Proteins in Motion
Towards Filming Bio-Molecular Processes at the Nano-Scale
- Subsequent HS-AFM images (9.4 x 9.4 nm, 10 fps) of an individual aquaglyceroporin (E. coli GlpF) embedded in a lipid membrane.
- Fig. 1: a, Topography view of a purple membrane. b, Consecutive AFM images of a single antibody. c and d, Scheme depicting the walking of an antibody molecule. Red arrow indicates conformational change of one Fab arm. White arrow indicates same conformation.
- Fig. 2: Hsp70 oligomerization. a, A mixture of Hsp70 monomers and dimers (circles) were observed. b, Superimposition of Hsp70 with the crystal structure of the bacterial homologue DnaK. c, High resolution imaging of Hsp70 reveals a dynamic nature of Hsp70 dimers as NBD1 moves around NBD2. At the lower row trimerization was observed. d, Schematic model of a Hsp70 dimer.
- Fig. 3: a, GlpF X-ray structure and HS-AFM time series of a single GlpF tetramer embedded in a supported lipid bilayer. b, X-ray structure of GlpFs loop C with stabilizing H-bonds and salt bridges (yellow lines); Radial stiffness map determined from HS-AFM imaging. c, AqpZ X-ray structure and HS-AFM time series of a single AqpZ tetramer embedded in a supported lipid bilayer. d, X-ray structure of AqpZs loop C; Radial stiffness map determined from HS-AFM imaging.
For a long time, observing dynamic biological processes with the atomic force microscope (AFM) was out of reach because it took minutes to acquire an image, which is much too slow to follow processes that typically occur in the millisecond to second time regime. Within the last decade, high-speed AFM has removed this restriction and allows to ‘watch’ biological processes with sub-molecular resolution in real-time, label-free and under physiological conditions.
High-Speed Atomic Force Microscopy
In ‘Atomic force microscopy’ (AFM), initially invented to visualize atoms, a sharp tip mounted on a micro cantilever scans horizontally, line-by-line, across a surface to generate a topographic image. This method has in the meantime developed into the high-resolution tool of choice for measurements under physiological conditions with respect to biological imaging, bio-molecular recognition and localization of specific binding sites. A tremendous variety of proteins, molecular complexes, molecular templates, and membranes has been investigated in the last two decades, revealing lateral resolution at the nano-scale – however, for a long time it suffered from a low imaging rate and was unable to capture fast dynamic processes. Several approaches have been pursued to overcome these limitations including sophisticated technical developments for increasing the imaging speed . An impressive breakthrough was accomplished by revealing how the motor protein myosin V uses its ‘legs’ to ‘walk’ on actin with at 100 ms time-scale . The current version of High-speed AFM (HS-AFM) is capable of recording movies of the interactions and the dynamic structural changes of bio-molecules at nm spatial and a few ms temporal resolution. Our report focuses on the investigation of dynamic biological phenomena at the nano-scale, including antibody-antigen interactions, protein clustering/aggregation/oligomerization, and internal protein dynamics.
Filming Conformational Changes of Antibodies during Recognition
With the recent developments of HS-AFM, directly filming dynamic structural changes of proteins under physiological solutions has become possible.
In pioneering immunological studies with HS- AFM recently carried out in our lab  (fig. 1), we discovered that antibodies were rapidly “walking” on bacterial and viral membranes (fig. 1a). Bivalent attachment of an antibody to surface antigens lead to mechanical strain, originating from a mismatch between the intrinsic symmetry of the antibody and the antigens (fig. 1b). This resulted in accelerated dissociation and immediate rebinding which caused random ‘walking’ of the antibodies (fig. 1c and d). As a result, antibody clusters are formed that may constitute nucleation sites for the docking of phagocytes and the complement system to the Fc stem, a prerequisite for immunological clearance. This phenomenon depicts a completely novel mechanism of the dynamic behavior of antibodies on native antigenic membranes.
