Metabolic click-labeling with a fucose analog reveals pectin delivery, architecture, and dynamics in Arabidopsis cell walls

Abstract

Polysaccharide-rich cell walls are a defining feature of plants that influence cell division and growth, but many details of cell-wall organization and dynamics are unknown because of a lack of suitable chemical probes. Metabolic labeling using sugar analogs compatible with click chemistry has the potential to provide new insights into cell-wall structure and dynamics. Using this approach, we found that an alkynylated fucose analog (FucAl) is metabolically incorporated into the cell walls of Arabidopsis thaliana roots and that a significant fraction of the incorporated FucAl is present in pectic rhamnogalacturonan-I (RG-I). Time-course experiments revealed that FucAl-containing RG-I first localizes in cell walls as uniformly distributed punctae that likely mark the sites of vesicle-mediated delivery of new polysaccharides to growing cell walls. In addition, we found that the pattern of incorporated FucAl differs markedly along the developmental gradient of the root. Using pulse-chase experiments, we also discovered that the pectin network is reoriented in elongating root epidermal cells. These results reveal previously undescribed details of polysaccharide delivery, organization, and dynamics in cell walls.

The primary cell walls of plants consist of a complex polysaccharide-rich network containing cellulose, hemicellulose, pectin, and structural proteins. Cellulose, which is the primary load-bearing component of the wall, is hydrogen-bonded to hemicelluloses, and this composite is embedded in a pectin matrix (1, 2). The cell wall must simultaneously maintain the structural integrity of the plant cell by resisting the osmotic pressure necessary for turgor-mediated cell growth, and remain sufficiently dynamic to expand along with the growing cell. These requirements are satisfied by the controlled synthesis, deposition, and alteration of cell-wall components. Cellulose is synthesized at the plasma membrane (3), whereas hemicelluloses and pectins are synthesized intracellularly and secreted into the wall via an incompletely characterized vesicle-mediated trafficking pathway (2).

In contrast to proteins, for which genetically encoded fluorescent tags have provided many insights into localization and function, detailed characterization of cell-wall polysaccharide structure and dynamics in intact plants has proven difficult. Chemical (4, 5) and biochemical (6–8) techniques can provide primary structural information for extracted cell-wall polymers. Imaging approaches, including transmission electron microscopy (9, 10), Raman microspectroscopy (11), and studies using a growing array of lectins, carbohydrate-binding modules, and carbohydrate-specific antibodies (12) have provided information regarding the location or abundance of cell-wall polysaccharides in various cell types (13). However, these imaging techniques suffer from limitations, including long sample-preparation times, fixation artifacts, lack of temporal information, lack of polymer specificity (12), and masking of antibody epitopes (14). In addition, the large size of protein-based probes relative to the pore size of the plant cell wall (15) might prevent the efficient labeling of cell-wall components. As a result, many developmental changes in cell-wall architecture remain uncharacterized.

Studies using a cellulose-specific dye (16) and the incorporation of fluorescent xyloglucan oligosaccharides into plant cell walls (17) have highlighted the potential of small-molecule probes to image cell-wall polysaccharide networks. The development of sugar analogs compatible with click chemistry (18) presents a unique opportunity to extend the suite of small-molecule cell-wall imaging tools. These sugar analogs are modified with an azido or alkynyl group and can be covalently coupled by a copper-catalyzed [2+3] cycloaddition “click” reaction to a fluorophore possessing the opposite functional group. This reaction is bio-orthogonal, proceeds rapidly at room temperature, and has been used to image glycoconjugates in multiple biological systems (18, 19), but to date has not been applied to plant cell walls.

In this study, we demonstrate that the sugar analog fucose alkyne (FucAl) is metabolically incorporated into Arabidopsis cell walls and that the resulting polysaccharides can be efficiently labeled by copper-catalyzed click chemistry. We also provide evidence that FucAl is differentially incorporated into cell-wall components and that the majority of the FucAl is present in pectic arabinogalactans.

Results

FucAl Is Incorporated into Living Plant Tissue and Can Be Fluorescently Labeled via Click Chemistry.

