DNA was one of the first biological molecules visualized by atomic force microscopy (AFM). It continues to be imaged by AFM for studies of DNA structure, topology, dynamics, and interaction with proteins. With a few exceptions, early AFM images showed DNA as a long featureless polymer with no indication of its underlying helical structure. However, with enhanced force control and sharp AFM tips, it has been possible to resolve, in buffer solution, the two oligonucleotide strands of the Watson-Crick double helix for single DNA molecules that were physisorbed on a mica substrate.1,2 Recent advances in AFM have made such studies more achievable.3 In particular, Bruker’s exclusive PeakForce Tapping® technology has enabled routine highresolution imaging of the DNA double helix at quantifiable imaging forces, without the need for specialized probes or restrictive AFM designs.
The introduction of TappingMode™ in the early 90s led to a significant increase in the use of AFM for biological research.4-14 In TappingMode, the probe oscillates at its fundamental resonance frequency, and the vertical position of the tip (or sample) is continuously adjusted to maintain a constant amplitude of oscillation as the probe scans across a surface. This constant amplitude is usually set slightly below the amplitude of the freely oscillating probe at some microns away from the sample surface. The probe oscillation essentially represents a tapping motion, with the probe continuously moving in and out of contact with the surface. The intermittent nature of the tip-sample contact reduced the shear forces associated with the previously used contact mode AFM. This puts less stringent demands on how rigidly the sample of interest is attached to a hard substrate, reducing the need of fixation and thus allowing the sample to be imaged under more physiologically relevant conditions.
Unfortunately, despite the advantages that TappingMode offers for studying the structure of biological samples, it has been criticized for ultimately providing lower-resolution images of biomolecules than contact mode imaging.15 Key to obtaining high-resolution AFM images is the ability to control the tip-sample interaction forces during imaging. For the setpoint amplitude to be an accurate measurement of the tip-sample forces, the free oscillation amplitude (at some microns above the surface) needs to remain constant. For TappingMode in liquid this is often not the case, since the cantilever amplitude not only depends on the cantilever resonance, but also on its convolution with mechanical resonances of the fluid cell (the so-called ‘forest of peaks’).16 As the liquid in the fluid cell changes shape, volume and composition throughout an experiment, these resonances shift. This can result in changes to the force applied between the tip and the sample, as the free amplitude of the cantilever changes. It can therefore be difficult to accurately determine and control the imaging force during a TappingMode experiment.
Together, these features of PeakForce Tapping enable direct and precise control of the tip-sample interaction force, facilitating imaging in fluid environments at forces of 100 pN or less. This helps protect both the AFM probe and the sample from potential damage and is one of the key factors in enabling high-resolution imaging. Additionally, imaging in PeakForce Tapping is considerably quicker than other force-distance curve-based imaging modes. As PeakForce Tapping operates at much higher frequencies (1-2 kHz) it is capable of performing thousands of force curves per second. PeakForce Tapping images can therefore be acquired at traditional imaging mode scan rates at high pixel resolution (=512 x 512 pixels).
In addition to protecting delicate samples and tips from damage by maintaining low imaging forces, PeakForce Tapping has also made imaging in fluid easier and effectively more consistent by eliminating the need to tune the cantilever. Unlike TappingMode, PeakForce Tapping does not operate at the resonance frequency of the AFM probe, such that cantilever tuning is simply not needed. PeakForce Tapping technology has also facilitated the self-optimizing ScanAsyst® imaging mode. In ScanAsyst, auto-optimization of the imaging setpoint prevents setpoint drift, which commonly occurs in other AFM operating modes, such as TappingMode and contact mode, due to resonance peak shifting and/or cantilever deflection drift. This auto-optimization of the imaging force at the point of each tip-sample interaction enables PeakForce Tapping to acquire high-resolution images more routinely than contact mode or TappingMode. Together with auto-optimization of other parameters in ScanAsyst mode, such as gain and scan rate, PeakForce Tapping now results in faster, more consistent data, regardless of the user skill level.
DNA is another highly suitable sample benchmark for PeakForce Tapping. It has been extensively imaged by AFM and was one of the first samples used to demonstrate the potential of TappingMode for imaging biomolecules.30-33 DNA is made up of two polynucleotide strands that form a double helix. B-DNA, the “Watson-Crick” form of DNA, exhibits a right-handed helix with a helical repeat (pitch) of ~3.6 nm, with major and minor grooves of widths ~2.2 nm and ~1.2 nm, respectively. The vast majority of DNA images in the AFM literature display DNA molecules as featureless strands. Recent developments in AFM technology, however, have facilitated the visualization of the DNA double helix as a tilted, double-banded structure repeating along the molecule using specialized instruments.1,2 Here we will show a method by which the secondary structure of DNA can be imaged using PeakForce Tapping and standard cantilevers.3
As with all AFM studies conducted in fluid environments, sample preparation is central to successfully imaging the DNA double helix. As such, the DNA plasmid must first be adsorbed on a suitable surface. One of the most commonly used substrates for AFM imaging is mica: Its planar structure can be readily cleaved using sticky tape, revealing an atomically flat and clean surface. However, at neutral pH, mica has an overall negative surface charge, which does not favor adsorption of the also negatively charged DNA. Several methods have been developed to overcome this, all of which essentially act to functionalize the mica to create a positive interface to which the DNA can attach.
