| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Institute of Food Research, Norwich Research Park, Colney, Norwich, UK
1Correspondence: Institute of Food Research, Norwich Research Park, Colney, Norwich NR4 7UA, UK. E-mail: patrick.gunning{at}bbsrc.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Key Words: flow cytometry • AFM • antimetastasis activity • polygalacturonic acid • RGI
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
2
3)
and in immunomodulation (4)
. The suggested hypothesis (1)
for the role of modified pectin is that the product contains structural elements that can bind to, and inhibit the function of, the protein galectin-3 (Gal3), which is considered as a diagnostic marker and a target protein for cancer treatment (5)
.
Gal3 is a member of a family of evolutionary-conserved proteins widely found in a range of species from lower invertebrates to mammals. The mammalian lectin Gal3 contains a 
14 kDa carbohydrate recognition domain (CRD) with an affinity for β-galactoside residues (6
7)
. The globular CRD is attached to a long, thin "collagen-like" tail, resulting in a "tadpole-like" structure revealed by electron microscopy (8)
. Solid-phase binding assays and inhibition assays have identified a number of natural ligands that will bind to galectins (9)
. These molecules include galactose, lactose, polylactosamine, and N-acetyllactosamine (LacNAc). Published (10
11
12)
X-ray data on complexes of Gal3-lactose, Gal3-LacNAc, and Gal3-LacNAc derivatives reveal the binding of the terminal galactose residue within the primary binding site. Hydrogen bonding to O4 and O6 of the galactose and O3 of the glucose has been shown to be important in stabilizing the ligand lactose within the CRD. The carbohydrates galactose, lactose, polylactosamine, and LacNAc are considered as natural inhibitors, and high-affinity carbohydrate-based inhibitors have been developed through the production of LacNAc (12)
and galactose (13)
derivatives. A degree of flexibility in the substitution at C3 of the galactose has allowed the development of higher-affinity ligands (13)
.
The various functions that have been attributed to Gal3 suggest that it plays important roles in several stages of cancer progression and metastasis. For cancer cells to detach from primary tumors, they have to attach to the extracellular matrix and then invade the blood vessels. Basement membrane in which epithelial cells are embedded is rich in such glycoconjugates as laminin and fibronectin. These proteins, through binding to Gal3, allow Gal3 to promote adhesion of the cells. Breast cancer cells expressing Gal3 have been shown to adhere to laminin (14)
, and transfection of Gal3 into human breast cancer cells was found to enhance in vitro invasion of Matrigel, a matrix derived from solubilized basement membrane (15)
. Thus, Gal3 is able to play important roles in detachment of cells from primary tumors and attachment of cells in the development of secondary tumors. Furthermore, Gal3 induces both in vitro and in vivo tumor-related angiogenesis (16)
, a mechanism providing a pathway for tumor cells to escape from the primary tumor and enter the bloodstream. Within blood vessels, tumor cells can enhance their survival through forming emboli (cell aggregates) that are protective within the hostile host environment. Gal3 expressed at cell surfaces can mediate self-aggregation of cells when induced by asialofectuin (17)
. Apoptosis is a natural mechanism occurring during cell differentiation and in the treatment of tumors by chemotherapy or radiotherapy. Resistance to apoptosis allows tumor cells to avoid programmed cell death induced by detachment from the extracellular matrix and to survive defense mechanisms triggered by the host immune system as the tumor cells circulate through the bloodstream. Gal3 plays several roles in cancer apoptosis, showing pro- or antiapoptopic effects dependent on the nature of the stimulus (18
19
20)
.
