Plant cell wall imaging by metabolic click‐mediated ... 2016 Plant cell wall imaging by... ·...

TECHNICAL ADVANCE

Plant cell wall imaging by metabolic click-mediated labellingof rhamnogalacturonan II using azido 3-deoxy-D-manno-oct-2-ulosonic acid

Marie Dumont1, Arnaud Lehner1, Boris Vauzeilles2,3,4, Julien Malassis5, Alan Marchant6, Kevin Smyth5,†, Bruno Linclau5,

Aur�elie Baron2, Jordi Mas Pons2, Charles T. Anderson7, Damien Schapman8, Ludovic Galas8, Jean-Claude Mollet1 and

Patrice Lerouge1,*1Laboratoire Glycobiologie et Matrice Extracellulaire V�eg�etale (Glyco-MEV), EA 4358, IRIB, VASI, Normandie Universit�e,

76821, Mont-Saint-Aignan, France,2Institut de Chimie des Substances Naturelles (ICSN) UPR CNRS 2301, 91198, Gif-sur-Yvette, France,3Institut de Chimie Mol�eculaire et des Mat�eriaux d’Orsay (ICMMO) UMR CNRS 8182, Universit�e de Paris Sud, 91405, Orsay,

France,4Click4Tag, Zone Luminy Biotech, Case 922, 163 Avenue de Luminy, 13009, Marseille, France,5Chemistry, University of Southampton, Southampton SO17 1BJ, UK,6Centre for Biological Sciences, University of Southampton, Southampton SO17 1BJ, UK,7Department of Biology and Center for Lignocellulose Structure and Formation, The Pennsylvania State University,

University Park, PA, USA, and8PRIMACEN, IRIB, Normandie Universit�e, 76821, Mont-Saint-Aignan, France

Received 3 June 2015; revised 23 November 2015; accepted 27 November 2015; published online 16 December 2015.

*For correspondence (e-mail [email protected]).†Present address University of Bristol, Life Sciences Building, 24 Tyndall Avenue, Bristol, UK.

SUMMARY

In plants, 3-deoxy-D-manno-oct-2-ulosonic acid (Kdo) is a monosaccharide that is only found in the cell wall

pectin, rhamnogalacturonan-II (RG-II). Incubation of 4-day-old light-grown Arabidopsis seedlings or tobacco

BY-2 cells with 8-azido 8-deoxy Kdo (Kdo-N3) followed by coupling to an alkyne-containing fluorescent

probe resulted in the specific in muro labelling of RG-II through a copper-catalysed azide–alkyne cycloaddi-

tion reaction. CMP-Kdo synthetase inhibition and competition assays showing that Kdo and D-Ara, a precur-

sor of Kdo, but not L-Ara, inhibit incorporation of Kdo-N3 demonstrated that incorporation of Kdo-N3 occurs

in RG-II through the endogenous biosynthetic machinery of the cell. Co-localisation of Kdo-N3 labelling with

the cellulose-binding dye calcofluor white demonstrated that RG-II exists throughout the primary cell wall.

Additionally, after incubating plants with Kdo-N3 and an alkynated derivative of L-fucose that incorporates

into rhamnogalacturonan I, co-localised fluorescence was observed in the cell wall in the elongation zone of

the root. Finally, pulse labelling experiments demonstrated that metabolic click-mediated labelling with

Kdo-N3 provides an efficient method to study the synthesis and redistribution of RG-II during root growth.

Keywords: rhamnogalacturonan-II, click chemistry, Kdo-N3, cell wall, Arabidopsis, BY-2 cells, technical advance.

