Quantitative single molecule imaging in living cells and animals

Speaker: Christof Gebhardt

Department: Institute of Biophysics, Universität Ulm

Subject: Quantitative single molecule imaging in living cells and animals

Location: Bionanoscience Department, TU Delft

Date: 07-04-2017

Author: Mirte Golverdingen

Life is an interplay of inanimate biomolecules, Christof Gebhardt tries to understand how these biomolecules function in isolation, interaction and what happens with them when we are ill. To understand this the research to reaction kinetics, their spatial distribution, stoichiometry and forces can help. Moreover, biomolecules can be better understood if we manipulate single biomolecules in vitro and in vivo.

Super resolution methods in combination with fluorescent proteins as markers can help us to understand biomolecules better. Gebhardt uses super resolution in combination with GFP. Next to green colours, other colours of the fluorescent protein are developed. However, single GFP expression for single molecule detection is hard to visualize using fluorescent microscopy because of the fluorescent background.

HILO illumination is an imaging method that only measures a small part of the cell, in this way the background can be reduced.  By adding a mirror to the HILO set-up, the side of the object can be imaged and in this way, you have more space to alter the cell at the top (See Figure 1). Moreover, imaging from the side reduces the contrast. The 3D image of the cell can be imaged by moving the objective up and down. By using other tags than GFP, Gebhardt was able to use HILO for in vitro, in cellula, in vivo and in silico studies.


Figure 1: Scheme of the reflected light-sheet principle (not drawn to scale). A laser beam is focused by an objective to form a vertical light sheet that is reflected by 90° off an AFM cantilever next to a cell in a Petri dish. Fluorescence is detected by a second high-NA objective. Three-dimensional optical sectioning is achieved through vertical displacement of the sample. Adapted from: Gebhardt, J. C. M., et. al (2013). Single-molecule imaging of transcription factor binding to DNA in live mammalian cells. Nature methods, 10(5), 421-426.

Gebhardt for instance, studied the transcription factor and DNA (TF-DNA) interaction times in chromatin organisation and transcription regulation. The transcriptional repressor CTCF interacts with the chromatin, however, it is not known how CtCF interacts with the chromatin during the cell cycle. Moreover, the influence of the regulator role of TF-DNA interaction time on transcription repression is a research topic of the research group of Gebhardt.

Kinetic information from the data can be obtained by researching video images of with an HoloAR labelled nuclear receptors the diffusion coefficient, the DNA bound tag and the DNA residence time of molecules can be obtained. To overcome the problem of photobleaching that occurs after continuous illumination, dark times were used. By measuring the DNA residence time of different TFs, Gebhardt showed that every transcription factor has another DNA residence time. There can be made a differentiation between specific (1/10/100 s) and unspecific (<1) DNA binding residence times.

CtCF chromatin interaction kinetics

The 3D organization of chromatin exists of nucleosomes, looped gene loci and chromosome territories. The chromatin architecture influences gene expression, differentiation, development, DNA replication and DNA repair. However, the control of the DNA loops is still unclear.

Chromatin topology is mediated by architectural proteins. Gebhardt was able to calculate the topologically associating domain, this shows how likely it is in a loop if chromatins interact with each other. There are certain TFs that is able to interact with the architectural proteins. CtCF is one of these transcription factors, therefore, it is seen as a potential ‘master weaver’. However, it is not clear if CtCF is activating, repressing or insulating the architectural proteins. CtCF contains 11 zinc finger motives that are able to dimerize the DNA. It has more than 100000 target sites on the DNA and it recruites cohesion. Gebhardt his study focusses on the kinetics of the CtCF TF, how long does it bind to the chromatin? And is the TF stable or mobile?

He was able to design a method by using fluorescent proteins that shows a wide spread binding kinetics. By developing a method based on the measured fluorescence, Gebhardt showed that CtCF shows three dissociation states. The residence times of these states are: <1s, <10s and ~1000s. The last two are specific, while the first one is specific and occurring in G1 and G2 phase of the cell cycle.

