Characterization of an arabinogalactan-protein from suspension culture of <Emphasis Type="Italic">Echinacea purpurea</Emphasis>

Suspension cultures of Echinacea purpurea have been established in MS medium supplemented with 2,4-D and an arabinogalactan-protein (AGP) was purified from the secreted soluble polymers by precipitation with ethanol, followed by precipitation with β-glucosyl Yariv reagent. It revealed typical features of AGPs: a high amount of polysaccharide (90% w/w) with the dominating monosaccharides galactose and arabinose and some glucuronic acid, and a small protein moiety (10% w/w) with the main amino acids Ala, Hyp, Glx, Ser, Asx and Thr. Linkage- and NMR-analyses showed the polysaccharide part to be composed of a branched core-polysaccharide of 3-, 6- and 3,6-linked Galp residues with terminal Araf, Arap, Galp and GlcAp residues. Compared to an AGP from pressed juice of the aerial parts of Echinacea purpurea, differences particularly in terminal arabinose mono- and oligosaccharides in arabinogalactan (AG) side branches could be detected. Testing of different AGP-antibodies with both AGPs confirmed the results of the analytical investigations. Binding of AGPs from plant and cell cultures to LM2, a monoclonal AGP-antibody reacting with a GlcA containing epitope, was comparable. The reactivity of a monoclonal antibody raised against the AGP from the plant recognizing a galactan epitope was also nearly similar with both AGPs. In contrast, polyclonal antibodies raised against the AGP from the plant and directed against an Araf-containing epitope of the AG side branches showed nearly no cross reactivity with the AGP from cell culture.     

Different arabinogalactan proteins are present in carrot (Daucus carota) cell culture medium and in seeds

Arabinogalactan proteins (AGPs) were isolated by Yariv phenylglycoside precipitation from the medium of carrot ( Daucus carota L.) cell cultures and from carrot seeds. The isolates showed a different composition of AGPs. The medium AGPs contained an arabinose poor AGP fraction that had relatively high levels of glucuronic acid and rhamnose. In contrast the seed AGPs only contained arabinose and galactose-rich AGP fractions that had low levels of glucuronic acid. Linkage analysis on all fractions showed that most of the arabinose residues were terminally linked and that almost all galactose was present in the 1,3-, 1,6- and 1,3,6- form. The strongly branched type II arabinogalactans are characteristic of the carbohydrate part of AGPs. AGP characteristic amino acid residues as Hyp, Pro, Glx, Ser, Gly, Asx, Ala, Leu and Thr were detected in three different fractions.     

Sucrose induces arabinogalactan protein secretion by <Emphasis Type="Italic">Beta vulgaris</Emphasis> L. cell suspension cultures

The aim of this study was to determine if the increase of the initial sucrose concentration (ISC) improves cell growth and arabinogalactan protein (AGP) secretion of Beta vulgaris L. cultures. ISC tested were 43.8, 87.6 and 131.4 mM. Cell growth and specific growth rate were improved increasing the ISC. Cell cultures grown with ISC 43.8 mM were fed with sucrose, and cellular growth was enhanced twofold, revealing the stimulatory effect of sucrose on cell growth. The AGP secretion was stimulated, increasing the ISC. This event was partially associated with the exponential growth phase of the culture. AGP precipitation with Yariv reagent of cell cultures inhibited cell growth without changes in viability. The assay of sucrose feeding confirmed the relationship between cell growth and AGP secretion. These observations suggest that AGPs may be required for cell division. The increase of AGP secretion by ISC coincided with a higher cellular aggregation, suggesting a possible role of AGP as cellular adhesion molecules. To determine whether AGP secretion is also stimulated by an osmotic effect, mannitol was fed to raise the osmotic potential from 23.78 to 95.97 mOsm kg⁻¹. Mannitol was not used for cell growth, but AGP secretion was stimulated sixfold in relation to the control. These results are important for understanding the possible factors involved in the AGP secretion of plant cell culture and that may be considered to improve the AGP production.     

