Single-Molecule Force Microscopy: A Potential Tool for the Mapping of Polysaccharides in Plant Cell Walls

Plant cells are surrounded by a polysaccharide-rich cell wall that, as well as being a supporting structure (O’Neill & York, 2003), plays important roles in plant growth and development, and in the protection of plants from both biotic and abiotic stresses (Bowles, 1990). Plant cell walls are also of global economic importance, with the cell walls of food crops being of great nutritional value, while those of agricultural crops are important as a renewable resource for the textile and building industries, and increasingly as a sustainable source of fuel (O’Neill & York, 2003).


Introduction
Plant cells are surrounded by a polysaccharide-rich cell wall that, as well as being a supporting structure (O'Neill & York, 2003), plays important roles in plant growth and development, and in the protection of plants from both biotic and abiotic stresses (Bowles, 1990).Plant cell walls are also of global economic importance, with the cell walls of food crops being of great nutritional value, while those of agricultural crops are important as a renewable resource for the textile and building industries, and increasingly as a sustainable source of fuel (O'Neill & York, 2003).

Plant cell wall structure
The primary plant cell wall is a three-dimensional assembly of the polysaccharides, cellulose, pectin and hemicelluloses together with water, minerals and some structural glycoproteins.Atomic force microscopy (AFM) of Arabidopsis thaliana leaf cell walls showed the cellulose exists as microfibrils (Davies & Harris, 2003).AFM of living celery parenchyma tissue showed the cellulose microfibrils exist in highly ordered parallel array (Thimm et al., 2000).While AFM on isolated cell walls from celery and cucumber hypocotyls that were kept hydrated showed the cellulose microfibrils were undulating in a roughly parallel manner (Thimm et al., 2009;Marga et al., 2005).Moreover, each cellulose microfibril is surrounded by matrix material (presumably pectin and hemicelluloses) that keeps the celluloses uniformly spaced apart (Marga et al., 2005;Thimm et al., 2000;Thimm et al., 2009).Small angle X-ray scattering of hydrated celery collenchyma cell walls also showed uniform spacing of cellulose microfibrils (Kennedy et al., 2007).Solid-state 13 C nuclear magnetic resonance spectroscopy (NMR) indicated that in mung bean hypocotyls less than 10% of the surface of cellulose microfibrils has xyloglucan adhering to it (Bootten et al., 2004) and a recent three-dimensional solid-state NMR study of Arabidopsis thaliana cell walls supported this finding and also showed somewhat more pectin than xyloglucan adhered to cellulose (Dick-Perez et al., 2011).Indeed, Zykwinska et al. (2007) have previously shown arabinan and galactan side chains of pectin are bound to cellulose.This leads to two possible models for the primary cell wall.In the first cellulose microfibrils are held apart at a uniform distance by a limited number of xyloglucan crosslinks that are reinforced by more indirect crosslinks via the pectin matrix (Bootten et al., 2004;O'Neill & York, 2003).It is pertinent to note that xyloglucan has been found covalently linked to pectin in cell walls (Popper & Fry, 2008).In the alternative model the cellulose micofibrils are embedded in a matrix of pectin and hemicelluloses some of which adhere strongly to the cellulose but others interact with those bound to cellulose and the interactions diminish with the distance from cellulose (Cosgrove, 2001;Talbott & Ray, 1992).In other words a layer of other cell wall polysaccharides, of decreasing density, surrounds each cellulose rod and there is no direct cross-linking between cellulose microfibrils.However, it is important to note the structures of the primary plant cell walls detailed above are just models, developed from a wealth of information provided by studies that have conducted chemical analyses of isolated cell walls or independently imaged isolated, or more rarely intact cell walls (Thimm et al., 2009).Hence cell wall architecture and chemistry have yet to be fully reconciled.
The chemical composition of plant cell walls is usually studied by first isolating cell walls from the plant tissues to be analysed, and then sequential extraction techniques are used to fractionate the main polysaccharide groups that make up the cell wall (Thimm et al. 2009).While isolated cell wall polysaccharides can be analysed using chemical methods or by cytochemical staining to visualise the polysaccharide types and gather information about their quantity and location, they do not give a true picture of the cell wall in its undisturbed state (Carpita et al., 1996;Thimm et al., 2002;Vian & Roland, 1991).