FucAl was chosen to investigate click chemistry-mediated labeling of plant cell-wall components because several of these glycans are fucosylated at known positions (20–22). Arabidopsis Col-0 seedlings were incubated in liquid Murashige and Skoog (MS) mineral medium containing 2.5 μM FucAl for 4 h, labeled with Alexa-488 or Alexa-594 azide by the copper-catalyzed reaction, and observed by epifluorescence microscopy. FucAl-treated seedlings exhibited robust fluorescence throughout the root tissue (Fig. 1 A and C), whereas control seedlings treated with 0.01% DMSO did not exhibit appreciable fluorescence when labeled with either fluorophore (Fig. 1 B and D).

Fig. 1.

FucAl incorporation in Arabidopsis seedlings. (A–D) Both 0.1 μM Alexa 488-azide (A) and 0.1 μM Alexa 594-azide (C) label seedlings treated with 2.5 μM FucAl for 4 h, but not seedlings treated with 0.01% DMSO for 4 h (B and D). (E and F) FucAl incorporation, but not labeling, is dependent on viability. Seedlings were treated with 2.5 μM FucAl for 4 h before (E) or after (F) fixation in 4% paraformaldehyde for 30 min, followed by labeling with 0.1 μM Alexa 488-azide. Images were recorded with a 10× objective on an epifluorescence microscope using fluorescence filter sets (SI Materials and Methods) or brightfield. Images were collected using identical exposure settings and were not contrast-enhanced. (Scale bar, 100 μm.)

To assess whether seedling viability is required for FucAl incorporation, seedlings were treated with 2.5 μM FucAl before or after fixation with 4% paraformaldehyde and labeled with Alexa-488 azide. Seedlings treated with FucAl, then fixed before fluorophore labeling, were indistinguishable from unfixed seedlings (compare Fig. 1E with Fig. 1A), whereas plants fixed before FucAl incorporation exhibited very little fluorescence (Fig. 1F). These results suggest that the FucAl-associated fluorescence observed in root tissue is a result of FucAl incorporation via metabolic activity rather than nonspecific adsorbtion of the sugar analog to the seedling.

FucAl Is Likely Incorporated into Cell Walls via the Fucose Salvage Pathway.

GDP-l-fucose is the substrate for fucosyltransferases in plants and animals and can be synthesized de novo from GDP-d-mannose (23) or generated from free fucose via a salvage pathway (24), suggesting that exogenous fucose might compete with FucAl for incorporation. To test this hypothesis, we assessed the ability of various monosaccharides to inhibit FucAl incorporation. As an initial test for nonspecific metabolic effects, seedlings were incubated in liquid MS with 25 mM monosaccharides for 4 h, washed, and plated on solid MS lacking monosaccharide for 24 h, after which root growth was measured as an assay for seedling viability. Of the sugars tested, only mannose inhibited root elongation after incubation (Fig. S1) and was excluded from subsequent experiments. Seedlings were then treated for 4 h with FucAl with or without the addition of 25 mM nontoxic monosaccharides and labeled, and the resulting FucAl-associated root-tip fluorescence was quantified. Treatment with FucAl plus 25 mM fucose led to a 59% reduction in mean root-tip fluorescence compared with seedlings treated with FucAl alone, and resulted in a mean fluorescence level identical to that for seedlings treated with DMSO alone (Fig. S1). Unexpectedly, seedlings treated with FucAl in the presence of excess glucose or galactose exhibited increased root-tip fluorescence compared with plants treated with FucAl alone (Fig. S1).

To determine the subcellular location of FucAl-associated fluorescence, seedlings were treated with FucAl for 24 h, labeled with Alexa 488-azide, and subjected to plasmolysis in 0.8 M mannitol. After plasmolysis, FucAl-associated fluorescence remained localized mainly to cell borders with minor fluorescence in the plasma membrane and cytoplasm (Fig. S2), indicating that FucAl labels a component of the plant cell wall.

Biochemical Analysis of FucAl Incorporation into Arabidopsis Cell Walls.

