users to generate, modify, and visualize the 3D cell packings used in this study (available at cellpack.allencell.org ). Together, these results and tools illustrate how spatial modeling can transform descriptive imaging data into testable, interpretable hypotheses of intracellular organization.

There is increasing interest in quantitatively combining phenotypic information and existing mathematical frameworks for cell state dynamics. We used hiPSC-derived endothelial cells as a controllable model system that adopts two distinct phenotypes at different magnitudes of shear stress. We trained a diffusion autoencoder model on timelapse imaging data to extract interpretable morphological features from brightfield. In this feature space, we found that observed phenotypic states correspond to stable fixed points of an inferred dynamical system. Cell migration coherence at these stable fixed points further distinguishes the states. Imaging cells across intermediate shear stresses revealed a regime of bistability, indicating that the shear-stress-dependent cell state transition occurs as a bifurcation. In N-terminally truncated VE-cadherin-expressing cells, we found that orientation and coherent migration were preserved, but morphology and migration speed were not. We demonstrate how machine learning and dynamical systems can be used in tandem to quantitatively characterize phenotypic cell states.

Cell states are increasingly conceptualized as attractors of high-dimensional dynamical systems, yet quantitative approaches for integrating phenotypic information into this framework remain limited. Here, we take an image-based approach that combines unsupervised machine learning (ML) with timelapse imaging to extract and characterize the temporal dynamics of morphological features. Using a cell line with endogenously tagged VE-cadherin, we acquired brightfield and fluorescence timelapse images of human induced pluripotent stem cell-derived endothelial cell (hiPSC-EC) monolayers, which adopt distinct phenotypes at two different magnitudes of shear stress in terms of their morphology, behavior, and VE-cadherin organization. To quantify these phenotypic cell states without segmentation, we trained a diffusion autoencoder to predict VE-cadherin signal from brightfield images. We identified interpretable ML-based features representing cell orientation, elongation, and local density. Treating these variables as dimensions of a morphological state space, we estimated a data-driven vector field and found that the two observed phenotypic cell states correspond to stable fixed points of the inferred dynamical system. Mapping measured cell migration coherence onto this space further distinguished the states. Imaging cells across intermediate shear stresses revealed a regime of bistability in which both states coexist, indicating that the shear-stress-dependent transition between endothelial cell states occurs as a bifurcation of the inferred dynamical system. Finally, we applied this method to study an N-terminal truncation of VE-cadherin, finding that mutated cells preserve alignment and coherent migration, but exhibit altered morphology and increased migration speed. This work demonstrates the applicability of a dynamical systems approach to quantitatively characterize morphological aspects of cell state from interpretable ML-based features.

A key challenge in understanding subcellular organization is quantifying interpretable measurements of intracellular structures with complex multi-piece morphologies in an objective, robust and generalizable manner. Here we introduce a morphology-appropriate representation learning framework that uses three-dimensional rotation-invariant autoencoders and point clouds. This framework is used to learn representations of complex shapes that are independent of orientation, compact and interpretable. We apply our framework to intracellular structures with punctate morphologies (for example, DNA replication foci) and polymorphic morphologies (for example, nucleoli). We explore the trade-offs in the performance of this framework compared to image-based autoencoders by performing multi-metric benchmarking across efficiency, generative capability and representation expressivity metrics. We find that the proposed framework, which embraces the underlying morphology of multi-piece structures, can facilitate the unsupervised discovery of subclusters for each structure. We show how this approach can also be applied to phenotypic profiling using a dataset of nucleolar images following drug perturbations.

