Nanostructure of Native Pectin Sugar Acid Gels Visualized by Atomic

Biomacromolecules 2004, 5, 334-341

334

Nanostructure of Native Pectin Sugar Acid Gels Visualized by Atomic Force Microscopy† Marshall L. Fishman,*,‡ Peter H. Cooke,§ and David R. Coffin‡ Crop Conversion Science and Engineering Research Unit, and Microbial Biophysics and Residue Chemistry and Core Technologies Research Unit, Eastern Regional Research Center, Agricultural Research Service, U.S. Department of Agriculture, Wyndmoor, Pennsylvania 19038 Received September 10, 2003; Revised Manuscript Received November 25, 2003

Height and phase shift images of high methoxyl sugar acid gels (HMSAG) of pectin were obtained by atomic force microscopy in the tapping mode. Images revealed that pores in these gels were fluid and flattened out when measured as a function of time. These images revealed for the first time the structure of adsorbed sugar on pectin in the hydrated native gels and how the pectin framework is organized within these gels. Segmentation of images revealed that the underlying pectin framework contained combinations of rods, segmented rods, and kinked rods connected end to end and laterally. The open network of strands was similar to pectin aggregates from 5 mM NaCl solution imaged earlier by electron microscopy (Fishman et al., Arch. Biochem. Biophys. 1992, 294, 253). Area measurements revealed that the ratio of bound sugar to pectin was in excess of 100 to 1 (w/w). Furthermore, images indicated relatively small differences in the organization of native commercial citrus pectin, orange albedo pectin, and lime albedo pectin gels at optimal pH as determined in this study. The findings are consistent with earlier gel strength measurements of these gels. In addition, values of gel strength were consistent with values of molar mass and viscosity of the constituent pectins in that they increased in the same order. Finally, we demonstrated the advantage of simultaneous visualization of height and phase shift images for observing and quantitating the nanostructure of relatively soft gels which are fully hydrated with a buffer. Introduction Pectin is a heterogeneous polysaccharide located in the cell walls of many plants.1 The most abundant constituent in pectin is homogalacturonan (HG) which is (1f4) linked, R-D-galacturonic acid and its methyl ester. Some HGs contain pendent R-D-xylose units. Another constituent present in pectin is rhamnogalacturonan I which contains arabinan, galactan, and arabinogalactan side chains. These constituents account for most of the monosaccharide units present in pectin preparations. Several important functions in cells have been attributed to pectin. These functions include governing cell wall porosity, modulating cell wall pH and charge, regulating intercell adhesion at the middle lamella, and signaling to plant cells the presence of foreign bodies such as symbiotic organisms, pathogens, and insects.1 Furthermore, there are many reports on the activity of pectin fragments as nutraceuticals.2 Some of these activities include immunostimulation, anti-metastasis, hypoglycemic, and cholesterol lowering effects. The ability of pectin to serve as a gelling and texturizing agent in unprocessed and processed foods is its most † Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. * To whom correspondence should be addressed. Tel: 215-233-6450. Fax: 215-233-6559. E-mail: [email protected]. ‡ Crop Conversion Science and Engineering Research Unit. § Microbial Biophysics and Residue Chemistry and Core Technologies Research Unit.

10.1021/bm0300655

important property as an article of commerce.3 High methoxyl food gels are formed typically from a hot solution (ca. 95-98 °C) of 0.2-0.4% high methoxyl pectin, 65 g of sucrose, and sufficient citric acid to lower the pH to the range of 3-3.6. These gels are commonly referred to as high methoxyl sugar acid gels (HMSAG). One school of thought is that the structure of these gels can be modeled as polymeric random coils cross-linked to form network structures.4 Flory5 first developed the theory for the formation of polymer networks to describe the gelation of synthetic polymers which behave like random coils in solution. According to Flory’s theory, polymer networks arise when a critical number of bridging molecules with a functionality of three or more are introduced into a polymer solution. The networks form via junction points. By way of contrast, pectin and certain other polysaccharides are thought to form cross-links at junction zones held together by cooperative interactions.6 It has been proposed that the stability of these junction zones depends on the specific energy of the interactions and the number of units and molecules cooperatively involved in the junction zones. Shear modulus studies7 near the sol-gel transition revealed that the junction zones of HMSAG of pectin were stabilized by a combination of hydrogen bonds and hydrophobic interactions. About 2/3 of the junction zone interaction energy was derived from hydrogen bonding and the remaining 1/3 derived from hydrophobic interactions. Furthermore, it has been argued that the requirement for the presence of high sugar content to lower water activity and