<Emphasis Type="Italic">Physcomitrella patens</Emphasis> arabinogalactan proteins contain abundant terminal 3-<Emphasis Type="Italic">O</Emphasis>-methyl-<Emphasis Type="SmallCaps">l</Emphasis>-rhamnosyl residues not found in angiosperms

A biochemical investigation of arabinogalactan proteins (AGPs) in Physcomitrella patens was undertaken with particular emphasis on the glycan chains. Following homogenization and differential centrifugation of moss gametophytes, AGPs were obtained by Yariv phenylglycoside-induced precipitation from the soluble, microsomal membrane, and cell wall fractions. Crossed-electrophoresis indicated that each of these three AGP fractions was a mixture of several AGPs. The soluble AGP fraction was selected for further separation by anion-exchange and gel-permeation chromatography. The latter indicated molecular masses of ∼100 and 224 kDa for the two major soluble AGP subfractions. The AGPs in both of these subfractions contained the abundant (1,3,6)-linked galactopyranosyl residues, terminal arabinofuranosyl residues, and (1,4)-linked glucuronopyranosyl residues that are typical of many angiosperm AGPs. Unexpectedly, however, the moss AGP glycan chains contained about 15 mol% terminal 3-O-methyl-l-rhamnosyl residues, which have not been found in angiosperm AGPs. This unusual and relatively nonpolar sugar, also called l-acofriose, is likely to have considerable effects on the overall polarity of Physcomitrella AGPs. A review of the literature indicates that the capacity to synthesize polymers containing 3-O-methyl-l-rhamnosyl residues is present in a variety of bacteria, algae and lower land plants but became less common through evolution to the extent that this sugar has been found in only a few species of angiosperms where it occurs as a single residue on steroidal glycosides.     

Salt stress upregulates periplasmic arabinogalactan proteins: using salt stress to analyse AGP function

Arabinogalactan proteins (AGPs) are implicated in cell expansion by unknown mechanisms, thus AGP content and cell-expansion rate might be correlated. We used Yariv reagent to quantify release rates and distribution of AGP at the cell surface of tobacco BY-2 cells: plasma membrane (M); soluble periplasmic AGPs released by cell rupture (S); cell wall (W); and growth medium (Gsink). In contrast to earlier reports, we observed massive upregulation of AGPs in salt-stressed cells, and hence the absence of a simple, direct cause-and-effect relationship between growth rate and AGP release. There was a more subtle connection. A dynamic flux model, M-->S-->W-->Gsink, indicated that turnover was nondegradative, with little free diffusion of AGPs trapped in the pectic matrix of nonadapted cells where transmural migration of high molecular-weight AGPs occurred mainly by plug flow (apposition and extrusion). In contrast, however, an up to sixfold increased AGP release rate in the slower-growing salt-adapted cells indicated a greatly increased rate of AGP diffusion through a much more highly porous pectic network. We hypothesize that classical AGPs act as pectin plasticizers. This explains how beta-D-glycosyl Yariv reagents might inhibit expansion growth by crosslinking monomeric AGPs, and thus mimic an AGP loss-of-function mutation.     

Molecular cloning of cDNAs encoding the protein backbones of arabinogalactan-proteins from the filtrate of suspension-cultured cells of Pyrus communis and Nicotiana alata

This paper reports the isolation of cDNAs encoding the protein backbone of two arabinogalactan-proteins (AGPs), one from pear cell suspension cultures (AGP Pc 2) and the other from suspension cultures of Nicotiana alata (AGP Na 2). The proteins encoded by these cDNAs are quite different from the 'classical' AGP backbones described previously for AGPs isolated from pear suspension cultures and extracts of N. alata styles. The cDNA for AGP Pc 2 encodes a 294 amino acid protein, of which a relatively short stretch (35 amino acids) is Hyp/Pro rich; this stretch is flanked by sequences which are dominated by Asn residues. Asn residues are not a feature of the 'classical' AGP backbones in which Hyp/Pro, Ser, Ala and Thr account for most of the amino acids. The cDNA for AGP Na 2 encodes a 437 amino acid protein, which contains two distinct domains: one rich in Hyp/Pro, Ser, Ala, Thr and the other rich in Asn, Tyr and Ser. The composition and sequence of the Pro-rich domain resembles that of the 'classical' AGP backbone. The Asn-rich domains of the two cDNAs described have no sequence similarity; in both cases they are predicted to be processed to give a mature backbone with a composition similar to that of the 'classical' AGPs. The study shows that different AGPs can differ in the amino acid sequence in the protein backbone, as well as the composition and sequence of the arabinogalactan side-chains. It also shows that differential expression of genes encoding AGP protein backbones, as well as differential glycosylation, can contribute to the tissue specificity of AGPs.     