CtCF has low- and high-affinity sites on the DNA and its dynamic interactions (~4s) are changing the loops within topological associating domains (TADs) or initiate abortive loop formation. CtCF chromatin interaction depends on the current phase of the cell-cycle. In the S-phase, much less high (stable) resistance times are measures, so in the S-phase an overexpression of CtCF blocks cells. It could be that replication dissolves CTCF-chromatin interactions during S-phase. In the M-phase: CtCF is excluded from chromatin, this may be caused by phosphorylation.

By studying the interaction of CtCF with the DNA, Gebhardt et al. showed that CtCF-chromatin interaction is cell-cycle dependent. Kinetic information on the interaction of biomolecules with DNA can help us understand the complex interactions in the cell better. I liked that he was able to adapt the traditional way of imaging with HILO to a more advanced and better adapted method. Moreover, the discovery of the tree dissociation states of CtCF with their corresponding residence times invites us to research the dissociation states of other transcription factors, giving more insight to the complex cell dynamics.


3D organisation of the mammalian genome and its function

Speaker: Wouter de Laat
Department: Bionanoscience
Subject: 3D organisation of the mammalian genome and its function
Location: Delft University of Technology
Date: 2016-02-11
Author: Romano van Genderen

The seminar started by debunking a common myth, namely that 97% of all DNA, which is often considered “junk DNA” actually consists of all sorts of regulative elements and switches, which are turned off and on in specific tissues, causing the morphological changes. But not only the genetic information on this non-coding plays a regulative role, also the role it plays in changing the 3D organisation of the coding DNA plays a role.

The first experiment he presented that showed evidence for this was done by the Spitz lab. They showed that by moving a so-called transposon cassette, a piece of DNA that can be inserted in the genome, across the genome, the expression of specific proteins can be influenced. Sometimes a small step with the cassette leads to a large change in expression, while a large step does not make a huge difference. This was used to show that these two sites in the large step are separate on the genome, but topologically very close.

Another experimental method he presented was the 3C (chromosome conformation capture) method and other derived methods. 3C is based on formaldehyde linking the proteins on the DNA together into a sort of knot, then digesting them to cut the knot out. This is ligated and read to show which parts of the DNA are close together in space. He showed an example where this method was used to study the haemoglobin protein. Also he showed some derivative methods like 4C and Hi-C.

Afterwards he explained the basis of higher-scale genome folding in mammals. The main point he wanted to introduce were the so-called topologically active domains, also called TAD’s. These are loops in the DNA, separated from one another by a protein called CTCF. These domains also act as functional domains with promotors and enhancers relatively close by. A next step in folding is made by sorting the genome on the basis of activity. The active domains bind one another and so do the inactive domains. Afterwards, the inactive domains move towards the nuclear periphery. He showed a practical application of this by explaining how leukaemia originates.

Next he elaborated on the CTCF protein, which forms the anchor of a chromatin loop. The most important property of the protein is that it has a direction. In order for it to function properly and form a loop, the two proteins must not be in the same direction. This is better explained in the following image:


Image 1: Image explaining the directionality of CTCF. Source: E. de Wit et al. CTCF binding polarity determines chromatin looping, Mol Cell, 60 (2015), pp. 676–684

Finally, he showed how 3C methods could be used for haplotyping, so to differentiate the maternal, paternal and foetal genome. This is because the more advanced 4C method can detect translocations in the genome. Because these C methods are based on location and not on sequence, a small difference in genome, like a few more SNP’s, are ignored. Using this haplotyping, a non-invasive form of prenatal diagnosis can be done using the fragments of foetal DNA found in the mother’s blood.

The thing I found most striking about this presentation is the situation that occurred during the questions. There was one person who was more of a proponent of the topological supercoiling model than the TAD model. Their discussions showed that sometimes one model is not sufficient to explain all the experiments.