Pectins are important structural components of the cell walls of land plants. Of all the cell wall polysaccharides, pectin has the most complex structure. Essentially, pectin contains a "smooth" backbone of homogalacturonan, which contains highly substituted rhamnogalacturonan 1 (RGI) units (termed "hairy regions") (21)
. A schematic picture of pectin structure is shown in Fig. 1
. This is a composite structure based on the average chemical composition of the extracted pectin. Homogalacturonan backbones are composed of (1
4)-linked 
-D-galacturonic acid residues (GalpA) in which some of the carboxyl groups are methyl-esterified. The more complex RGI units consist of a repeating disaccharide [(14)--D-GalpA-(12)--L-Rhap-(1-)] of galacturonic acid and rhamnose (Rhap) residues. A proportion of the Rhap residues contain neutral sugar side chains linked at C4, consisting mainly of -L-arabinose (Araf) and/or β-D-galactose (Galp) residues. No RGI regions have been fully characterized, and the structure and composition of the RGI regions will vary for different plant species. The main types of neutral sugar side chains that are found in RGI are shown in Fig. 1
. Pertinent features of the RGI regions relevant to this study are the arabinogalactan side chains. There are 2 basic types of arabinogalactans linked to C4 of the backbone rhamnose residues: essentially based on linear (14)-β-D-galactans and branched (13,6)-β-D-galactans. The (14)-β-D-galactans may be linear unsubstituted chains or may contain predominately (13)--L-Araf side chains. All the arabinogalactans contain terminal galactose residues, but the most likely candidates for binding to Gal3 are linear (14)-β-D-galactans; the simple linear structures might be expected to better facilitate binding within the Gal3 CRD.
|
The modification of pectin to produce material used in published clinical studies (22
23)
involves an alkaline treatment followed by an acid treatment. Alkali treatment of pectin causes β-elimination reactions, which will result in a depolymerization of the polysaccharide backbone (24)
. Alkali treatment will de-esterify the homogalacturonan regions and generate lower molecular weight oligomers of polygalacturonic acid (PGA), plus RGI regions attached to polygalacturonic acid oligomers (Fig. 2a
). Acid treatment preferentially cleaves neutral sugars and is expected to lead to the release of modified hairy regions of the pectin chain (25)
and preferential removal of arabinose residues. Thus, acid treatment would be expected to generate RGI regions lacking arabinan side chains and with reduced arabinose substitution of the arabinogalactan side chains, plus free arbinogalactan and galactan chains (Fig. 2b
). The potential importance of arabinogalactans or galactans is supported by published evidence (26
27)
for the anticancer activity of arabinogalactans. Recent comparative studies (28)
of pectins from various dietary sources have shown that galectin inhibitory behavior is enhanced in pectic polysaccharides with higher arabinose and galactose contents, indicating the importance of arabinogalactans in the inhibition of Gal3-mediated cellular interactions.
|
Atomic force microscopy (AFM) has emerged (29
30)
as a powerful tool for measuring intramolecular and intermolecular forces associated with biological systems. In the present study, the intention is purely to establish whether pectin fragments can bind specifically to Gal3. Thus, AFM has been used to complement flow cytometry and microscopy studies to directly probe for specific binding of pectin fragments to Gal3. The simplest molecular fragment derivable from pectin that might be expected to bind to Gal3 is a galactan that can be obtained by -L-arabinofuranosidase treatment of isolated RGI regions containing arabinogalactan side chains.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
-L-arabinofuranosidase to remove the bulk of the arabinose residues. The quoted sugar composition of the galactan is 88% galactose, 6% galacturonic acid, 3% arabinose, and 3% rhamnose. Quoted sugar composition for the RGI extract is 62% galacturonic acid, 20% rhamnose, 12% galactose, 3.3% arabinose, and 1% xylose. PGA samples were prepared by a partial depolymerization of citrus pectin with polygalacturonase and were also purchased from Megazyme. The quoted sugar composition is 96% galacturonic acid with trace amounts of galactose (1%), arabinose (0.2%), rhamnose (1.2%), xylose (1.0%), and glucose (0.3%). The samples were dialysed before coupling them to substrates. All other materials were purchased from Sigma Chemicals Ltd. (Poole, UK) unless otherwise stated. Recombinant human Gal3 was purchased as a Gal3-lactose complex. The BSA used was fraction V.