Dynamic turnover of arabinogalactan proteins in cultured Arabidopsis cells

Arabinogalactan proteins (AGPs) are very large proteoglycans thought to have more of a signaling than a structural role when secreted into the plant cell wall. AGPs are also the first known family of abundant plant proteins synthesized with glycosylphosphatidylinositol(GPI) anchors. Nascent cellular Arabidopsis AGPs, still bearing an intact GPI anchor, and AGPs copiously discharged into the culture medium after phospholipase-cleavage of their anchor were each represented by more than 15 seemingly homologous molecular species of increasing size. In washed cells ³H-ethanolamine was slowly incorporated into each AGP’s GPI anchor via phosphatidylethanolamine. Pulse labeling of AGPs by ³H-acetate and by ³H-galactose was much more rapid, allowing labeled AGP detection in the growth medium within 1 h. HPLC analysis of the radiolabel distribution in AGPs secreted within 1–8 h revealed a sharp preference for the larger molecular species. After several hours a population of smaller radioactive AGP species began to appear in the medium. Following certain manipulations of the cells newly secreted AGP species measured by HPLC on a relative mass basis formed a pattern surprisingly different from the radioactivity pattern, although larger species still dominated. Thus Arabidopsis cells appear capable of releasing higher mass AGP species apparently stored in cell wall sites along with a unique mixture of freshly synthesized AGPs in combinations potentially active in signaling.     

Arabinogalactan proteins are involved in cell aggregation of cell suspension cultures of <Emphasis Type="Italic">Beta vulgaris</Emphasis> L.

Arabinogalactan proteins (AGPs) are glycoproteins present at cell surfaces. Although exact functions of AGPs remain elusive, they are implicated in plant growth and development. The aim of this study was to evaluate the role of AGPs in the process of cell aggregation of Beta vulgaris L. suspension cultures. It was observed that B. vulgaris suspension cultures accumulated AGPs in parallel form to its cell growth. The AGPs maximum content in the stationary phase was 0.330 mg g⁻¹ dry weight (DW) in the cell wall (CW) and 1.534 mg g⁻¹ DW in the culture medium (CM), generating cell aggregates >500 μm (93.21% DW). The addition of tunicamycin (TM) caused a reduction of AGPs content in CW and CM of 46 and 64%, respectively. These changes were associated with inhibition of growth and the reduction of the cell aggregates >500 μm (50.0% DW). When TM was removed from the CM, cell growth, aggregation, and AGPs content on CW and CM were recovered. Precipitation of AGPs with Yariv reagent generated a reduction of 61.14% of AGPs content in CW and a total inhibition of AGPs secretion in CM. This Yariv treatment generated a reduction in the cell aggregates >500 μm of 51.31% of DW. When the Yariv reagent was removed from the culture, cells did not recover their AGPs accumulation. In addition, cell cultures did not recover their ability to grow and aggregate. These results indicate that AGPs are molecules required in the cellular aggregation process of B. vulgaris L. suspension cultures.     

Arabinogalactan-protein epitopes in somatic embryogenesis of Daucus carota L.

Two monoclonal antibodies (ZUM 15 and ZUM 18) directed against carrot (Daucus carota L.) seed arabinogalactan proteins (AGPs) were used to isolate specific AGP fractions. For both carrot and tomato (Lycopersicon esculentum Mill.) seed AGPs analyzed by crossedelectrophoresis, the ZUM 15 and ZUM 18 AGP fractions showed one identical peak. However, the Rf values for the two species were different: 0.82 for carrot seed AGPs and 0.52 for tomato seed AGPs. When the fractionated AGPs (carrot or tomato) were added to carrot cell lines they had a dramatic effect on the culture. One AGP fraction (ZUM 15 AGPs) was able to induce vacuolation of embryogenic cells. Those cells failed to produce embryos. The other AGP fraction (ZUM 18 AGPs) increased the percentage of embryognic cells from about 40% up to 80% within one week and this subsequently resulted in the formation of more embryos on hormone-free medium. This activity was higher than that of unfractionated carrot seed AGPs, while the optimum concentration was 50-fold lower. Since both ZUM 18 AGPs (carrot or tomato) yielded identical responses it can be concluded that neither the Rf value nor the source are essential for biological activity. The dose-response curve of ZUM 18 AGPs showed a sharp optimum. When the AGPs that also bound to the antibody ZUM 15 were removed, the dose-response curve of the remaining AGPs (containing only the ZUM 18 epitope, not the ZUM 15 epitope) resembled a saturation curve. Regardless of its concentration, the fraction in which AGP molecules contained both epitopes showed no appreciable embryogenesis-promoting activity. The biological activity of AGPs was therefore determined by the presence of embryogenesis-enhancing and-inhibiting epitopes. The inhibiting and enhancing epitopes can be located on separate molecules or one single AGP molecule.     

