Structural Information, Resolution, and Noise in High-Resolution Atomic Force Microscopy Topographs

AFM has developed into a powerful tool in structural biology, providing topographs of proteins under close-to-native conditions and featuring an outstanding signal/noise ratio. However, the imaging mechanism exhibits particularities: fast and slow scan axis represent two independent image acquisition axes. Additionally, unknown tip geometry and tip-sample interaction render the contrast transfer function nondefinable. Hence, the interpretation of AFM topographs remained difficult. How can noise and distortions present in AFM images be quantified? How does the number of molecule topographs merged influence the structural information provided by averages? What is the resolution of topographs? Here, we find that in high-resolution AFM topographs, many molecule images are only slightly disturbed by noise, distortions, and tip-sample interactions. To identify these high-quality particles, we propose a selection criterion based on the internal symmetry of the imaged protein. We introduce a novel feature-based resolution analysis and show that AFM topographs of different proteins contain structural information beginning at different resolution thresholds: 10 Å (AqpZ), 12 Å (AQP0), 13 Å (AQP2), and 20 Å (light-harvesting-complex-2). Importantly, we highlight that the best single-molecule images are more accurate molecular representations than ensemble averages, because averaging downsizes the z-dimension and “blurs” structural details.Abbreviations: 2D, two-dimensional; 3D, three-dimensional; ACV, auto-correlation value; AFM, atomic force microscopy; AQP0, aquaporin-0; AQP2, aquaporin-2; AqpZ, aquaporin-Z; bR, bacteriorhodopsin; CCV, cross-correlation value; CTF, contrast transfer function; DPR, differential phase residual; EM, electron microscopy; FRC, Fourier ring correlation; FSC, Fourier shell correlation; IS, internal symmetry; LH2, light-harvesting-complex 2; RMSD, root mean-square deviation; SD, standard deviation; SNR, signal/noise ratio; SSNR, spectral signal/noise ratio     

Phase imaging of moving DNA molecules and DNA molecules replicated in the atomic force microscope.

Phase imaging with a tapping mode atomic force microscope (AFM) has many advantages for imaging moving DNA and DNA-enzyme complexes in aqueous buffers at molecular resolution. In phase images molecules can be resolved at higher scan rates and lower forces than in height images from the AFM. Higher scan rates make it possible to image faster processes. At lower forces the molecules are imaged more gently. Moving DNA molecules are also resolved more clearly in phase images than in height images. Phase images in tapping mode AFM show the phase difference between oscillation of the piezoelectric crystal that drives the cantilever and oscillation of the cantilever as it interacts with the sample surface. Phase images presented here show moving DNA molecules that have been replicated with Sequenase in the AFM and DNA molecules tethered in complexes with Escherichia coli RNA polymerase.     

Tobacco mosaic virus as an AFM tip calibrator

The study of high-resolution topographic surfaces of isolated single molecules is one of the applications of atomic force microscopy (AFM). Since tip-induced distortions are significant in topographic images the exact AFM tip shape must be known in order to correct dilated AFM height images using mathematical morphology operators. In this work, we present a protocol to estimate the AFM tip apex radius using tobacco mosaic virus (TMV) particles. Among the many advantages of TMV, are its non-abrasivity, thermal stability, bio-compatibility with other isolated single molecules and stability when deposited on divalent ion pretreated mica. Compared to previous calibration systems, the advantage of using TMV resides in our detailed knowledge of the atomic structure of the entire rod-shaped particle. This property makes it possible to interpret AFM height images in term of the three-dimensional structure of TMV. Results obtained in this study show that when a low imaging force is used, the tip is sensing viral protein loops whereas at higher imaging force the tip is sensing the TMV particle core. The known size of the TMV particle allowed us to develop a tip-size estimation protocol which permits the successful erosion of tip-convoluted AFM height images. Our data shows that the TMV particle is a well-adapted calibrator for AFM tips for imaging single isolated biomolecules. The procedure developed in this study is easily applicable to any other spherical viral particles.     

Chaperonins GroEL and GroES: views from atomic force microscopy.