Understanding the lateral organization of proteins in lipid membranes is of key importance, as molecular associates often trigger complex cellular processes. However, the dynamics underlying this process cannot be easily obtained by other structural techniques than HS-AFM: A recent example for such an application is the organization and orientation of the stress inducible heat shock protein 70 (Hsp70) on tumor (e.g. melanoma) cell surfaces, where its clustering facilitates Hsp70-mediated endocytosis . As a possible factor that may govern clustering of Hsp70 we identified oligomerization: In line with previous biochemical data, dimers and higher order oligomers of recombinant human Hsp70 were visualized by high resolution HS-AFM (fig. 2a). At higher magnification, the nucleotide-binding domains (NBDs, fig. 2b,c) of dimers and trimers were clearly identified. The substrate-binding domains (SBDs) were poorly resolved, due to their lack of association to the surface and their flexible connection to the NBD via the linker domain. Analysis of the relative positions of NBDs (fig. 2c) revealed a fairly stable, yet flexible dimer structure (sketch fig. 2d). A dynamic formation of trimers was also evident (fig. 2c). This example demonstrates how HS-AFM gains direct structural-kinetic data at the molecular level that are required to understand processes at the cellular level and beyond.
Direct Imaging of Protein Domain Elasticity
Membrane proteins represent one of the main pharmaceutical targets. The flexibilities of extracellular loops crucially determine ligand binding and activation of membrane receptors. Classical experimental methods like NMR, X-ray diffraction, and neutron scattering are unable to provide quantitative flexibility values of individual protein subdomains. Accurate predictions are difficult to compile, as small changes in amino acid position and orientation relative to both the membrane surface and to neighboring moieties may significantly alter the interaction forces with other loops and, therefore, its flexibility. By directly assessing the topographical variability of individual protein subdomains by means of HS-AFM, we recently quantified flexibility values . We directly visualized the thermal motion of the main periplasmic loops of two structurally highly homologous aquaporins, E. Coli AqpZ and GlpF in their native lipid environment at sub-second timescale (fig. 3a and c). Analysis of individual loop positions over time allowed for the determination of their configuration space, from which the free energy landscape and the protein flexibilities underlying loop motions were derived. Mapping the obtained stiffness values onto the crystal structures (fig. 3b and d) revealed the role of salt bridges and H-bonds as stabilizing elements in protein structures: (i) Loops appear stiffer along the directions of potential bonds connecting them to other domains of the proteins, whereas they appear less stiff in perpendicular direction; (ii) a higher amount of bonds increases the overall stiffness of the loops. Suchlike information is invaluable for the improvement of in silico docking studies. Matching the flexibility map with the network of polar bonds from structural models offers the possibility to validate results obtained with sub-nm resolution under non-physiological conditions – as with X-ray crystallography or electron microscopy. Such flexibility maps may also ease the prediction of the conformational transitions that membrane transporters and channels undergo when they interact with their substrates and ligands.
Tremendous progress was accomplished since the development of a biocompatible HS-AFM, which provides up to 10 ms temporal and 1 nm spatial resolution at low invasiveness. Proteins in its native biological action were readily captured at video-rate, yielding novel insights into their working mechanisms. As this new technique significantly contributed to the progress in bio-science in the last years, biological researchers may be inspired to consider the use of HS-AFM in their studies to directly observe molecular and cellular phenomena at unprecedented spatio-temporal resolution.
 T. Ando, T. Uchihashi and S. Scheuring: Filming biomolecular processes by high-speed atomic force microscopy, Chem. Rev. 114, 3120-3188 (2014) DOI: 10.1021/cr4003837
 N. Kodera, D. Yamamoto, R. Ishikawa and T. Ando: Video imaging of walking myosin V by high-speed atomic force microscopy, Nature 468, 72-76 (2010) DOI 10.1038/nature09450
 J. Preiner, N. Kodera, J. Tang, A. Ebner, M. Brameshuber, D. Blaas, N. Gelbmann, H. J. Gruber, T. Ando and P. Hinterdorfer: IgGs are made for walking on bacterial and viral surfaces, Nature communications 5, (2014) DOI 10.1038/ncomms5394
 B. Nimmervoll, L. A. Chtcheglova, K. Juhasz, N. Cremades, F. A. Aprile, A. Sonnleitner, P. Hinterdorfer, L. Vigh, J. Preiner and Z. Balogi: Cell surface localised Hsp70 is a cancer specific regulator of clathrin-independent endocytosis, FEBS Lett. (2015) DOI 10.1016/j.febslet.2015.07.037
 J. Preiner, A. Horner, A. Karner, N. Ollinger, C. Siligan, P. Pohl and P. Hinterdorfer: High-speed AFM images of thermal motion provide stiffness map of interfacial membrane protein moieties, Nano Lett. 15, 759-763 (2014) DOI 10.1021/nl504478f