Imaging the plant cell wall of higher plants
While imaging techniques such as transmission electron microscopy (TEM), field emission scanning electron microscopy (FESEM) and contact mode atomic force microscopy (AFM) can provide valuable information on the appearance, arrangement and structural dimensions of cell wall components (Hansma et al., 1997;Li, 1999;Thimm et al., 2002;Thimm et al., 2009), only limited chemical information can be obtained.For example, when using AFM, the contrast mechanism used to obtain an image is the force between the tip and the sample, and many samples produce the same force, thus it is difficult to reliably determine the chemical groups that are imaged.However, single-molecule force spectroscopy (SMFS), a technique where AFM tips are modified by the addition of specific molecules has the potential to provide both structural and chemical information (Mueller et al., 2009a(Mueller et al., & 2009b)).

Contact mode AFM
Contact mode AFM is a non-destructive imaging technique by which images with nanometer resolution can be obtained, and is therefore ideal for studying cell wall structure.In this mode a sharp stylus or tip is scanned across the surface of a sample and the force between the tip and sample is measured at a series of points.A colour or grey scale is assigned to each force, and then a map of colours displays an ``image'' of the surface (Thimm et al., 2000).This form of AFM has been used to examine isolated cell wall material Single-Molecule Force Microscopy: A Potential Tool for the Mapping of Polysaccharides in Plant Cell Walls 283 from plants including apple, water chestnut, potato, carrot (Kirby et al., 1996a(Kirby et al., & 1996b) ) cotton (Pesacreta et al., 1997) and celery (Thimm et al., 2009), and in the "intact walls" of celery parenchyma cells (Thimm et al., 2002) and single Zinnia elegans tracheary elements (Lacayo et al., 2010).For a recent review of the application of AFM and other scanning probe microscopy techniques potentially useful for the study of plant cell walls see Yarbrough et al. (2009).While providing useful structural information this technique does not provide any chemical information regarding the structures observed.Ducker et al. (1992) demonstrated that the forces between colloid particles could be measured by attaching a colloid particle to the end of an AFM cantilever.In biological systems, this approach has been adapted to estimate bond strength between different biological molecules such as ligands and receptors (You & Yu, 1999), antigen antibody binding (Nyquist et al., 2000), receptor binding (Raiteri et al., 1999), and the molecular assembly of biomolecules (Kossek et al., 1998, Perrin et al., 1999).

Chemical and single-molecule AFM
The principle of sensor force measurement involves attaching one type of molecule to the AFM tip and binding another molecule to a surface.The tip is moved towards the surface until the molecules interact.This has been demonstrated for biotin-streptavidin (Florin et al., 1994), avidin-biotin (Moy et al., 1994), for cell-cell interactions (Antonik et al., 1997), celladhesion proteoglycans (Dammer et al., 1995;Misevicz, 1999) and antigen-antibody systems (Dammer et al., 1996;Hinterdorfer et al., 1996).This technique can also be used to gain chemical information from the surfaces of structural components isolated from cells or from the surfaces of living cells (Mueller et al. 2009a(Mueller et al. & 2009b)).For example, Gad et al. (1997) used gold-coated tips functionalized with concanavalin A to map the cell wall polysaccharides of living microbial cells.
Single-molecule force sensors can also be used to resolve topographical information (Muller et al., 2009a(Muller et al., & 2009b)).Force mapping is a force sensing technique where a small area is raster force probed.Each force curve can potentially be identified and a force map generated, showing for example the distribution of cell wall polysaccharides as mentioned above.When using single-molecule force microscopy, the probe is not in physical contact with the surface, but instead uses the attractive/repulsive part of the force curve, whereby the cantilever is purely force driven.We used a variant of this technique, with sugars attached to the tip of the cantilever, in an attempt to map the distribution of polysaccharides in the parenchyma cell walls of celery epidermal peels, produced as described in Thimm et al. (2000).In the following sections we present some preliminary findings regarding the development of sensors to both image and to determine the chemical characteristics of celery parenchyma cell walls.