The Escherichia coli chaperonins, GroEL and GroES, as well as their complexes in the presence of a nonhydrolyzable nucleotide AMP-PNP, have been imaged with the atomic force microscope (AFM). We demonstrate that both GroEL and GroES that have been adsorbed to a mica surface can be resolved directly by the AFM in aqueous solution at room temperature. However, with glutaraldehyde fixation of already adsorbed molecules, the resolution of both GroEL and GroES was further improved, as all seven subunits were well resolved without any image processing. We also found that chemical fixation was necessary for the contact mode AFM to image GroEL/ES complexes, and in the AFM images. GroEL with GroES bound can be clearly distinguished from those without. The GroEL/ES complex was about 5 nm higher than GroEL alone, indicating a 2 nm upward movement of the apical domains of GroEL. Using a slightly larger probe force, unfixed GroEL could be dissected: the upper heptamer was removed to expose the contact surface of the two heptamers. These results clearly demonstrate the usefulness of cross-linking agents for the determination of molecular structures with the AFM. They also pave the way for using the AFM to study the structural basis for the function of GroE system and other molecular chaperones.     

Atomic force microscopy of long and short double-stranded, single-stranded and triple-stranded nucleic acids.

Atomic force microscopy (AFM, also called scanning force microscopy) is proving to be a useful technique for imaging DNA. Thus it is important to push the limits of AFM imaging in order to explore both what types of DNA can be reliably imaged and identified and also what substrates and methods of sample preparation are suitable. The following advances in AFM of DNA are presented here. (i) DNA molecules as short as 25 bases can be seen by AFM. The short single-stranded DNAs imaged here (25 and 50 bases long) appeared globular in the AFM, perhaps because they are all capable of intramolecular base pairing and because the DNAs were in a Mg(ll) buffer, which facilitates intramolecular cross-bridging. (ii) AFM images in air of short double-stranded DNA molecules, 100-200 bp, gave lengths consistent with A-DNA. (iii) AFM images of poly (A) show both short bent lumpy molecules with an apparent persistence length of 40 nm and long straight molecules with an apparent persistence length of 600 nm. For comparison, the apparent persistence length for double-stranded DNA from phX-174 under the same conditions was 80 nm. (iv) Structures believed to be triple- stranded DNA were seen in samples of poly(dA.poly(dT) and poly (dG).poly(dC). These structures were twice as high as double-stranded DNA and the same width. (v) Entire molecules of lambda DNA, approx. 16 micron long, were imaged clearly in overlapping scans. (vi) Plasmid DNA was imaged on oxidized silicon, although less clearly than on mica.     

A Quantitative Method to Track Protein Translocation between Intracellular Compartments in Real-Time in Live Cells Using Weighted Local Variance Image Analysis

The genetic expression of cloned fluorescent proteins coupled to time-lapse fluorescence microscopy has opened the door to the direct visualization of a wide range of molecular interactions in living cells. In particular, the dynamic translocation of proteins can now be explored in real time at the single-cell level. Here we propose a reliable, easy-to-implement, quantitative image processing method to assess protein translocation in living cells based on the computation of spatial variance maps of time-lapse images. The method is first illustrated and validated on simulated images of a fluorescently-labeled protein translocating from mitochondria to cytoplasm, and then applied to experimental data obtained with fluorescently-labeled hexokinase 2 in different cell types imaged by regular or confocal microscopy. The method was found to be robust with respect to cell morphology changes and mitochondrial dynamics (fusion, fission, movement) during the time-lapse imaging. Its ease of implementation should facilitate its application to a broad spectrum of time-lapse imaging studies.     

Imaging mixed lipid monolayers by dynamic atomic force microscopy.

Phase imaging with tapping mode atomic force microscopy (AFM) and force modulation microscopy were used to probe the mechanical properties of phase-separated lipid monolayers made of a mixture (0.25:0.75) of the surface-active lipopeptide surfactin and of dipalmitoylphosphatidylcholine (DPPC). The pi-A isotherms and the result of a molecular modeling study revealed a loose, 2-D liquid-like organization for the surfactin molecules and a closely packed, 2-D solid-like organization for DPPC molecules. This difference in molecular organization was responsible for a significant contrast in height, tapping mode phase and force modulation amplitude images. Phase imaging at light tapping, i.e., with a ratio of the set-point tapping amplitude with respect to the free amplitude A(sp)/A(0) approximately 0.9, showed larger phase shifts on the solid-like DPPC domains attributed to larger Young's modulus. However, contrast inversion was observed for A(sp)/A(0)<0.7, suggesting that at moderate and hard tapping the image contrast was dominated by the probe-sample contact area. Surprisingly, force modulation amplitude images showed larger stiffness for the liquid-like surfactin domains, suggesting that the contrast was dominated by contact area effects rather than by Young's modulus. These data emphasize the complex nature of the contrast mechanisms of dynamic AFM images recorded on mixed lipid monolayers.     

