Encyclopedia of Biophysics

Living Edition
| Editors: Gordon Roberts, Anthony Watts, European Biophysical Societies

Atomic Force Microscopy (AFM) for Topography and Recognition Imaging at Single-Molecule Level

  • Memed Duman
  • Yoo Jin Oh
  • Rong Zhu
  • Michael Leitner
  • Andreas Ebner
  • Peter HinterdorferEmail author
Living reference work entry
DOI: https://doi.org/10.1007/978-3-642-35943-9_496-1
  • 257 Downloads

Synonyms

Definition

High-resolution atomic force microscopy (AFM) topography has become a powerful bioanalytical tool when utilized for single-molecule force spectroscopy or simultaneous “topography and recognition imaging” (TREC). These modes allow for mapping of specific ligand-binding sites on biological samples under physiological conditions with nanometer resolution.

Introduction

Atomic force microcopy (AFM) is a version of scanning probe microscopy (SPM) that is extensively used in life sciences because it can be operated in physiological salt solution. In AFM imaging, a sharp probe tip mounted on a microcantilever scans over the specimen line by line, whereby the topographic image of the sample surface is generated by “feeling” rather than “looking.” Tips with high sharpness provide high resolution, and cantilevers with low spring constants allow precise control of the forces between tip and sample. These forces cause the cantilever to bend or deflect. The deflection is monitored by a laser that is focused on the backside of the very end of the cantilever, and from there, it is reflected onto a split photodiode. The movement of the laser spot on the photodiode causes an electric signal change that is fed into a feedback loop for keeping the exerted force on the AFM tip constant. Within the feedback loop, voltage is applied to a piezoelectric scanner, which moves the tip or the sample up and down. The vertical movements of the piezo allow a computer to generate a map of surface topography (Fig. 1a).
Fig. 1

Principal operation of various AFM and its modes. Schematic representation of (a) commercial atomic force microscopy, (b) contact mode imaging, and (c) tapping mode imaging

In the last decade, AFM has been widely applied to biological samples ranging from single molecules, such as proteins and nucleic acids, to macromolecular assemblies and whole cells, due to the fact that sub-nanometer resolution is achieved under physiological conditions without extensive sample preparation or labeling (Horber and Miles 2003; Francis et al. 2009; Muller 2008). Topographical imaging of these samples can be done in various modes in accordance with the experimental requirements, which can be divided into contact and dynamic mode AFM imaging.

Contact Mode Imaging

In contact mode imaging, the tip is in permanent physical contact with the sample surface (Fig. 1b). The scanner traces the tip over the surface, and the contact forces cause the cantilever to bend upon changes in topography. The deflection signal is held at a constant value by the feedback loop; thus, constant force is exerted on the specimen. With an AFM operated in the contact mode, topographic images with a vertical resolution of less than 0.1 nm and a lateral resolution of about 0.2 nm have been obtained (Bhushan and Marti 2010). Contact mode AFM imaging of soft biological samples requires delicate adjustment of the force set point, so as to optimize image contrast and reduce damage to the sample.

Dynamic Mode Imaging

Force-sensitive samples such as weakly attached cells can be pushed away by contact mode imaging, and continuous contact with the tip might damage living cells (You et al. 2000). Therefore, dynamic force microscopy (DFM) imaging methods like tapping mode (Putman et al. 1994) and magnetic AC mode (MAC mode) AFM (Han et al. 1997) imaging are more preferable for imaging weakly attached and/or soft cell samples.

In tapping mode, the AFM tip is oscillated near its resonance frequency. While it scans over the surface, it touches the sample only intermittently at the end of its downward movements, which result in an amplitude reduction upon sample contact (Fig. 1c). In order to apply a constant force on the sample, the amplitude reduction upon intermittent tip–surface contact is held constant in the feedback loop. Since the tip touches the sample surface at the very end of its downward movement only, the lateral forces acting on the sample are greatly reduced and result in less sample damage.

In magnetic AC mode (MAC mode) AFM, the oscillation of the magnetically coated cantilever is caused by an alternating magnetic field generated in a coil beneath the sample stage (Han et al. 1996). This allows for a sensitive adjustment of oscillation frequency and of the related phase response, a prerequisite for high-resolution imaging (Raab et al. 1999). MAC mode AFM is a powerful imaging technique to reveal membrane features such as lamellipodia, cytoskeleton fibers, F-actin filaments, and small globular structures on soft and weakly attached cells with a resolution of 5 nm on fixed and 20 nm on living cells under physiological conditions.