In Arabidopsis cell walls, fucose is present in the hemicellulose xyloglucan (6), the pectins rhamnogalacturonan-I (RG-I) (25) and rhamnogalacturonan-II (RG-II) (26), arabinogalactan proteins (AGPs) (21), and N-linked glycoproteins (22). Each of these cell-wall components has a characteristic extraction pattern (27) and can be analytically digested by a specific set of enzymes (8). To identify cell-wall glycans that were modified by FucAl, seedlings were treated with FucAl and labeled with Alexa-594 azide. Labeled seedlings were homogenized, extracted with water and 70% ethanol to remove small molecules, and extracted with 1:1 chloroform:methanol to remove lipids and some proteins. FucAl-dependent fluorescence was not extracted during alcohol insoluble residue (AIR) preparation, and the remaining AIR contained considerable Alexa 594 fluorescence. AIR was extracted with 50 mM CDTA (trans-1,2-diaminocyclohexane-N,N,N′,N′-tetraacetic acid) to remove Ca²⁺-crosslinked uronic acid-rich pectins, resulting in an eightfold increase in solubilized FucAl-dependent fluorescence over DMSO-incorporated controls (Fig. 2A and Table S1). These results suggest that FucAl is incorporated into CDTA-extractable pectin. CDTA extraction solubilized 14.5 ± 2.5% of the total AIR fluorescence and 31.3 ± 4.2% of the total uronic acids, suggesting that both CDTA-resistant FucAl-labeled material and additional pectins remained in the AIR pellet. Further extraction with 8 M urea, which releases tightly bound pectins and some hemicelluloses (27), also removed some FucAl-associated fluorescence (Fig. 2A). Further chemical extraction to remove remaining pectins and hemicelluloses was uninformative because of the fact that the alkaline solutions required for these extraction steps destroyed Alexa 594 fluorescence.

Fig. 2.

Extraction and characterization of FucAl-labeled cell wall components. (A) Fluorescently labeled FucAl is enriched in CDTA and urea extracts of cell walls. Seedlings were treated with 25 μM FucAl or 0.1% DMSO for 24 h and labeled with Alexa 594-azide. The labeled seedlings were homogenized and cellular components were sequentially extracted. Fluorescence at 585 nm (F₅₈₅) of 200-μL aliquots of each extract was measured, and the ratio of FucAl-treated to DMSO-treated F₅₈₅ was calculated. (B) Cell walls prepared from labeled FucAl-treated seedlings were digested with 2.5 U of PME/PG, 1,5-Ara, 1,4-Gal, 1,5-Ara/1,4-Gal, 1,3-Gal, and XEG. Fluorescence was measured as in A and standardized to no enzyme controls (Materials and Methods). Error bars represent SEM from three independent experiments. (C–H) Solubilized material from the indicated digests (red) or no enzyme control (blue) was fractionated on a Sephadex G-75 column, and F₅₈₅ of 1 mL fractions was measured. Chromatograms are representative of three repetitions of each experiment. Note different scales on y axes of graphs. v₀ = void volume, v_i = included volume.

To further characterize FucAl-labeled cell-wall glycans, AIR from FucAl-treated plants was analytically digested with specific polysaccharide-degrading enzymes, and the products of these digests were separated by gel-filtration chromatography. Treatment with a pectin methylesterase (PME)/endo-polygalacturonase (PG) mixture, which cleaves pectin backbones, solubilized twice as much FucAl-associated fluorescence from AIR as buffer alone (Fig. 2B). PME/PG digests of plant cell-wall polysaccharides have a characteristic molecular weight distribution when subjected to gel-filtration chromatography, with RG-I eluting in the void volume followed by RG-II and smaller oligogalacturonides (28). When the PME/PG-solubilized material was subjected to gel-filtration chromatography, FucAl-associated fluorescence consistently eluted as a single peak in the void volume (Fig. 2C), suggesting that FucAl labels a high molecular-weight cell wall component that is most likely RG-I.

RG-I contains arabinan, galactan, and arabinogalactan side chains that can be degraded by endo-1,5-arabinanase (1,5-Ara), endo-1,4-galactanase (1,4-Gal), or a combination of these enzymes. Because RG-I arabinogalactan side chains are terminally fucosylated (25), we reasoned that digestion of FucAl-labeled material with these enzymes would release smaller fluorescently labeled fragments that could be resolved by gel filtration. Digestion of FucAl-labeled AIR with 1,5-Ara, 1,4-Gal, or a combination of these enzymes released twice as much FucAl-associated fluorescence as buffer alone, indicating that these enzymes solubilize FucAl-labeled glycans (Fig. 2B and Table S2). However, the gel-filtration profiles of the digestion products differed markedly: 1,5-Ara or 1,4-Gal alone released material that eluted in the void volume, similar to the products of the PME/PG digest (Fig. 2 D and E), whereas treatment with both enzymes resulted in the appearance of a second fluorescent peak that eluted near the included volume of the column (Fig. 2F), indicating that this enzyme mixture produced a fluorescently labeled low molecular-weight degradation product consistent with arabinogalactan side-chain cleavage.