This article not subject to U.S. Copyright. Published 2004 by the American Chemical Society Published on Web 01/21/2004

Nanostructure of Native Pectin Sugar Acid Gels

low pH to reduce charge-charge repulsions, indicates the importance of chain-chain interactions (i.e., junction zones) in the gelation of HMSAG of pectin.4,8 X-ray diffraction studies9 on oriented pectin fibers in the solid state indicate that pectin can form helical structures and that these helices form adjacent dimers. This work has been cited as evidence for ordered helical structures as the basis of junction zones.7 Nevertheless, the number of polysaccharides involved in junction zones is an unresolved question.4 Electron microscopy has played an important role in imaging the ultrastructure of pectate,10 apple pectin,11 onion pectin,12 and peach pectin13,14 dried from aqueous solution. More recently atomic force microscopy (AFM) has been employed to image pectin extracted from tomato.15,16,17 Interestingly, there appears to be some similarities in results between images13,14 of alkaline soluble peach pectin (ASPP) which were obtained from rotary shadowing of platinum replicas by conventional electron microscopy and AFM images of alkaline soluble tomato pectin. As in the case of tomatoes,15 log-normal distributions of contour lengths were found for peach pectin.13,14 Furthermore, contour lengths of images from histograms ranged from 30 to 390 nm for tomatoes, whereas histogram contour lengths ranged from 10 to 330 nm in the case of peaches. In both cases, branched structures were observed. For tomatoes, it was suggested that these branched structures were single molecules based on height measurements, whereas in the case of peaches, branched structures disappeared when samples were dried from 50% aqueous glycerol solutions rather than 0.005 M NaCl. Instead, separate rods, segmented rods, and kinked rods were observed. Thus, it was concluded that the branched structures imaged were comprised of rods, segmented rods, and kinked rods held together by noncovalent interactions. Moreover, in these same studies, it was found that infinite networks and circular networks (i.e., microgels) were imaged when ASPP was dissolved in water. Also it was found that components of intact networks had comparable contour length distributions to the dissociated components imaged from samples which had been dried from aqueous 50% glycerol. One may conclude from these studies that, under appropriate conditions of ionic strength and pH, linear pectin molecules could undergo aggregation to form branched structures which, following cross-linking of a critical number of molecules, can form circular networks. At sufficiently high concentration, these microgels could coalesce into an infinite network, namely a gel. The microgel also could be thought of as a circular blob, i.e., a subunit of a pectin network.18 Thus, at least, in the case of ASPP, we have imaged from dilute solution all of the structures required to form a gel as suggested by de Gennes.18 In the case of pectin, one possibility for the formation of a branched aggregated structure is an intermolecular hydrogen bond between a terminal carboxyl and a side chain hydroxyl. Another possibility is an intermolecular hydrogen bond between a terminal carboxyl and a hydroxyl at the 4 position of rhamnose in the galacturonate backbone. Presumably, this type of aggregation would go undetected by AFM height measurements in the absence of dissociating solvents. It is also possible that there are branched pectins in which

