Transcriptional dynamics in developing embryos at the single molecule level
During animal development the fate of individual cells in the early embryo are decided with exquisite spatial and temporal precision. At the molecular level these decisions are regulated by highly dynamic interactions between transcription factors (TFs), the chromatin architecture, and the transcriptional machinery. In the past decade, research leveraging live cell single molecule imaging methods has made significant progress in characterizing the nature of these dynamics. However, due to technological limitations, these experiments have been limited to monolayer cell cultures which cannot reconstitute the complexity of a developing embryo.
Video of mitosis in the early Drosophila embryo using Lattice Light Sheet Microscopy
To address these limitations, we have built a lattice light-sheet microscope (LLSM) (Chen et al., 2014) which overcomes the technical barriers and allows for single molecule imaging inside of living embryos and thick tissue. A lattice light-sheet microscope uses a combination of optical technologies to generate an exquisitely thin (less than a micrometer) sheet of light which is used to illuminate sample. The thin nature of the illumination sheet provides a dramatic improvement in contrast while minimizing the light exposure to the specimen and enables the visualizing of single molecules within embryos while keeping them alive.
Main body of the Lattice light-sheet microscope at UC Berkeley
Armed with the capability to perform single molecule imaging in live developing embryos we decided to first study the dynamics of the iconic morphogen transcription factor Bicoid (in collaboration with Hernan Garcia’s lab) which is responsible for spatial patterning along the anteroposterior axis during early development in Drosophila embryos developing. Although this morphogen is considered as one of the best understood systems that controls gene expression during development, we discovered a new and unanticipated mechanism that modulates Bicoid binding.
Initial experiments revealed that Bicoid interacts with DNA in a much weaker fashion than other transcription factors and also that it binds at significant levels to its DNA targets in regions of the embryo where it exists at vanishingly low concentrations. This is surprising as the classical models on morphogen function do not predict that this should be happening. In order for this binding to be possible we reasoned that a mechanism must exist to facilitate this process.
To shed light on a potential mechanism we examined where individual molecules of Bicoid are binding in space and time within nuclei and found that they bind in a clustered manner forming local hubs of high concentration. We eamed with Prof. Mike Eisen’s lab (Molecular & Cell Biology) to perform genome wide analysis of Bicoid binding and found that the DNA sequences that Bicoid likely binds in the hubs are usually in close vicinity of those bound by another recently discovered protein called Zelda. This led the team to hypothesize that Zelda is mediating the formation of the Bicoid hubs and when they generated embryos that lack Zelda they found that the Bicoid hub formation was indeed abolished
Bicoid Hubs do not form in embryos lacking Zelda
These discoveries add to a growing body of evidence that challenges the classical ideas of precise concentration readouts of morphogen gradients and calls for a revision of these models to take into account how interactions with other proteins can lead to the formation of localized hubs or microenvironments where gene expression is regulated.
We are now using single molecule imaging to shed further light on the nature of local high concentration hubs of transcription factors. We are especially interested in how these hubs are formed, their composition, and how general of a phenomenon they are.