A tapping mode atomic force microscope (AFM) method for the visualization of plasmid DNA, cytoplasmic proteins, and DNA-protein complexes is described. The method includes alternate approaches for preparing samples for AFM imaging following biochemical manipulation. DNA containing specific protein interacting regions are observed in near-physiologic buffer conditions.
Atomic force microscopy (AFM) allows for the visualizing of individual proteins, DNA molecules, protein-protein complexes, and DNA-protein complexes. On the end of the microscope’s cantilever is a nano-scale probe, which traverses image areas ranging from nanometers to micrometers, measuring the elevation of macromolecules resting on the substrate surface at any given point. Electrostatic forces cause proteins, lipids, and nucleic acids to loosely attach to the substrate in random orientations and permit imaging. The generated data resemble a topographical map, where the macromolecules resolve as three-dimensional particles of discrete sizes (Figure 1) 1,2. Tapping mode AFM involves the repeated oscillation of the cantilever, which permits imaging of relatively soft biomaterials such as DNA and proteins. One of the notable benefits of AFM over other nanoscale microscopy techniques is its relative adaptability to visualize individual proteins and macromolecular complexes in aqueous buffers, including near-physiologic buffered conditions, in real-time, and without staining or coating the sample to be imaged.
The method presented here describes the imaging of DNA and an immunoadsorbed transcription factor (i.e. the glucocorticoid receptor, GR) in buffered solution (Figure 2). Immunoadsorbed proteins and protein complexes can be separated from the immunoadsorbing antibody-bead pellet by competition with the antibody epitope and then imaged (Figure 2A). This allows for biochemical manipulation of the biomolecules of interest prior to imaging. Once purified, DNA and proteins can be mixed and the resultant interacting complex can be imaged as well. Binding of DNA to mica requires a divalent cation 3,such as Ni2+ or Mg2+, which can be added to sample buffers yet maintain protein activity. Using a similar approach, AFM has been utilized to visualize individual enzymes, including RNA polymerase 4 and a repair enzyme 5, bound to individual DNA strands. These experiments provide significant insight into the protein-protein and DNA-protein biophysical interactions taking place at the molecular level. Imaging individual macromolecular particles with AFM can be useful for determining particle homogeneity and for identifying the physical arrangement of constituent components of the imaged particles. While the present method was developed for visualization of GR-chaperone protein complexes 1,2 and DNA strands to which the GR can bind, it can be applied broadly to imaging DNA and protein samples from a variety of sources.
1. Preparing DNA and protein samples to be imaged free of contaminants
2. Mounting AFM probe
3. Locating cantilever tip, aligning laser, and adjusting photodetector
4. Positioning AFM head and tuning cantilever
5. Imaging mica sample surface
6. Imaging biomolecules of interest
7. Representative Results:
Examples of AFM images are presented in Figure 4. Mica substrate (A) provides a molecularly flat surface onto which DNA and protein can adsorb. Imaging mica prior to biomolecule samples provides a negative control and assessment of imaging noise. It also provides a level of assurance that the cantilever is properly tuned and subsequent sample imaging will be successful. Double-stranded plasmid DNA (B, D), presumptively supercoiled, is readily identified by its asymmetric appearance and uniform depositing on the previously unremarkable mica substrate. Protein complexes of discreet particle sizes (C, E) are also uniquely distinguishable from the mica substrate. Particle size differences indicate sample heterogeneity, and can be useful to approximating protein complex stoichiometry or biochemical activity. The general diagonal shape and consistent orientation of the protein particles observed in Panel C is an imaging artifact, as proteins would expectedly be oriented on the mica substrate in a random fashion. Possible causes of the observed artifact is a physical tip abnormality or AFM imaging at too rapid of a scan rate. While tip convolution (illustrated in Figure 1B) prevents absolute length measurement calculations in the x and y axes, height measurement (z axis) and relative x and y measurements may be nonetheless useful for estimating biophysical properties of the imaged biomolecules.
Figure 1: Schematic presentation of tapping mode atomic force microscopy (AFM). A, the microscope. At the end of the cantilever is a sharp tip that oscillates up and down as it scans over the surface of a mica substrate to which biomolecular complexes adsorb. As the scanner moves in the x and y directions, a laser beam is reflected from the back of the cantilever onto a position-sensitive photodiode detector to map the vertical (z) distance the tip moves as it passes over the biomolecular complexes seated on the mica. B. Distortion of the image in the x and y directions caused by tip convolution. The nominal radius of a conventional silicon nitride tip is larger than the particles to be imaged, and the edge of the tip contacts the sample as it traverses the surface. Tip convolution results in the imaged particles appearing larger in the x and y directions, but not the z direction. This effect can be minimized by the use of tips with a smaller nominal tip radius.
