A method for investigating the structure of a protein photoreceptor using atomic force microscopy (AFM) is described in this paper. PeakForce Quantitative Nanomechanical Property Mapping (PF-QNM) reveals intact protein dimers on a mica surface.
Atomic force microscopy (AFM) uses a pyramidal tip attached to a cantilever to probe the force response of a surface. The deflections of the tip can be measured to ~10 pN by a laser and sectored detector, which can be converted to image topography. Amplitude modulation or “tapping mode” AFM involves the probe making intermittent contact with the surface while oscillating at its resonant frequency to produce an image. Used in conjunction with a fluid cell, tapping-mode AFM enables the imaging of biological macromolecules such as proteins in physiologically relevant conditions. Tapping-mode AFM requires manual tuning of the probe and frequent adjustments of a multitude of scanning parameters which can be challenging for inexperienced users. To obtain high-quality images, these adjustments are the most time consuming.
PeakForce Quantitative Nanomechanical Property Mapping (PF-QNM) produces an image by measuring a force response curve for every point of contact with the sample. With ScanAsyst software, PF-QNM can be automated. This software adjusts the set-point, drive frequency, scan rate, gains, and other important scanning parameters automatically for a given sample. Not only does this process protect both fragile probes and samples, it significantly reduces the time required to obtain high resolution images. PF-QNM is compatible for AFM imaging in fluid; therefore, it has extensive application for imaging biologically relevant materials.
The method presented in this paper describes the application of PF-QNM to obtain images of a bacterial red-light photoreceptor, RpBphP3 (P3), from photosynthetic R. palustris in its light-adapted state. Using this method, individual protein dimers of P3 and aggregates of dimers have been observed on a mica surface in the presence of an imaging buffer. With appropriate adjustments to surface and/or solution concentration, this method may be generally applied to other biologically relevant macromolecules and soft materials.
Atomic force microscopy (AFM) has become a very important tool for investigating the structural and mechanical properties of surfaces, thin films, and single molecules since its invention in 1986 (Figure 1).1-3 Using a liquid-cell, the method has become particularly useful in studies of biological macromolecules and even living cells in a physiologically relevant environment.4-10 Tapping-mode AFM has traditionally been used for imaging soft materials or loosely bound molecules to the surface, since contact-mode AFM is typically unsuitable due to the damage caused by the lateral forces exerted on the sample by the cantilever.11 Tapping-mode AFM substantially reduces these forces by having the tip intermittently touch the surface rather than being in constant contact. In this mode, the cantilever is oscillated at or near its resonant frequency normal to the surface. Similarly to contact-mode AFM, topography is analyzed by plotting the movement of the z-piezo as a function of xy (distance).
The cantilever dynamics can be quite unstable at or near resonance; therefore, they are very challenging to automate outside of a “steady-state” situation. Specifically, these dynamics depend on both the sample properties and scanning environment. For a soft molecule adsorbed to a hard(er) surface, a well-tuned feedback loop for the molecule may lead to feedback oscillation for the surface. Operation in fluid further complicates the tuning of the cantilever. Changes in temperature or fluid levels require constant readjustment of set point, gains, and other imaging parameters. These adjustments tend to be very time-consuming and challenging for users.