Functionalization of silica particles and glass surfaces
Silica beads with a mean diameter of 6.8 µm were purchased from Bangs Laboratories (Fishers, IN, USA). The beads were acid washed to regenerate surface hydroxyl groups (30 min in 1:1 HCl:MeOH) and cleaned before functionalization by washing and then spinning down in a microcentrifuge through the following series: ultrapure water (3x1 ml, 18.2 M
(Elga Ltd, Marlowe, UK), very low-residue ethanol (3x1 ml (Fluka Chemicals, Poole, UK), and toluene (3x1 ml, dried over molecular sieve 4 Å). After each centrifugation cycle, the supernatant was discarded. Functionalization was carried out following the protocol described by Bhatia et al. (31)
. All of the silica beads were functionalized with an initial layer of 3-mercaptopropyltrimethoxy silane (MTS). For the AFM studies, the silanized beads were then functionalized with 4-(4-N-maleimidophenyl) butyric acid hydrazide hydrochloride (MPBH, Pierce, Rockford, IL, USA), which is a carbohydrate coupling agent. For the flow cytometry studies, the silanized beads were functionalized with N-
-maleimidobutyryloxy succinimide ester (GMBS, Pierce, Rockford, IL, USA), a protein coupling agent that allowed attachment of Gal3 or BSA to the silica beads. It is possible to couple either Gal3 or the Gal3-lactose complex to the glass beads. Dialysis or washing can be used to break the complex either before coupling to or after coupling to the substrate. Use of the complex is preferable because it reduces the risk of the coupling reaction blocking the binding site on the protein. Covalent attachment of Gal3 onto GMBS-derivatized beads was carried out by incubation of the beads in 0.1 mg ml–1 solutions of the lectin in PBS for 1 h at room temperature, followed by rinsing (3x1 ml PBS). Before the binding experiments, Gal3-functionalized beads were incubated in a 10 mg ml–1 solution of glycine to amine-cap any unreacted succinimide groups, and the derivatized beads were then rinsed in PBS.
For the AFM studies, MPBH-functionalized beads were glued to the end of silicon nitride AFM cantilevers (long thin NP; Veeco Inc., Santa Barbara, CA, USA) by using a tiny amount of 2-part epoxy (40 min cure time; Permatex Inc., Hartford, CT, USA). Once the epoxy resin had cured, the cantilever-bead assemblies were incubated with 0.1 mg ml–1 galactan solutions in PBS for 1 h at room temperature. The assemblies were rinsed in PBS and then incubated for 30 min in a 10 mg ml–1 glucose solution in PBS to sugar-cap any unreacted hydrazide sites on the bead, followed by rinsing in PBS. The fully functionalized tips were inserted into the liquid cell of the AFM and maintained under PBS solution. When the binding experiments were complete, each functionalized tip was examined by scanning electron microscopy (SEM) to assess the quality of the attachment of the bead to the cantilever. The procedure for the covalent coupling of galactans to the silica beads was verified by visualizing the attachment of fluorescent-labeled galactans.
The functionalization of glass slides (optical microscopy and AFM studies) with Gal3, Gal3-lactose complexes, or mixtures thereof with BSA followed the same procedure as for the silica beads, with GMBS used as the heterobifunctional crosslinker. Selected regions of the functionalized slides were incubated with 0.1 mg ml–1 solutions of the relevant proteins in PBS for 1 h at room temperature, followed by washing in PBS. The surfaces were then incubated in a 10 mg ml–1 solution of glycine to amine-cap any unreacted succinimide groups on the glass and then rinsed in PBS before being inserted into the liquid cell of the AFM.
Fluorescent labeling
Fluorescent labeling of the pectin fragments (galactan, RGI, and PGA) was achieved by using 5-(4,6-dichlorotriazinyl) aminofluorescein (5-DTAF, Invitrogen, Paisley, UK): 5-DTAF was added to the polysaccharide solution (20 mg ml–1) in borate buffer (pH 10.8) to a concentration of 2 mg ml–1 and incubated at room temperature overnight. Unbound DTAF was removed by extensive dialysis followed by separation on Sephadex G-50 gel filtration columns, eluted with PBS. The binding of DTAF-labeled galactan, PGA, or RGI to Gal3-derivatized beads were imaged in both bright-field and fluorescence modes on an inverted microscope fitted with a UPlanAPO x40, 0.85 NA objective lens (IX-70; Olympus, Tokyo, Japan). Epifluorescence micrographs were obtained after placing a dichroic filter block in the optical path (UM41001 HQ: FITC; Chroma Technology Corp., Rockingham, VT, USA).