Force Spectroscopy

In live cells, molecular recognition events between receptors and their cognate ligands initiate many important biological processes, including genome replication, gene expression, enzymatic activity, immune responses, and other cellular processes. A number of techniques are presently available to investigate inter- and intramolecular interaction forces of biomolecules at the single-molecule level. These techniques span a measurable force window ranging from entropic forces at several femto Newtons (1 fN = 10−15 N) up to the rupture of covalent bonds at several nano Newtons (1 nN = 10−9 N) (Florin et al. 1994; Lee et al. 1994; Hinterdorfer et al. 1996). Since AFM offers one of the smallest force sensors, it is currently the only technique that allows for mapping and analysis of single receptor molecules with a lateral resolution at the nanometer (nm) scale.

In contrast to conventional AFM imaging modes, single-molecule force spectroscopy (SMFS) requires an upgraded sensor design. Here, a single biological ligand molecule has to be bound to the outer apex of the AFM tip by using a flexible linker in order to study and quantify binding interactions to surface-bound receptors (Ebner et al. 2008).

The force between a ligand-functionalized AFM tip and target surface with attached receptor molecules is monitored in a so-called force–distance cycle (Fig. 2) by moving the tip down (“trace”) and up (“retrace”) at constant vertical speed and at a fixed lateral position. In the beginning of the approach phase period (Fig. 2, trace points 1–3), the cantilever deflection remains zero because neither the tip nor the tip-bound sensor molecule reaches the sample surface (Fig. 2, trace point 1). After tip–surface contact (Fig. 2, trace point 2), the cantilever bends upward, consistent with a repulsive force that linearly increases with progressive “negative distance” (Fig. 2, trace point 3). Subsequent tip–surface retraction (Fig. 2, retrace points 4–5) first leads to relaxation of the cantilever deflection until the repulsive force drops to zero. Upon further retraction, the binding between the ligand on the tip and the receptor on the surface exerts a progressive force that bends the cantilever downward (Fig. 2, retrace point 4). Since the ligand is tethered to the tip surface through a flexible polymer linker, the shape of the retract curve shows parabolic-like (nonlinear) characteristics (retrace, point 4), reflecting an increase of the spring constant of the linker during extension. The downward bending of the retracting cantilever continues to rise until the ligand–receptor complex finally dissociates at a certain critical force called unbinding force, fu. Unbinding of the complex is indicated by a sharp spike in the retraction curve that reflects an abrupt jump of the cantilever to its resting position (retrace, point 5). Specificity of binding is usually demonstrated by block experiments in which free ligands are added into the solution to bind to the receptor sites on the sample surface, or by adding free receptor molecules that can attach to the ligand on the AFM tip. After such specific block, the retrace curve will look very much like the trace curve (Fig. 2, inset), i.e., only repulsive forces but no downward deflection of the cantilever is observed.
Fig. 2

Schematics of a force–distance cycle

Recently, SMFS has been applied to investigate the microbial surface nanostructures in native environments (Beaussart et al. 2014; Dufrêne 2015). One of the latest studies revealed the detailed information about the adhesion of curliated bacteria to fibronectin (FN) (Oh et al. 2016). For this purpose, AFM cantilever tips were modified with different fibronectin (FN) constructs (i.e., multi-domain full-length fibronectin (FN), isolated domain III (FN III), or a peptide with the core RGD sequence (RGD)), and a silicon chip surface was coated with bacteria curli protein CsgA monomers (Fig. 3a, b, c). Conjugation of FN constructs to APTES modified AFM cantilevers and CsgA monomers to surface were achieved using heterobifunctional NHS–PEG–aldehyde and NHS–PEG–maleimide linkers (Ebner et al. 2007; Oh et al. 2016). Hence, it was possible to reveal information about binding epitope of FN constructs for CsaA monomer. In most of the force–distance curves recorded with different FN constructs, single-molecule unbinding signatures were observed (Fig. 3a, b, c). From these forces, experimental probability density functions (PDFs) were constructed, in which the maxima represent the most probable measured unbinding force, whereas the widths reflect the stochastic nature of the unbinding process. The maxima values for each construct were remarkably similar (Fig. 3d) (RGD (51 ± 19 pN), FN III (43 ± 16 pN), and FN (57 ± 23 pN)), implying that binding of FNs to CsgA might occur through the same binding epitope. Blocking experiments, in which the RGD peptide was added into the solution, not only showed the specificity of binding but also revealed that binding of all FN constructs to the CsgA monomer is RGD-dependent (Fig. 3e).
Fig. 3

