INTRODUCTION
Characterization of the structural and physical properties of microbial cell surfaces is a continuously expanding field of microbiology.[1-10] Studying cell surfaces is important not only in basic research to elucidate their functions (cell shape, protection, molecular sieve, molecular recognition, cell adhesion, and cell aggregation), but also in medicine (fouling of implants, microbial infections) and biotechnology (cellular interactions in fermentation technology, removal of heavy metals). Electron microscopy has long been the technique of choice for probing cell surface structure down to the molecular level.[1,2,4,6,7] Yet the requirement of vacuum conditions may raise the question of whether the information obtained is always relevant to the real, hydrated state. In parallel, a variety of approaches have been developed to probe the cell surface physical properties, including chemical analysis of wall constituents, specific binding studies, selective degradation by enzymes, cell wall mutants, modifications by antibiotics, and surface analysis using various physical techniques.[1,4,5,10] Many of these techniques provide averaged information obtained on a large ensemble of cells; that is, they are not suited for mapping the nanoscale distribution of physical properties on individual cells.
Since the late 1980s, atomic force microscopy (AFM)[11] has been increasingly used in biology and it is now established as a versatile tool to address the structure, properties, and functions of biological structures, going from single molecules to lipid membranes and living cells.[12-17] Basically, AFM offers two main advantages over conventional microscopy techniques. First, it provides images of the specimen with nanometer (subnanometer) resolution, in real-time and under physiological conditions. Second, because the instrument works by sensing the force between a very sharp tip and the sample surface, this principle can be exploited to measure molecular interactions and physical properties on a local scale. These capabilities provide a range of novel opportunities in microbiology.[18] For example, the following questions can now be addressed with AFM. Does the surface nanostructure of untreated living cells correlate with that observed by electron microscopy? How does surface structure change with time during dynamic processes such as cell growth and cell division? How do external agents such as enzymes and antibiotics affect the cell surface architecture? Do surface properties such as hydrophobicity, charge, and elasticity vary across the surface of a single cell? What are the intermolecular forces involved in molecular recognition and cellular interactions? What is the elasticity of single-cell surface macromolecules? The intent of this article is to survey recent achievements brought by AFM in microbiology, particularly emphasizing studies on whole cells. Rather than providing an exhaustive review of the literature in the area, the paper discusses imaging and force spectroscopy applications using a selection of recent data.
IMAGING
Image Recording and Sample Preparation
Atomic force microscopy images are created by raster-scanning a sharp tip across the sample, while sensing the force experienced by the tip. To this end, the sample is mounted on a piezoelectric scanner that ensures three-dimensional positioning with high resolution. The force is monitored by attaching the tip to a soft cantilever and by measuring the bending or ”deflection” of the cantilever. The larger the cantilever deflection, the higher the force that will be experienced by the tip. Nowadays, most instruments use an optical method to measure the cantilever deflection with high resolution. A laser beam is focused on the free end of the cantilever and the position of the reflected beam is detected by a position-sensitive photodiode.
In contact mode, the most common imaging mode, images can be acquired in two ways, i.e., either in the constant-height mode in which the cantilever deflection is recorded while the sample is horizontally scanned, or in the constant-deflection mode in which the sample height is adjusted to keep the deflection of the cantilever constant using a feedback loop. In the latter case, the feedback output can be used to display a true height image, which gives calibrated height information about the sample morphology. Alternatively, the error signal can also be employed to generate a so-called deflection image that does not provide quantitative height information but is more sensitive to fine surface details. Besides contact mode, several other imaging modes have been developed. Among these, tapping mode atomic force microscopy in which the tip is oscillated at a given frequency is very useful for imaging soft samples.