Biomacromolecules, Vol. 5, No. 2, 2004 335

branches are attached to the pectin backbone either by covalent links or by hydrogen bonding. In any case, atomic force microscopy is a powerful tool in that it offers the possibility of resolving the structure of pectin gels under native, hydrated conditions. Presently, three modes of AFM operation have been employed.19 These are contact, noncontact, and tapping. In all three modes, a sharp probe tip attached to a cantilever scans the sample. In contact-mode AFM, the probe tip is allowed to skim across the surface of the sample and remains in contact with the sample surface through an adsorbed layer of fluid at the sample surface. In AFM noncontact-mode, the probe tip is not in contact with the adsorbed layer of fluid on the sample surface but oscillates above it while scanning above the sample. In AFM tapping mode, an oscillating probe tip taps at high frequency (ca. 350 kHz) the sample while scanning it. Research concerning AFM in contact-mode and noncontact-mode have been reviewed elsewhere.20,21 In this study, we have chosen AFM in tapping mode22 to characterize HMSAG from pectin because these gels are very soft. Advantages of this method are that low forces and minimal damage enable it to image soft samples in air. Furthermore, lateral forces are practically eliminated so that deformation of the sample is minimal and lateral resolution is on the order of 1-5 nm. Materials and Methods Materials. Fresh albedo was obtained from Florida early Valencia oranges23 and Florida tropical seedless limes. Immediately upon arrival in the laboratory, the flavedo was stripped manually from the skin with a potato peeler, followed by removal of the albedo with a paring knife. After cutting the albedo into small pieces (ca. 1 mm2) it was stored at -20 °C in sealed polyethylene bags until extraction. Commercial citrus pectin (MexPec 1400, degree of methylesterification, 71%) was supplied by the Grinsted Division of Danisco-Cultor (Kansas City, KS) in dry powder form and used as received. Extraction. The method of extracting pectin from albedo has been described in detail previously.23 Briefly, microwave heating was performed in a model MDS-2000 microwave sample preparation system (CEM Corp., Matthews, NC). Samples were irradiated with 630 W of microwave power at a frequency of 2450 MHz. For each experiment, six equally spaced cells were placed in the sample holder, a rotating carousel. One vessel was equipped with temperature and pressure sensing devices which measured and controlled the temperature and pressure within the cell. Time of irradiation was 3 min followed by rapid cooling in a cold water bath to room temperature. The maximum allowed pressure level within the cell was set at 52 ( 2 psi, and the maximum temperature within the cell was set at 195 °C. Cells were loaded with 1 g of albedo dispersed in 25 mL of HCl, pH 2. Solubilized pectin was precipitated with 70% isopropyl alcohol (IPA) and washed once with 70% IPA and once with 100% IPA. Finally, the sample was vacuum-dried at room temperature and stored in screw capped vials until further use.

336

Biomacromolecules, Vol. 5, No. 2, 2004

Fishman et al.

Preparation of Gels and Determination of Gel Strength. Gels were prepared by dissolving 0.7 g of the pectin to be tested in 102 g of a 0.01 M citrate buffer. The pH of the buffer was controlled by the molar ratio of citric acid and sodium citrate. The molar citric acid:sodium citrate ratio was 88:12 for pH 3.00, 83:17 for pH 3.23, and 77:23 for pH 3.53. These were the pHs of maximum gel strength at break for commercial citrus pectin (CCP), orange albedo pectin (OAP), and lime albedo pectin (LAP), respectively. The beaker containing the pectin solution was placed in a 9598 °C water bath, and 189.0 g of sucrose was added with stirring. The mixture was heated for 10 min and then poured into three 45 mL weighing bottles (40 mm o.d. × 50 mm h.) to which Scotch tape had been placed around the top to allow the gelling solution to be poured to a level above the rim of the bottle. The solutions were allowed to cool at ambient room temperature for 20 h. The tape was removed, and the part of the gel above the rim of the bottle also was removed using a cheese cutter. The final composition (w/ w) of the gels was 0.25% pectin, 65% sucrose and 35% of buffer and their pH ranged from 3.00 to 3.53 depending on the source of pectin. Gel strengths were determined with a Stable Micro Systems TA-XT2 Texture Analyzer. Probe speed was 1.0 mm/s, and the 0.5 in. diameter Delrin probe was allowed to penetrate to a depth of 15 mm. The maximum in the forcepenetration curve was taken as the gel strength at break. Three replicates were run for each sample. Atomic Force Microscopy. A thin (1 mm) slice of transparent gel was cut manually from a dollop of gel with a stainless steel razor blade. A freshly cleaved 10 mm diameter disk of mica was applied to the cut surface of the gel. After 5-10 min, the disk was peeled off the gel surface and mounted in a Multimode Scanning Probe microscope with a Nanoscope IIIa controller, operated as an atomic force microscope in the tapping mode (Veeco Instruments, Santa Barbara, CA). The thin layer of pectin adhering to the mica surface was scanned with the AFM operating in the intermittent contact mode using tapping mode etched silicon probes (TESP). The spring constants for these probes were 20100 N/m and the nominal tip radius of curvature was 5-10 nm. The cantilever controls, namely drive frequency, amplitude, gains, and amplitude set point ratio (rsp) were adjusted to give a phase image with the clearest image details. Values of rsp used in this study were about 0.95. The definition of rsp is given by eq 1 rsp ) Asp/A0