Identification of three potent hydroxyproline O‐galactosyltransferases in Arabidopsis

Arabinogalactan proteins (AGPs) are plant‐specific extracellular glycoproteins implicated in a variety of processes during growth and development. AGP biosynthesis involves O‐galactosylation of hydroxyproline (Hyp) residues followed by a stepwise elongation of the complex sugar chains. However, functionally dominant Hyp O‐galactosyltransferases, such that their disruption produces phenocopies of AGP‐deficient mutants, remain to be identified. Here, we purified and identified three potent Hyp O‐galactosyltransferases, HPGT1, HPGT2 and HPGT3, from Arabidopsis microsomal fractions. Loss‐of‐function analysis indicated that approximately 90% of the endogenous Hyp O‐galactosylation activity is attributable to these three enzymes. AGP14 expressed in the triple mutant migrated much faster on SDS‐PAGE than when expressed in wild‐type, confirming a considerable decrease in levels of glycosylation of AGPs in the mutant. Loss‐of‐function mutant plants exhibited a pleiotropic phenotype of longer lateral roots, longer root hairs, radial expansion of the cells in the root tip, small leaves, shorter inflorescence stems, reduced fertility and shorter siliques. Our findings provide genetic evidence that Hyp‐linked arabinogalactan polysaccharide chains are critical for AGP function and clues to how arabinogalactan moieties of AGPs contribute to cell‐to‐cell communication during plant growth and development.     

Immunochemical comparison of membrane-associated and secreted arabinogalactan-proteins in rice and carrot

Arabinogalactan-proteins (AGPs) occurring in suspension-cultured rice (Oryza saliva L.) cells, their conditioned medium and at the rice root apex were investigated using monoclonal antibodies and the AGP-binding -glucosyl Yariv reagent ( GlcY). A monoclonal antibody, LM2, was generated that recognized an acidic carbohydrate epitope common to two soluble AGPs occurring in the conditioned medium of proliferating rice cells, membrane-associated AGPs (rmAGP) in the cultured cells and two AGPs at the rice root apex. In addition, LM2 recognized AGPs secreted by suspensioncultured carrot (Daucus carota L.) cells. The two AGPs of the rice culture medium, srAGP1 and srAGP2, were discriminated by their mobilities during sodium dodecyl sulfate-polyacrylamide gel electrophoresis, reaction with GlcY, the presence of arabinogalactan epitopes and anion-exchange chromatography. The association of rmAGP with the plasma membrane was investigated by Triton-X-114/aqueous partitioning of both microsomal and plasma-membrane preparations and rmAGP was found to partition into the detergent phase, indicating that AGPs are hydrophobic plasma-membrane proteins in rice. This was in contrast to plasma-membrane AGPs of suspension-cultured carrot cells that partitioned into the aqueous phase. At the rice root apex most of the AGP was associated with the microsomal fraction and also partitioned into the detergent phase, although a distinct highmolecular-mass AGP entered the aqueous phase.Abbreviations AGP arabinogalactan-protein - GlcY -glucosyl Yariv reagent - ELISA enzyme-linked immunosorbent assay We gratefully acknowledge support from the Leverhulme Trust, the UK Biotechnology and Biological Sciences Research Council and the Royal Society.     

Carrot arabinogalactan proteins are interlinked with pectins

Cell wall extracts from a carrot cell culture and tap roots were obtained by sequential extraction with water, EDTA buffer solution and cold sodium hydroxide solution. Arabinogalactan proteins (AGPs) were isolated from the extracts and from the medium of the cell culture and analysed for their molecular weight distribution and carbohydrate composition. Copper ions were used to separate the Yariv positive fractions into AGP fractions with a high and a low level of galacturonic acid (GalA). The GalA rich AGP fractions were incubated with pectin methylesterase and polygalacturonase. This enzyme incubation released GalA fragments from the AGP fractions as monitored by HPAEC and MALDI-TOF MS. At least part of carrot AGPs from the medium and cell walls may be covalently linked to pectin containing a homogalacturonan structural element.     