Single-molecular force spectroscopy experiments on surface-bound CsgA monomers. Typical force–distance curves recorded using AFM cantilever tips functionalized with (a) RGD, (b) FN III, and (c) FN. Monomeric CsgA was tethered to silicon chip surfaces (sketches in (a–c)) via a flexible poly(ethylene glycol) (PEG) chain. Likewise, the RGD peptide, FN III, or FN were flexibly linked to AFM tips. The PEG linker that connected the molecules to the AFM tips and probe surfaces ensured sufficient motional freedom for unconstrained interaction measurements. Sketches on the left side show bond formation and the downward deflection of the AFM cantilever during retraction. Red arrows in force–distance curves indicate a bond load increase, and circles mark bond rupture. (d) PDFs of unbinding forces at a retraction velocity of 500 nm/s. For each PDF, 1000 force curve measurements were recorded. (e) Binding probability (defined as the percentage of force experiments displaying unbinding events) for RGD (n = 3000, 3 different tips), FN III (n = 3000, 3 different tips), and FN (n = 6000, 6 different tips). Addition of RGD peptides into the measurement solution (blocked) resulted in a significant drop of the binding probability: from 16% to 7% for RGD (n = 3000, 3 tips), from 10% to 3% for FN III (n = 3000, 3 tips), and from 12% to 6% for FN (n = 3000, 3 tips). (Figures were adapted from Oh et al. 2016)

In order to reveal the distribution of molecular bonds involved in the adhesion of curliated bacteria to fibronectin in real time, two experimental strategies were followed: (i) FN-modified cantilevers were allowed to interact with living bacteria coupled to surfaces (Fig. 4a), and (ii) cantilevers modified with bacteria measured the adhesion strength of single bacteria with FN-conjugated surfaces (Fig. 4b). From the measured adhesion traces, a histogram of the de-adhesion work for the dissociation of RGD, FN III, and FN from bacterial mutant strains that overexpressed CsgA (CsgA (+)) was constructed (Fig. 4c). Up to four tip-adorned RGD and FN III molecules interacted with the bacterial membrane to accomplish the overall adhesion process. In contrast, wild-type (WT) FN showed a broad de-adhesion work distribution, with the most probable value being about seven- to eightfold the work quantum required for single RGD de-adhesion (Fig. 4c). The adhesion strength of single bacteria with a different degree of curliation, i.e., wild type (WT), CsgA knock-out (CsgA (−)), and overexpressed CsgA (CsgA (+)) mutant strain, to FN was studied using AFM tips functionalized with bacterial cells (microbial cell force spectroscopy). The resulting de-adhesion rupture work was broadly distributed for all bacteria. The value of 10,950 ± 2710 pN·nm for the CsgA (+) mutant (Fig. 4d) can be taken as the upper threshold value for these bacterial cells to resist detachment from external forces.
Fig. 4

(a) AFM tips contained RGD, FN III, and FN; surface-bound bacteria were WT, CsgA(−), and CsgA(+). (b) AFM tip-less cantilevers containing CsgA(+), CsgA(−), and WT. Surfaces were coated with FN. (c) Histogram of de-adhesion work for the dissociation of RGD, FN III, and FN from the surface of CsgA(+) at a pulling velocity of 1000 nm/s. Each distribution contains calculated adhesion work from 1000 force curves. The most probable de-adhesion work values from the maxima were fitted with multi-Gaussian distributions. Gray bars indicate work quanta. Single RGD required a de-adhesion work of 552 ± 17 pN·nm, and the second and third peaks were at 840 ± 10 pN·nm and 1234 ± 8 pN·nm. FN III showed maxima at 580 ± 1, 856 ± 11, and 1326 ± 6. FN required much higher de-adhesion work, and the most probable value was 3610 ± 30 pN·nm. (d) Histogram of de-adhesion work upon dissociation of bacterial mutant (WT, CsgA(−), CsgA(+)) from the FN surface (n = 1000 for each mutant). The y-axis was normalized with respect to unbinding probability. (Figures were adapted from Oh et al. 2016)

This study revealed that curliated E. coli and fibronectin form dense quantized and multiple specific bonds (approximately 10 bonds) with high tensile strength as a first step in bacterial infection, which allow bacteria to resist detachment from host cells. Single molecule and microbial cell force spectroscopy are powerful tools for unraveling the quantitative mechanisms that govern bacteria–host cell interaction.

Single molecule force spectroscopy under physiological conditions reveals dynamic information on specific recognition that is inaccessible via other methods, such as NMR and X-ray crystallography. In a recent study, interaction forces between the serotonin transporter (SERT) and the S- and R-enantiomers of citalopram (CIT) were measured to elucidate the number and the mechanism of the ligand-binding site(s) in SERT (Zhu et al. 2016). SERT is a member of the neurotransmitter/sodium symporter (NSS) family and a common target of antidepressants (such as citalopram) and illicit drugs (such as cocaine). In order to address S- and R-CIT on SERT, AFM tips were functionalized with azido-terminated flexible PEG linker and S- or R-CIT through click chemistry (Lewis et al. 2004) (Fig. 5a). SMFS measurements were applied by performing force–distance cycles on living CHOK1 cells that expressed human SERT.
Fig. 5