(1)

Asp is the oscillating amplitude in contact with the sample whereas A0 is the freely oscillating amplitude (out of contact amplitude). Set point amplitudes approaching 1 correspond to light normal forces, i.e., soft tapping.24 Images were analyzed by software version 5.12 rev. B which is described in the Command Reference Manual supplied by the manufacturer. Length, widths and areas of strands and pores were determined by particle analysis. To image objects of interest (OOI), in this case strands and pores, the OOI must be separated from other objects by a process called threshold segmentation. To accomplish this, we chose

a threshold value for pixel intensity which removed or masked the intensity of background pixels and allowed the pixel brightness of the OOIs to remain undiminished. Prior to particle analysis, lowpass filtering was applied to reduce background noise and high pass filtering was applied to highlight the OOI which are delineated from the background as areas of rapidly changing height or phase. Results and Discussion For purposes of comparison, gels were made at the pH value which gave optimum gel strength at break. We found that lime albedo pectin (LAP) gave an optimum gel strength of 9.1 KPa (117 g) at pH 3.53; orange albedo pectin (OAP) gave a value of 8.3 KPa (107 g) at pH 3.23; and commercial citrus pectin (CCP) gave a value of 6.7 KPa (87 g) at pH 3.00.26 The weight average molar masses, Mw, for LAP, OAP, and CCP were 311 000, 373 000, and 254 000, respectively; the degrees of methyl esterification were 88%, 78%, and 77%, respectively; and % anhydrogalacturonic acid contents were 76%, 76%, and 81%, respectively. The general features found for all three types of gel are illustrated in Figure 1. One micrometer-square areas contained a complex network of curvilinear strands and junctions. Dimensions of strands and pores are given in Tables 1-3. In Figure 2, parts A and B, are height (A) and phase-shift (B) images of a native HMSAG made from CCP. These images were obtained simultaneously within minutes of pulling a thin layer of pectin gel off a slab of pectin with a freshly cleaved mica surface. When measuring gels in the soft tapping mode, the amplitude of the oscillating probe tip may decrease or increase by approaching the surface of the sample. This change in amplification measures changes in sample height at the surface of the sample and produces an image of the sample surface topography. At the same time, interaction between the sample and the probe tip may increase the phase-shift of the probe motion relative to an oscillator which drives the probe near its mechanical resonance frequency.22 Phase imaging is sensitive to elastic and/ or adhesive forces interacting with the probe tip.27 In that the gels are soft and sticky, phase shift changes are expected to measure changes in adhesion as the probe tip traverses the sample surface and also produces an image of the surface topography. Amplification (height) changes and phase shifts (adhesion changes) are registered as bright and dark regions in AFM images. It is of interest to note that strands in both images in Figure 2 are essentially identical in appearance except that in the height image strands are registered as dark or low areas whereas strands in the phase-shift images are registered as bright or adhesive areas. This phenomenon will occur when traversing an area of the sample which causes the probe tip amplitude and phase-shift to simultaneously increase or decrease. For example, in Figure 2 when the probe tip traverses a strand, its amplitude and phase-shift increase. This indicates that initially strands lie in crevices because they are below the surface of the interstitial material or pores (Figure 2A). Furthermore, the strands are more adhesive than the interstitial material (Figure 2B). If the area containing strands was above the surface of the pore material and more adhesive than the pore material, the probe tip height



Figure 1. Phase-shift image of an acid gel made from 0.25% orange albedo pectin (OAP), 65% sucrose, and 35% 10 mM sodium citrate buffer solution. Areas of curvilinear strands held together by junction zones and branch points shifted the phase of the cantilever thereby defining a primary network surrounding irregular pore areas with little or no phase-shift. Also pores were often divided by a network of fine or secondary strands. Table 1. Dimension of 100 Largest Gel Segmentsa strands

a

pores

sample

area (nm2)

length (nm)

width (nm)

area (nm2)

length (nm)

width (nm)

CCP OAP LAP

1598(126) 1734(84) 1107(208)

118(3) 119(4) 100(6)

27.6(2.5) 28.6(2.7) 19.8(1.2)

703(98) 629(37) 514(127)

70.1(4.9) 64.7(3.8) 56.8(6.0)

14.2(2.8) 12.8(0.6) 11.6(2.5)

Numbers in parentheses are standard deviations of triplicate analysis.