INTRODUCTION

Expression of target proteins fused to green fluorescent

protein tags is routinely used in cell biology to study pro-

tein localisation in living cells. However, this efficient strat-

egy cannot be extended to glycans and lipids. The use of

bio-orthogonal chemical reporters offers an alternative for

in vivo labelling of biomolecules without the requirement

for genetic manipulation. In this method, a chemical repor-

ter is incorporated into the target biomolecule using the

endogenous biosynthetic machinery of the cell. This repor-

ter can then be covalently coupled in a highly selective and

efficient reaction at room temperature to an exogenously

delivered probe. The copper-catalysed azide–alkyne

© 2015 The AuthorsThe Plant Journal © 2015 John Wiley & Sons Ltd

437

The Plant Journal (2016) 85, 437–447 doi: 10.1111/tpj.13104

cycloaddition, a type of so-called ‘click’ chemistry, is the

most widely used click-mediated labelling method in biol-

ogy (Rostovtsev et al., 2002; Tornøe et al., 2002). An azide

chemical reporter is usually selected for the step of incor-

poration into biomolecules because of its small size, meta-

bolic stability and the lack of reactivity with naturally

occurring biomolecules. Labelling is then performed by

copper-catalysed linkage to reporter molecules carrying

terminal alkynes. The study of cell-surface glycans has

benefited greatly from the emergence of click chemistry

for use in cell biology. These experiments have mostly

been performed in vertebrates and bacteria by metabolic

glycan labelling of cells, using a modified monosaccharide

containing a bio-orthogonal chemical reporter. Chemical

ligation to a fluorescent probe then enables the visualisa-

tion of glycoconjugates that accumulate at the cell surface

(Chang et al., 2009; Dehnert et al., 2011; Dumont et al.,

2012; Mas Pons et al., 2014).

Although click chemistry has been widely used in animal

and bacterial biology, this approach has received limited

attention in plant biology to date. Click chemistry was

recently used to label cells undergoing DNA synthesis

while following a Golgi protein marker fused to a fluores-

cent protein (Bourge et al., 2015). Several recent studies

have reported the imaging of plant cell wall components

using a click chemistry approach and two groups have

recently reported the imaging of plant cell wall lignification

using different azido- and alkynyl-tagged monolignol ana-

logues (Bukowski et al., 2014; Tobimatsu et al., 2014; Pan-

dey et al., 2015). In addition, an alkyne derivative of L-

fucose (Fuc-Al) has been used for cell wall labelling in Ara-

bidopsis thaliana (Anderson et al., 2012). In this study, the

authors demonstrated that seedlings fed with the peracety-

lated alkyne derivative of L-fucose incorporated this L-Fuc

analogue into cell wall pectic rhamnogalacturonan-I (RG-I)

but not into xyloglucan, the main fucose-containing cell

wall polysaccharide. Chemical ligation to a fluorescent

azide-containing probe enabled labelling of primary cell

walls of the roots of Arabidopsis seedlings. This fucose–alkyne-based labelling strategy was also used to investi-

gate the secretion of RG-I in mutants lacking the kinesin

FRAGILE FIBER 1 (FRA1) that is defective in the secretion

of cell wall polysaccharides (Zhu et al., 2015).

Pectins are composed of a-1,4-linked homogalacturonan

(HG) and RG-I, which consists of arabinan and galactan

chains linked to a [?2)-a-L-Rhap-(1?4)-a-D-GalpA-(1?]

repeating disaccharide backbone. A third pectic molecule,

rhamnogalacturonan II (RG-II), represents only about 2–4%of the primary cell wall in plants and is found in the pri-

mary walls of angiosperms, gymnosperms, lycophytes and

pteridophytes (O’Neill et al., 2004; Bar-Peled et al., 2012).

RG-II is composed of a HG backbone that is substituted

with four oligosaccharide side chains (Figure 1a) (O’Neill

et al., 2004). At least 12 different glycosyl residues are pre-

sent in RG-II, including several rare sugars such as aceric

acid, apiose, 3-deoxy-D-lyxo-hept-2-ulosaric acid and 3-

deoxy-D-manno-oct-2-ulosonic acid (Kdo) (O’Neill et al.,

2004). Despite its complex structure, RG-II is evolutionarily

conserved in the plant kingdom and is present in the pri-

mary cell wall predominantly in the form of a dimer that is

cross-linked by a borate di-ester bond between two apiosyl

residues (Kobayashi et al., 1996; O’Neill et al., 1996). The

demonstration that RG-II exists in primary walls as a dimer

was a major advance in the understanding of the struc-

ture–function relationship of this cell wall polysaccharide.

Dimer formation results in the covalent cross-linking of

two HG chains involved in the in muro formation of a pec-

tic network. This network contributes to the mechanical

properties of the primary wall and is required for normal

plant growth and development. Boron deficiency or defects

in RG-II biosynthesis result in altered plant growth and

changes in cell wall architecture, demonstrating the crucial

function of boron-mediated RG-II cross-linking in generat-

ing the covalently cross-linked pectic network (Fleischer

et al., 1998, 1999; O’Neill et al., 2001; Voxeur et al., 2011;

Dumont et al., 2014). Nevertheless, the precise function of

RG-II remains unclear to date. Moreover, despite its impor-

tance in primary cell wall synthesis and elongation, little is

known about RG-II biosynthesis and in planta localisation

and turnover. Carbohydrate-binding proteins, such as lec-

tins and antibodies, have been demonstrated to be highly

valuable tools for the investigation of the complexity and

diversity of RG-I, HG and hemicelluloses at the cell and tis-

sue levels (Lee et al., 2011). With the exception of rabbit

polyclonal and Fab antibodies reported by Matoh et al.