Two populations of unbinding events. (a) The cantilever (force sensor) tip was functionalized with S-CIT via a flexible cross linker. Identification of two binding sites. (b) Both S-CIT and (c) R-CIT show two peaks in PDF curves of unbinding forces. Control experiments on cells lacking SERT revealed negligible binding activity for both S- and R-CIT tips (black lines in a and b). (d) The unbinding forces of the two PDF peaks were plotted against the force loading rate for S- and R-CIT. This shows similar binding strength at the S2 site (corresponding to the first peak in force PDF) for S- and R-CIT, but reveals distinct binding forces at the S1 site (the second peak) for the two enantiomers. (e) Force measurements on WT SERT in buffer without Na+(Li+ buffer) display a single peak (corresponding to the first peak in the presence of Na+). (f) Force measurements on mutant SERT-G402H (point mutation in vestibular S2 site) show a single peak (corresponding to the second peak of WT SERT). (Figures were reproduced with permission (License Number 4175241120605) from Zhu et al. 2016)

Experimental probability density functions (PDFs) of forces were generated from many unbinding forces (>200). Two distinct populations of unbinding forces were observed for both S- and R-CIT-modified tips (Fig. 5b, c). In order to reveal whether the measured interactions were specific, the same experimental procedure was applied to CHOK1 cell lacking SERT and binding activity was largely reduced (Fig. 5b, c, black lines).

The fitted curves in the specific unbinding force versus force loading rate graph show similar binding strength and dissociation rates for the weaker binding site of both enantiomers. However, much higher binding forces for S-CIT were observed for the stronger binding site (Fig. 5d). Further, force spectroscopy experiments showed other important characteristics of SERT. While there were two distinct populations of characteristic binding strengths of citalopram for SERT in Na+-containing buffer, in Li+-containing buffer, SERT showed only a single peak in the PDF (Fig. 5e). The absence of the second peak points out that the central S1 site is Na+ dependent. Moreover, the vestibular mutant SERT-G402H merely displayed the higher force population, which confirmed that the weaker unbinding events of the wild type SERT originate from the vestibule S2 site (Fig. 5f).

This nanopharmacological approach of single molecule force spectroscopy has the potential for exploring transient binding sites in clinically relevant membrane transporters.

Simultaneous Topography and Recognition Imaging

Identification and localization of specific binding sites of biological sample surfaces with high spatial accuracy is an important objective in life science. Microscopy techniques, such as epifluorescence microcopy, photo-activated localization microscopy (PALM), stimulated emission depletion microscopy (STED), single particle tracking, single dye tracing, or scanning electron microscopy, have the drawbacks of limited resolution, lack of topographic information, and/or inapplicability under physiological conditions. AFM renders possible high-resolution topographical images at the nanometer scale, combined with single-molecule recognition under ambient conditions. The fastest and most straightforward method in this respect is the so-called TREC mode. The name stands for simultaneous acquisition of “Topography and RECognition.” In this AFM mode, the surface of a biological specimen is scanned with a biofunctionalized tip at regular imaging speed, yielding a map of specific ligand-binding sites together with a topographic image (Stroh et al. 2004a, b; Hinterdorfer and Dufrene 2006).

The operating principle of TREC is based on the MAC mode, in combination with a ligand that is covalently bound to the AFM tip via a flexible PEG linker of defined length (∼6 nm, Stroh et al. 2004a, b). This functionalized cantilever is oscillated close to its resonance frequency while scanning across a surface exhibiting binding sites (receptors) for the ligand. In recognition imaging, the cantilever oscillation amplitude is divided into two parts (i.e., lower and upper parts with respect to the baseline of the oscillation) and processed in different paths by using a specially designed electronic circuit called TREC box. While the lower part of the signal is used for generation of the topography image, the upper part reflects recognition events and gives the recognition image (Fig. 6a). Figure 6b shows the results of full scan lines with a bare and an antibody-functionalized AFM tip (top and bottom panels, respectively). In both panels, the lower envelope of the cantilever oscillation provides the information of the surface topography. The upper envelope of cantilever oscillation reflects transient binding to immobilized antigens on the sample surface; thus, it is only seen in the lower panel, measured with an antibody-functionalized tip.
Fig. 6

Principle of TREC (a). The cantilever oscillation is divided into two parts in the TREC box. While the envelope of the upper part yields the recognition image, the lower part provides for the topography image (b). The characteristic oscillation signal of a bare (upper panel) and a HyHEL-5 antibody-functionalized AFM tip (lower panel) on a sample surface containing lysozyme molecules is shown. (Figures were adapted from Stroh et al. 2004a)

By using a cantilever with a low Q-factor (the Q-factor divided by its resonance frequency represents the “memory” (characteristic time constant) of the cantilever), the amplitude reduction in the lower part of the oscillation (originating from a change in the topography) is sufficiently separated in time from amplitude reductions in the upper part of the oscillation (originating from the linker stretching, which only occurs during molecular recognition between the ligand and the receptor). Consequently, only the lower part of the sinusoidal oscillation is fed into the feedback loop and is thereby held constant to obtain the unbiased surface topography. The upper part of the oscillation, solely containing information on recognition between ligand and receptor, is recorded to generate a recognition image simultaneously to the topography image.

