Cryogenic optical localization provides 3D protein structure data with Angstrom resolution

Starting with efforts in scanning nearfield optical microscopy in the 1980s, many laboratories around the world have contributed to pushing the resolution of optical imaging beyond the Abbe limit [1]. A particularly fruitful outcome has been the invention of "super-resolution microscopy", culminating in the award of the Nobel Prize for Chemistry in 2014. In one of these approaches, a single fluorescent molecule is imaged and its point-spread function localized with a precision that is limited only by the signal-to-noise ratio of the detection method. The challenge then is to address the individual molecule separately, so the signals of the neighbouring molecules do not overlap. In 2002, an ETH team led by Vahid Sandoghdar demonstrated 3D nanometre resolution of two molecules spaced by about ten nanometres [2]. In that work, individual molecules were selected via their narrow resonances at liquid helium temperature. The results were not however encouraging for biological samples because very narrow transitions occur only in aromatic hydrocarbon molecules, which are not water soluble. Sandoghdar resumed work on cryogenic super-resolution microscopy in 2010. This time, the aim was to use conventional dyes and simply exploit the slower photo-chemistry at low temperatures: reduced bleaching should result in more emitted photons per molecule and thus a higher shot-noise-limited signal-to-noise ratio. While the existing super-resolution methods typically reach a resolution of a few tens of nanometres at room temperature, we aimed at two orders of magnitude improvement down to the Å level. In the first phase of this ambitious MPL-based project, we resolved dye molecules separated by several nanometres on DNA double strands [3]. Finally, in 2015 we reached Å resolution in studying the protein structure. In recent work we examined the conformational state of the cytosolic Per-ARNT-Sim domain from the histidine kinase CitA in collaboration with Christian Griesinger and his group in Göttingen as well as the four binding sites of biotin in streptavidin (see figure). By employing algorithms from electron tomography, we reconstructed the 3D arrangement of the label molecules from their 2D projections [4], leading to the name "cryogenic optical localization in 3D" (COLD) for this method. COLD brings fluorescence microscopy to its fundamental limit dictated by the size of the label molecule.

Figure

[1] S. Weisenburger et al., Contemp. Phys. 56, 123 (2015).

[2] C. Hettich et al., Science 298, 385 (2002).

[3] S. Weisenburger et al., ChemPhysChem. 15, 763 (2014).

[4] S. Weisenburger et al., Nature Methods 14, 141 (2017).

Cryogenic optical localization provides 3D protein structure data with Angstrom resolution

S. Weisenburger, D.  Boening, B. Schomburg, K.  Giller, S.  Becker, C.  Griesinger, and V. Sandoghdar, Nature Methods 14, 141 (2017)

In press: 

Max Planck Society - Research News: Ein tiefer Blick ins Protein (2017)

FAU aktuell: Ein tiefer Blick ins Protein (2017)