(1998) and Williams et al. (1996), respectively, most of the

efforts to obtain specific antibodies against RG-II for

immunocytochemistry studies have been unsuccessful. As

a consequence, appropriate tools for studying the synthe-

sis and distribution of RG-II within the plant cell are lack-

ing.

As well as being a constituent of RG-II in plants, Kdo is

also present in the inner core of the lipopolysaccharide

(LPS) of Gram-negative bacteria. As in bacteria, in plants

this sugar is synthesised in the cytosol by the isomerisa-

tion of D-ribulose-5-P (D-Rul-5-P) into D-arabinose-5-P (D-

Ara-5-P), followed by its coupling to phosphoenol pyruvate

(PEP) to yield Kdo-8-phosphate (Kdo-8-P) (Matsuura et al.,

2003; Delmas et al., 2008) (Figure 1b). After dephosphory-

lation of Kdo-8-P and coupling to CMP by a CMP-Kdo syn-

thetase (Royo et al., 2000), the activated CMP-Kdo sugar

nucleotide is transported into the Golgi apparatus and

transferred to the RG-II backbone, probably by sialyl-like

transferases (Deng et al., 2010; Dumont et al., 2014). 8-

Azido 8-deoxy Kdo (Kdo-N3; Figure 1c) was recently used

to detect living bacteria through the labelling of LPS

(Dumont et al., 2012). This study demonstrated that Kdo-

N3 was efficiently utilised by the bacterial glycosylation

© 2015 The AuthorsThe Plant Journal © 2015 John Wiley & Sons Ltd, The Plant Journal, (2016), 85, 437–447

438 Marie Dumont et al.

machinery. Moreover, modification of the C8 position of

Kdo with an azido group prevents reverse metabolism by

Kdo-8-P phosphatase and incorporation of the azido moi-

ety into other cell metabolites, which might result in a loss

of labelling specificity.

In this paper we report the metabolic labelling of plant

cell walls using Kdo-N3 (Figure 1c). We show that the azido

Kdo analogue can be metabolically integrated into the pec-

tin-rich cell walls of 4-day-old light-grown Arabidopsis

seedlings and Nicotiana tabacum L. cv Bright Yellow-2 (to-

bacco BY-2) cells and can then be efficiently coupled to an

alkyne-containing fluorescent probe via a copper-catalysed

azide–alkyne cycloaddition, allowing for specific metabolic

labelling of RG-II.

RESULTS

Metabolic click-mediated labelling of plant tissues was

investigated with the 8-azido Kdo analogue because this

monosaccharide is specific for RG-II and the Kdo C-8

hydroxy group is not involved in linkage to the rhamnose

residue in the Kdo-containing side chain of RG-II (Fig-

ure 1a). Four-day-old light-grown Arabidopsis seedlings

were incubated for 16 h in liquid Murashige and Skoog

(MS) medium containing from 0 to 50 lM Kdo-N3. Seed-

lings were then transferred for 1 h to a solution of Alexa

Fluor� 488-alkyne, a membrane-impermeable probe, in

copper-catalysing conditions before observation of fluo-

rescence in cell walls by confocal microscopy. As shown

in Figure 2(a–g), seedlings treated with Kdo-N3 exhibited

a strong labelling of their root cell walls. The highest

labelling intensity was obtained with 50 lM Kdo-N3 (Fig-

ure 2d–g). X–Z projections of z-series showed that the

labelling is restricted to the root surface (Figure 2f), indi-

cating that the fluorescent probe did not diffuse within

the root tissues. The labelling occurs throughout all the

primary root epidermal cell walls in the division and elon-

gation zones of seedling roots, as well as in root hairs

(Figure 2b) and the developing root hair bulges (Fig-

ure 2e). Labelling experiments were also carried out on

suspension cultured tobacco BY-2 cells. As illustrated in

(a)

(b) (c)

Figure 1. Schematic representation of RG-II and

biosynthetic pathway of CMP-Kdo. (a) Structure of

rhamnogalacturonan II (RG-II) with the location of

3-deoxy-D-manno-oct-2-ulosonic acid (Kdo) indi-

cated in a dashed box.