SEM
Silica bead cantilever assemblies were coated with a 5 nm layer of gold by using an Emitech 550 Plasma coater (Quorum Technologies, Newhaven, UK). SEM images were obtained on a Stereoscan 360 SEM (Leica Microsystems, Milton Keynes, UK), operating at 10 kV and a working distance of 12 mm.
Flow cytometry
Samples for flow cytometry were prepared by incubating the fluorescently labeled carbohydrate with Gal3-derivatized, BSA-derivatized, or glycine-capped control beads. Incubation was carried out for 1 h at room temperature (22°C) at pH 7.2 (PBS buffer) or pH 12.0 (PBS+NaOH) as required. Following incubation, the samples were washed with the relevant buffer (3x1 ml) before resuspending in PBS at pH 7.2 for flow cytometry. Bead samples were analyzed with an inFlux flow cytometer (Cytopeia, Seattle, WA, USA). In all cases, a minimum of 5000 events were acquired. The median fluorescence intensity of beads, selected on their forward and side-scatter characteristics, was measured with a photomultiplier tube by using a 528/38 nm bandpass filter. Binding of fluorescent DTAF-labeled galactan, RGI, or PGA was measured by quantifying the fluorescence signal of individual beads. For the RGI samples, any aggregation of the beads on binding of the fluorescently labeled carbohydrate was monitored by measuring the levels of monomers, dimers, and trimers as well as their associated fluorescence. Data analysis was performed with FlowJo software (Tree Star, Ashland, OH, USA).
Dose-response studies
DTAF-galactan
Gal3-derivatized beads were incubated with the relevant concentration of DTAF-galactan in PBS for 1 h at room temperature (22°C), followed by washing in PBS (3x1 ml).
Lactose inhibition
Gal3-derivatized beads were added to lactose solutions of varying concentration containing a fixed concentration of DTAF-galactan of 0.8 mg ml–1 in PBS. The samples were incubated for 1 h at room temperature (22°C) and then washed in PBS (3x1 ml).
Molecular weight determination
An estimate of galactan molecular weight was obtained by gel permeation chromatography by using a series of TSK-Gel columns (G3000PW, G4000PW, G6000PW; Anachem Ltd., Luton, UK). The sample was dissolved at 2 mg ml–1 in 0.2 M sodium phosphate buffer at pH 6.8 (including 0.04% sodium azide as a preservative). Then, 100 µl of sample was injected onto the columns at a flow rate of 0.5 ml min–1 (eluted with the same buffer), and sample peaks were monitored with a refractive index detector. Molecular weight was calculated based on a comparison of elution times with those obtained for a series of pullulan standards (molecular weight range 180–853,000, part no. 2090-0100; Polymer Laboratories Ltd., Church Stretton, UK).
NMR
The 13C NMR spectrum of the galactan sample (40 mg ml–1 in D2O) was measured at 300 K on a Bruker Avance 600 spectrometer equipped with a TCI cryoprobe (Bruker, Ettlingen, Germany), operating at 150.9 MHz for 13C. The spectrum (24,000 scans) was acquired by using a 30° pulse angle, 0.9 s acquisition time, and 2 s relaxation delay, and 3 Hz line broadening was applied before Fourier transformation. The mean length of the galactan chains was calculated from a ratio of peak heights by taking a mean of 2 signals from the galactose residues in the β(14)-linked galactan chain (78.49 and 75.34 ppm) vs. a mean of 2 signals from the nonreducing end of the galactose residue (C-4 at 69.45 ppm and C-5 at 75.94 ppm).