The capability of TREC as a functional investigation technique at the nanoscale was successfully demonstrated in one recent study (Leitner et al. 2016). Here, the authors used TREC to map charge distributions of differently composed and treated Multiple Ionic Polymer Layer (SMIL) coatings for capillary zone electrophoresis (CZE). To investigate the negatively charged coating, AFM cantilever was turned to single-molecule sensor by modifying it with positively charged protein avidin. Coupling of avidin to AFM tip was achieved by a well-established protocol (Ebner et al. 2008). As seen in Fig. 7a, there are three basic steps: (1) aminofunctionalizaiton with APTES, (2) binding of a heterobifunctional NHS–PEG(18)–acetal linker, and (3) coupling the sensing protein avidin.
Fig. 7

Molecular Recognition Force Spectroscopy. (a) Tip chemistry: An APTES-coated tip (1) is functionalized with a heterobifunctional NHS–PEG(18)–Acetal crosslinker (2) via amide formation. The deprotected acetal residue of the linker is used to tether the adhesion sensor molecule avidin (or streptavidin) to the tip (3). (b) Square capillary with SMIL coating. Insert provides a detailed SMIL stratification. Molecular adhesion measurements of PDADMAC-DS-PDADMAC-55% PAMAMPS SMIL-coated capillaries before (c) and after (d) NaOH treatment and (e) PDADMAC-DS-PDADMAC-DS SMIL-coated capillary. Freshly prepared SMIL capillaries were imaged using an avidin-functionalized tip (c, top), resulting in a smooth surface showing protrusions of 1e5 nm (c, middle) and an RMS roughness of 1.43 nm. In contrast, the adhesion image reveals a rather homogenously distributed adhesion (c, bottom). Although the surface topography of NaOH-treated SMIL-coated capillaries (d, top) show only minor changes in their surface topography (d, middle) (RMS roughness of 1.56 nm), the adhesion map (d, bottom) appears significantly more heterogeneous. Surfaces with a final DS layer (e, top) showed increased size of polymers (e, middle) and more heterogeneous adhesion properties (e, bottom) compared to 55% PAMAMPS as final layer. Z-range for the adhesion images is set to 50 mV. The scaling for the cross-section of the topographical images is 9 nm and for the adhesion images 60 mV. (Figures were reproduced with permission (License Number 4175240921493) from Leitner et al. 2016)

To prove the electrostatic nature of interaction, molecular recognition force spectroscopy experiments were initially performed and demonstrated that the avidin-functionalized AFM tips can be used as electrostatic sensors for recognition imaging. TREC mode of AFM imaging was applied to three different SMIL coatings (Fig. 7b). While the topography of the terminal 55% PAMAMPS layer appeared rather flat with an Root Mean Square (RMS) average of the roughness 1.43 ± 007 nm (Fig. 7c), NaOH rinsing of this layer increased the roughness slightly to 1.56 ± 008 nm (Fig. 7d). However, the highest surface roughness (2.1 ± 0.2 nm) was measured on the terminal DS layer (Fig. 7e), which can be explained by the branched structure of DS.

In TREC measurements, 55% PAMAMPS layer showed the homogeneously distributed adhesion (red and yellow areas in Fig. 7a, lower image), but the surface appeared more heterogeneous after NaOH treatment (Fig. 7b, lower image). Thus, it is evident that the adhesive properties can be altered by the NaOH treatment. Moreover, different charge distribution pattern was observed for the PDADMAC-DS-PDADMAC-DS layer (Fig. 7e, lower images). The high adhesion areas were from nano-domains and distributed heterogeneously on the surface.

These results show that recognition imaging offers a promising tool for the functional investigation yielding in parallel information about the morphology and the adhesive properties of the coating layer of electrophoresis capillaries.