A system for single-molecule force microscopy of celery (Apium graveolens L.) parenchyma cell walls using glycoside probes
4.1 The celery cell wall Thimm et al. (2002) isolated the primary walls of celery parenchyma cells and their polysaccharide components were characterized by glycosyl linkage analysis, crosspolarization magic-angle spinning solid-state 13 C nuclear magnetic resonance (CP/MAS 13 C NMR) and X-ray diffraction.The cell walls consisted of mainly cellulose (43 mol%) and pectic polysaccharides (51 mol%), made up of rhamnogalacturonan (28 mol%), arabinan (12 mol%) and galactan (11 mol%), with only smalls amounts of xyloglucan (2 mol%) and xylan (2 mol%) detected in the cell walls.Solid-state 13 C NMR signals were consistent with the constituents identified by glycosyl linkage analysis and allowed the walls to be divided into three domains, based on the rigidity of the polymers.Cellulose (rigid) and rhamnogalacturonan (semi-mobile) polymers responded to the CP/MAS 13 C NMR pulse sequence and were distinguished by differences in proton spin relaxation time constants.The arabinans, the most mobile polymers, responded to single-pulse excitation (SPE), but not CP/MAS 13 C NMR. From solid-state 13 C NMR of the cell walls the diameter of the crystalline cellulose microfibrils was determined to be approximately 3 nm while X-ray diffraction of the cell walls gave a value for the diameter of approximately 2 nm.

Sensor design
Glucose and galactose were chosen as sensor molecules as cellulose is a polymer of glucose and is one of the main structural components in the celery cell wall, while galactose is an important component of celery pectins (Thimm et al., 2002).The two main types of probes used in conventional AFM experiments are either silicon-nitride (pyramidal) or silicon (ultralever).Sugar groups can be covalently attached directly to the silanol or silolamine surface groups on commercial tips, or the commercial tip can be coated in a thin film that is convenient for further reactions.Gold is a useful film because a broad spectrum of groups can be attached to gold via thiol linkages.Thiols are well known to adsorb under mild conditions onto gold surfaces (Biebuyck & Whitesides, 1994;Fritz et al., 1996;Liu et al., 1994, Nyquist et al., 2000) and thiol adsorption is strong enough to use sugar-thiols as sensors, for these reasons we chose to use gold-coated tips.
Most interactions in biological systems depend on the orientation of the molecules.Therefore, forces between the sensor and the surface can only be detected if the sensor is able to align itself unhindered relative to the surface molecule.A carbon spacer chain was introduced to ensure that the sensor molecule had enough flexibility to sample different orientations relative to the sample.The chain-length of the spacer between the sugar ringsystem and the probe determines its flexibility or ability to rotate freely, and a spacer of C4carbons between the interacting sugar ring and the thiol-group was found to be sufficient for probing plant cell wall extracts and surfaces (Thimm, 2000).

Synthesis of glycoside-sensors and attachment of glycoside-sensors to goldcoated cantilevers
The synthetic approach for the derivatisation of thiobutyl-glycosides with a C4-spacer (butyl-spacer) can be found in Thimm (2000) and the probes used are shown in Figure 1.
To avoid contamination of the gold-coated AFM cantilevers, all sensor attachment procedures were conducted in a laminar flow cabinet, in 9 cm polystyrene petri dishes.A glass-slide was cut to fit inside a petri dish and 30 min prior to sensor attachment, distilled ethanol was poured around the glass-slide and the dish was covered to saturate the atmosphere with ethanol vapour.The cantilevers were placed onto the glass-slide and 3-4 drops of a dilute ethanolic solution (1 to 5 mmol) of the thio-glycosides was placed onto the gold-coated cantilever and the petri dish was covered.If the sample appeared to be drying Fig. 1.The probes used in the experiments detailed below out within the incubation time, a drop of distilled ethanol was added.The surfaces were then carefully washed with ethanol and then deionised water, and stored wet.Force experiments were conducted where glycoside-sensors were used to probe gold-coated surfaces, to which glycosides were attached, to the determine if the sensors were sufficiently coated with thio-glycosides and to determine the functionality of the system (Thimm, 2000).