Table 2. Range of Lengths between Cross-Links (nm) for 100 Largest Gel Segmentsa sample

average

one sigma minimum

CCP OAP LAP

118(3) 119(4) 100(6)

58(5) 52(5) 58(5)

a

one sigma maximum 178(5) 186(5) 142(5)


would decrease and the phase-shift would increase when the tip traversed strands of the gel. In that case, strands would be bright in amplitude and in phase-shift images. One set of images was obtained (not shown) in which areas containing

Table 3. Dimension of Entire Field of Gel Segmentsa strands sample CCP OAP LAP a

area (nm2)

length (nm)

pores width (nm)

area (nm2)

length (nm)

width (nm)

260(9) 28.0(0.7) 7.88(0.22) 207(29) 22.0(6.0) 7.80(0.76) 254(40) 27.8(1.7) 7.76(0.61) 190(20) 25.1(1.4) 7.49(0.45) 183(32) 25.1(0.8) 6.57(0.44) 152(33) 21.7(2.5) 6.91(0.73)


strands in both images are bright and therefore the strands at the surface of the gel are above the remainder of the gel. In cases where height images were obtained initially (e.g., Figure 2A), contrast gradually diminished with time which indicated that the surface was becoming smoother.

338


Fishman et al.

Figure 2. Matching images of (A) height and (B) phase-shift of an acid gel made from commercial citrus pectin (CCP). The organization of thin clefts or depressions in the surface of the gel in (A) corresponds to the arrangement of phase-shifted or adhesive areas in (B).

In many cases, height images were almost flat initially and barely visible whereas phase-shift images showed sharp delineation between strands and pore areas (images not shown). In these instances the strands are brighter than the background in the phase-shifted images. The combination of a phase-shift image with good contrast and a height image with poor contrast indicated that the surface was almost flat with the strands slightly above the surface of the interstitial material. Overall, there were many more images from all three sources of pectin in which the amplitude image was almost flat with little contrast and yet the phase image showed well-defined stranded and pore areas than there were images in which both phase and height images showed welldefined stranded and pore areas. Possibly, this behavior depended on the relative strength of interactions between strands and constituents in the pores. Pectin is a heterogeneous polysaccharide in which some monosaccharide regions may be more hydrophilic than others and some may be more branched than others. Another possibility is that the behavior was related to differences between the viscoelasticity of pectin and interstitial fluid. Figure 3 contains HMSAG phase-shift images of CCP, OAP, and LAP. Table 1 contains average dimensions of 100 of the largest gel segments and 100 of the largest pores obtained from phase images. Each dimension is the average obtained from 3 images. For strands, there is no significant difference in average area, length, and width between CCP and OAP, whereas all 3 dimension are smaller for LAP. For pores, there is no significant difference among all 3 dimensions for the 3 sources of gel. Comparable dimensions of strands are larger than those of pores. Height measurements on tomato pectin by AFM suggest that individual pectin molecules have diameters in the range of 0.5-0.8 nm.15 In this study, width measurements on large strands gave average values of about 28 nm for CCP and OAP gels and about 20 nm for LAP gels. If we assume that the width of a sugar molecule is about 0.5 nm, as many as 55 sugar molecules could be immobilized