In addition, treatment with an exo-1,3-galactanase, which cleaves linkages that are present in both RG-I and AGP side chains (25), also enhanced the solubilization of fluorescence from AIR, albeit at lower levels than the above-mentioned enzymes (Fig. 2B). Interestingly, this digestion resulted in a similar molecular weight distribution of FucAl-associated fluorescence as that observed for the 1,5-Ara/1,4-Gal mixture (Fig. 2G). In contrast, treatment with xyloglucan-specific endoglucanase (XEG) did not enhance the solubilization of fluorescence from AIR (Fig. 2B), and low levels of fluorescence were detected when the digestion products were separated by gel filtration (Fig. 2H); this enzyme specifically acts on xyloglucan (6), suggesting that FucAl is not primarily incorporated into this polymer.

Genetic Analysis of FucAl Incorporation into Arabidopsis Cell Walls.

To independently characterize the identity of the FucAl-labeled components, Arabidopsis mutants containing lesions in genes encoding enzymes which fucosylate side chains in xyloglucan (29), N-linked glycans (22), and AGPs (21) (Table S3) were treated with FucAl and labeled. After incorporation and labeling, mutant roots were observed by epifluorescence microscopy and compared with wild-type roots by quantifying the mean fluorescence intensity in each root tip for ≥ 30 seedlings. Incorporation of FucAl was not significantly lower than that in wild-type controls in mur2-1, mur3-2, and xxt1;xxt2 seedlings (Fig. S3). These mutants do not produce fucosylated xyloglucan (29–31), which normally accounts for 50% of the fucose in plant cell walls (29). Similarly, FucAl incorporation was not significantly reduced in cgl1-1, cgl1-2, cgl1-3, fut11, fut12, and fut13 mutants, which lack normally fucosylated N-linked glycans (Fig. S3) (22, 32–34). Finally, FucAl incorporation into fut4;fut6 double-mutant seedlings was identical to wild-type (Fig. S3); the products of these two genes have recently been shown to fucosylate AGPs (21). Taken together, the above results indicate that the majority of FucAl is incorporated into pectic RG-I in Arabidopsis rather than xyloglucan, N-linked glycans, RG-II, or AGPs.

Newly Synthesized FucAl-Containing Glycans Are Delivered to the Inner Face of the Cell Wall at Discrete Locations.

To investigate the dynamics of FucAl incorporation in subcellular detail, seedlings were treated with FucAl for increasing time periods before labeling, and the resulting fluorescence pattern in root elongation-zone epidermal cells was observed by spinning disk confocal microscopy. After 1 h of incorporation, labeling was largely restricted to actively growing cells in the division and elongation zones of the root. FucAl-associated fluorescence manifested as small, distinct punctae (Fig. 3A), and the density of these punctae increased after 2 h of incorporation (Fig. 3B). After 4 h of incorporation, labeled FucAl was evident along the entire root length and in the elongation zone was more evenly distributed across the cell wall (Fig. 3C). Labeling was homogeneous after 8 and 12 h of incorporation (Fig. 3 D and E), indicating that FucAl-containing glycans were either progressively delivered to every part of the cell wall or became evenly distributed in the cell wall after delivery. Fluorescent labeling was much weaker in control seedlings incubated with DMSO for up to 12 h (Fig. 3F), indicating that FucAl was responsible for the observed patterns of fluorescence. These results suggest that FucAl-labeled glycans are rapidly and continually delivered to the cell wall, but are initially delivered at discrete locations that might represent fusion sites of RG-I–containing vesicles with the plasma membrane.

Fig. 3.

Time-course of FucAl incorporation in elongating root cells. (A–E) Four-day-old seedlings were treated with 2.5 μM FucAl for the indicated times, labeled with Alexa 488-azide, and z series of elongation-zone root epidermal cells were recorded using a spinning disk confocal microscope with a 1.4 NA 100× oil-immersion objective. Images are contrast enhanced maximum projections of the z series. (F) Control seedlings treated with DMSO for 12 h and labeled with Alexa 488-azide show background fluorescence. (G) Maximum projections of z series of elongation-zone root epidermal cells from a 4-d-old seedling treated with 2.5 μM FucAl for 2 h, labeled with Alexa 488-azide, and stained with 0.01% S4B for 30 min. In the merged image on the right, labeled FucAl is green and S4B is red. Lower image shows an x-z projection through the dotted line in a merged z projection and shows that FucAl labeling lies below S4B labeling. (Scale bars, 10 μm.)