Simultaneous topography and recognition imaging can be also used for more complex systems like membranes and cells. In a recent study, localization and distribution of protein tyrosine kinase-7 (PTK7) receptor complexes on acute lymphoblastic leukemia cells (Jurkat T cells), which are cancerous cells, were investigated using TREC imaging (Leitner et al. 2017). In order to address PTK7 receptors on cells, AFM tip was functionalized with PTK7-specific SH-sgc8c or NH2-sgc8c aptamers via NHS–PEG–PDP and NHS–PEG–acetal, respectively (Fig. 8a). SMFS experiments were applied to determine the affinity between sgc8c aptamer and PTK7 receptor. The aptamer bound with high probability (38.3 ± 7.48%) and high specificity. The determined kinetics off-rate (koff = 5.16 ± 0.19 s−1) indicates low dissociation of sgc8c–PTK7 complex.
Fig. 8

DNA sequence of the used sgc8c aptamer. (a) Tip chemistry. (1) Inert silicon nitride cantilevers are amino-functionalized using APTES gas phase silanization. The heterobifunctional crosslinker NHS–PEG–PDP (2a) or NHS–PEG–Acetal (2b) is coupled allowing binding NH2-terminated (4) DNA aptamers [after deprotection of the acetal group (3b)] or SH-terminated sgc8c aptamers (3a), respectively. TREC experiments: (b) Schematic of TREC setup. The upper part of the oscillation is used to gain the recognition image, whereas the lower part is influenced by the sample topography. (c) Topography and (d) simultaneously acquired recognition image on a T-cell membrane using sgc8c-functionalized tips. Whereas, (g) the recognition spot is completely abolished as a result of blocked (e) PTK7 receptors; (f) after addition of free aptamers, the topography remains unchanged. Scale bar for all AFM images is 500 nm. (Figures were adapted from Leitner et al. 2017)

TREC were experiments applied to map the distribution of PTK7 at the nanometer scale under physiological conditions. During TREC experiments, the modified AFM cantilever was oscillated close to its resonance frequency at a lateral scan rate of 0.5–1.0 Hz. Aptamer can bind to PTK7 receptor in downward swing of the oscillation, which causes a linker stretching and reduction of the oscillation amplitude in the upward direction (Fig. 8b). In topography images, the Jurkat cell showed a rather smooth surface. A high number of unique dark spots (recognition sites) were observed in the recognition image, indicating the position of aptamer-binding sites (Fig. 8c, d).

To prove that the detected recognition spots were the result of the interactions between sgc8c aptamer and PTK7 receptor, free sgc8c aptamers were injected into the sample solution to saturate the cells with sgc8c molecules (Fig. 8e). After scanning the saturated cells using the same tips with the same imaging parameters, the dark spots in the recognition image (Fig. 8g) were completely abolished due to blockage of the PTK7 receptors, while the topography image (Fig. 8f) remained unchanged. TREC imaging revealed that the PTK7 receptors were homogeneously distributed on the Jurkat cells. Statistical analysis showed that the PTK7 molecules form small receptor clusters with a surface density of 325 ± 12 PTK7 receptors per μm2.

In conclusion, simultaneous topography and recognition imaging may provide new insights into the detection of cancer markers and reveals potential for future clinical diagnostics.

Moreover, TREC has successfully been exploited to understand the mechanism of invariant natural killer T (iNKT) cell polarization, stimulated by recognition of different structural glycolipids of T-cell receptor (TCR) (Duman et al. 2013). CD1d molecule, a monomorphic major histocompatibility complex class I-like molecule, presents different types of glycolipids to invariant natural killer T (iNKT) cells that play an important role in immunity to infection and tumors, as well as in regulating autoimmunity. The recently accomplished study using TREC imaging allows obtaining the distribution and the localization of CD1d molecules on THP1 cells loaded with three different glycolipids (a-GalCer, C20:2 and OCH12) at single-molecule resolution. In this study, CD1d–glycolipid complex on THP1 cell surface was recognized by a magnetically coated AFM tip, functionalized with a biotinylated iNKT-cell receptor (Fig. 9a).
Fig. 9

Simultaneous topography and recognition imaging on THP1 cells: (a) coupling scheme of a biotinylated iNKT TCR to an AFM tip via a flexible aldehyde–PEG–NHS crosslinker to sense and visualize glycolipid-loaded CD1d molecules on THP1 cells. (b) AFM topography and epifluorescence images of THP1 cells (scale bar is 2 μm with height scale ranging from 0 to 2 μm). (c–f) Nanomapping of CD1d molecules on THP1 cells loaded with different glycolipids: representative topographical and recognition images of CD1d molecules when they pulsed with α-GalCer, C20, and OCH12 for 16 h and without any glycolipid (as a control), respectively. Histograms (c–e) represent the area distribution of CD1d microdomains corresponding recognition images detected in four different areas of the same cell (threshold = −0.3 V). (Figures were reproduced with permission (License Number 4175251092050) from Duman et al. 2013)

