Transcription factors (TFs) have a fundamental role is regulating gene expression. The basic model, based on numerous biochemical analyses, has determined where TFs bind (usually at specific sites at or near promoters), when they bind the DNA (at a resolution of minutes/hours) and what do they do there (induce/repress transcription. Duh!). However, much is yet unknown. One aspect that is fairly unknown is the dynamics of how TFs search for their binding sites, bind them and later dissociate, particularly at the single molecule level. To explore this, the Transcription Imaging Consortium (TIC) at Janelia Research Center (JRC) (it used to be Janelia Farm, but the “farm” part was removed from the name. oh well) applied sophisticated imaging techniques to measure the dynamics of two TFs, SOX2 and OCT4 in the nuclei of live embryonic stem (ES)cells. Their results were published in Cell almost a year ago.
First, how do you detect a single molecule? I’ve discussed many times in this blog how we can visualize single mRNA molecules by using MS2-like systems. These systems allow multiple fluorescent proteins (FPs) on a single mRNA molecule, thus enhancing signal to noise ratio (SNR). However, it’s not advisable to attach multiple FPs to a single protein – it will probably make this protein inactive, if not worse.
So, they had no choice but to fuse the Sox2 and Oct4 to a single fluorescent tag. Out of the endless choices available these days, the TFs were fused to a HaloTag ( developed by Promega). HaloTag is a system which consists of HaloTag protein (a mutated form of a bacterial haloalkane dehalogenase, DhaA) and a modified ligand molecule. This ligand consists of a reactive linker and a functional group (such as biotin or a fluorescent dye). Once the ligand binds the protein, it forms a covalent bond. HaloTag can be fused to the protein of choice, just like any FP, except it is “dark” unless the ligand is applied. Furthermore, the ligand binds only existing proteins. This means that after the ligand is washed away, any nascent protein that is translated will be “dark”. The fluorescent dyes are, supposedly, brighter than the common FPs.
Still, there should be hundreds to thousands of Sox2-Halo proteins in the nucleus; how can we distinguish single molecules? -With a nice trick. Since these are TFs, they are supposed to bind DNA and, therefore, stay immobile for at least a short amount of time. A 10 millisecond (ms) exposure time was too short – too many molecules, even just Halo protein alone, were visible as diffraction limited spots. However, with a 500 ms exposure time, fast moving molecules only look like a blur, which greatly reduced the background. Thus, only immobile proteins did not appear as a blur. Furthermore, only a 2D cross-section of the nucleus was visualized, thus any molecules above or below the plane of imaging also blended into the background. Last, they used a low excitation power, limiting the number of emitting molecules. Using this approach, they found a two-component residence time (i.e. “immobile” molecules) of TFs on the DNA. Most molecules had a residence time of ~0.8 seconds, but some had an average residence time of ~12 seconds. The interpretation is that Sox2 “samples” the DNA in search of the specific binding sequence. This sampling takes, on average, 0.8s. Once bound, the protein stays on the DNA for 12 seconds. A control experiment with Sox2 without the DNA binding domain showed only the short residence time.
This data, from in vivo experiments, was then verified in a “clean” in vitro experimental set-up of single molecule binding. In this set-up, fluorescently-labeled DNA molecules are attached to a glass slide. The protein in question is then applied on the DNA and the residence time of the Halo-tagged protein on the DNA is measured by total internal reflection fluorescence (TIRF) microscopy. [TIRF allows imaging only to ~100nm hight above the glass surface, thus imaging only bound molecules and greatly reducing background. TIRF is also very useful in imaging the under surface of adherent cells]. This assay allowed examining different DNA lengths or mutated binding sites. The numbers obtained by this method were very consistent with the in vivo results.
One of the cooler stuff described in this paper is the multi-focal microscope. This microscope has the capacity to simultaneously take images in 9 focal planes, thus allowing really fast single molecule 3D imaging.
[I heard a talk about the construction of this microscope. Wasn’t easy… The grating had to be lithographed very precisely, at the 100 micron accuracy, i think. This was a one-color microscope, but I know they are developing a two color microscope. because of the difference in wave length of green and red, the grating for the green is not suitable for the red. Anyway, very cool microscope]
Using this method, they developed a kinetic model for Sox2 diffusion, search and binding to the DNA. Based on their results, their model suggests that Sox2 “samples” the DNA on average for 84 times before finding a target binding site. The search duration is ~377 seconds. This means it take a Sox2 transcription factor molecule ~6 minutes to find its target, but it only remains bound for several seconds , presumably to stimulate transcription.
Biochemical results suggest there are ~7000 binding sites for Sox2. They used FCS measurements to calculate how many Sox2 molecules there are in an ES cell nucleus. Their calculations suggest 115,000 molecules! Thus, on average, every Sox2 binding site is sampled every 24 second by a Sox2 molecule, meaning that the TF binding sites are occupied at least half the time.
In summary – this paper used several imaging techniques to get new insights on the behavior of TFs in live cells. Their results are really interesting and shed light on the dynamics of TF behavior, on the effect of TF expression level and diffusion rates and other aspects I did not mention here.
Chen J, Zhang Z, Li L, Chen BC, Revyakin A, Hajj B, Legant W, Dahan M, Lionnet T, Betzig E, Tjian R, & Liu Z (2014). Single-molecule dynamics of enhanceosome assembly in embryonic stem cells. Cell, 156 (6), 1274-85 PMID: 24630727