Reconstruction of low-resolution molecular structures from simulated atomic force microscopy images

BackgroundAtomic Force Microscopy (AFM) is an experimental technique to study structure-function relationship of biomolecules. AFM provides images of biomolecules at nanometer resolution. High-speed AFM experiments produce a series of images following dynamics of biomolecules. To further understand biomolecular functions, information on three-dimensional (3D) structures is beneficial.MethodWe aim to recover 3D information from an AFM image by computational modeling. The AFM image includes only low-resolution representation of a molecule; therefore we represent the structures by a coarse grained model (Gaussian mixture model). Using Monte-Carlo sampling, candidate models are generated to increase similarity between AFM images simulated from the models and target AFM image.ResultsThe algorithm was tested on two proteins to model their conformational transitions. Using a simulated AFM image as reference, the algorithm can produce a low-resolution 3D model of the target molecule. Effect of molecular orientations captured in AFM images on the 3D modeling performance was also examined and it is shown that similar accuracy can be obtained for many orientations.ConclusionsThe proposed algorithm can generate 3D low-resolution protein models, from which conformational transitions observed in AFM images can be interpreted in more detail.General significanceHigh-speed AFM experiments allow us to directly observe biomolecules in action, which provides insights on biomolecular function through dynamics. However, as only partial structural information can be obtained from AFM data, this new AFM based hybrid modeling method would be useful to retrieve 3D information of the entire biomolecule.     

Simultaneous height and adhesion imaging of antibody-antigen interactions by atomic force microscopy.

Specific molecular recognition events, detected by atomic force microscopy (AFM), so far lack the detailed topographical information that is usually observed in AFM. We have modified our AFM such that, in combination with a recently developed method to measure antibody-antigen recognition on the single molecular level (Hinterdorfer, P., W. Baumgartner, H. J. Gruber, K. Schilcher, and H. Schindler, Proc. Natl. Acad. Sci. USA 93:3477-3481 (1996)), it allows imaging of a submonolayer of intercellular adhesion molecule-1 (ICAM-1) in adhesion mode. We demonstrate that for the first time the resolution of the topographical image in adhesion mode is only limited by tip convolution and thus comparable to tapping mode images. This is demonstrated by imaging of individual ICAM-1 antigens in both the tapping mode and the adhesion mode. The contrast in the adhesion image that was measured simultaneously with the topography is caused by recognition between individual antibody-antigen pairs. By comparing the high-resolution height image with the adhesion image, it is possible to show that specific molecular recognition is highly correlated with topography. The stability of the improved microscope enabled imaging with forces as low as 100 pN and ultrafast scan speed of 22 force curves per second. The analysis of force curves showed that reproducible unbinding events on subsequent scan lines could be measured.     

Simulation atomic force microscopy for atomic reconstruction of biomolecular structures from resolution-limited experimental images

Atomic force microscopy (AFM) can visualize the dynamics of single biomolecules under near-physiological conditions. However, the scanning tip probes only the molecular surface with limited resolution, missing details required to fully deduce functional mechanisms from imaging alone. To overcome such drawbacks, we developed a computational framework to reconstruct 3D atomistic structures from AFM surface scans, employing simulation AFM and automatized fitting to experimental images. We provide applications to AFM images ranging from single molecular machines, protein filaments, to large-scale assemblies of 2D protein lattices, and demonstrate how the obtained full atomistic information advances the molecular understanding beyond the original topographic AFM image. We show that simulation AFM further allows for quantitative molecular feature assignment within measured AFM topographies. Implementation of the developed methods into the versatile interactive interface of the BioAFMviewer software, freely available at www.bioafmviewer.com, presents the opportunity for the broad Bio-AFM community to employ the enormous amount of existing structural and modeling data to facilitate the interpretation of resolution-limited AFM images.     

Atomic force microscopy of supported planar membrane bilayers.

Membrane bilayers of dipalmitoyl phosphatidylcholine (DPPC) and dipalmitoyl phosphatidylethanolamine (DPPE) adsorbed to a freshly cleaved mica substrate have been imaged by Atomic Force Microscopy (AFM). The membranes were mounted for imaging by two methods: (a) by dialysis of a detergent solution of the lipid in the presence of the substrate material, and (b) by adsorption of lipid vesicles onto the substrate surface from a vesicle suspension. The images were taken in air, and show lipid bilayers adhering to the surface either in isolated patches or in continuous sheets, depending on the deposition conditions. Epifluorescence light-microscopy shows that the lipid is distributed on the substrate surfaces as seen in the AFM images. In some instances, when DPPE was used, whole, unfused vesicles, which were bound to the substrate, could be imaged by the AFM. Such membranes should be capable of acting as natural anchors for imaging membrane proteins by AFM.     