Fluorescence microscopy was also used prior to TREC imaging to select proper THP1 cells according to its CD1d expression (Fig. 9b). Topography images showed cellular globular membrane features with heights ranging from ~20 to ~70 nm (Fig. 9cf). Different sizes of microdomains were observed as recognition spots, detected from the amplitude reduction arising from interactions between iNKT TCR and CD1d–glycolipid complex. When the THP1 cells were pulsed with a-GalCer and C20:2, the analyzed recognition images (Fig. 9c, d) revealed that the CD1d–a-GalCer and CD1d–C20:2 complexes formed microdomains with dimensions (area) from ~250 to ~10,000 nm2 (mean ± SD, 2219 ± 989, n = 523) and were distributed nonuniformly. However, the distribution and the area size of the CD1d–glycolipid complexes clearly changed when the cells were incubated with OCH12. The area of the recognition spots was increased up to 30,000 nm2 (mean ± SD, 8197 ± 6925, n = 455), and the distribution of the spots changed and appeared as they were connected to each other (Fig. 9e). When the control group of cells, not pulsed with any glycolipids, were used, no recognition spots were observed (Fig. 9f). It means that the iNKT TCR-functionalized AFM tips select the CD1d–glycolipid complexes specifically.

The specificity of these measurements was also confirmed by applying blocking experiments in which free anti-CD1d antibody was added into the liquid cell while acquiring topography and recognition images. Addition of free antibody leads to the disappearance of almost all binding events in the recognition image, whereas no change in the topography image was detected.

Overall, TREC successfully revealed the distribution and localization of CD1d–glycolipid complexes on THP1 cell with single-molecule resolution which is the smallest size of recognition spots (about 25 nm, corresponding to a single CD1d-binding site) under physiological conditions.

Summary

Simultaneous topography and recognition (TREC) imaging is a combination of high-resolution AFM topography imaging with single-molecule force microscopy. This powerful AFM technique not only yields fine structural details about topography but also senses biochemical composition of native biological samples under physiological conditions. TREC shows single molecular interactions and thus allows one to visualize, identify, and quantify local receptor-binding sites and assign their locations to topographical features of a cell surface. For these reasons, TREC is a promising tool for the identification and location of receptor-binding sites on cells, organelles, and other subcellular structures.

