ABSTRACT
Faithful genome transmission in dividing cells requires that the two copies of each chromosome's DNA package into separate but physically linked sister chromatids. The linkage between sister chromatids is mediated by cohesin, yet where sister chromatids are linked and how they resolve during cell cycle progression has remained unclear. In this study, we investigated sister chromatid organization in live human cells using dCas9-mEGFP labeling of endogenous genomic loci. We detected substantial sister locus separation during G2 phase irrespective of the proximity to cohesin enrichment sites. Almost all sister loci separated within a few hours after their respective replication and then rapidly equilibrated their average distances within dynamic chromatin polymers. Our findings explain why the topology of sister chromatid resolution in G2 largely reflects the DNA replication program. Furthermore, these data suggest that cohesin enrichment sites are not persistent cohesive sites in human cells. Rather, cohesion might occur at variable genomic positions within the cell population.
Subject(s)
Cell Cycle , Chromatids/metabolism , Adaptor Proteins, Signal Transducing/metabolism , Cell Cycle Proteins/metabolism , Cell Line , Cell Survival , Chromatin/metabolism , Chromosomal Proteins, Non-Histone/metabolism , DNA Replication Timing , Fluorescent Dyes/metabolism , Genetic Loci , Genome, Human , Humans , Interphase , Models, Biological , CohesinsABSTRACT
Supervised machine learning is a powerful and widely used method for analyzing high-content screening data. Despite its accuracy, efficiency, and versatility, supervised machine learning has drawbacks, most notably its dependence on a priori knowledge of expected phenotypes and time-consuming classifier training. We provide a solution to these limitations with CellCognition Explorer, a generic novelty detection and deep learning framework. Application to several large-scale screening data sets on nuclear and mitotic cell morphologies demonstrates that CellCognition Explorer enables discovery of rare phenotypes without user training, which has broad implications for improved assay development in high-content screening.
Subject(s)
High-Throughput Screening Assays/methods , Statistics as Topic/methods , Algorithms , Animals , Biological Variation, Population/genetics , High-Throughput Screening Assays/statistics & numerical data , Humans , Machine Learning , Numerical Analysis, Computer-Assisted , Phenotype , SoftwareABSTRACT
Eukaryotic cells store their chromosomes in a single nucleus. This is important to maintain genomic integrity, as chromosomes packaged into separate nuclei (micronuclei) are prone to massive DNA damage. During mitosis, higher eukaryotes disassemble their nucleus and release individualized chromosomes for segregation. How numerous chromosomes subsequently reform a single nucleus has remained unclear. Using image-based screening of human cells, we identified barrier-to-autointegration factor (BAF) as a key factor guiding membranes to form a single nucleus. Unexpectedly, nuclear assembly does not require BAF's association with inner nuclear membrane proteins but instead relies on BAF's ability to bridge distant DNA sites. Live-cell imaging and in vitro reconstitution showed that BAF enriches around the mitotic chromosome ensemble to induce a densely cross-bridged chromatin layer that is mechanically stiff and limits membranes to the surface. Our study reveals that BAF-mediated changes in chromosome mechanics underlie nuclear assembly with broad implications for proper genome function.