Cell wall fractionation
As discussed in section 3 plant cell walls can be studied by selectively fractionating the polysaccharide groups that make up the wall, using sequential extractions, so that each fraction can be studied in more detail (Melton & Smith, 2005).Water-soluble pectins are extracted with boiling water or cold HEPES buffer, while less-methylated pectins, held in the wall with calcium bridges, are solubilised with calcium chelating agents such cyclohexanetrans-1,2-diaminetetraacetic acid (CDTA).Esterified pectins, and some hemicelluloses are extractable with weak alkali (sodium carbonate, Na 2 CO 3 ).Hemicelluloses and some branched pectins are extracted with 1M KOH, while 4M KOH is used to extract more hemicelluloses, including xylogucans, and additional branched pectins (Siddiqui, 1989a and1989b).The insoluble final residue is largely cellulose, with any remaining hemicelluloses and some complex pectins (Percy et al., 1997, Redgwell et al., 1997;Thimm et al., 2009).
We used force mode AFM, using the glycoside-sensors detailed in the last section, to probe the cell wall components isolated/extracted from celery parenchyma cells as described by Thimm et al. (2009).The components probed were the final insoluble residue largely composed of cellulose, water-soluble pectins (referred to as the HEPES fraction) and the esterified pectins, and some hemiceluloses (referred to as the carbonate fraction).Briefly a dilute (1-4 mg/mL) 20 % ethanolic suspension of each component was adsorbed onto freshly cleaved mica and allowed to dry in air for up to 20 min.To avoid floating particles the sample was careful rinsed with 1-2 mL dionized water (Thimm et al., 2009).The cell wall components were then probed using glycoside-sensors and force curves generated for the cell wall components detailed above.

The final insoluble residue (cellulose)
To determine if the force curves associated with the final residue were largely due to the chemical properties of cellulose, a standard (Whatman cellulose, CF11) was probed and the force curves compared to those of the insoluble final residue.The force curves for both the Gal:R 1 =OH, R 2 =H; Glc:R 1 =H, R 2 =OH www.intechopen.comAtomic Force Microscopy Investigations into Biology -From Cell to Protein 286 CF11 and the final insoluble residues were similar (Figure 2).The force curves showed jumps starting at 8 to 8.3 nm distance and less defined jumps at 3 to 3.1 nm and 2 to 2.3 nm distance.The force curves generated when the final insoluble residue was probed with either a galactoside or glucoside sensor showed no clear differences (Figure 3).

Esterified pectins and some hemicelluloses (the carbonate fraction)
Figure 5 shows a force-distance plot for a glucoside-sensor vs. the carbonate fraction.Two distinct jumps were observed at 16.5 nm ± 1 (with a displacement force of 0.2 nM ± 0.1) and at 2.9-3.2nm (0.1 nN ± 0.05).Two weak repulsions at 5.4 ± 0.1 and 9.2 nm ± 0.7 were also observed, but these were often not well resolved.

Mixture of fractions
Force mapping experiments were also conducted on mixtures of the three cell wall components in an attempt to simulate an in situ system.This experiment attempted to determine if cell wall polysaccharides could be identified, by their force curves, in a complex mixture and if the results could then be used to generate a polysaccharide distribution map.A mixture of the three cell wall components (1:1:1 w/w) were adsorbed onto a mica sheet and the sheet was force mapped using a glucoside-sensor.Figure 6 shows a polysaccharide distribution map of such a mixture, showing the distribution of the three cell wall components, identified based on their force curves.
For one sample of mixed components, 242 force curves were generated.Of these, 78 counts (32 %) were for cellulose, possibly indicating that cellulose could readily be detected on the www.intechopen.comsurface, possibly as a result of swollen microfibrils that are larger and hence were more easily detected that the other cell wall components.The HEPES fraction was detected 44 times (18 %).This result was unexpectedly low and may have indicated preferred adsorption of the HEPES polysaccharides to the mica surface.Mica (muscovite) is a potassium aluminium silicate hydroxide fluoride that exposes silicate oxygen atoms at the surface, resulting in a negative charge in water.This charge might induce the preferred adsorption of particular polysaccharides from the polysaccharide mixture.Once the first species is adsorbed onto mica, their properties might drive the www.intechopen.com

