Quantitative imaging of mammalian transcriptional dynamics: from single cells to whole embryos

Abstract

Probing dynamic processes occurring within the cell nucleus at the quantitative level has long been a challenge in mammalian biology. Advances in bio-imaging techniques over the past decade have enabled us to directly visualize nuclear processes in situ with unprecedented spatial and temporal resolution and single-molecule sensitivity. Here, using transcription as our primary focus, we survey recent imaging studies that specifically emphasize the quantitative understanding of nuclear dynamics in both time and space. These analyses not only inform on previously hidden physical parameters and mechanistic details, but also reveal a hierarchical organizational landscape for coordinating a wide range of transcriptional processes shared by mammalian systems of varying complexity, from single cells to whole embryos.

Fig. 1.

Optical techniques useful for imaging the mammalian cell nucleus in space and time. a Single-molecule tracking (SMT) using epi-illumination, in which fluorescently labeled molecules within the laser focus (purple oval) are excited and their movements followed over time; a few representative single-molecule trajectories are depicted. b Fluorescence correlation spectroscopy (FCS), which analyzes the fluctuations in fluorescence intensity as molecules move in and out of the laser focus to obtain quantitative information on their dynamics; a representative intensity fluctuation trace (inset) and the autocorrelation function curve calculated from the trace are shown. c Photobleaching-based imaging techniques, depicting a small region of the nucleus (green) that has been selectively photobleached (dark green box); a typical FRAP curve is shown here (red arrowhead denotes photobleaching)

Fig. 2.

Temporal modes of organizing mammalian transcriptional dynamics. a Modulating transcription through TF binding and target search, in which a TF molecule undergoes “facilitated diffusion” by partitioning its movement between free 3D diffusion (purple) and transient 1D sliding along the DNA (blue) until the specific target sites are located and to which the TF stays bound for a long time (red). b An example of the physiological consequence of TF binding dynamics, in which the long-lived bound fraction of Sox2 in a blastomere of a four-cell embryo predicts the bias with which this blastomere will contribute to the inner mass of the embryo subsequently. c Modulating transcription through pulsatile production or “bursts” of mRNAs, as a consequence of stochastic switching of the gene between the “on” and ”off” states. d Widely different Nanog expression among a population of mouse ES cells as revealed by smFISH. In contrast to the sparse Nanog molecules present in the two lower cells, multiple spots where bursts of Nanog transcription took place (indicated by red arrowheads) are discernible in the top cell. Dotted lines delineate nuclear boundaries. Adapted from [66] (a, b) with modifications

Fig. 3.

Spatial modes of organizing mammalian transcriptional dynamics. a A super-resolution map of RNAP II distribution inside a mammalian cell nucleus. Inset shows a zoom-in area illustrating the co-existence of both isolated RNAP II molecules (left) as well as transient clusters (“transcription factories”, right) that coordinate the expression of spatially disparate genes. b When the angles between successive translocation steps of TF molecules during target search are measured by SMT, P-TEFb (blue) exhibits substantial asymmetry (evidenced by the strong bias toward 180° in the angular distribution and the negative asymmetry coefficient), indicating a propensity to “back-stepping” due to spatial constraints of the search process. In contrast, c-Myc (orange) explores the nuclear space more or less unhindered with no preferred directionality (evidenced by the near-zero asymmetry coefficient). c Modulating transcription through oscillatory nucleo-cytoplasmic translocation of NF-κB; single-cell snapshots of nuclear NF-κB level at representative time points are depicted in insets. Adapted from [12] (a), [69] (b), and [101] (c) with modifications

Fig. 4.

Organizing mammalian transcriptional dynamics through modulating concentration, oligomerization, or epigenetic states. a Intranuclear concentration of p53 can exhibit either pulsatile (top) or sustained (bottom) changes when perturbed with different stimuli; cells with pulsatile response recover from DNA damage whereas those with sustained response enter senescence. b Regulating TF mobility and function through the dynamic interconversion among its monomeric, dimeric, and tetrameric forms. Depicted here is the case of STAT3, whose dimers must first translocate to the nucleus and bind DNA before forming tetramers, which could then amplify or repress STAT3 dimer-mediated transcription. c Immunofluorescence staining of four-cell mouse embryos reveals distinct differences in H3R26me2 levels (top); cells with a higher level (orange) exhibit a larger fraction of Sox2 engaged in long-lived DNA-binding than those with a lower level (blue) (bottom). Adapted from [118] (a), [124] (b), and [66] (c) with modifications