Peak Force Quantitative Nanomechanical Property Mapping (PF-QNM), like tapping mode AFM, avoids lateral interactions by intermittently contacting the sample (Figure 2).12-15 However, PF-QNM operates in non-resonant mode and frequencies much lower than tapping-mode AFM. This eliminates the tuning challenges of tapping-mode AFM, particularly those exacerbated by the presence of fluid. With PF-QNM, images are collected by taking a force response curve at every point of contact. With the addition of ScanAsyst software,15 adjustment of the scanning parameters can be automated and a high-resolution image obtained in a matter of minutes by even inexperienced users. Once the user becomes more familiar with the AFM, any or all of the automated parameters may be disabled at any time which permits the experimentalist to fine tune the image quality manually. Since its inception, PF-QNM has been applied to map bacteriorhodopsin, a membrane protein, and other native proteins at the submolecular level.16-18 For bacteriorhodopsin, there is a direct correlation between protein flexibility and X-ray crystallographic structures.12 PF-QNM has been utilized to investigate living cells with high resolution.19,20 Furthermore, PF-QNM data has elucidated important connections between structure and mechanics within the erythrocyte membrane that are critical for cell integrity and function.21
We have employed scanning probe microscopy (SPM) methods,22 including AFM,23 to study the structure of red-light photoreceptors called bacteriophytochromes (BphPs).24,25 They consist of a light sensing module covalently linked to a signaling-effector module such as histidine kinase (HK).26 The light-sensing module typically contains a bilin chromophore which undergoes structural transformation upon absorption of a photon, with a series of structural changes reaching the signaling-effector module and leading to a global transformation of the entire protein.24,27-29 Based on this transformation, there are two distinct light absorbing states of BphPs, a red and far-red light absorbing state, denoted as Pr and Pfr. Pr is thermally stable, dark-adapted state for most BphPs.28 The molecular basis of Pr/Pfr photoconversion is not entirely understood due to limited structural knowledge of these proteins. With the exception of one structure from D. radiodurans,30 all published X-ray crystallographic structures of these proteins are in the dark-adapted state and lack effector domain. The intact BphPs are too large to be effectively studied by Nuclear Magnetic Resonance (NMR) and are notoriously difficult to crystallize in their intact form (particularly in the light-adapted state) for X-ray crystallography. BphPs have recently been engineered as infrared fluorescent protein markers (IFP’s).31 Structural characterization of these proteins can further aid in effective IFP design.32-36
The focus of this article is to present a procedure for imaging of BphPs using liquid-cell AFM via PF-QNM. The method is demonstrated by studies of the light-adapted state of the bacteriophytochrome RpBphP3 (P3) from the photosynthetic bacterium R. palustris. The AFM procedure presented here is convenient and straightforward approach for imaging of proteins as well as other biological macromolecules. With this method, structural details of individual molecules can be collected in a short period of time, similar to an upper-level science course laboratory session. Through measuring cross-sections and completing further dimensional analyses, experimental data can be compared to useful computational models.37-42
1. Computer and Microscope Set Up
2. Preparation of Mica Surface
3. Fluid Cell Assembly and Imaging of Mica Surface
4. Protein Deposition and Preparation of Sample for Imaging
5. Imaging Proteins on Mica
Representative AFM images of a photoreceptor protein, P3, in its light-adapted state are presented in Figures 3 and 4. A freshly-cleaved mica substrate (Figure 3A) is a suitable, flat surface for protein adsorption. Collecting an image of clean mica as a negative control is important for several reasons. First, it insures the liquid cell is clean and no residual materials from previous experiments will contaminate the surface. Second, it tests the quality of the probe. If the probe is dirty or deformed, this will appear in the image as streaks or present itself as noise. Finally, a clean mica image confirms the probe can be appropriately tuned for future experiments.
Single photoreceptor protein dimers can be observed using PeakForce QNM if the surface coverage is appropriate (Figure 3B,C). Arrows signal individual dimers; the asterisk denotes a protein aggregate. The concentration of the protein sample, deposition time, and the ionic strength of the imaging buffer impact the surface coverage.23 For a high distribution of single molecules, very dilute solutions with short deposition times are required. For example, doubling the deposition time yields a high distribution of even larger aggregates. If the protein concentration is increased to 0.100 mg/ml, a thin film of photoreceptors is observed without modifying the deposition time or buffer ionic strength.
Using image analysis software, the cross-sections of the proteins may be measured after the image is flattened (Figure 4). The cross-sections provide xy-data with corresponding height measurements. These measurements can be used to provide structural details, protein/surface interactions, mechanical strength, etc. The xy-data may be convoluted by the AFM probe; however, this effect can be minimized by using a sharper probe. In fact, the choice of probe is a delicate balance between an acceptable amount of convolution and the resistance of the imaged material to damage by a sharp probe. Tip convolution may also be subtracted from the image, if necessary, by using appropriate software.