Theoretical Limits on Errors and Acquisition Rates in Localizing Switchable Fluorophores

A variety of recent imaging techniques are able to beat the diffraction limit in fluorescence microcopy by activating and localizing subsets of the fluorescent molecules in the specimen, and repeating this process until all of the molecules have been imaged. In these techniques there is a tradeoff between speed (activating more molecules per imaging cycle) and error rates (activating more molecules risks producing overlapping images that hide information on molecular positions), and so intelligent image processing approaches are needed to identify and reject overlapping images. We introduce here a formalism for defining error rates, derive a general relationship between error rates, image acquisition rates, and the performance characteristics of the image processing algorithms, and show that there is a minimum acquisition time irrespective of algorithm performance. We also consider algorithms that can infer molecular positions from images of overlapping blurs, and derive the dependence of the minimum acquisition time on algorithm performance.     

Fuzzy Logic Based Edge Detection in Smooth and Noisy Clinical Images

Edge detection has beneficial applications in the fields such as machine vision, pattern recognition and biomedical imaging etc. Edge detection highlights high frequency components in the image. Edge detection is a challenging task. It becomes more arduous when it comes to noisy images. This study focuses on fuzzy logic based edge detection in smooth and noisy clinical images. The proposed method (in noisy images) employs a 3×3 mask guided by fuzzy rule set. Moreover, in case of smooth clinical images, an extra mask of contrast adjustment is integrated with edge detection mask to intensify the smooth images. The developed method was tested on noise-free, smooth and noisy images. The results were compared with other established edge detection techniques like Sobel, Prewitt, Laplacian of Gaussian (LOG), Roberts and Canny. When the developed edge detection technique was applied to a smooth clinical image of size 270×290 pixels having 24 dB ‘salt and pepper’ noise, it detected very few (22) false edge pixels, compared to Sobel (1931), Prewitt (2741), LOG (3102), Roberts (1451) and Canny (1045) false edge pixels. Therefore it is evident that the developed method offers improved solution to the edge detection problem in smooth and noisy clinical images.     

High-performance confocal system for microscopic or endoscopic applications

We designed a high-performance confocal system that can be easily adapted to an existing light microscope or coupled with an endoscope for remote imaging. The system employs spatially and temporally patterned illumination produced by one of several mechanisms, including a micromirror array video projection device driven by a computer video source or a microlens array scanned by a piezo actuator in the microscope illumination path. A series of subsampled "component" video images are acquired from a solid-state video camera. Confocal images are digitally reconstructed using "virtual pinhole" synthetic aperture techniques applied to the collection of component images. Unlike conventional confocal techniques that raster scan a single detector and illumination point, our system samples multiple locations in parallel, with particular advantages for monitoring fast dynamic processes. We compared methods of patterned illumination and confocal image reconstruction by characterizing the point spread function, contrast, and intensity of imaged objects. Sample 3D reconstructions include a diatom and a Golgi-stained nerve cell collected in transmission.     

Camera-based visibility estimation: Incorporating multiple regions and unlabeled observations

This paper investigates image processing and pattern recognition techniques to estimate atmospheric visibility based on the visual content of images from off-the-shelf cameras. We propose a prediction model that first relates image contrast measured through standard image processing techniques to atmospheric transmission. This is then related to the most common measure of atmospheric visibility, the coefficient of light extinction. The regression model is learned using a training set of images and corresponding light extinction values as measured using a transmissometer.The major contributions of this paper are twofold. First, we propose two predictive models that incorporate multiple scene regions into the estimation: regression trees and multivariate linear regression. Incorporating multiple regions is important since regions at different distances are effective for estimating light extinction under different visibility regimes. The second major contribution is a semi-supervised learning framework, which incorporates unlabeled training samples to improve the learned models. Leveraging unlabeled data for learning is important since in many applications, it is easier to obtain observations than to label them. We evaluate our models using a dataset of images and ground truth light extinction values from a visibility camera system in Phoenix, Arizona.     