Single-Molecule Force Microscopy:
A Potential Tool for the Mapping of Polysaccharides in Plant Cell Walls 289 aggregation of the next layer.It is possible that the HEPES fraction adheres more quickly to the mica than the other cell wall components.The other cell wall components could then form a second or third layer on top of the HEPES fraction and thus it would be less accessible to the sensor, which probes only the surface.
The carbonate fraction was only detected slightly more frequently (21 %) than the HEPES fraction.This pectic material might show a similar behaviour as the HEPES fraction and hence compete for mica surface space.Adsorption to mica is possibly through the formation of electrostatic bonds with the uronic acids of the HEPES fraction.Greater branching and methyl-esterification of the carbonate fraction could result in a reduced ability to compete for mica, even though the sodium carbonate extraction might have caused saponification of methyl-esters to some degree.Another possible explanation might be that the three cell wall components aggregate in solution prior to adsorption to the mica.
A large number (70 counts (29 %) of 242 force curves) of force curves could not be identified.One factor contributing to this was the surface of the polysaccharide mixture seemed to be very rough for force experiments.During raster scanning, forces were probed in regions consisting of gaps or raised regions of the surface.Another explanation might be that forces that could not be assigned are a direct result of intermolecular aggregation of the three cell wall components in solution, resulting in new unidentified force profiles.The distribution maps for the three cell wall components showed an even distribution covering most of the scanned region.No accumulation of individual cell wall components at particular areas was observed.

Imaging using force mode
Figure 7 shows an AFM deflection image of a celery epidermal peel in situ, captured in force mode, using a glucoside sensor.Force mode was used to image the cell in order to prevent or minimize contamination of the sensor by reducing physical contact with the surface.The force mode images of the cell wall of celery parenchyma epidermal peels are not as sharp or defined as previously published contact mode images (Thimm et al., 2000).

Force mapping the surface in situ
Prior to force mapping the cell wall, a force mode image was captured and was used to guide the probe (reading x, y coordinates in nm) when force mapping the surface in situ.For the image shown in Figure 7, 126 force curves were generated, from different points on the sample and the structural identity of each point was determined by comparison to the forces curves generated from the isolated cell wall components.For each component of the wall, an example of a force curve from the in situ experiments is shown below, compared to the isolated cell wall components, cellulose (Figure 8), HEPES (Figure 9) and carbonate (Figure 10).Of the 126 force curves generated in situ using the glucoside-sensor, 53 (42 %) could be attributed to cellulose, 9 (7 %) could be attributed to HEPES fraction polysaccharides and 13 (10 %) could be attributed to carbonate fraction polysaccharides.Of the 126 curves, 51 (41 %) could not be identified.The large number of forces that could not be assigned could be due to the presence of non-polysaccharide material, to aggregated polysaccharides or to Fig. 7. AFM force mode deflection image of celery epidermal peels in situ captured using a glucoside sensor (scan size 2 μm, scale bar 200 nm, height scale 5 nm) Fig. 8. Representative force-distance plots, using a glucoside-sensor, showing two over laid force curves.One for celery parenchyma cells in situ and the second for the cell wall final residue (mainly cellulose).polysaccharide components of the cell wall for which no reference force curves were available.Most of the force curves that could not be identified showed a strong exponential repulsive behaviour, which could be due to the presence of high molecular weight polymers or aggregates.By overlaying data points of the force maps for each identified cell wall component, with the force mode AFM deflection image in Figure 7, an image of the wall in situ with some chemical information could be generated (Figure 11).Force curves identified as belong to the final cell wall residue did appear to be localised more frequently on the raised microfibril like structures clearly seen in Figure 11.However, no distinct pattern of localisation of the three cell wall components to any particular area of the image was consistently observed.