A key challenge in understanding subcellular organization is quantifying interpretable measurements of intracellular structures with complex multi-piece morphologies in an objective, robust and generalizable manner. Here we introduce a morphology-appropriate representation learning framework that uses 3D rotation invariant autoencoders and point clouds. This framework is used to learn representations of complex multi-piece morphologies that are independent of orientation, compact, and easy to interpret. We apply our framework to intracellular structures with punctate morphologies (e.g. DNA replication foci) and polymorphic morphologies (e.g. nucleoli). We systematically compare our framework to image-based autoencoders across several intracellular structure datasets, including a synthetic dataset with pre-defined rules of organization. We explore the trade-offs in the performance of different models by performing multi-metric benchmarking across efficiency, generative capability, and representation expressivity metrics. We find that our framework, which embraces the underlying morphology of multi-piece structures, facilitates the unsupervised discovery of sub-clusters for each structure. We show how our approach can also be applied to phenotypic profiling using a dataset of nucleolar images following drug perturbations. We implement and provide all representation learning models using CytoDL, a python package for flexible and configurable deep learning experiments.

Summary: To investigate the fundamental question of how cellular variations arise across spatiotemporal scales in a population of identical healthy cells, we focused on nuclear growth in hiPS cell colonies as a model system. We generated a 3D timelapse dataset of thousands of nuclei over multiple days, and developed open-source tools for image and data analysis and an interactive timelapse viewer for exploring quantitative features of nuclear size and shape. We performed a data-driven analysis of nuclear growth variations across timescales. We found that individual nuclear volume growth trajectories arise from short timescale variations attributable to their spatiotemporal context within the colony. We identified a strikingly time-invariant volume compensation relationship between nuclear growth duration and starting volume across the population. Notably, we discovered that inheritance plays a crucial role in determining these two key nuclear growth features while other growth features are determined by their spatiotemporal context and are not inherited.

Autophagosome-lysosome fusion is required to clear cellular waste and maintain neuronal homeostasis. Computational modeling and microscopy elucidate the dynamics of autophagosome-lysosome interaction in axons. Each autophagosome fuses with only a few lysosomes, and autophagosomal maturation is limited by slow kinetics of inner membrane breakdown.

Macroautophagy is a homeostatic process required to clear cellular waste including aggregated proteins and dysfunctional organelles. Neuronal autophagosomes form constitutively in the distal tip of the axon and are actively transported toward the soma, with cargo degradation initiated en route. Cargo turnover requires autophagosomes to fuse with lysosomes to acquire degradative enzymes; however, the timing and number of these fusion events in the axon have proven difficult to detect using microscopy alone. Here we use a quantitative model, parameterized and validated using data from live and fixed imaging of primary hippocampal neurons, to explore the autophagosome maturation process on a cellular scale. We demonstrate that retrograde autophagosome motility is independent from lysosomal fusion, and that most autophagosomes fuse with only a few lysosomes by the time they reach the soma. Furthermore, our imaging and model results highlight the two-step maturation of the autophagosome: fusion with a lysosome or late endosome is followed by the slow degradation of the autophagosomal inner membrane before actual cargo degradation can occur. Together, rigorous quantitative measurements and mathematical modeling elucidate the dynamics of autophagosome-lysosome interaction and autophagosomal maturation in the axon.

Eukaryotic cells face the challenging task of transporting a variety of particles through the complex intracellular milieu in order to deliver, distribute, and mix the many components that support cell function. In this review, we explore the biological objectives and physical mechanisms of intracellular transport. Our focus is on cytoplasmic and intra-organelle transport at the whole-cell scale. We outline several key biological functions that depend on physically transporting components across the cell, including the delivery of secreted proteins, support of cell growth and repair, propagation of intracellular signals, establishment of organelle contacts, and spatial organization of metabolic gradients. We then review the three primary physical modes of transport in eukaryotic cells: diffusive motion, motor-driven transport, and advection by cytoplasmic flow. For each mechanism, we identify the main factors that determine speed and directionality. We also highlight the efficiency of each transport mode in fulfilling various key objectives of transport, such as particle mixing, directed delivery, and rapid target search. Taken together, the interplay of diffusion, molecular motors, and flows supports the intracellular transport needs that underlie a broad variety of biological phenomena.