The above results suggest that FucAl constitutes a marker for newly synthesized RG-I. To test this hypothesis, seedlings were treated with FucAl for 2 h, labeled with Alexa-488 azide, and stained with Pontamine Fast Scarlet 4B (S4B), a fluorescent dye that labels cellulose throughout root epidermal cell walls (16). Dual-color confocal imaging and x-z projections of z series revealed that the labeled FucAl was present at the inner face of the cell wall adjacent to the plasma membrane, whereas S4B staining occurred throughout the wall (Fig. 3G). These results indicate that nascent FucAl-containing glycans are delivered to the inner face of the cell wall and support the above hypothesis.

Subcellular Distribution of FucAl-Containing Glycans Differs Over the Course of Root Epidermal Cell Development.

Arabidopsis roots represent a developmental gradient in which cells originate by division near the root tip, then elongate and undergo differentiation (35). To examine the developmental profile of FucAl incorporation in roots, seedlings were treated with FucAl for 12 h, labeled with Alexa-488 azide, and imaged by confocal microscopy from the root tip through the differentiation zone of the root. Projections of contiguous z series were used to construct mosaics of FucAl-treated roots (Fig. 4A), revealing distinct developmental fluorescence patterns. Cells that were newly divided or beginning to elongate exhibited diffuse labeling (Fig. 4D) that was consistent with the 12-h incorporation time point in Fig. 3E. However, cells in the early differentiation zone contained large intracellular globular bodies that were distributed throughout the cells (Fig. 4C). Epidermal cells in the late differentiation zone showed diagonal fibrils that spanned the width of the cells with intense fluorescence at root-hair primordia (Fig. 4B). Quantification of each pattern using cell length to determine developmental stage (Table S4) indicated that younger cells exhibited the diffuse staining pattern, whereas older and longer cells exhibited a gradient of patterns starting with globular fluorescence and progressing to fibrillar fluorescence. These results suggest that the spatial pattern of FucAl-containing glycan delivery and distribution in the cell wall changes over developmental time. Except for an alteration in the subcellular distribution of FucAl in fut1-1 elongation-zone epidermal cells, spinning disk microscopic analysis of representative mutants treated with FucAl revealed largely similar developmental patterns (Fig. S4), providing additional evidence that these mutations do not alter the global pattern of FucAl incorporation.

Fig. 4.

Developmentally distinct patterns of FucAl incorporation in Arabidopsis roots. Four-day-old light-grown Arabidopsis seedlings were treated with 2.5 μM FucAl for 4 h, labeled with 0.1 μM Alexa-488 azide, and imaged by confocal microscopy. (A) Mosaic of maximum projections of a contiguous z series collected starting at the root tip and progressing into the late differentiation zone. (Scale bar, 50 μm.) (B–D) Representative images collected in the late differentiation zone (B), the early differentiation zone (C), and elongation zone (D). (Scale bars 10 μm.) Arrowhead in (B) indicates bright fluorescence associated with a root-hair primordium.

Spatial Organization of a Pulse of FucAl-Labeled Glycan Changes Over Time.

To investigate the possibility of using FucAl as a tool to examine cell-wall dynamics in vivo, Arabidopsis seedlings were grown on media containing FucAl or click-labeling reagents. The toxicity of each compound was determined by a root-length bioassay after 7 d of growth. FucAl was nontoxic at a concentration 10-fold higher than necessary to label plants for microscopy. However, the CuSO₄ and ascorbic acid concentrations used to perform click-labeling were highly toxic (Fig. S5), suggesting that these reagents and the labeling reaction are not compatible with live-cell imaging.

To circumvent this issue, we performed pulse-chase experiments to track the dynamics of a subpopulation of FucAl-labeled material over time. Seedlings were treated with FucAl for 1 h, chased by plating on medium lacking FucAl for 0, 4, 8, 12, or 24 h, and labeled. After 1 h of incorporation, the FucAl incorporation pattern was micropunctate, similar to that observed in time-course experiments (compare Fig. 5B with Fig. 3A). After chasing for 4 and 8 h, the FucAl incorporation pattern became homogeneous (Fig. 5 C and D), suggesting further incorporation of FucAl into the cell walls and spreading of the incorporated glycans. Strikingly, after a 12-h chase, longitudinal striations in the labeled cell walls became evident (Fig. 5E), and these structures became more pronounced with a more heterogeneous distribution after a 24-h chase (Fig. 5 F and G). After chasing for 12 or 24 h, we also observed bright regions of fluorescence along the rootward borders of some epidermal cells (85/142 cells at 12 h and 58/87 cells at 24 h from three experiments). Overall, these results suggest that FucAl-containing glycans reorient along the longitudinal axis of the cell during elongation.