laterally to a single large pectin molecule. Even if as indicated by studies in dilute solution13,28 that as many as 4 pectin molecules are aggregated in a strand, the total number of laterally immobilized sugars would change only slightly. The number average length between cross-links for 100 of the largest strands was about 118 nm for CCP and OAP and about 100 nm for LAP (Table 1). Our earlier transmission electron microscope study of rotary-shadowed peach pectin14 revealed that the number average contour length of pectin cross-links in circular microgel networks which were dissolved in water was about 70 nm. Dissociating these networks in 50% aqueous glycerol gave individual components with a number average length of about 71 nm. These averages were close to the value of 74 nm found by HPSEC for peach pectin dissolved in 0.05 M NaCl. The values found for peach pectin are about 60% as large as average pectin cross-links in CCP or OAP gels or about 70% as large as pectin cross-links in LAP gels. Differences in length between peach and citrus pectin could arise from differences due to the type of fruit, variety, method of extraction, or instrumental sensitivity. Table 2 contains the upper and lower length values for segments which are within ( one standard deviation of the average, i.e., about 68% of the distribution. Outside these limits, the smallest length measured was about 22 nm and the longest value was about 640 nm. As mentioned previously, in the case of peaches, the highest value measured was about 330 nm and the lowest about 10 nm. Table 3 contains average dimensions of the strands and pores for the entire field. These numbers were obtained from 3 phase images for each kind of gel. In the case of strands, fields were averaged from 3004, 2848, and 4222 objects for CCP, OAP, and LAP, respectively. For pores, fields were averaged from 4551, 4947, and 5852 objects for CCP, OAP, and LAP, respectively. Comparison of the data in Tables 1 and 3 reveals that there are many more barely visible small segments in the sample than there are large segments. The ratio of strand to interstitial


Figure 3. Comparison of phase-shift images of acid gels from three sources. (A) commercial citrus pectin, (B) orange albedo pectin, (C) lime albedo pectin.

area for CCP, OAP, and LAP is 0.83, 0.77, and 0.87, respectively, for the entire field. The percentage of the whole field area which is polymer is 45%, 43%, and 46% for CCP, OAP, and LAP, respectively. This last finding is rather remarkable if one considers that the ratio of sucrose to pectin


(w/w) is 260:1. If one assumes that the strands are comprised of a pectin framework holding bound or immobilized sucrose and the bound sucrose is proportional to the strand area, then the ratio of bound sucrose to pectin is in the range of 112:1 to 120:1. In Figure 4A-C are images which were obtained by electronically thinning strands of phase images such as those shown in Figures 1, 2B, and 3. The strands in these images were thinned to a width of 1 pixel. Parts A-C of Figure 4 are images of CCP, OAP, and LAP, respectively. These images reveal that the presumed pectin framework in the absence of sucrose is a partially cross-linked network (PCN) in that many of the potential cross-linking moieties have linked only at one end. Visual examination reveals that the density of molecules in the (PCN) is somewhat greater for LAP gels than for OAP or CCP gels. The greater number of objects in LAP gels over CCP or OAP gels is consistent with a higher pectin density for LAP as compared with either CCP or OAP. Interestingly there are several similarities between the images in Figure 4 and previously rotaryshadowed images of peach pectin13,14 from dilute aqueous solutions of 0.005 M NaCl. In the cases of both the peach and the citrus, the networks are a collection of mostly open branched structures. All but a few of the branches are trifunctional, and all structures are comprised of rods, segmented rods, kinked rods, or combinations of these linked end to end. Although some of the branched points could be construed as arising from junction zones, a significant number of branch points occur at right angles. These could arise from the joining of individual functional groups such as hydrogen bond formation between oxygen from carboxyl groups, OH groups, or glycosidic bonds and hydrogens from carboxyls and OH groups. This picture is in sharp contrast to the appearance of the gels when the adsorbed sugars are visualized, e.g., Figures 1-3. In this case, most branched points appear to be formed from “cooperative junctions whose stability depends on the specific energy of the linkages and the number of units cooperatively bound.” 4 One might infer from these images that the adsorbed sugar molecules function as a coating which aids in strengthening branch points, possibly through cooperative interactions. Recently, Lofgren et al.28 imaged by transmission electron microscopy thin sections of HMSAG fixed with glutaraldehyde, stained with ruthenium red, and dehydrated. They imaged areas which were in the range of 0.5-4 µm2, whereas our fields were 1 µm2. They described the image to contain “open network structures with many pores above 0.5 µm”. They also describe that “the strands were aggregated in bundles or loose aggregates and branched in an irregular manner”. Furthermore, they suggested that there was a “tendency toward parallel alignment” of strands and that strands appeared stiff and straight. Our direct images of native gels (Figure 4A-C) revealed a uniformly distributed PCN of strands in which strands appear branched and stiff. As mentioned previously, in addition to straight rodlike strands, our images reveal curved and kinked strands (segmented rods and kinked rods). Rees and Wight predicted that pectin molecules would kink at points where rhamnose was located in the pectin backbone