Conclusions
The use of force curves generated by single molecule sensors is potentially a useful way that information about the chemical nature, the strength of interactions, and possibly the molecular arrangement of complex biological structures can be deduced (Gad et al., 1997;Hansma et al., 1997;Li et al., 1998;Li et al., 1999;Marzalek et al., 1998;Osada et al., 1999;Okabe et al., 2000;Rief et al., 1997).To date a few attempts have been made to force probe intact plant cell walls in situ.The approach taken in our investigation demonstrates the potential for using single molecule force microscopy to study plant cell wall chemistry and structure and opens up a variety of possible uses, depending on the type of sensor chosen.By probing, using the glucoside-sensors, the three celery cell wall components (HEPES fraction, carbonate fraction and the final residue) in isolation and a mixture of them, we have shown that individual cell wall components can be distinguished in a complex mixture.
When force mode imaging, using glucoside-sensors, was combined with force curve identification, it was possible to map the distribution of cell wall components in celery epidermal peels in situ.However, while a cell wall component distribution map of celery epidermal peels in situ was obtained, showing the chemical information overlaid over a topographical AFM force mode image, the HEPES and carbonate fraction polysaccharides did not show any regular or ordered distribution.Only force curves identified as belong to the final cell wall residues showed some localisation, appearing more frequently on the raised microfibril-like structures.
For celery, the chemical resolution of the force map was limited to the number of calibration standards produced.We used only three polysaccharide extracts (HEPES fraction, carbonate fraction and the final cellulose residue).The quality and/or purity of the fractions used as standards is unknown, because it is not known how many different types of polysaccharides are present in, for example, the HEPES or carbonate fraction.More sophisticated extraction and purification methods could be used to produce cell wall standards to generate force curves that could be interpreted in more detail.
By increasing the number of calibration standards a larger number of forces could be identified.Ideally polysaccharides that are highly purified or synthesised, to give molecules of known composition, molecular arrangement and linkage type (e.g.xyloglucan, xylan) could be used.A different approach might be to mimic polysaccharide structures in the plant cell wall by synthesising identified structural elements of, for example, polysaccharide sidechains of pectin (galactans and arabinans).Their conformation, strength and fingerprint forces could then be studied systematically.
In the present study, a number of forces could not be identified.One explanation for this was the lack of calibration standards, but another important factor might be the interaction of polysaccharides in solution and in situ.Aggregation of polysaccharides has been observed for xyloglucan and cellulose as well as pectin sidechains and cellulose (Pauly et al., 1999;Whitney et al. 1995Whitney et al. & 1999;;Zykwinska et al., 2007) and this could influence the force curve observed.Force measurements of calibration standards (e.g.hemicelluloses such as xyloglucan and xylan, pectins such as homogalacturonan, arabinan, galactan or arabinogalactan) need to be made of polysaccharides mixed in solution prior to adsorption on to a mica surface.In addition, further experiments need to be conducted to determine what influences polysaccharide aggregation might have on the observed force curves.

Fig. 4 .
Fig. 4. A force-distance plot for a glucoside-sensor vs. the HEPES fraction.No differences were observed between the glucoside-and galactoside-sensors.Jumps are indicated by an *.

Fig. 5 .
Fig. 5.A representative force-distance plot for a glucoside-sensor vs. the carbonate fraction.Jumps are indicated by an *.

Fig. 6 .
Fig.6.A force curve distribution map of three cell wall components adsorbed onto a mica sheet and force mapped using glucoside-sensor.

Fig. 9 .
Fig. 9. Representative force-distance plots, using a glucoside sensor, showing two over laid force curves.One for celery parenchyma cells in situ and the second for HEPES fraction.

Fig. 10 .
Fig. 10.Representative force-distance plots, using a glucoside-sensor, showing two over laid force curves.One for celery parenchyma cells in situ and the second for the carbonate fraction.

Fig. 11 .
Fig. 11.A map of a celery epidermal cell wall in situ, showing the cell wall component distribution (cellulose, circle; HEPES, square; carbonate, diamond; unassigned forces, X).