Fig. 5.

Pulse-chase analysis of FucAl incorporation. (A) Pulse-labeling protocol. Four-day-old light-grown Col-0 seedlings were treated with 2.5 μM FucAl for 1 h, washed, plated on MS plates lacking FucAl for increasing time periods, and labeled with Alexa 488-azide. (B–F) Maximum projections of z series of root elongation-zone epidermal cells treated with a 1-h pulse of FucAl and chased for the indicated times. Some cells displayed bright fluorescence at the rootward edge after a 12- or 24-h chase (arrowheads in E and F). (Scale bar, 10 μm.) (G) Maximum projection of z series of root differentiation-zone epidermal cells treated with a 1-h pulse of FucAl and chased for 24 h.

Discussion

In this study, we show that a click chemistry-compatible sugar analog, FucAl, is metabolically incorporated into Arabidopsis root epidermal cell walls and that a significant fraction of the incorporated FucAl is present in RG-I. Click-labeling of incorporated FucAl allowed us to identify RG-I delivery sites, map RG-I localization during root development, and investigate the dynamics of RG-I subpopulations in elongating cell walls.

FucAl incorporation only occurs in living tissue and can be reduced by competition with excess fucose, indicating that FucAl likely enters cellular metabolism via the fucose salvage pathway. However, the addition of glucose or galactose led to increased FucAl incorporation; one possible explanation for this result is that excess glucose or galactose causes the down-regulation of de novo GDP-l-fucose synthesis, resulting in GDP-l-FucAl comprising a larger fraction of the substrate pool for the relevant fucosyltransferases. Whereas the fucosyltransferase reactions that should lead to FucAl incorporation most likely occur in the Golgi, we observed the majority of FucAl-labeling in the cell wall (Fig. S2). This result could be because of a lack of membrane permeability of Alexa 488- and Alexa 594-azide caused by the negative charge of Alexa fluorophores (36).

In animal cells, FucAl is incorporated into N-linked glycoproteins (19). However, several lines of evidence indicate that, in Arabidopsis, the majority of FucAl is incorporated into RG-I. First, the finding that mutants defective in xyloglucan and N-linked glycan fucosylation do not display obvious reductions in FucAl incorporation levels (Fig. S3) or global alterations in FucAl incorporation patterns (Fig. S4) when compared with wild-type plants suggests that the majority of FucAl is not incorporated into these polymers. Second, the solubilization of FucAl-associated fluorescence by CDTA extraction and PME/PG digestion suggests that pectins contain the labeled FucAl. In addition, labeled material was solubilized by the pectinolytic enzymes 1,5-Ara, 1,4-Gal, 1,5-Ara/1,4-Gal, and 1,3-Gal. However, only 1,5-Ara/1,4-Gal and 1,3-Gal produced low molecular-weight fragments, as detected by gel-filtration chromatography; these linkages are present in side chains associated with RG-I and AGPs (37–39). The possibility remains that a portion of incorporated FucAl is present in the side chains of AGPs, which share common linkages with RG-I side chains; but given that PME/PG solubilizes the labeled material, the simplest interpretation of our results is that the majority of the label resides in RG-I. Finally, the fact that the FucAl-associated component is initially delivered to the cell wall at discrete locations and is reorganized as a coherent network suggest that a single type of carbohydrate, or a group of tightly associated carbohydrates, are being labeled. It is entirely possible that other cell-wall components incorporate FucAl at lower levels, and the full characterization of FucAl incorporation will likely require both more extensive carbohydrate analyses and the identification of the fucosyltransferases responsible for transferring FucAl to cell-wall polymers. Furthermore, linkage analysis suggests that fucose is attached to the arabinogalactan side chain of RG-I by a unique linkage (25) that is not present in xyloglucan, RG-II, N-linked glycoproteins, or AGPs, potentially explaining the observed distribution of FucAl incorporation.