340


Figure 4. Electronically thinned strands from phase images of gels. (A) commercial citrus pectin, CCP; (B) orange albedo pectin, OAP; (C) lime albedo pectin, LAP. Images appear to reveal pectin framework in the absence of sucrose. Joined structures within the framework are combinations of rods (R), segmented rods (SR), and kinked rods (KR) (Figure 4A).

based on NMR measurements in combination with modeling studies.29 As indicated by differences in height images, i.e., strands that are above, below, and flush with the surface of

Fishman et al.

the interstitial fluid, the gel appears quite fluid. In contrast, the components of the PCN’s appear quite stiff. In fact, the networks may be quite deformable in that they are partially cross-linked and held together by secondary forces, e.g., hydrogen bonds and van der Waals forces, which may break when subjected to pressure and reform when the pressure is released. This could be the basis of pectin’s unique ability to form spreadable gels. In Table 1, the average pore length for 100 of the largest pores from the three kinds of gels ranged from about 57 to 70 nm, the average pore width from about 12 to 14 nm, and the average pore area from about 514 to 703 nm2. Table 3 contains averages of all pores in a 1 µm2 area. In this case, average lengths are about 23 nm, widths about 7.5 nm, and areas about 183 nm2. Also worthy of note is that comparison of images of structures in a 1 µm2 area with those in a 25 µm2 area reveals that in both cases structures are uniformly distributed. In any study involving imaging by AFM, one must consider the effect of probe sample interactions on the molecular dimensions which are obtained. An important artifact which requires consideration is over- or underestimating molecular dimensions of the molecules under examination. The tapping mode probes used in this study are pyramidal in shape. If the molecule which is being measured sits on a surface, the side of the probe will cause the tip to rise before it reaches the leading edge of the molecule and to fall after it passes over the trailing edge of the molecule. In that case the measured dimension will be larger than the actual dimension. This effect has been termed “probe broadening”.20 If the molecule sits in a cleft or a well below the surface, the tip will not reach the bottom of the cleft until after it has passed the leading edge of the molecule. Furthermore, the tip will rise prior to reaching the trailing edge of the molecule. In that case the measured dimension will be smaller than the actual dimension. This effect could be termed “probe narrowing”. In Figure 2, we clearly demonstrate that the strands are in a cleft below the surface of the interstitial fluid (pores). Thus, it is possible that we have underestimated the area and width of the strands and overestimated the area and width of the pores. The flatness of the sample and the high set point ratio (about 0.95) employed in this study should work to mitigate the effects of these artifacts. Another possible source of artifacts here is sample dehydration during imaging. Because of the high concentration of solids (sucrose) in the gel, we expect dehydration to be minimal due to the low vapor pressure of water under these conditions. Also the entirety of the phase shift images and the strands in the height images remain relatively unchanged with time and sample replication. A third source of artifacts, namely, probe tip contamination, seems to have been avoided by timely changing of tips. Conclusion AFM images indicated relatively small differences in the nanostructure of optimal pH, native CCP, OAP, and LAP gels. These data are consistent with earlier gel strength, molar