These pictures may be directly compared to previously published images collected using tapping-mode AFM.20 The measured dimensions of P3 on mica using PeakForce QNM are within 5% of values obtained with tapping-mode AFM.
Figure 1. Liquid-Cell Atomic Force Microscopy. (A) Laser, cantilever, probe, and surface alignment. The probe and cantilever are affixed inside the liquid cell. (B) Liquid cell. The probe, cantilever, and sample are completely immersed in the fluid. (C) Complete assembly. The liquid cell is attached to a micrometer. The picture shows the optical head and scanner in relationship to the liquid cell. Please click here to view a larger version of this figure.
Figure 2: PeakForce Quantitative Nanomechanical Property Mapping. This is an illustration for a small peak force. (A) Approach (B) Jump to contact (C) Peak Force (D) Adhesion (E) Retract. The figure was reproduced and adapted with permission.15
Figure 3. Atomic Force Microscopy of P3 on Mica. (A) Image of clean mica (1 x 1 µm). (B) Image (1 µm x 1 µm) of P3 applied to mica. The arrows indicate examples of single protein dimers. The asterisk denotes an aggregate of protein dimers. (C) Image (170 x 170 nm) of two protein dimers with a three-dimensional inset of a single dimer. The caret designates which protein dimer is present in the inset. All images were taken in the presence of the imaging buffer (protein concentration = 0.0010 mg/ml for deposition). Please click here to view a larger version of this figure.
Figure 4. Cross-Sectional Analysis of P3 on Mica. AFM image of P3 on mica (170 x 170 nm) with cross-sectional height/distance measurements noted for two dimers. Please click here to view a larger version of this figure.
AFM is a scanning probe microscopy method fully capable of imaging proteins and other biological macromolecules in physiologically relevant conditions. In comparison to X-ray crystallography and NMR, one limitation of AFM is its inability to achieve the same resolution, particularly lateral resolution. When using AFM to analyze a molecule on any surface, the impact of the surface and the probe on the image of the molecule must also be considered when data are analyzed. Deconvoluting of the probe’s impact on the acquired image may be completed with appropriate software. Developing theoretical models to compare to experimental outcome can also prove to be quite useful.
In this paper, AFM images of a bacterial red-light photoreceptor, RpBphP3 (P3), are presented (Figure 3). While the descriptions of the microscope assembly and the software operation are specific to the particular instrument model employed, the mica and protein sample preparation procedures are general and may be applied to any AFM experiment.
A few critical steps of this protocol must be followed in order to obtain the desired protein coverage on the mica surface. First, the diluted protein solution must be gently mixed before it is applied to the mica surface (step 4.3). Due to the high molecular weight of P3, the depositing solution is quite colloidal; therefore, P3 settles to the bottom of the tube over time making the solution non-homogeneous. A gentle stir remedies this issue. Second, once the protein is applied, the surface must be rinsed to remove any loosely bound molecules (step 4.5 – 4.6). If loosely bound P3 remains on the surface, the molecules will be either released into the solution when the liquid cell is filled with buffer solution and/or picked up by the AFM probe during the experiment. Both of these events hinder a successful imaging experiment. If the liquid-cell becomes contaminated with P3, it must be disassembled and thoroughly cleaned. If the AFM probe picks up a molecule, it typically must be replaced or image quality is sacrificed. Finally, the method of rinsing the protein sample is critical to obtaining excellent images. Mica is a non-selective surface to P3; therefore, the molecules should be found to be oriented randomly. The method of rinsing described in the protocol is one way to avoid rearranging the molecules on the surface into one single orientation. The key is to always keep the sample wet and to avoid holding the surface at angle and literally spraying the surface with buffer. If the sample is rinsed by dipping, while holding it parallel to the basin of a Petri dish, the distortion of P3 on the surface by force currents of the buffer solution is minimized.
