Functional analysis of biological matter across dimensions by atomic force microscopy (AFM): from tissues to molecules and, ultimately, atoms

For a detailed understanding of biological tissues and proteins and their dynamical processes the 3D structures of the components involved must be known. Most of the structural data have been obtained through the combination of three major techniques: X-ray crystallography, NMR and TEM. These three methods enable the determination of the structure of biological macromolecules at near atomic resolution and each of those was developed over many years to perfection. Nevertheless each one has its advantage and limitations and all of them may be considered today as useful and complementary tools (Schabert and Engel 1995).
In addition to those established techniques, probably the most spectacular advances have been achieved with the AFM, which is the first imaging device allowing direct correlation between structural and functional states of biomolecules in their physiological environment over a range of scaling very much comparable to the capabilities of EM. Moreover, the AFM has the striking possibility to observe biological matter with high resolution in space, time and applied force and with a striking signal-to-noise ratio that typically allows to directly using unprocessed data.
Using time-lapse AFM, it is possible to monitor the structural and conformational changes with various biological processes such as the assembly/ disassembly mechanisms of proteins. The AFM tip can also be used to manipulate and control single molecules with forces in the piconewton range. Two interesting examples are the mechanical unfolding of proteins and the measurement of the actin-myosin interaction force.
Theoretically the AFM enables to image, probe and manipulate biomolecules in a fully native state at submolecular resolution. However, at the moment this is only feasible for reconstituted transmembrane proteins that are forming 2D crystals in vitro. Imaging of a huge variety of other interesting specimens, e.g. DNA, motor proteins, living cells or the sponge like nuclear pore complex, only reveal macromolecular or even a poorer resolution and protocols for investigating and analyzing other types of biological samples, e.g. connective tissues (cartilage, tendon, bone) or the plaque particles in the coronary arteries, so far are not worked out. In this doctoral thesis I refined or developed several new preparation protocols applicable for the investigation on a variety of biological or medical relevant samples:
In a first project we applied high resolution AFM imaging to the luminal and cytoplasmic face of the asymmetric unit membrane (AUM) of the urine bladder epithelium that forms 2D crystalline plaques in vivo. Those urothelial plaques are meant to be involved in a variety of pathological processes, like bladder cancer that ranges among the five most frequent cancer diseases. Moreover, the investigation of AUM in the AFM could offer a basic understanding of the plaque forming process at the innermost part of the bladder, which potentially might also give new insights into a wide variety of structural and dynamical changes occurring in a lipid membrane in situ.
In a second project we succeeded to establish a protocol for the robust immobilization of different assemblies of the water soluble protein actin, a 43-kDa protein that plays a fundamental role in muscle contraction, cytoskeletal processes and motility. We achieved high-resolution AFM images of the G- and F-actin arrays that hold promise for detailed structural analyses at the molecular level of morphological changes induced by chemical effectors. This project aims ultimately to trace the muscle-dynamics in its different conformational steps throughout the ATPase/ cross-bridge cycle and at the level of single molecules, i.e. to trace the mechanical interaction of the molecular motor myosin with actin filaments.
In a third project we developed new preparation and evaluation protocols that should enable us to directly measure the critical biomechanical parameters of soft biological tissue at the micrometer to nanometer scale. Therefore AFM imaging and indentation-based AFM have enabled us to trace alterations of the tissue architecture at all relevant scales. As a first clinically relevant sample we imaged and mechanically tested articular cartilage in its normal state and compared it to a disease state caused by osteoarthritis. These results are paving new vistas for monitoring disease processes at the scale they are actually occurring, i.e. at the cellular to molecular level. Elasticity measurements at different scales on native articular cartilage provided us with a better understanding of cartilage biomechanics and offer new possibilities in mapping cartilage mechanical properties, which will also include the viscoelastic properties in further measurements.
In a next step, individual building blocks were specifically removed out of the tissue architecture by enzyme action. Those digestion experiments helped us to model or mimic alterations caused by pathological degradation processes. At this stage sufficiently smooth samples for high resolution imaging can only be prepared from frozen samples by employing cryo-microtomy. For a refinement of the procedure a protocol has to be developed for combining high resolution imaging with the mechanical probing at the specific sites of interest. Of course the final goal would be that all preparation steps could be done under near physiological conditions.