Force spectroscopy
The AFM used in this study was a TM Lumina (Veeco, Santa Barbara, CA, USA), which is a combined AFM-inverted optical microscope. The AFM head is fitted with a small charge-coupled device camera that views the sample from one side, which, in combination with the liquid cell and the inverted optical microscope, permitted observation of the sample under liquid (PBS) from both above and below. This feature enabled the placement of the AFM tip onto regions of the glass slides that had been functionalized with attached proteins. All binding measurements on functionalized glass surfaces were carried out under PBS. For the inhibition studies, lactose was added into the liquid cell to a concentration of 3.3 mg ml–1. The experimental data were captured in a so-called layered imaging mode (at a rate of 10 µm s–1 in the Z direction and at a scan rate of 1 Hz). In this mode, the instrument ramps the Z piezo element of the scanner by a predetermined amount at each sample point over a selected scan area and records the subsequent deflection of the cantilever as it is pushed into, then retracted away from, the sample surface. This process produces a matrix of curves of cantilever deflection vs. distance that relates to the image coordinates.
Analysis of force-distance curves
The raw data on cantilever vs. piezo deflection were collected and analyzed automatically by using in-house software. Raw data, consisting of 625 curves of cantilever deflection vs. piezo displacement captured by the AFM, were converted into force vs. distance curves by using experimentally determined values for INVOLS (which quantifies the optical response of the cantilever when pressed against a hard surface) and the spring constant, k, based on thermal noise analysis (32
33)
. The force-distance curves were then inspected for adhesive events. Adhesion was monitored by quantifying the hysteresis between the approach and retract portions of the force vs. distance data at the pull-off point. Rupture forces were calculated by fitting a straight line to the "off" region of the retract curve and calculating the maximum rupture force in the hysteresis region of the curve. The data on rupture force were then plotted as a histogram showing the distribution of measured rupture forces.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The extent of binding of the fluorescently labeled galactan, RGI, and PGA to Gal3-functionalized silica beads was examined by flow cytometry. Figures 3
and 4
show the fluorescence intensity data obtained from these experiments.
|
|
Underivatized beads, or functionalized beads not exposed to the fluorescently labeled galactan, show no detectable fluorescence (data not shown). To account for any nonspecific binding of the DTAF carbohydrates to derivatized beads, beads were prepared that were derivatized with MTS and GMBS and then terminated with glycine. As an example, incubation of these beads with DTAF-galactan led to a low degree of detectable fluorescence because of nonspecific adsorption of the labeled galactan onto the beads (Fig. 3
). The beads coated with Gal3 showed significant fluorescence following incubation with DTAF-galactan, demonstrating that the fluorescently labeled galactan bound to the Gal3-derivatized beads to a far greater extent (Fig. 3
).
Figure 4A
shows the relative affinity of the 2 principle constituents of pectin, PGA and RGI, compared with galactan toward Gal3-derivatized beads. For these studies, incubation of the Gal3-coated beads with the fluorescently labeled carbohydrate was carried out at pH 12. The beads were then washed in the appropriate buffer and resuspended in PBS at pH 7.2 for the flow cytometry studies. The binding of labeled galactan to Gal3 was found to be similar when incubation was carried out at either pH 7.2 or 12, suggesting that the low level of galacturonic acid in the galactan extract had little effect on the binding. The results confirm the microscopic observations; they show no significant binding of PGA, a very low degree of specific binding for RGI, and significant specific binding of galactans to Gal3. Despite the very low level of binding of RGI, the data obtained for this sample indicated that the RGI caused aggregation of the Gal3-derivatized beads. Figure 4B
quantifies the aggregation of the Gal3-derivatized beads on attachment of fluorescently labeled RGI in terms of the changes in fluorescence for monomers, dimers, and trimers. (Figure 4B
lists the percentage of aggregated beads for each category above the respective class of aggregate). It is notable that the changes in fluorescence intensity scale with the number of beads in the aggregate; this effect is not seen following incubation of the Gal3 beads with DTAF-PGA or DTAF-galactan. The extent of aggregation indicates that the binding of RGI is very poor, leaving unbound galactan chains. The fact that the fluorescent intensity scales with the number of beads within the aggregate demonstrates, however, that RGI can act as a multivalent ligand and cross-link Gal3-derivatized beads.