mass, and viscosity data which also showed relatively small differences in the physical properties of these same three kinds of gels. Furthermore, these images appear to have revealed for the first time the structure of the adsorbed sucrose on the pectin in the gels and the pectin framework after electronic stripping of pixels related to sucrose. The skeletonized pectin framework appears similar to partially cross-linked networks imaged from solution at low ionic strength by electron microscopy. Nevertheless, the adsorption of sucrose and its extent of adsorption require confirmation by independent methods. Finally, the value of simultaneous height and phase shift imaging was demonstrated for visualizing and quantitating the nanostructure of soft gels in which the surface of pores is flush with the surface of strands. References and Notes (1) Carpita, N.; McCann, M. C. In Biochemistry and Molecular Biology of Plants; Buchanan, B., Ed.; American Society of Plant Physiologists: Rockville, MD, 2000; pp 52-108. (2) Yamada, H. In Pectins and Pectinases; Voragen, A. G. J., Visser, J., Eds.; Elsevier: Amsterdam, 1996; pp 173-190. (3) Rolin, C. In Industrial Gums; Whistler, R. L., BeMiller, J. N., Eds.; Academic Press: San Diego, CA, 1993; pp 257-293. (4) Rinaudo, M. In Pectin and Pectinases; Voragen, A. G. J., Visser, J., Eds.; Elsevier: Amsterdam, 1996; pp 21-33. (5) Flory, P. J. In Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953. (6) Dea, I. C. M. In Industrial Gums; Whistler, R. L., BeMiller, J. N., Eds.; Academic Press: San Diego, CA, 1993; pp 21-52. (7) Oakenfull, D. G. In The Chemistry and Technology of Pectin; Walter, R. H., Ed.; Academic Press: San Diego, CA, 1991; pp 87-108. (8) Morris, E. R.; Gidley, M. J.; Murray, E. J.; Powell, D. A.; Rees, D. A. Int. J. Biol. Macromol. 1980, 2, 327-330. (9) Walkinshaw, M. D.; Arnott, S. J. Mol. Biol. 1981, 153, 1075-1085. (10) Leeper, G. F. J. Texture Studies 1973, 4, 248-253. (11) Hanke, D. E.; Nortcote, D. H. Biopolymers 1975, 14, 1-17.

Biomacromolecules, Vol. 5, No. 2, 2004 341 (12) McCann, M. C.; Wells, B.; Roberts, K. J. Cell Sci. 1990, 96, 323334. (13) Fishman, M. L.; Cooke, P.; Levaj, B.; Gillespie, D. T. Arch. Biochem. Biophys. 1992, 294, 253-260. (14) Fishman, M. L.; Cooke, P.; Hotchkiss, A.; Damert, W. Carbohydr. Res. 1993, 248, 303-316. (15) Round, A. N.; MacDougall, A. J.; Ring, S. G., Morris, V. J. Carbohydr. Res. 1997, 303, 251-253. (16) Morris, V. J.; Gunning, A. P.; Kirby, A. R.; Round, A., Waldron, K.; Ng, A. Int. J. Biol. Macromol. 1997, 21, 61-66. (17) Round, A. N.; Rigby, N. M.; MacDougall, A. J.; Ring, S. G.; Morris, V. J. Carbohydr. Res. 2001, 331, 337-342. (18) de Gennes, P. G. Scaling Concepts in Polymer Physics; Cornell University Press: Ithaca, NY, 1979. (19) Anon. Scanning Probe Microscopy Training Notebook; Digital Instruments/Veeco Metrology Group, Inc.: Santa Barbara, CA. (20) Morris, V. J.; Kirby, A. R.; Gunning, A. P. Atomic Force Microscopy for Biologists; Imperial College Press: London, 1999. (21) McIntire, T. M.; Brant. D. M. In Techniques in Glycobiology; Townsend, R. R., Hotchkiss, A. T.; Eds.; Marcel Dekker: New York, 1997; pp 187-288. (22) Magonov, S. N. E. V.; Whangbo, M. H. Surface Sci. Lett. 1997, 375, L385-L391. (23) Fishman, M. L.; Chau, H. K.; Hoagland, P.; Ayyad, K. Carbohydr. Res. 2000, 323, 126-138. (24) Kanchanasopa, M.; Manias, E.; Runt, J. Biomacromolecules 2003, 4, 1203-1213. (25) Fishman, M. L.; Chau, H. K.; Coffin, D. R.; Hotchkiss, A. T. In AdVances in Pectin and Pectinase Research; Voragen, F., Schols, H., Visser, R., Eds.; Kluwer Accademic Publishers: Dordrecht, The Netherlands, 2003; pp 107-122. (26) Wang, L. Surf. Sci. 1999, 429, 178-185. (27) Fishman, M. L.; Pepper, L.; Pfeffer, P. E.; In Water Soluble Polymers: Beauty with Performance; Glass, J. E., Ed.; American Chemical Society: Washington, DC, 1986; 213, pp 57-70. (28) Lofgren, C.; Walkenstrom, P.; Hermansson, A. M. Biomacromolecules 2002, 3, 1144-1153. (29) Rees, D. A.; Wight, A. W. J. Chem. Soc. (B) 1971, 1366-1372.

BM0300655

Nanostructure of Native Pectin Sugar Acid Gels Visualized by Atomic

Recommend Documents