Performance of deep learning restoration methods for the extraction of particle dynamics in noisy microscopy image sequences.

Particle tracking in living systems requires low light exposure and short exposure times to avoid phototoxicity and photobleaching and to fully capture particle motion with high-speed imaging. Low-excitation light comes at the expense of tracking accuracy. Image restoration methods based on deep learning dramatically improve the signal-to-noise ratio in low-exposure data sets, qualitatively improving the images. However, it is not clear whether images generated by these methods yield accurate quantitative measurements such as diffusion parameters in (single) particle tracking experiments. Here, we evaluate the performance of two popular deep learning denoising software packages for particle tracking, using synthetic data sets and movies of diffusing chromatin as biological examples. With synthetic data, both supervised and unsupervised deep learning restored particle motions with high accuracy in two-dimensional data sets, whereas artifacts were introduced by the denoisers in three-dimensional data sets. Experimentally, we found that, while both supervised and unsupervised approaches improved tracking results compared with the original noisy images, supervised learning generally outperformed the unsupervised approach. We find that nicer-looking image sequences are not synonymous with more precise tracking results and highlight that deep learning algorithms can produce deceiving artifacts with extremely noisy images. Finally, we address the challenge of selecting parameters to train convolutional neural networks by implementing a frugal Bayesian optimizer that rapidly explores multidimensional parameter spaces, identifying networks yielding optimal particle tracking accuracy. Our study provides quantitative outcome measures of image restoration using deep learning. We anticipate broad application of this approach to critically evaluate artificial intelligence solutions for quantitative microscopy.

Statistical mechanics of chromosomes: in vivo and in silico approaches reveal high-level organization and structure arise exclusively through mechanical feedback between loop extruders and chromatin substrate properties.

The revolution in understanding higher order chromosome dynamics and organization derives from treating the chromosome as a chain polymer and adapting appropriate polymer-based physical principles. Using basic principles, such as entropic fluctuations and timescales of relaxation of Rouse polymer chains, one can recapitulate the dominant features of chromatin motion observed in vivo. An emerging challenge is to relate the mechanical properties of chromatin to more nuanced organizational principles such as ubiquitous DNA loops. Toward this goal, we introduce a real-time numerical simulation model of a long chain polymer in the presence of histones and condensin, encoding physical principles of chromosome dynamics with coupled histone and condensin sources of transient loop generation. An exact experimental correlate of the model was obtained through analysis of a model-matching fluorescently labeled circular chromosome in live yeast cells. We show that experimentally observed chromosome compaction and variance in compaction are reproduced only with tandem interactions between histone and condensin, not from either individually. The hierarchical loop structures that emerge upon incorporation of histone and condensin activities significantly impact the dynamic and structural properties of chromatin. Moreover, simulations reveal that tandem condensin-histone activity is responsible for higher order chromosomal structures, including recently observed Z-loops.

Polymer perspective of genome mobilization.

Chromosome motion is an intrinsic feature of all DNA-based metabolic processes and is a particularly well-documented response to both DNA damage and repair. By using both biological and polymer physics approaches, many of the contributing factors of chromatin motility have been elucidated. These include the intrinsic properties of chromatin, such as stiffness, as well as the loop modulators condensin and cohesin. Various biological factors such as external tethering to nuclear domains, ATP-dependent processes, and nucleofilaments further impact chromatin motion. DNA damaging agents that induce double-stranded breaks also cause increased chromatin motion that is modulated by recruitment of repair and checkpoint proteins. Approaches that integrate biological experimentation in conjunction with models from polymer physics provide mechanistic insights into the role of chromatin dynamics in biological function. In this review we discuss the polymer models and the effects of both DNA damage and repair on chromatin motion as well as mechanisms that may underlie these effects.

The regulation of chromosome segregation via centromere loops.

Biophysical studies of the yeast centromere have shown that the organization of the centromeric chromatin plays a crucial role in maintaining proper tension between sister kinetochores during mitosis. While centromeric chromatin has traditionally been considered a simple spring, recent work reveals the centromere as a multifaceted, tunable shock absorber. Centromeres can differ from other regions of the genome in their heterochromatin state, supercoiling state, and enrichment of structural maintenance of chromosomes (SMC) protein complexes. Each of these differences can be utilized to alter the effective stiffness of centromeric chromatin. In budding yeast, the SMC protein complexes condensin and cohesin stiffen chromatin by forming and cross-linking chromatin loops, respectively, into a fibrous structure resembling a bottlebrush. The high density of the loops compacts chromatin while spatially isolating the tension from spindle pulling forces to a subset of the chromatin. Paradoxically, the molecular crowding of chromatin via cohesin and condensin also causes an outward/poleward force. The structure allows the centromere to act as a shock absorber that buffers the variable forces generated by dynamic spindle microtubules. Based on the distribution of SMCs from bacteria to human and the conserved distance between sister kinetochores in a wide variety of organisms (0.4 to 1 micron), we propose that the bottlebrush mechanism is the foundational principle for centromere function in eukaryotes.

Three-Dimensional Thermodynamic Simulation of Condensin as a DNA-Based Translocase.