To establish further whether the increased level of binding of the galactan is specific to the presence of the Gal3 on the beads and occurs as the result of a carbohydrate-lectin interaction, additional data were obtained for BSA-derivatized beads incubated with DTAF-galactan. Figure 3
reveals negligible binding to BSA-derivatized beads, and the measured fluorescence is similar in level to that observed for the control beads capped with glycine.
To identify the extent of DTAF-galactan binding to Gal3-derivatized beads, we have determined a dose-response curve for binding and also quantified the effects of lactose inhibition on binding. Figure 5
shows a dose response curve for the binding of DTAF-galactan to Gal3-derivatized beads. The maximum was chosen to correspond to the amounts of pectin fragments used in the data presented in Fig. 4
. This level of DTAF-galactan did not cause any significant aggregation of the beads, and no significant aggregation was observed over the range of concentrations used in the present studies.
|
Figure 6
shows the effect of lactose addition on the binding of DTAF-galactan to Gal3-coated beads. The data reveal that progressive addition of lactose reduced the level of DTAF-galactan bound to the Gal3-coated beads. Even at levels of lactose of some 20 mg ml–1 (galactose equivalence 10 mg ml–1), the lactose did not completely inhibit the binding of the 0.8 mg ml–1 DTAF-galactan (galactose equivalence 0.7 mg ml–1). Thus, the binding of DTAF-galactan appears to be more than 10-fold greater than that of lactose. Lactose binds to Gal3 with a Kd 0.51 mM.
|
The flow cytometry data provide an excellent ensemble measurement of the level and specificity of binding but cannot easily be used to determine the magnitude of the binding force. Further evidence for specific binding between Gal3 and galactans can be obtained by using force spectroscopy measurements. In the force spectroscopy studies, AFM has been used to probe binding between a silica bead derivatized with galactans and Gal3 immobilized on glass slides. Fluorescence microscopy was used to verify the procedure for the covalent coupling of galactans to the silica beads. Figure 7a
shows fluorescent-labeled galactans covalently coupled to MPBH-functionalized silica beads. In this image, the clustering of the beads results from fluid flow that frequently occurs beneath the coverslip during preparation of the slide. Control silica beads, which were not decorated with the coupling agents, exhibited no fluorescence following incubation with the labeled galactan but showed similar clustering. Having established the validity of the coupling reaction, the silica beads were derivatized with unlabeled galactans for the AFM binding studies. After binding studies, the modified bead-cantilever assemblies were imaged by SEM. Figure 7b
shows a functionalized bead nestling against the pyramidal tip.
|
Figure 8A
shows a typical force-distance curve obtained in PBS when the activated bead was used to force-map the region of the glass slide coated with Gal3. On retraction, a finite force was required to detach the functionalized silica bead from the activated glass slide. Figure 8B
shows the distribution of adhesive values obtained after analysis of 625 force-distance curves. Galactan-Gal3 interaction data yielded adhesive forces in the range of 0–3.3 nN with a modal adhesive force of 276 pN. As was observed with the flow cytometry data, addition of lactose was found to partially inhibit binding between the galactan and Gal3 (Fig. 8C
). The binding events then yielded a lower modal value of 79 pN.
|
The 13C NMR data (Fig. 9
) show that the galactans are connected to the RGI backbone and provide an estimate of the average chain length. The spectrum is characteristic of a linear (14)-β-D-linked galactan chain with minor peaks corresponding to the RGI backbone. There is no free reducing end for the galactan chains, and the C1 resonance at 104.18 ppm corresponds to the linkage of the galactan chain to the rhamnose in the RGI backbone (34)
. There are minor peaks for C4 (69.45 ppm) and C5 (75.94 ppm) for the nonreducing end of the galactan chain (34)
. The estimated average chain length is 22 galactose residues.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
. The alkaline treatment of pectin will de-esterify the homogalacturonan backbone and release lower-molecular-weight PGA fragments. Studies conducted on extracts of citrus pectin PGA using fluorescence microscopy and flow cytometry suggest that there is little evidence for specific binding of the PGA fragments with Gal3. This result is perhaps not surprising because the PGA is (14)-linked rather than β(14)-linked, and the replacement of the secondary hydroxyl group with the carbonyl group will interfere with the hydrogen bonding to substituents on C6 of the galactose within the primary galactose binding site in Gal3.