After the first AFM images of biological structures were published, it was soon realized that a crucial factor for successful imaging is specimen preparation. This led to the development of a number of immobilization methods. For single molecules, two-dimensional protein crystals, lipid membranes, and animal cells, immobilization protocols are generally based on physical adsorption onto flat substrates such as mica, glass, and silicon oxide (for a detailed description of these procedures, see Ref. [19]). However, for microbial cells, sample preparation is often challenging because of the cell geometry. Indeed, a microbial cell must be viewed as a fairly rigid spherical or rod-like particle of 1-5 mm size. This geometry usually leads to a very small cell-substrate contact area and therefore to cell detachment by the scanning tip. Presumably, this is the main reason why the potential of AFM in cellular microbiology had been neglected in the early age. However, the past few years have seen tremendous progress in the application of AFM to microbiological specimens owing to improvements in imaging conditions and sample immobilization procedures.
The selection of an appropriate imaging environment is critical for high-resolution imaging because it directly influences the force acting between tip and sample. In air, high-resolution imaging is often difficult as a result of strong adhesion forces resulting from the presence of a water layer on both tip and sample. This problem is eliminated when imaging in aqueous solution, which is actually the most relevant environment for microbiological specimens. By selecting appropriate buffer conditions (nature of electrolytes, ionic strength, pH), it is generally possible to maintain a very low applied force, typically in the range of 0.1-0.5 nN.[20] Hence it is certainly very useful to investigate the effect of salt composition, pH, and ionic strength on the image quality when high resolution is desired.
Imaging of Living Cells
By combining high spatial resolution with the possibility to work in liquids, AFM enables researchers to visualize the surface morphology of untreated living cells, thus providing information that is complementary to data obtained with classical microscopy techniques. As mentioned above, an important requirement is that the cells must be firmly attached to a solid substrate. This can be achieved by mechanically trapping the cells into a porous polymer membrane. Fig. 1 presents a three-dimensional height image showing a single Saccharomyces cerevisiae cell protruding from a pore of a polymer membrane. In these conditions, images could be repeatedly obtained in aqueous solution without detaching the cell or altering the surface morphology. Generally, the entire cell surface was very smooth, an observation that was consistent with previous chemical and electron microscopy studies. However, some cells showed circular protrusions, about 1 mm in diameter, which reflected the bud scars left on the surface after cell division.[21]
Provided that the cell is well attached, it is possible to acquire high-resolution images at different locations. For S. cerevisiae, topographic images were recorded to a lateral resolution of 2 nm, without significant modification of the surface morphology. The surface roughness (on 250 x 250 nm areas) was found to be smaller than 1 nm.[21] Interestingly, a very different surface architecture was observed for fungal spores. Fig. 2A shows that the surface of Aspergillus oryzae spores was covered by a layer of regularly arranged nanostructures, referred to as rodlets, that were several hundred nanometers in length and had a periodicity of 10 nm.[22] Presumably, these crystalline-like nanostructures play an important role in determining the biological functions of the spore (i.e., dispersion, protection).
Accordingly, AFM is a valuable technique for probing the surface architecture of native microbial cells. However, researchers must realize that high-resolution imaging often remains difficult, especially for bacterial surfaces. A first source of problems is related to sample softness. The above images were obtained on cells that possess thick, mechanically strong cell walls. In comparison, bacterial walls are thinner and more fragile, meaning that imaging may be destructive, especially in contact mode. For instance, the image contrast obtained for the surface of Lactococcus lactis was dominated by tip-sample interactions; that is, grooves oriented along the scanning direction were created by the scanning tip.[23] For some bacteria, the presence of flexible, loosely bound macromolecules or appendages at the surface may further limit the image resolution. The immobilization procedure may also be an issue. The porous membrane approach described here is only suited for spherical cells, indicating that alternative strategies must be used for other geometries. Consequently, further developments are necessary (nondestructive dynamic imaging modes, novel immobilization procedures) before researchers can achieve molecular resolution and perhaps monitor molecular conformational changes as is already the case with isolated cell surface layers made of two-dimensional protein crystals.[14]
Fig. 1 Atomic force microscopy height image (6 x 6 mm; z range: 1 mm) of a living S. cerevisiae cell immobilized in a porous polymer membrane.
