Fig. 1: Combining optogenetics and live imaging enables dissection of single-cell repression dynamics in a developing animal.

A Key questions regarding transcriptional repression. Left: Whether single-cell repression occurs in a gradual or switch-like fashion over time. Right: Whether repression is reversible. B Knirps represses even-skipped (eve) stripes 4 + 6 transcription in the fruit fly embryo. Top: Knirps is expressed in a bell-shaped domain during early embryogenesis. Bottom: Knirps specifies the position and sharpness of the inner boundaries of eve stripes 4 and 6. C Two-color tagging permits the simultaneous visualization of input protein concentration and output transcriptional dynamics in vivo. Maternally deposited EYFP molecules bind to Knirps-LlamaTag, resulting in increased nuclear fluorescence, which provides a real-time readout of the nuclear protein concentration. Maternally deposited MS2 coat protein (MCP) binds to MS2 stem-loops in the nascent RNA formed by RNAP molecules elongating along the body of the eve 4 + 6 reporter construct leading to the accumulation of fluorescence at sites of nascent transcript formation. LEXY tag is also fused to Knirps to allow for optogenetic manipulation of its nuclear concentration. D Representative frames from live-imaging data. The embryo is oriented with the anterior (head) to the left. Green and magenta channels correspond to Knirps repressor and eve 4 + 6 transcription, respectively. When Knirps concentration is low, eve stripe 4 + 6 is expressed in a broad domain, which refines into two flanking stripes as Knirps concentration increases. E Optogenetic control of nuclear protein export. Upon exposure to blue light, the nuclear export signal within the LEXY domain is revealed. As a result, the fusion protein is exported from the nucleus. F Fluorescence images of embryos expressing the Knirps-LEXY fusion undergoing an export-recovery cycle. G Relative nuclear fluorescence of the repressor protein over time (n = 55 nuclei). Half-times for export and recovery processes are estimated by fitting the fluorescence traces to exponential functions.