The other major component of pectin is the heavily substituted RGI region containing neutral sugars. In addition to arabinan side chains, this region contains galactan and arabinogalactan side chains with terminal galactose residues (Fig. 1
). These are potential candidates for binding to Gal3. RGI regions can contain both β(14)-linked and β(13,6)-linked galactan side chains. The latter galactans are β(13) linked in their backbone, and the branching at C6 would interfere with hydrogen bonding within the primary binding site on Gal3. The branching will sterically inhibit accommodation of terminal galactose residues within the binding site. Recent studies (35)
on the removal of neutral sugar side chains from extracts of RGI, enzymatically released from citrus pectin, have shown that a combination of the use of endo-1,5--L-arabinanase, endo-1, 4-β-D-galactanase, -L-arabinofuranosidase, and β-D- galactosidase can be used to debranch the RGI sample. This finding suggests that citrus pectins contain β(14)-D-arabinogalactan and/or galactan side chains. Thus, acid treatment of citrus pectin would enhance removal of arabinose from RGI regions and promote the release of linear galactans and linear galactans lightly substituted with arabinose (Fig. 2b
). The stereochemistry of the linear β(14)-D-arabinogalactan and β(14)-D-galactan side chains suggest that they are the most likely candidates for binding to Gal3. This conclusion is reinforced by experimental evidence for the anticancer activity of arabinogalactans (26
27)
.
Samples of RGI regions from citrus pectin are not readily available commercially. To investigate the possible role of components of RGI regions, however, we have investigated the potential binding of enzymatically-released RGI regions from potato pectin. These fragments are available commercially in reasonable quantities for experimental studies. Enzymatic studies (36)
of the structure of RGI regions from potato pectin suggest that the RGI regions contain (15)-linked arabinans and an abundance of β(14)-D-galactans, poorly substituted with arabinose. These samples provide a route to examining the potential importance of the β(14)-D-arabinogalactan in citrus pectin RGI regions. Even binding of arabinose at C3 on terminal galactose may not inhibit potential binding to Gal3 because there is a degree of flexibility in the accommodation of substituents at C3 within the primary binding site of the Gal3 CRD. The fluorescence microscopy and the flow cytometry data (Fig. 4A
) suggest a very small degree of specific binding of the intact RGI regions to Gal3 under conditions that will eliminate nonspecific charge-charge complexation.
Acid treatment of citrus pectin will preferentially remove arabinans and arabinose substituents from RGI regions and will lead to release of galactans and galactan oligomers lightly substituted with arabinose. To simulate the effects of the removal of arabinose, we have used potato galactans prepared by the treatment of potato RGI with -L-arabinofuranosidase. These samples will consist of modified RGI regions and are multivalent, unlike free galactans and arabinogalactan chains, which might be expected to be largely monovalent.
The specificity of the interaction between the pectin galactan and Gal3 is shown by the fact that from fluorescence microscopy and flow cytometry observations, the galactans were found to bind to immobilized Gal3 but not to immobilized BSA (Fig. 3
). Furthermore, the binding was found to be inhibited by the addition of lactose (Fig. 6
). This finding is confirmed by the effect of the addition of the inhibitor lactose on the binding events seen by force spectroscopy (Fig. 8C
). The present study confirms suggestions that pectin fragments will bind to Gal3. The findings are consistent with the molecular hypothesis (1)
for the observed anticancer action of modified pectin and suggest that bioactivity probably results from modified neutral sugar side chains (arabinogalactans) of the pectin. In the present study, we have used modified RGI regions to assess the role of the galactan side chains. Technically, the galactan samples used contain small amounts of galacturonic acid, arabinose, and rhamnose. It is possible, but unlikely, that these minor sugars could be responsible for the observed binding. However, we have already shown that PGA does not bind specifically to Gal3. Rhamnose is 6-deoxy-L-mannose; the different substitution pattern at C6, the decoration with arabinan and galactan side chains at C4, and the (12) linkage to galacturonic acid would interfere with the hydrogen bonding at the primary binding site in the CRD. The galactan or modified RGI also contains terminal arabinose, but the arabinose is present in the furanose form, which would be incompatible with the Gal3 CRD. Thus, the present study identifies the importance of galactans and the likely role of linear β(14)-linked galactans or β(14)-linked galactans lightly substituted with arabinose.