Roles of arabinogalactan proteins in cotyledon formation and cell wall deposition during embryo development of <Emphasis Type="Italic">Arabidopsis</Emphasis>

Arabinogalactan proteins (AGPs) are a class of highly glycosylated, widely distributed proteins in higher plants. In the previous study, we found that the green fluorescence from JIM13-labeled AGPs was mainly distributed in embryo proper and the basal part of suspensor but gradually disappeared after the torpedo-stage embryos in Arabidopsis. And (β-d-Glc)₃ Yariv phenylglycoside (βGlcY), a synthetic reagent that specifically binds to AGPs, could inhibit embryo development. In this study, as a continuous work, we investigated the AGP functions in embryo germination, cotyledon formation, and cell wall deposition in Arabidopsis embryos by using immunofluorescent, immunoenzyme, transmission electron microscopy (TEM), and Fourier transform infrared spectroscopy (FTIR) techniques. The results showed that 50 μM βGlcY caused inhibition of embryo germination, formation of abnormal cotyledon embryos, and disorder of cotyledon vasculature. Compared with the normal embryos in vitro and in vivo, the AGPs and pectin signals were quite weaker in the whole abnormal embryos, whereas the cellulose signal was stronger in the shoot apical meristem (SAM) of abnormal embryo by calcofluor white staining. The FTIR assay demonstrated that the cell wall of abnormal embryos was relatively poorer in pectins and richer in cellulose than those of normal embryos. By TEM observation, the SAM cells of the abnormal embryos had less cytoplasm, more plastid and starch grains, and larger vacuole than that of normal embryos. These results indicated that AGPs may play roles in embryo germination, cotyledon formation, cell wall cellulose and pectin deposition, and cell division potentiality during embryo development of Arabidopsis.     

Arabinogalactan-proteins from the suspension culture medium and plasma membrane of rose cells

Arabinogalactan-proteins (AGPs) that bind to beta-glycosyl Yariv antigens have been purified from the culture medium and plasma membrane of "Paul's Scarlet" rose cells. Starting from culture medium or from plasma membrane vesicles prepared by aqueous two-phase partitioning, the purification procedure involved Yariv antigen-induced precipitation and subsequent chromatographic steps. Two fractions, AGP-(a) and AGP-(b), were obtained from the culture medium, and one AGP fraction was obtained from the plasma membrane. The glycosyl compositions of all three fractions were dominated by arabinosyl and galactosyl residues and included glucuronosyl and other minor residues. Methylation analysis showed that AGP-(a) and AGP-(b) were both highly branched 3,6-galactans with terminal arabinofuranosyl substituents. The amino acid compositions of all three AGPs were high in alanine, hydroxyproline, and serine and/or threonine. The amino-terminal sequence of AGP-(b) contained an alanine-hydroxyproline repeat. While sharing general structural similarity, the AGPs from the plasma membrane and the culture medium were distinguishable by composition and by size and charge, with the plasma membrane AGPs being larger and more negatively charged than the culture medium AGPs.     

Effects of Yariv dyes, arabinogalactan-protein binding reagents, on the growth and viability of Brazilian pine suspension culture cells

Arabinogalactan-proteins (AGPs) are a family of highly glycosylated hydroxyproline-rich glycoproteins implicated in several aspects of plant growth and development. (β-d-glucosyl)₃ Yariv phenylglycoside (β-GlcY), commonly known as Yariv reagent, selectively binds AGPs. We treated cell suspension cultures of Araucaria angustifolia, the Brazilian pine, with β-GlcY and observed inhibition of biomass increase in a culture medium with 50 μM β-GlcY. However, the growth was not inhibited by (α-d-galactosyl)₃ Yariv phenylglycoside (α-GalY) which does not bind AGPs. Fluorescein diacetate staining of cells indicated that β-GlcY severely affected cell viability. However, cell swelling, bursting and release of cellular contents, all characteristics of necrotic cell death, were not observed in β-GlcY-treated cells. Instead, programmed cell death (PCD) structural changes such as cytoplasmic shrinkage and condensation were observed in β-GlcY-treated cells. In addition, callose accumulation, which is another marker of PCD, was also observed in β-GlcY-treated cells. The use of both, Ac-VEID-CHO, an inhibitor of caspase-like proteolytic activity related to PCD, and phenyl methyl sulphonyl fluoride (PMSF), a protease inhibitor known to suppress PCD, in the culture medium did not reverse the growth inhibition caused by β-GlcY. These data indicate that the β-GlcY-induced inhibition of Araucaria cell’s growth is related to AGP perturbation, and also that this growth inhibition is due to increased cell death not driven by necrosis.     