Cross-References

References

  1. Beaussart A, Baker AE, Kuchma SL, El-Kirat-Chatel S, O’Toole GA, Dufrene YF (2014) Nanoscale adhesion forces of Pseudomonas aeruginosa type IV pPili. ACS Nano 8:10723–10733CrossRefGoogle Scholar
  2. Bhushan B, Marti O (2010) Scanning probe microscopy-principle of operation, instrumentation, and probes. In: Bhushan B (ed) Handbook of nanotechnology, 3rd edn. Springer, Heidelberg, pp 573–617CrossRefGoogle Scholar
  3. Dufrêne YF (2015) Sticky microbes: forces in microbial cell adhesion. Trends Microbiol 23:376–382CrossRefGoogle Scholar
  4. Duman M, Chtcheglova LA, Zhu R, Bozna BL, Polzella P, Cerundolo V, Hinterdorfer P (2013) Nanomapping of CD1d-glycolipid complexes on THP1 cells by using simultaneous topography and recognition imaging. J Mol Recognit 26:408–414CrossRefGoogle Scholar
  5. Ebner A, Wildling L, Kamruzzahan ASM, Rankl C, Wruss J, Hahn CD, Hölzl M, Zhu R, Kienberger F, Blaas D, Hinterdorfer P, Gruber HJ (2007) A new, simple method for linking of antibodies to atomic force microscopy tips. Bioconjugate Chemistry 18: 1176–1184CrossRefGoogle Scholar
  6. Ebner A, Wildling L, Zhu R, Rankl C, Haselgrubler T, Hinterdorfer P, Gruber HJ (2008) Functionalization of probe tips and supports for single-molecule recognition force microscopy. Top Curr Chem 285:29–76CrossRefGoogle Scholar
  7. Florin EL, Moy VT, Gaub HE (1994) Adhesion forces between individual ligand-receptor pairs. Science 264:415–417CrossRefGoogle Scholar
  8. Francis LW, Lewis PD, Wright CJ, Conlan RS (2009) Atomic force microscopy comes of age. Biol Cell 102:133–143PubMedGoogle Scholar
  9. Han W, Lindsay SM, Jing TW (1996) A magnetically driven oscillating probe microscope for operation in liquids. Appl Phys Lett 69:4111–4113CrossRefGoogle Scholar
  10. Han W, Lindsay SM, Dlakic M, Harrington RE (1997) Kinked DNA. Nature 386:563CrossRefGoogle Scholar
  11. Hinterdorfer P, Dufrene YF (2006) Detection and localization of single molecular recognition events using atomic force microscopy. Nat Meth 3:347–355CrossRefGoogle Scholar
  12. Hinterdorfer P, Baumgartner W, Gruber HJ, Schilcher K, Schindler H (1996) Detection and localization of individual antibody-antigen recognition events by atomic force microscopy. Proc Natl Acad Sci 93:3477–3481CrossRefGoogle Scholar
  13. Horber JK, Miles MJ (2003) Scanning probe evolution in biology. Science 302:1002–1005CrossRefGoogle Scholar
  14. Lee GU, Chrisey LA, Colton RJ (1994) Direct measurement of the forces between complementary strands of DNA. Science 266:771–773CrossRefGoogle Scholar
  15. Leitner M, Stock LG, Traxler L, Leclercq L, Bonazza K, Friedbacher G, Cottet H, Stutz H, Ebner A (2016) Mapping molecular adhesion sites inside SMIL coated capillaries using atomic force microscopy recognition imaging. Anal Chim Acta 930:39–48CrossRefGoogle Scholar
  16. Leitner M, Poturnayova A, Lamprecht C, Weich S, Snejdarkova M, Karpisova I, Hianik T, Ebner A (2017) Characterization of the specific interaction between the DNA aptamer sgc8c and protein tyrosine kinase-7 receptors at the surface of T-cells by biosensing AFM. Anal Bioanal Chem 409:2767–2776CrossRefGoogle Scholar
  17. Lewis WG, Magallon FG, Fokin VV, Finn M (2004) Discovery and characterization of catalysts for azide− alkyne cycloaddition by fluorescence quenching. J Am Chem Soc 126:9152–9153CrossRefGoogle Scholar
  18. Muller DJ (2008) AFM: a nanotool in membrane biology. Biochemistry 47:7986–7998CrossRefGoogle Scholar
  19. Oh YJ, Brenner MH, Gruber HJ, Cui Y, Traxler L, Siligan C, Park S, Hinterdorfer P (2016) Curli mediate bacterial adhesion to fibronectin via tensile multiple bonds. Sci Rep 6:33909CrossRefGoogle Scholar
  20. Putman CAJ, Van der Werf KO, De Grooth BG, Van Hulst NF, Greve J (1994) Tapping mode atomic force microscopy in liquid. Appl Phys Lett 64:2454–2456CrossRefGoogle Scholar
  21. Raab A, Han W, Badt D, Smith-Gill SJ, Lindsay SM, Schindler H, Hinterdorfer P (1999) Antibody recognition imaging by force microscopy. Nat Biotechnol 17:901–905CrossRefGoogle Scholar
  22. Stroh CM, Ebner A, Geretschlager M, Freudenthaler G, Kienberger F, Kamruzzahan AS, Smith-Gill SJ, Gruber HJ, Hinterdorfer P (2004a) Simultaneous topography and recognition imaging using force microscopy. Biophys J 87:1981–1990CrossRefGoogle Scholar
  23. Stroh CM, Wang H, Bash R, Ashcroft B, Nelson J, Gruber H, Lohr D, Lindsay SM, Hinterdorfer P (2004b) Single-molecule recognition imaging microscopy. Proc Natl Acad Sci U S A 101:12503–12507CrossRefGoogle Scholar
  24. You HX, Lau JM, Zhang S, Yu L (2000) Atomic force microscopy imaging of living cells: a preliminary study of the disruptive effect of the cantilever tip on cell morphology. Ultramicroscopy 82:297–305CrossRefGoogle Scholar
  25. Zhu R, Sinwel D, Hasenhuetl PS, Saha K, Kumar V, Zhang P, Rankl C, Holy M, Sucic S, Kudlacek O, Karner A, Sandtner W, Stockner T, Gruber HJ, Freissmuth M, Newman AH, Sitte HH, Hinterdorfer P (2016) Nanopharmacological force sensing to reveal allosteric coupling in transporter binding sites. Angew Chem 55:1719–1722CrossRefGoogle Scholar

Copyright information

© European Biophysical Societies' Association (EBSA) 2018

Authors and Affiliations

  • Memed Duman
    • 1
    • 2
  • Yoo Jin Oh
    • 1
  • Rong Zhu
    • 1
  • Michael Leitner
    • 1
  • Andreas Ebner
    • 1
  • Peter Hinterdorfer
    • 1
    Email author
  1. 1.Institute for BiophysicsJohannes Kepler University of LinzLinzAustria
  2. 2.Institute of Science, Nanotechnology and Nanomedicine DivisionHacettepe UniversityAnkaraTurkey

Section editors and affiliations

  • Nils G. Walter
    • 1
  1. 1.Department of ChemistryThe University of MichiganAnn ArborUSA