Patterns and Regulation of Organelle Interactions and Cell Anatomy Revealed by SXT: a High-Throughput Study in S. cerevisiae and Cross-Modality Synthesis of the Whole-Cell Imaging Literature

Résumé fourni par la source

The bridge between the cellular meso- and whole-cell scales is an under-appreciated level of biological organization. The past decade has seen massive progress in unraveling the complex interconnectivity of organelles through contact sites and functional crosstalk. However, how the entire organelle network spatially arranges within the cytoplasm is only beginning to be explored. What are the structural patterns and physical rules governing the state of the organelle network? How do structural interdependencies influence organelle morphology and overall cell anatomy, and how do these interdependencies shift in dysfunctional states? Addressing such questions requires minimal-artifact volumetric data reconstructing whole cells with multiple identifiable organelles, with sufficient quantitative throughput to analyze statistical variation in morphological features and perform robust comparisons of cell populations or conditions. However, whole-cell reconstruction data is costly and time consuming to generate, typically limiting single-study sample size. We address this challenge with two complementary approaches: a high-throughput study in budding yeast, and a synthesis of whole-cell imaging studies across the literature. We have created the largest-yet dataset of whole-cell images of S. cerevisiae using soft X-ray tomography in collaboration with the National Center for X-ray Tomography. Custom automated segmentation enabled us to reconstruct the nucleus, vacuole, lipid droplet for 490 cells across three strains at 30-35nm isotropic resolution. This combination of spatial and quantitative resolution enables precise characterization of mutant phenotypes by comparison with control strains [1]. For example, we find that vac14, classically described as an enlarged-vacuole phenotype, in fact has a vacuole size variability and overall cell enlargement phenotype in unbudded cells [2, 3]. To what extent is cell anatomy a reflection of organelle confinement within the limited space of the crowded cytoplasm? 3D morphometric statistical analysis and integrated modeling showed that vacuole enlargement (vac14) progressively breaks the linear scaling of the nucleus and lipid droplets, and drives nuclear displacement and flattening. We developed a pipeline to create triangulated meshes of cell and organelle surfaces using GAMer2, enabling quantitative whole-organelle shape analysis in unprecedented detail [4, 5]. Joint 3D shape analysis of apposing vacuole-nucleus pairs shows correlated curvature at the interface in some of these cells. These results suggest that asymmetric deformations may arise due to vacuole-nucleus packing. We adapted the Allen Institute's integrated modeling pipeline to yeast cells in order to quantify the key constrained and variable features in yeast and to determine the effects of vacuole enlargement on whole-cell anatomy [6]. Dimensionality reduction of joint spherical harmonics-based shape representations of cell-nucleus-vacuole complexes point to relative organelle size, positioning, shape, and orientation as principal components of cell anatomical variation in wild-type and large-vacuole mutant cells. Our working model is that physical packing of the nucleus and vacuole in the limited cytoplasmic volume plays a significant role in establishing organelle morphology and global cell patterning. Ongoing analyses of these co-varying geometric features, and biophysical deformation simulations based on observed shapes in interacting organelle pairs analysis between WT and large-vacuole mutants, will refine our model of the interplay between packing, scaling, and contact aspects of the vacuole-nucleus relationship. Second, we adapted interdisciplinary evidence synthesis methods systematically curate the whole-cell volumetric imaging literature, including all studies that report whole-cell data with at least two organelles, through 3 rounds of screening involving 3 independent reviewers, resulting in 89 top hits and 38 “borderline” studies between 2004-2024 [7]. A broad characterization, or “scoping review”, of bibliometrics, study design, and reporting practices shows accelerating technological development and research output. We find high variability in study design and reporting practices, including imaging modality, model organism, cellular contexts, organelles imaged, and analyses. Due to the laborious, low-throughput nature of most volumetric imaging methods, we find trends toward small sample sizes (<10 cells) and small cell types. We show that at least 13 distinct imaging modalities have successfully generated whole-cell data, with FIB-SEM and SXT the most commonly represented across studies. X-ray-based techniques most consistently achieved isotropic resolution compared to electron- and light-based methods, and SXT boasted the largest sample sizes, with resolution of 60nm or better. Thus, SXT emerges as the standout modality for cell anatomy imaging – consistently offering isotropic resolution high enough to visualize detailed membrane morphologies at whole-organelle scales, with quantitative power sufficient for statistical analysis and modeling, and widespread use across the whole-cell imaging community. Our curated dataset now enables future aggregate and comparative analyses to potentially reveal larger patterns and generate more generalized hypotheses. Ongoing work is now merging common quantitative analyses across studies, including volumetric ratios of organelles and inter-organelle contact analyses to identify common and context-specific whole-cell anatomical patterns and changes in organelle relationships. This work establishes a growing dataset of whole-cell imaging literature and data, and motivates a call for standardized whole-cell imaging study design, reporting, and data sharing practices.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.