Arabinogalactan-protein epitope Gal4 is differentially regulated and localized in cell lines of hybrid fir (<Emphasis Type="Italic">Abies alba</Emphasis> × <Emphasis Type="Italic">Abies cephalonica</Emphasis>) with different embryogenic and regeneration potential

Arabinogalactan proteins (AGPs) are important proteoglycans regulating somatic embryogenesis in diverse plant species. Embryogenic cells of somatic embryos are covered by special extracellular cell wall layer called extracellular surface matrix network (ECMSN) at their early developmental stages. Here we show that highly embryogenic cell line AC78 of hybrid fir (Abies alba × Abies cephalonica) differs from very low-embryogenic cell line AC77 in the abundance, subcellular localization and deposition of subset of secreted AGPs. A specific AGP epitope containing Gal residues and reacting to Gal4 antibody is secreted and deposited into ECMSN, which covers the surface of the embryogenic cells showing high embryogenic and regeneration capacity in the cell line AC78. On the other hand, this Gal4 AGP epitope was not secreted and/or found on the surface of meristematic cells showing low embryogenic and regeneration capacity in the cell line AC77, as well as on the surface of non-embryogenic suspensor cells and callus cells in both cell lines AC77 and AC78. As a positive control, we have used another AGP epitope LM2 (containing glucuronic acid) showing no significant differences in these two Abies hybrid lines. This study defines specific AGPs containing β-(1→6)-galactotetraosyl group as a first molecular component of ECMSN covering embryogenic cells in gymnosperms. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Fractionation and Structural Characterization of Arabinogalactan-Proteins from the Cell Wall of Rose Cells

Arabinogalactan-proteins (AGPs) have been purified from Paul's Scarlet rose (Rosa sp.) cell walls. As estimated by gel permeation chromatography, the apparent molecular masses of the two major cell-wall AGP fractions were 130 and 242 kD. Since the 130-kD AGP had a ratio of arabinose/glucuronic acid that was 12 times higher than that of the 242-kD AGP, the fractions were named cell-wall AGP1 (CW-AGP1) and glucuronogalactan-protein (GGP), respectively. CW-AGP1 and GGP contained predominantly t-arabinofuranosyl residues; 3-linked, 6-linked, and 3,6-branched galactopyranosyl residues; and 4-linked and t-glucuronopyranosyl residues. The 1H-nuclear magnetic resonance spectra of CW-AGP1 and GGP showed that the arabinofuranosyl and galactopyranosyl residues were predominantly in [alpha]- and [beta]-anomeric configuration, respectively, and that GGP contained a few O-acetyl residues. The protein moieties of CW-AGP1 and GGP were both rich in hydroxyproline and alanine but differed in the percentage of various amino acids, including hydroxyproline, alanine, serine, and glycine. Cell-wall AGPs bound to ([beta]-D-glucosyl)3 Yariv phenylglycoside, but the stoichiometry of binding was about 6 times greater in GGP than in other Rosa AGPs. GGP seems to be peculiar to the cell wall, since no similar molecule was found in the culture medium.     

Yariv reagent treatment induces programmed cell death in Arabidopsis cell cultures and implicates arabinogalactan protein involvement

Arabinogalactan proteins (AGPs) are a family of highly glycosylated, hydroxyproline-rich glycoproteins implicated in various aspects of plant growth and development. (beta-D-glucosyl)3 and (beta-D-galactosyl)3 Yariv phenylglycosides, commonly known as Yariv reagents, specifically bind AGPs in a non-covalent manner. Here (beta-D-galactosyl)3 Yariv reagent was added to Arabidopsis thaliana cell suspension cultures and determined to induce programmed cell death (PCD) by three criteria: (i) DNA fragmentation as detected by terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) of DNA 3'-OH groups; (ii) inter- nucleosomal DNA fragmentation as visualized by genomic Southern blotting; and (iii) structural changes characteristic of PCD including cytoplasmic shrinkage and condensation, chromatin condensation and nuclear membrane blebbing. These findings implicate AGP involvement in PCD in plants, presumably by perturbation of AGPs located at the plasma membrane-cell wall interface.