Fig. 2 High-resolution deflection images acquired under water for the surface of A. oryzae spores: (A) dormant spore (500 x 500 nm); (B) spore after ~ 10 h of germination (1 x 1 mm).
Following Dynamic Processes
Of particular interest is the possibility to study the changes of cell surface structure associated with dynamic processes such as cell growth and germination. An example of this is the germination of fungal spores. For A. oryzae spores,[22] high-resolution images revealed that dramatic changes of the surface morphology occurred upon germination, the rodlet layer (Fig. 2A) changing into a layer of soft material (Fig. 2B). On close examination, this material showed streaks oriented in the scanning direction, suggesting that soft, loosely bound material was interacting with the scanning tip. These direct observations were in good agreement with previous structural and chemical studies showing that germination results in the disruption of the proteinaceous rodlet layer and reveals inner spore walls that are rich in polysaccharides. As we shall see below, morphological changes were directly correlated with profound modifications of molecular interactions as measured by force spectroscopy.
Fig. 3 Series of deflection images (4.2 x 4.2 mm; insets: 2.5 x 2.5 mm) recorded in real time for a single S. cerevisiae cell prior to and after contact with a protease solution for 6, 9, 15, 19, 30, 40, 50, and 60 min.
Real-time imaging also enables to follow the progressive effect of external agents such as solvents, ions, chemicals, enzymes, and antibiotics on the cell surface. In this context, the surface morphology of yeast cells was investigated at fixed time intervals following addition of protease and amyloglucosidase solutions.[21] Progressive changes of the cell surface topography were clearly observed after protease addition (Fig. 3). With time, the surface became eroded and showed large depressions, about 500 nm in diameter, surrounded by protruding edges, about 50 nm in height (estimated from height images). By contrast, no modification of the cell surface was noted upon addition of amyloglucosidase, which was consistent with the cell wall biochemical composition. These real-time experiments have a strong potential in cell wall enzyme digestion studies as well as in medicine, where they may be used to investigate the effect of antibiotics on microbial cell walls.
FORCE SPECTROSCOPY
Principle of Force Spectroscopy
The basic idea behind force spectroscopy is that rather than scanning the tip in the x and y directions, the cantilever deflection is recorded as a function of the vertical displacement of the sample, i.e., as the sample is pushed toward the tip and retracted from it. Using appropriate corrections, the raw voltage vs. displacement curve is converted into a force vs. distance curve, which then enables the measurement of molecular interactions and physical properties.[24,25] Interestingly, spatially resolved information can be obtained by recording a so-called ”force-volume image” consisting of arrays of force-distance curves.[25]
Fig. 4 Example of force-distance curve.
A typical force-distance curve obtained for a nonde-formable surface in liquid is shown in Fig. 4. When the tip is far away from the sample surface, there is no force acting on the tip and the cantilever is not deflected (label 1). As the tip approaches the surface, the cantilever may bend upward as a result of repulsive forces (label 2) until the tip jumps into contact when the gradient of attractive forces exceeds the spring constant plus the gradient of repulsive forces (label 3). When the force is increased in the contact region, different behaviors may be observed depending on the sample stiffness. On a hard, nondeform-able surface, the plot is a vertical line (label 3), while on a soft surface, elastic indentation will occur, leading to a different shape. In the latter case, analyzing the in-contact part of the curve with appropriate theoretical models enables quantitative determination of the sample elasticity. Upon retracting the tip from the surface, the curve often shows a hysteresis referred to as the adhesion ”pull-off” force (label 4), which can be used to estimate the surface energy of solids or the binding forces between complementary biomolecules. In the presence of long, flexible macromolecules, elongation forces usually develop nonlinearly as a result of the deformation of the molecules. In this case, it is common to present the force curve as the positive pulling force vs. extension. Theoretical models from statistical mechanics enable the elasticity of the molecules to be deduced from the force-extension profile. In the next sections, the various applications offered by force spectroscopy in microbial cell surface research are surveyed.