All of the parts to the liquid cell, probe, scanner, and optical head are very delicate and must be handled with care. The springs that hold the scanner and optical head together are very strong. It is imperative to keep a hand on the scanner when assembling or dissembling any of the pieces. For PF-QNM with ScanAsyst software, as described in the protocol, it is essential to turn off the automated software control of the Z-limit and set this value to at least 500 nm if the original scan size is 1 µm (step 3.16). This protects the scanner if the surface has become unexpectedly rough or has become contaminated. Once a single image has been collected and it is confirmed that the imaged area is smooth, the Z-limit may be adjusted downward manually to achieve higher resolution images. It is crucial to readjust the Z-limit back to at least 500 nm when the scanner is moved to a new area of the surface or if the tip is retracted for any reason.
While this paper has been focused on using PF-QNM to obtain high-quality topographical images, the method itself can provide a library of additional information about the mechanical strength and flexibility of surface structures. This feature can lead to further insights about functionality by simply selecting the appropriate channels during the experiment.16,18 With appropriate changes to the concentration of the depositing solution and/or the surface, this method can be general for using PF-QNM to investigate biological macromolecules by liquid-cell AFM. With the automated help of software, students in upper-level science courses with no prior experience with AFM can complete a liquid-cell AFM experiment much more quickly than with the traditional tapping-mode AFM. Students can experience the basics of AFM and get a taste of the interdisciplinary research made possible by this powerful technique within a typical 3 – 4 hr laboratory section time constraint.
The authors have nothing to disclose.
The NSF-MRI program (CHE: 1229103) is acknowledged for funding the purchase of new control electronics, software, liquid cells, and other equipment needed to assemble a dual AFM/STM. We acknowledge the shared facilities at The University of Chicago NSF-MRSEC program (DMR-0820054) for assistance with AFM instrumentation, training, and imaging, and for instrument time made available by the Materials Research Facilities Network (DMR-0820054). We particularly thank Dr. Qiti Guo, Dr. Justin Jureller, and Prof. Ka Yee Lee for welcoming our students before the funding of the NSF-MRI proposal that brought the necessary instrumentation to our campus. We acknowledge funding from a Title III STEM grant (ID: P031C110157) awarded to Northeastern Illinois University that provided summer research stipends for students and faculty as well as support for consumable supplies. Finally, we acknowledge Bruker-Nano, Inc. for continued instrumental support and for permission to reproduce a plot showing the mechanism of PeakForce QNM and to use the word ScanAsyst to describe the automated software.
Name of Material/Equipment | Company | Catalog # | Comments |
Tris-HCl | Fisher Scientific | O4997-100 | |
NaCl | Acros Organics | 7647-14-5 | |
MgCl2 | Acros Organics | 7791-18-6 | |
Multimode 8 AFM | Bruker-Nano | 492-008-011 | equipped with Nanoscope V controller and J scanner |
Probe | Bruker-Nano | SNL-10, ScanAsyst-Fluid+ | |
Tapping Mode Fluid Cell | Bruker-Nano | MTFML | |
Mica V-4 Grade | SPI supplies | 1150503 | 25 x 25 x .26 mm |
Sample support disk | nanoSurf | BT02236 | |
Petri dish | Plasta-Medic, Inc. | 100 mm x 15 mm | |
micropipettors | Denville Scientific | XL 3000i | |
RpBphP3 | Prepared according to cited references | ||
Nanoscope software | Bruker-Nano | ||
Fiber Optic Light | Digital Instruments Inc. | F0-50 | |
Pelco AFM Disc Gripper | Ted Pella Inc | 1668 | 12 mm |
1 ml syringe | McKesson | 102-ST1C | |
Eppendorf tubes | Denville Scientific | C2171 | |
The Pymol Molecular Graphic System v.1.5.0.1 | Schrodinger, LLC |