Super-resolution Live Cell Imaging to Study DNA Looping in 3D

The human genome contains ~20,000 genes that encode the proteins responsible for cellular and organismal function. Dysregulation of gene expression is a frequent cause of disease; for example, overexpression of an oncogene may result in cancer.

However, it has been challenging to understand how gene expression is regulated in mammals, because the key regulatory regions within the genome – called enhancers – tend to be located far away from the genes they regulate. This has lead researchers to investigate the mechanism of DNA looping, which brings distant regions of DNA close together.

Dr. Anders Sejr Hansen (far left) and lab members (NAMES NEEDED).

From left to right: Dr. Anders Sejr Hansen and lab members Hugo B. Brandão, Asmita Jha and Michele Gabriele.

Dr. Anders Sejr Hansen and his team at the Massachusetts Institute of Technology, USA, seek to understand the 3D folding of the genome in order to understand how distal enhancers find and loop to their target genes and regulate their expression. Recently, they were able to publish new data in this area of discovery using live cell super-resolution microscopy with ZEISS LSM 900 with Airyscan 2.

The dynamics of DNA loops

Previously, DNA looping has been studied with genomics methods that can only generate a static snapshot. Thus, the time-dimension was missing. It was not known whether DNA loops were dynamic or stable structures.

Diagram of CTCF/cohesin mediated DNA loop

Diagram of CTCF/cohesin mediated DNA loop

In M. Gabriele et al., Dr. Hansen and his team use super-resolution live cell imaging with ZEISS LSM 900 with Airyscan 2 to directly visualize a specific type of DNA loop – namely, DNA loops that are mediated by the proteins CTCF and cohesin – in living cells, for the first time. Their experiments using microscopy sought to answer two key questions:

1) Are these loops stable or dynamic structures?

2) Are these loops rare or present in most single cells?

Super-resolution microscopy of dynamic and short-lived DNA loops

Dr. Hansen’s team used localization microscopy implemented on the ZEISS LSM 900 point-scanning confocal with super-resolution Airyscan detector to visualize these CTCF/cohesin loops. DNA loop anchors were fluorescently labeled using DNA arrays, called TetO and ANCHOR3, which were molecularly engineered to be specifically located near the CTCF sites at the base of DNA loops. The fluorescent labels, TetR-3x-mScarlet and EGFP-OR3, bind to these arrays and can be visualized with fluorescence microscopy.

Using 3D localization microscopy, the 3D coordinates of each DNA loop anchor were determined and then the distance between loop anchor sites is used as a metric for loop status (high 3D distance, loop unlikely; low 3D distance, loop more likely). By collecting z-stacks over time, they tracked the 3D distance between the loop anchors over time.

Live-imaging of DNA looping. Representative XY maximum intensity projections of a typical 3D imaging field of view containing a single mESC colony from the C36 cell line. The two anchors of the DNA loop are labeled in green and magenta and the movie runs for just over 2 hours (365 timepoints with 20 seconds between frames).

Their experiments found that CTCF/cohesin loops are both dynamic and rare, with median lifetime of just 10-30 minutes and present only ~3-6.5% of the time. These results suggest that models for how CTCF/cohesin loops regulate gene expression may need to be revisited. If the fully looped state is so rare, they suggest it is unlikely to be the most functionally important state.

The ZEISS LSM 900 confocal microscope with super-resolution Airyscan detector was crucial as it combines excellent optical sectioning with fast 3D imaging at a very high signal-to-noise ratio.

Dr. A.S. Hansen, MIT

What’s next

In this study, they focused on only one DNA loop in one cell type. Therefore, Dr. Hansen says that future studies will be necessary to test the generality of their results. They anticipate that the new experimental, imaging, and computational methods developed in the paper will facilitate such future studies.

Read Next