As early as 1995, Mou et al. resolved the pitch of B-DNA by AFM as a periodic modulation of 3.4 ±0.4 nm.34 In their study, DNA was adsorbed onto the surface of a cationic supported lipid bilayer, deposited on a mica substrate. Interestingly, the pitch of the DNA was only observed when the DNA strands were densely and uniformly packed on the bilayer surface, and not where bilayers were populated by individual isolated DNA strands. The researchers concluded that this close packing limited the movement of the molecules, supported by the knowledge that DNA is a highly dynamic molecule, having both translational and rotational movement.35 The resolution obtained on DNA thus depends on the degree of adhesion and immobilization of the DNA molecules on the substrate. Providing an alternative for cationic lipid surfaces, mica can be chemically modified with 3-aminopropyltriethoxysilane (APTES) or 1-3-aminopropylsilatrane (APS), to give a positive interface with which the DNA can interact.36
Employing the same DNA immobilization strategy as Leung et al., our goal was to resolve the helical structure of loosely bound DNA using the low and precisely controlled imaging forces enabled by PeakForce Tapping mode, as achieved by Pyne et al.3 To demonstrate that this type of spatial resolution is not specific to a particular AFM system or probe, we carried out PeakForce Tapping experiments on the MultiMode 8, Dimension FastScan Bio, and BioScope Resolve™ atomic force microscopes (see figure 3) using ScanAsyst Fluid+, MSNL-F, FastScan-D, and ScanAsyst Fluid-HR probes, which all have standard silicon tips. PeakForce Tapping imaging on the MultiMode 8 in 10 mM HEPES, 1 mM NiCl2, pH 7.4 revealed corrugations along the DNA strand that correspond to the major and minor grooves of the DNA double helix (see figure 4A). A high-resolution image obtained on the BioScope Resolve (inset of figure 4A) was obtained under the same imaging conditions while operating on an inverted light microscope and using ScanAsyst Fluid-HR probes. This image illustrates the widths of the alternating major and minor grooves, at 2.2 nm and 1.2 nm, respectively. To analyze the mobility of the surface-bound DNA, continuous high-speed TappingMode imaging was performed on the plasmid DNA immobilized on the mica surface in 1mM NiCl2 (see figure 4B), using the FastScan Bio atomic force microscope and FastScan-D probes that have a small cantilever but a standard silicon tip.
The time series of high-speed images illustrates that while some parts of the DNA strand remain immobile under continuous imaging, other parts move over the surface. Therefore, while the Ni2+ immobilizes the DNA sufficiently to enable high-resolution imaging ofthe helical structure, it also allows for some degree of rotational and translational movement of the individual strands. Height variations were also observed in the topography along the length of the DNA, possibly indicating twisting of the DNA strand (see figure 5A). This would also suggest that the low concentration of Ni2+ allows the DNA to maintain a more physiologically relevant structure on the mica surface.
As stated earlier, one of the keys to obtaining high-resolutionimages of the DNA double helix is precise and continuous control of the force applied to the sample. PeakForce Tapping has the unique advantage over other intermittent contact modes in that the imaging force is easily quantified at all times. Figure 5B(i-iii) shows the effect of force on AFM topography using PeakForce Tapping mode on the Multimode 8 with MSNL-F probes. To best illustrate how the DNA is compressed with increasing tip-sample force, the height scale is kept the same for all images. At the minimum possible applied peak force of 39 pN, the measured height of the plasmid is close to the 2 nm diameter of DNA, as derived from its crystal structure (corresponding AFM height profiles shown in figure 5B(iv)). There is, however, very little corrugation visible along the length of the DNA strand in corresponding high-resolution images, shown in the inset of figure 5B(i), which may be due to difficulties in tracking the molecule at these low forces. At 70 pN of applied force, a 20% compression of the molecule occurs, reducing the measured height of the plasmid to ~1.6 nm. At this force the corrugation is most visible, as shown in the inset of figure 5B(ii). Beyond 100 pN, the major and minor grooves become less clear (figure 5B(iii)) and the measured heights reduce to <1.5 nm, similar to earlier TappingMode AFM experiments in liquid.34,38 At this point, the sample is also at significant risk of being dislocated from the mica surface, demonstrated by the movement of the molecule as indicated by the white arrow. Figure 5B(v) shows that the measured height agrees with the diameter of the DNA for applied forces of around 50 pN or less, while slightly more force may need to be applied to accurately resolve the secondary structure, as evident in the PeakForce Tapping images.
Figure 6A shows a high-resolution image of a DNA plasmid imaged by PeakForce Tapping on the FastScan Bio using FastScan-D probes at low force. This image shows corrugation corresponding to the double helix. To further investigate this structure, the scan size was reduced to image the smaller area highlighted by the white box. High resolution images of this smaller scan area are shown in figure 6B in which the major and minor grooves of the strand are clearly shown. The double helix structure is clearly visible in both the trace and retrace images, with the scan direction indicated by the white arrows, as well as in a number of subsequent scans that are shown in time order. Interestingly, the major and minor grooves show variations in depth along the strand, which are reproduced between trace and retrace scans and in subsequent images (see figure 6C). This demonstrates that not only can PeakForce Tapping resolve the submolecular features of the DNA double helix, but that it is also able to reproduce image variations in this helical structure.
PeakForce Tapping mode provides precise force control and easy quantification of the tip-sample interaction force, enabling imaging at forces of less than 100pN to obtain high-resolution images of soft biological samples in fluid environments. The high-resolution imaging capability of PeakForce Tapping mode is demonstrated by resolving the major and minor groves of the DNA double helix on individual plasmids using Bruker’s MultiMode 8, Dimension FastScan Bio, and BioScope Resolve atomic force microscopes. The ability to reliably achieve this type of submolecular resolution consistently, without the need for specialized probes or dedicated AFM designs, is helping to redefine the high-resolution imaging performance of atomicforce microscopes for biological samples.