Cell wall extension results in the coordinate separation of parallel microfibrils: evidence from scanning electron microscopy and atomic force microscopy

Enlargement of the cell wall requires separation of cellulose microfibrils, mediated by proteins such as expansin; according to the multi-net growth hypothesis, enlargement passively reorients microfibrils. However, at the molecular scale, little is known about the specific movement of microfibrils. To find out, we examined directly changes in microfibril orientation when walls were extended slowly in vitro under constant load (creep). Frozen-thawed cucumber hypocotyl segments were strained by 20-30% by incubation in pH 4.5 buffer or by incubation of heat-inactivated segments in alpha-expansin or a fungal endoglucanase (Cel12A). Subsequently, the innermost layer of the cell wall was imaged, with neither extraction nor homogenization, by field-emission scanning electron microscopy (FESEM) and atomic force microscopy (AFM). AFM images revealed that sample preparation for FESEM did not appreciably alter cell wall ultrastructure. In both FESEM and AFM, images from extended and non-extended samples appeared indistinguishable. To quantify orientational order, we used a novel algorithm to characterize the fast Fourier transform of the image as a function of spatial frequency. For both FESEM and AFM images, the transforms of non-extended samples were indistinguishable from those of samples extended by alpha-expansin or Cel12A, as were AFM images of samples extended by acidic buffer. We conclude that cell walls in vitro can extend slowly by a creep mechanism without passive reorientation of innermost microfibrils, implying that wall loosening agents act selectively on the cross-linking polymers between parallel microfibrils, rather than more generally on the wall matrix.     

Enzymatic degradation of monolayer for poly(lactide) revealed by real-time atomic force microscopy: effects of stereochemical structure, molecular weight, and molecular branches on hydrolysis rates

The influences of the stereochemical structure, the molecular weight, and the number of molecular branches for poly(lactide) (PLA) on enzymatic hydrolysis rates of PLA monolayers were studied by atomic force microscopy (AFM) and the Langmuir-Blodgett (LB) technique. Monolayers of six kinds of PLA with different molecular weights, stereochemical structure, and numbers of molecular branches were prepared by LB techniques and then characterized by AFM in air. The PLA molecules covered homogeneously with a silicon substrate and did not form lamellar crystals in the monolayer. We determined the initial hydrolysis rate of PLA monolayers in presence of proteinase K by volumetric analysis from the continuous AFM height images. The presence of D-lactyl unit reduced the hydrolysis rate of the monolayer. The hydrolysis rate for the linear PLLA samples increased with a decrease in the molecular weight. In contrast, the rates of erosion for branched PLLA monolayers were independent of the molecular weight of samples. The erosion rate of branched PLLA monolayers was found to be dependent on the average molecular weight of PLLA segment in branched molecules, not on the overall molecular weight of samples. From these results, furthermore, the hydrolysis mode of PLAs by proteinase K is discussed.     

Charting and unzipping the surface layer of Corynebacterium glutamicum with the atomic force microscope

Bacterial surface layers (S-layers) are extracellular protein networks that act as molecular sieves and protect a large variety of archaea and bacteria from hostile environments. Atomic force microscopy (AFM) was used to asses the S-layer of Coryne-bacterium glutamicum formed of PS2 proteins that assemble into hexameric complexes within a hexagonal lattice. Native and trypsin-treated S-layers were studied. Using the AFM stylus as a nanodissector, native arrays that adsorbed to mica as double layers were separated. All surfaces of native and protease-digested S-layers were imaged at better than 1 nm lateral resolution. Difference maps of the topographies of native and proteolysed samples revealed the location of the cleaved C-terminal fragment and the sidedness of the S-layer. Because the corrugation depths determined from images of both sides span the total thickness of the S-layer, a three-dimensional reconstruction of the S-layer could be calculated. Lattice defects visualized at 1 nm resolution revealed the molecular boundaries of PS2 proteins. The combination of AFM imaging and single molecule force spectroscopy allowed the mechanical properties of the Corynebacterium glutamicum S-layer to be examined. The results provide a basis for understanding the amazing stability of this protective bacterial surface coat.     

Dual-mode Imaging of Cutaneous Tissue Oxygenation and Vascular Function

Accurate assessment of cutaneous tissue oxygenation and vascular function is important for appropriate detection, staging, and treatment of many health disorders such as chronic wounds. We report the development of a dual-mode imaging system for non-invasive and non-contact imaging of cutaneous tissue oxygenation and vascular function. The imaging system integrated an infrared camera, a CCD camera, a liquid crystal tunable filter and a high intensity fiber light source. A Labview interface was programmed for equipment control, synchronization, image acquisition, processing, and visualization. Multispectral images captured by the CCD camera were used to reconstruct the tissue oxygenation map. Dynamic thermographic images captured by the infrared camera were used to reconstruct the vascular function map. Cutaneous tissue oxygenation and vascular function images were co-registered through fiduciary markers. The performance characteristics of the dual-mode image system were tested in humans.