Concerning clinically relevant applications, this work opens the possibility of an in vivo analysis of cartilage consistence in the knee or the investigation of plaque particles in the coronary arteries. This directly leads to the ultimate goal of in situ force mapping by a minimally invasive procedure, i.e. by incorporating the AFM into an arthroscope or into a heart catheter.
In summary: In the context of structural research in biology and medicine the AFM is a sensitive microscopical technique with exciting new possibilities to explore new areas of revealing interactions of molecules and merging them with those obtained in the related fields of biochemistry and genetics or by different techniques to gain a comprehensive understanding of cell function: motility, organelle dynamics, membrane transport, and regulation. Moreover, it might be very useful to integrate AFM and AFM-based techniques rather than separating the different techniques and disciplines. Structural, genetical and biochemical data might eventually be composed with the intermediate resolution of EM and AFM and atomic resolution achieved by X-ray crystallography and NMR. In a truly integrated approach the AFM should be complementary to – and not competing with – other microscopical techniques, diffraction methods and spectroscopies. The power of AFM is rooted in its unique capabilities of an extremely high dimensional resolution capability and its resolution in applying and/ or measuring forces that potentially allows the tracing of the fundamental biological processes of life, for example, the translocation of mRNA into protein by the ribosome, the production of molecular “fuel” (e.g., ATP) in mitochondria, and the translocation of ions and small molecules across membranes by channels and pumps. Further developments in sample preparation and instrumentation will open up new AFM capabilities in terms of high-resolution imaging, a better temporal resolution in time-lapse imaging, therefore leading to a real-time imaging and a more time efficient force-mapping and maybe a direct data processing.
A higher spatial resolution could be achieved by employing a more sensitive feedback system that requires lower loading forces in combination or in addition to novel and more efficient operational modes, like an improved and more sensitive tapping mode or even better, a real non-contact mode in liquid.
Currently new AFMs are designed that are specialized for faster recording of data (Ando et al. 2001; Viani et al. 1999). These instruments operate shorter cantilevers with low-noise characteristics that respond much faster than those currently in use. The new design allows much higher image speed for tapping mode in liquid and also drastically reduces the time for force mapping. Recently a scan rate of up to one frame per second and one to a few minutes for the force mapping have been achieved respectively. This is an important step to trace conformational changes of biological matter in their physiological states and studying the dynamic processes in real time. In combination with more specific software this eventually allows the recording frames at video speed in the future.
As stated in the introduction, AFM does not only provide the “eyes’’ for high resolution imaging but it also provides the “fingers’’ to measure and manipulate matter at the atomic and molecular level. This is of great relevance in the life sciences: Recently it has been demonstrated by dissecting DNA from a particular part of the chromosome that the probe can also be used as a nanoscalpel to dissect fragile biological samples (Thalhammer et al. 1997).
In this context, force-spectroscopy is another dedicated new methodology, which continuously is under development for investigation, the dynamical processes occurring during the unfolding of a biomolecule (Oesterhelt et al. 2000; Rief et al. 2000; Rief et al. 1997). It might be possible to determine forces generated by a biomolecule on the microsecond time scale. A faster and more sensitive force-spectroscopy based on AFM might allow studying biological processes with greater resolution in time and applied force.