The steric crowding within the RGI region appears to restrict binding to Gal3. Treatment to remove the arabinans and reduce the arabinose substitution of the galactans significantly enhances the binding of the galactans to Gal3. In these experiments, the galactan chains are still bound to the RGI backbone, which is indicated by the chemical composition of the sample and the measured pullulan-equivalent molecular mass of 272 kDa. The estimated molecular mass of chemically extracted potato RGI has been reported to be in the range 50–500 kDa (36)
. Although the estimated molecular mass for the galactan determined in our study will be a gross overestimate, because of the complex shape of the galactan and the stiffness of the β(14)-linked galactan chains relative to the highly flexible pullulan standards, the value indicates the presence of several galactan chains attached to the RGI backbone. The length of neutral sugar side chains in RGI regions will vary with the source, but the chain lengths may be as long as 20 sugar residues (37)
. The NMR data suggest that the main side chains are linear (14)-β-D-galactans linked to the RGI backbone. The estimated average chain length of 22 residues would correspond to a molecular mass of 4 kD, again suggesting the presence of several galactan chains attached to the RGI backbone in the galactan sample. The removal of arabinose appears to promote enhanced accessibility to the galactan chains and allows several individual galactan chains to bind to the same Gal3-derivatized bead, rather than cross-linking beads as seen with the addition of RGI.
The methods developed in the present study provide a basis for comparing the binding of isolated arabinogalactan and galactan side chains and thus for identifying the optimum binding of modified neutral sugar side chains. The ability to extract binding forces or binding energies from the force spectroscopy provides a basis for quantifying the effect of carbohydrate structure on the binding of modified hairy regions of the pectin chains to Gal3 at the molecular level.
In the present force spectroscopy studies, we have chosen to attach the carbohydrate to silica beads rather than directly to the AFM tip. We have shown previously (38)
that this approach is ideal for probing carbohydrate-lectin interactions on cell surfaces but tends to overestimate the binding force when used to monitor lectin-carbohydrate interactions where the lectin is immobilized on a flat surface, because of the possibility of multiple interactions. The measured modal values of the binding force for Gal3 immobilized on glass, however, could still be used to assess relatively the effects of chemical composition on the binding of modified neutral sugar side chains to Gal3. The flow cytometry data indicate that lactose inhibits the binding of galactans to Gal3 but allows a small residual binding to a reduced concentration of Gal3 molecules. This finding is consistent with the fact that the addition of lactose significantly reduces the number of binding events observed in the force spectroscopy studies and reduces the spread of measured adhesive values and the modal value of the measured adhesion. Lactose would be expected to reduce the number of binding events with the derivatized beads. Therefore, a modest concentration of lactose was added to eliminate any multiple binding between the galactan and the Gal3-derivatized surface. We thus believe that the observed binding events in the presence of lactose may be more representative of individual binding between galactan side chains and Gal3 molecules. The modal value of 79 pN is similar in magnitude to values reported for interactions between carbohydrate ligands and lectins. Touhami et al. (39
, 40)
reported a value of 96 ± 55 pN for single carbohydrate interactions between glucose and Concanavalin A and a similar value of 121 ± 53 pN for single interactions between glucose and individual cell-surface-expressed lectins on live yeast cells.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 14, 2008. Accepted for publication September 11, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
