Creating an Integrated Landscape of Human IPSC Nuclear States

The Allen Institute for Cell Science is developing a state space of stem cell structural signatures to understand the principles by which cells reorganize as they traverse the cell cycle and differentiate. To do this, we have developed a pipeline that generates high-replicate, dynamic image data of cell organization and activities in human induced pluripotent stem cell (hiPSC) lines (the Allen Cell Collection at www.allencell.org). Each of the ∼35 lines expresses an endogenous monoallelic EGFP-tagged protein that represents a particular cellular organelle or structure. Thousands of replicate high resolution 3D images are acquired for each structure, which are used for integrative data analysis, machine learning and computational modeling. We are applying these same approaches to creating an integrated state space of nuclear signatures. We have endogenously EGFP-tagged an initial set of 17 proteins representing key organizational landmarks at multiple spatial scales including at the level of the nucleus and nucleoplasm as a whole (nuclear lamina, nuclear pores, several nucleolar subcompartments, and nuclear speckles), chromatin structure (histone H2B, HP1-beta, and EZH2), key proteins in chromatin looping (CTCF, SMC1A), two types of chromatin loci (telomeres and DNA replication sites), RNA polymerase, and two pluripotency transcription factors. We are developing assays to image, quantify, segment and analyze features of these various nuclear proteins, including methods for tracking loci and creating machine-learning transfer functions to enhance resolution. The cell lines, image data and analyses are being integrated with genome-wide assays of chromatin architecture together with the 4D Nucleome Project NOFIC centers, and once fully characterized, will be distributed publicly. For example, to provide spatial constraints within the nucleus for more accurate modeling of chromatin organization based on genomic assays, we are applying deep-learning based “label-free” technology to generate integrated models of key nuclear landmarks.