By employing IT AFM the tip can be placed directly atop individual structural features to measure its biomechanical properties. In this doctoral thesis I have demonstrated that biomechanical information such as the elastic modulus can be strongly dependent on the experimental scale which is given by the size of the indenter. This could be especially interesting for putative clinical applications, such as obtaining clinical relevant information at fundamentally different levels of tissues architecture. The attainable spatial resolution is directly related to the shape and sharpness of the scanning probe tip. The apex of commercially available tips typically has typical radii of curvature of some nanometers (Czajkowsky et al. 2000) and will improve in the future in terms of sharpness and in reduction of cantilever noise. There may be a practical advantage in using other tips grown in situ onto cantilevers or carbon nanotubes using an electron microscope. Both types of tips possess many unique properties that make them ideal AFM probes. Their high aspect ratio provides faithful imaging of deep trenches, while good resolution is retained due to their nanometer-scale diameter. These geometrical factors also lead to reduce tip-sample adhesion, which therefore might allow a gentler imaging. New types of cantilevers might be more specific in terms of their physical or chemical functioning of the probe or on the other hand biochemically compatible for biomedical applications. Multifunctional cantilevers will provide the extension to a whole set of new experimental opportunities of monitoring possibly a wide range of multiple signals, like voltage, currents of ions or electrons and fluorescently labeled molecules (Engel et al. 1999).
Nanotubes are electrically conductive, which allow their use for the measurements of (surface-) potentials, and they can be modified at their ends with specific chemical or biological groups for high resolution functional imaging. Nanotubes could also be used as nanopipettes for drug delivery or as an electrical conductive tip for measuring local potentials, currents or trafficking of biomolecules across membranes.
Arrays of cantilevers (millipede) are under development that could be used for mechanically probing simultaneously neighboring regions of the specimen. This could allow mapping mechanical properties or imaging of larger regions within a short time or a more reliable diagnosis, i.e. different tissues or the plaques in the coronal arteries involved in heart diseases, all of them with an extremely high interest for medical research and treatment of diseases.
Combined AFM-SNOM (scanning near-field optical microscopy) - probe tips are potentially of great use and can be handled almost like the commonly used AFM probes. SNOM allows the detection of a conventional AFM height image and additionally an optical image with an improved optical resolution of more than one order of magnitude compared to conventional LM. SNOM therefore offers in addition to the topography also corresponding optical information of fluorescently labeled biological features beyond the diffraction limit of conventional light microscopy. One potential application might be the detection of the green fluorescent protein expressed by some cell strains. These tips can either be used as a detector to follow for example the specimen of green fluorescent proteins or as an effector to induce photoactive processes at the single molecule level.
One of our future projects is to detect the bidirectional trafficking through the nuclear pore complex by fluorescently labeled cargo. To achieve this we are developing a protocol to tightly seal the nuclear envelope of Xenopus oozytes over microfabricated pores. The goal is to measure in their complete functional state the transport rates of these nuclear gates at the level of single pores. In a comparison with genetically modified pores we aim to understand the regulation of cargoes between the nucleus and cytoplasm. The impacts of nano science in medically relevant applications foster the development of new devices and analytical tools for clinical practice. Due to the very high signal-to-noise ratio, AFM-based techniques offer the observation, probing and characterization of the chemical and mechanical properties of single molecules within a tissue in a physiologically relevant environment. Therefore pathological changes in human tissue can be registered and hopefully cured at an early stage of the disease progression. On the other hand by expanding our ability to characterize and control biological tissue in much more detail diagnostics and therapeutics could be revolutionized in terms of delivery of drugs in small amounts with a high resolution to the target.
The new possibilities opened by the AFM will also have an important impact for engineering and (quality-) control of surface properties of new biocompatible implants, tissue-engineered organs or novel biodegradable products that can be resorbed by the body after use. Progress in nanobiotechnology will promote the healing of natural tissue and will stimulate mimicking high-performance biocompatible nanosystems such as artificial active components or biological implants for organ and joint replacement. Thus new implants will be designed with nanoscale features that will replace degenerated, diseased or damaged tissues, e.g. arthritic cartilage or bone.
Since its invention the AFM has opened new avenues for a wide range of exciting applications ranging from basic structural research up to new developments for in situ applications in a clinical environment. However, diseases are often caused largely by damages either at the molecular or the cellular scale. For a local assessment today’s surgical tools are simply too big. In the future we should get tools that are molecular, both in their size and their precision; and then we will be able to intervene directly at the level where the damage occurs and correct it. Ultimately the dream of Richard Feynman and Albert R. Hibbs might become true that once nanoscale machines can be injected into the bloodstream for dialysis or drug delivery.