Chromatin dynamics and organization can be altered by condensin complexes. In turn, the molecular behavior of a condensin complex changes based on the tension of the substrate to which condensin is bound. This interplay between chromatin organization and condensin behavior demonstrates the need for tools that allows condensin complexes to be observed on a variety of chromatin organizations. We provide a method for simulating condensin complexes on a dynamic polymer substrate using the polymer dynamics simulator ChromoShake and the condensin simulator RotoStep. These simulations can be converted into simulated fluorescent images that are able to be directly compared to experimental images of condensin and fluorescently labeled chromatin. Our pipeline enables users to explore how changes in condensin behavior alters chromatin dynamics and vice versa while providing simulated image datasets that can be directly compared to experimental observations.

Geometric partitioning of cohesin and condensin is a consequence of chromatin loops.

SMC (structural maintenance of chromosomes) complexes condensin and cohesin are crucial for proper chromosome organization. Condensin has been reported to be a mechanochemical motor capable of forming chromatin loops, while cohesin passively diffuses along chromatin to tether sister chromatids. In budding yeast, the pericentric region is enriched in both condensin and cohesin. As in higher-eukaryotic chromosomes, condensin is localized to the axial chromatin of the pericentric region, while cohesin is enriched in the radial chromatin. Thus, the pericentric region serves as an ideal model for deducing the role of SMC complexes in chromosome organization. We find condensin-mediated chromatin loops establish a robust chromatin organization, while cohesin limits the area that chromatin loops can explore. Upon biorientation, extensional force from the mitotic spindle aggregates condensin-bound chromatin from its equilibrium position to the axial core of pericentric chromatin, resulting in amplified axial tension. The axial localization of condensin depends on condensin's ability to bind to chromatin to form loops, while the radial localization of cohesin depends on cohesin's ability to diffuse along chromatin. The different chromatin-tethering modalities of condensin and cohesin result in their geometric partitioning in the presence of an extensional force on chromatin.

SUMO-Targeted Ubiquitin Ligase (STUbL) Slx5 regulates proteolysis of centromeric histone H3 variant Cse4 and prevents its mislocalization to euchromatin.

Centromeric histone H3, CENP-ACse4, is essential for faithful chromosome segregation. Stringent regulation of cellular levels of CENP-ACse4 restricts its localization to centromeres. Mislocalization of CENP-ACse4 is associated with aneuploidy in yeast, flies and tumorigenesis in human cells; thus, defining pathways that regulate CENP-A levels is critical for understanding how mislocalization of CENP-A contributes to aneuploidy in human cancers. Previous work in budding yeast has shown that ubiquitination of overexpressed Cse4 by Psh1, an E3 ligase, partially contributes to proteolysis of Cse4. Here, we provide the first evidence that Cse4 is sumoylated by E3 ligases Siz1 and Siz2 in vivo and in vitro. Ubiquitination of Cse4 by Small Ubiquitin-related Modifier (SUMO)-Targeted Ubiquitin Ligase (STUbL) Slx5 plays a critical role in proteolysis of Cse4 and prevents mislocalization of Cse4 to euchromatin under normal physiological conditions. Accumulation of sumoylated Cse4 species and increased stability of Cse4 in slx5∆ strains suggest that sumoylation precedes ubiquitin-mediated proteolysis of Cse4. Slx5-mediated Cse4 proteolysis is independent of Psh1 since slx5∆ psh1∆ strains exhibit higher levels of Cse4 stability and mislocalization compared to either slx5∆ or psh1∆ strains. Our results demonstrate a role for Slx5 in ubiquitin-mediated proteolysis of Cse4 to prevent its mislocalization and maintain genome stability.

ChromoShake: a chromosome dynamics simulator reveals that chromatin loops stiffen centromeric chromatin.

ChromoShake is a three-dimensional simulator designed to find the thermodynamically favored states for given chromosome geometries. The simulator has been applied to a geometric model based on experimentally determined positions and fluctuations of DNA and the distribution of cohesin and condensin in the budding yeast centromere. Simulations of chromatin in differing initial configurations reveal novel principles for understanding the structure and function of a eukaryotic centromere. The entropic position of DNA loops mirrors their experimental position, consistent with their radial displacement from the spindle axis. The barrel-like distribution of cohesin complexes surrounding the central spindle in metaphase is a consequence of the size of the DNA loops within the pericentromere to which cohesin is bound. Linkage between DNA loops of different centromeres is requisite to recapitulate experimentally determined correlations in DNA motion. The consequences of radial loops and cohesin and condensin binding are to stiffen the DNA along the spindle axis, imparting an active function to the centromere in mitosis.

Spindle assembly checkpoint proteins are positioned close to core microtubule attachment sites at kinetochores.

Spindle assembly checkpoint proteins have been thought to reside in the peripheral corona region of the kinetochore, distal to microtubule attachment sites at the outer plate. However, recent biochemical evidence indicates that checkpoint proteins are closely linked to the core kinetochore microtubule attachment site comprised of the Knl1-Mis12-Ndc80 (KMN) complexes/KMN network. In this paper, we show that the Knl1-Zwint1 complex is required to recruit the Rod-Zwilch-Zw10 (RZZ) and Mad1-Mad2 complexes to the outer kinetochore. Consistent with this, nanometer-scale mapping indicates that RZZ, Mad1-Mad2, and the C terminus of the dynein recruitment factor Spindly are closely juxtaposed with the KMN network in metaphase cells when their dissociation is blocked and the checkpoint is active. In contrast, the N terminus of Spindly is ∼75 nm outside the calponin homology domain of the Ndc80 complex. These results reveal how checkpoint proteins are integrated within the substructure of the kinetochore and will aid in understanding the coordination of microtubule attachment and checkpoint signaling during chromosome segregation.