Background

The early embryo is transcriptionally silent and development is driven by maternally deposited mRNAs and proteins (Hamm and Harrison, 2018; Schulz and Harrison, 2019). In Drosophila, the nuclei exist in a syncytium and undergo rapid nuclear divisions, consisting only of synthesis and mitosis phases (Foe and Alberts, 1983). During the 14^th nuclear cycle (NC), the genome of the developing embryo is transcriptionally up-regulated during the major wave of zygotic genome activation (ZGA) (Hamm and Harrison, 2018; Schulz and Harrison, 2019). Prior to widespread ZGA at NC14, a small subset of genes, mostly encoding transcription factors, is transcribed during a minor wave of ZGA (McDaniel et al., 2019). This entire process from egg laying to widespread ZGA takes less than three hours, which has made it technically challenging to dissect the role various transcription factors play in ZGA. Further, while genetic techniques are available to maternally deplete essential transcription factors, this removes these factors from the embryo throughout early development, making it impossible to tease apart their roles during the minor and major waves of ZGA. Other systems have used techniques like Anchor-Away or proteasome targeting to deplete nuclear proteins (Haruki et al., 2008; Caussinus et al., 2011), however, these methods take too long to inactivate their targets to work on the timescale necessary to study in the early embryo. Further, these techniques are not reversible on the timescales needed to study transcription-factor activity in the early embryo. To address these limitations, we developed an optogenetic system to rapidly and reversibly inactivate the essential transcription factor Zelda (Zld) in the early embryo of Drosophila with nuclear-cycle resolution.

Optogenetic protocols require the addition of a photo-responsive protein tag to the protein of interest, either at the endogenous locus or via a transgene. We note that both published examples of optogenetic transcription-factor inactivation in Drosophila have used an N-terminal tag, though it remains unclear if tagging at the N-terminus is required for inactivation (Huang et al., 2017; McDaniel et al., 2019). Using Cas9-mediated genome engineering, we endogenously tagged the N-terminus of Zld with the blue-light responsive CRY2 tag (McDaniel et al., 2019). CRY2, originally characterized in Arabidopsis, is traditionally part of a two-component system, where CRY2 dimerizes with CIB1 upon exposure to blue light (Liu et al., 2008). Both components are available for use in Drosophila (Guglielmi et al., 2015). However, CRY2 alone has been demonstrated to be sufficient to inactivate transcription factors, albeit through different mechanisms; CRY2-tagged Bicoid remains on chromatin upon blue-light exposure, while CRY2-tagged Zelda dissociates from chromatin (Huang et al., 2017; McDaniel et al., 2019). The mechanism of inactivation is likely a conformational change upon blue-light exposure that disrupts the ability of the transcription factor to activate transcription. One benefit of the CRY2 system is its rapidity of inactivation; upon blue-light exposure, the CRY2-tagged protein is inactivated within seconds and, when the light is removed, refolds in ~5 min (Huang et al., 2017).

While we utilized the one component CRY2 system for our experiments, this system depends on conformational changes that may be protein specific. Thus, inactivation must be confirmed experimentally. If the single-component system fails to inactivate protein function, the two-component system is available for long-term protein inactivation via cell-membrane sequestration (Guglielmi et al., 2015). In the two-component system, CRY2 will dimerize with CIB1 within seconds, but takes over an hour to dissociate once the light is removed. In Drosophila, a cell membrane localized, GFP-tagged CIB1 has been developed and could be used to sequester a target protein away from chromatin using blue-light exposure when the nuclear envelope breaks down during mitosis (Guglielmi et al., 2015). This would have a similar sequestration-mediated effect as the recently described Jabba-trap system (Seller et al., 2019). Thus, depending on experimental needs, either the one or two-component system can be used.

Here, we will describe, in detail, how to use this versatile one-component system to reversibly inactivate a protein of choice with precise temporal resolution in the early embryo of Drosophila and harvest single embryos for downstream RNA-seq.