Previous studies (9, 10, 40) have indicated that pectin and hemicelluloses are delivered to the apoplast by Golgi-derived vesicles and exist in distinct cell-wall layers. In time-course experiments, FucAl-containing material initially appears at discrete micropunctae that are distributed across the length and width of the cell wall. These micropunctae likely represent the sites of fusion between FucAl-containing vesicles and the plasma membrane. In seedlings treated for longer times, the FucAl-containing material exhibits diffuse, globular, or fibrillar localization depending on the developmental stage of the cell. Intriguingly, the diagonal fibrillar staining pattern we observed in differentiated cells is reminiscent of the organization of both cortical microtubules (41) and newly synthesized cellulose (42) in roots, suggesting that pectin in these cells might be aligned with one or both of these polymers. These findings contrast with previous models in which pectin is distributed evenly across the x-y plane of the cell wall (1). Calcium crosslinked pectin networks correlate with increased cell wall rigidity (43), and the presence of FucAl-containing fibrils in differentiated cells that have ceased expanding might reflect this rigidification process. In addition, recent data indicate that root-hair tips are enriched in pectins (44), and the bright FucAl-associated fluorescence we observed at root-hair primordia is in agreement with these results.

Using pulse-labeling, we tracked the delivery pattern and spatial reorientation of FucA1-containing material in elongating cell walls. The evolution of longitudinal fibrils is similar to the observed reorientation of cellulose bundles in living cells (16), further suggesting that these two networks might interact during cell-wall expansion. In some cases, pectin has been found to be covalently linked to hemicellulose (45), which in turn interacts with cellulose via noncovalent interactions (2); these interactions might enable the coordinated movement of polymer networks during cell-wall expansion.

The finding that pulse-incorporated FucAl-associated fluorescence is brighter at the rootward borders of many cells after chasing for 12 or 24 h can potentially be explained in two ways: either these regions of the cell walls did not expand as much as other regions during the chase period, and thus did not dilute the FucAl, or more FucAl-containing material was initially deposited at the rootward borders of these cells. The former explanation is more likely, given that brighter fluorescence at the rootward borders of cells was not observed in seedlings continuously treated with FucAl for up to 12 h (Fig. 3A). In root epidermal tissue, auxin inhibits expansion and is transported between cells in a shootward direction (46), and the lack of FucAl dilution at the rootward edge of cells might occur as a result of this inhibition.

In future studies, recently developed nontoxic click labeling reagents (47) and an expanding set of sugar analogs compatible with click chemistry (18) could extend the observations reported here by enabling the fine-scale characterization of the intracellular trafficking, delivery kinetics, and apoplastic modification of RG-I and other cell-wall polysaccharides in living seedlings. The specificity of FucAl incorporation we observed suggests that structural variants of sugar analogs might be differentially incorporated by their cognate glycosyltransferases, providing specific probes for individual cell-wall polysaccharides. In combination with additional chemical, biochemical, and cell biological approaches, these studies should enhance our understanding of the structure and dynamics of plant cell walls, enabling the efficient use of this abundant and renewable resource.

Materials and Methods

To incorporate fucose alkyne, 4-d-old light grown Arabidopsis seedlings were transferred from MS agar plates (16) to MS containing 2.5 μM FucAI (Invitrogen), and incubated in constant light at 22 °C. After incorporation, the seedlings were washed three times in MS without FucAl and transferred to labeling solution (MS containing 1 mM CuSO₄, 1 mM ascorbic acid, 0.1 μM Alexa-488 azide) (Invitrogen) at 25 °C in the dark for 1 h, then seedlings were washed three times with MS before confocal imaging (16).

To measure incorporation of label into polysaccharides, the labeled seedlings were homogenized and sequentially extracted with water, 70% ethanol (EtOH), 1:1 chloroform:methanol, 50 mM CDTA, and 8 M urea. Fluorescence of each fraction was measured (excitation = 585 nm, emission = 635 nm) (Table S1). Extracted polysaccharides were digested with endo-polygalacturonase, endo-arabinanase, endo-galactanase, exo-1,3-galactanase, pectin methylesterase, or xyloglucan-specific endoglucanase, either singly or in combinations described in the text and Table S2. The molecular-weight distribution of the solubilized digestion products was determined by fractionating each supernatant on a Sephadex G-75 gel filtration column.

See SI Materials and Methods for additional details.