Figure 2: Illustration of immunoadsorbed protein and DNA sample preparation. A, Release of immunoadsorbed GR•hsp70 protein complex from the monoclonal antibody (mAb)-protein A-Sepharose (PAS) pellet. Incubating the immunopellet with a peptide containing the mAb epitope will facilitate the release of the GR•hsp70 protein complex from the pellet, allowing the complex to be collected from the supernatant and visualized by AFM. B, Preparation of DNA for AFM imaging requires the use of a divalent cation adsorption buffer. The divalent cation increases the affinity of DNA for the mica substrate.
Figure 3: Final arrangement of the AFM head, liquid cell probe holder, sample, specimen holder, and buffer meniscus. A, Care must be taken when depositing sample and additional buffer onto the mica substrate to avoid physical contact between the pipet tip and AFM head, liquid call, and mica substrate. B, Magnification of A. The buffer meniscus spans from the specimen disc to the liquid cell probe holder.
Figure 4: Atomic force micrographs of mica substrate (A), 2xGal4-2xGRE-luciferease plasmid DNA 8 (B, D), and GR•hsp70 protein complexes (C, E) in buffered solution. Mica serves as a control image to compare with the visualizations of DNA and protein. GR•hsp70 protein complexes were prepared by immunoadsorption of GR primed with hsp70 and then released from the antibody-bead pellet using a mAb competing antibody (illustrated in Figure 2A). DNA was prepared by conventional plasmid miniprep. Panels A, B, and C are of an equivalent magnification (scale bars = 200 nm), as are Panels D and E (scale bars = 40 nm).
AFM provides a unique microscopic technique capable of imaging individual uncoated biomolecules in aqueous and near-physiologic buffered solutions in real-time. This allows for the visualization of individual proteins and DNA molecules, as well as multiprotein complexes and protein-DNA complexes. Imaging macromolecular particles with AFM can be useful for assessing sample homogeneity and for identifying the physical arrangement of the constituent components of the observed particles. This approach of observing individual macromolecular particles can be a useful adjunct to conventional biochemical techniques, such as immunoprecipitation assays, polyacrylamide gel electrophoresis, and size exclusion column chromatography, which provide significant summative data representing the 103-1011 individual biomolecular complexes of interest present in a sample.
Research into multiprotein complex stoichiometry, biomolecular interactions, and cofactor requirements can all be investigated using AFM. The method presented here can be adapted to accommodate specific biophysical questions of interest. For example, using a liquid cell probe holder capable of exchanging buffer, it is be possible to assay cofactor requirements for protein complex formation. DNA-protein assemblies can be formed and dissociated in near-real-time and even while being imaged. DNA can be generated with specific sequences of interest (e.g. a presumptive response elements or novel protein binding sequence), mixed with the hypothesized binding protein, and imaged to provide direct evidence of intermolecular interactions.
Tapping mode AFM imaging is substantially less complicated if dry samples are utilized rather than samples in aqueous solutions, and linear DNA stands and closed-circle DNA plasmids have both been imaged in this way 3,9. The method presented here utilizes samples in buffered solutions in order to provide a more physiological imaging environment. Other noteworthy methods of imaging proteins should also be considered, including near field infrared microscopy 10. The complementary use of immunoadsorption in consort with AFM provides an opportunity to prepare a myriad of protein complexes, using well-established biochemical techniques, and then visualize them following release from the immunoadsorbing antibody-bead pellet. For example, immunoadsorbed GR has been assayed for its association with the molecular chaperone protein hsp70 1 as well as the dynein motor protein 2 using this approach in order to estimate complex size and stoichiometries. Particle size and biophysical features (e.g. rigidity) have been useful in ascertaining the identity of the biomolecules observed in AFM imaged samples 11,12. It is also possible to confirm biomolecule identity by the addition of an interacting biomolecule (e.g. a ligand or monoclonal antibody), that would bind to its target and cause a particle size increase if the target biomolecule is present.
The authors have nothing to disclose.
This work was funded by National Institutes of Health Grant GM086822. The authors would like to thank Drs. Alec Pakhomov & Paul Wallace (Univ. of Washington Nanotechnology User Facility (NTUF)) and Andrea Slade (Bruker AXS) for their expert technical assistance. DNA AFM imaging was conducted at the Univ. of Washington NTUF, a member of the National Nanotechnology Infrastructure Network. The 2xGal4-2xGRE-luciferease plasmid was kindly provided by the laboratory of Dr. Keith Yamamoto (Univ. of California, San Francisco).
Name of the reagent | Company | Catalogue number |
---|---|---|
Dimension 3100 | Bruker | Dimension 3100 |
Dimension Fluid Cell | Bruker | DTFML-DD-HE |
Sharp nitride lever (SNL) silicon nitride AFM probe | Bruker | SNL-10 |
Microlever sharp nitride lever (MSNL) silicon nitride AFM probe | Bruker | MSNL-10 |
NanoScope AFM instrument software | Bruker | 004-132-000 |
Metal AFM specimen discs | Ted Pella | 16208 |
Grade V1 Mica Discs, 12 mm | Ted Pella | 50-12 |