(b) Biosynthetic pathway of CMP-Kdo.

(c) Structure of the 8-azido 8-deoxy Kdo (Kdo-N3).


Rhamnogalacturonan-II imaging using click chemistry 439

Figure 2(h,i), strong labelling of the cell walls of tobacco

cells was observed.

Control experiments were performed to demonstrate the

specificity of the metabolic click-mediated labelling of the

cell wall. No labelling of cell walls was observed in seed-

lings treated with Kdo-N3 or the fluorescent probe alone

(Figure S1 in Supporting Information). To investigate

whether cell viability is required for incorporation of Kdo-

N3, seedlings were treated with the Kdo analogue before

or after fixation with 4% paraformaldehyde and then

(a) (b) (c)

(d) (e)

(h) (i)

(f)

(g)

Figure 2. Confocal microscope images of roots of 4-day-old light-grown Arabidopsis thaliana seedlings and tobacco BY-2 cells.

(a)–(e) Seedlings were incubated for 16 h in MS medium containing from 0 to 50 lM 8-azido 8-deoxy Kdo (Kdo-N3; Kdo, 3-deoxy-D-manno-oct-2-ulosonic acid)

and treated with Alexa Fluor� 488 alkyne in copper-catalysed conditions. In the upper part of the elongation zone (e), the emerging sites of root hairs are

strongly labelled. Scale bars = 100 lm (a, b, c) and 50 lm (d, e). (f) X–Z projections of z-series of epidermal cells showing that the labelling is restricted to the

root surface. Scale bars = 50 lm. (g) Overall view of a Kdo-N3-labelled root. Scale bar = 100 lm. (h), (i) Tobacco BY-2 cells were collected 3 days after subculture

and incubated in liquid MS medium containing 0 lM (h) or 100 lM Kdo-N3 (i) for 16 h and then labelled with 0.1 lM Alexa 488-alkyne for 1 h in copper-catalysed

conditions. Scale bars = 20 lm.

Inserts correspond to the transmitted light images.



labelled with Alexa Fluor� 488-alkyne. Seedlings treated

with Kdo-N3, then fixed before coupling to the fluorophore

tag, exhibited strong cell wall labelling. In contrast, no flu-

orescence of the root tissues was observed in seedlings

fixed prior to labelling, thus demonstrating that the incor-

poration of Kdo-N3 requires living tissue (Figure 3a,b).

Click-mediated labelling of Arabidopsis seedlings was then

performed in the presence of 2b-deoxy Kdo, an inhibitor of

cytosolic CMP-Kdo synthetase (Figure 1b) (Smyth et al.,

2013). In these conditions, a strong decrease in cell wall

labelling was observed compared with the control (Fig-

ure 4a). This observation supports the hypothesis that

incorporation of Kdo-N3 in the cell wall occurs through the

endogenous Kdo biosynthetic machinery. Together, these

data indicate that the fluorescence of click-labelled cell

walls in roots of Arabidopsis seedlings results from the

incorporation of Kdo-N3 into RG-II, the sole Kdo-containing

cell wall polysaccharide. This probably occurs through the

endogenous metabolism of living plant cells and not from

non-specific adsorption of Kdo-N3 to the cell wall.

D-Ara-5-P is the metabolic precursor of Kdo in plants

(Figure 1b). In the presence of D-Ara, metabolic click-label-

ling of Arabidopsis seedlings exposed to Kdo-N3 was

strongly inhibited, whereas no inhibition was observed

with L-Ara, the isomer found in several cell wall polysac-

charides (Figure 4b). Inhibition of labelling was also

observed in seedlings co-incubated with Kdo (Figure 4b).

This competitive inhibition observed with D-Ara and Kdo

further confirmed that incorporation of Kdo-N3 occurred

through the Kdo biosynthesis pathway. Root elongation

was not affected in these experiments (Figure 4c), indicat-

ing that inhibition of Kdo-N3 incorporation was not due to

the toxicity of D-Ara or Kdo.

To investigate the co-localisation of the observed Kdo-

N3 labelling with cellulose in primary cell walls, seedlings

were treated with Kdo-N3, labelled with Alexa Fluor� 488-

alkyne and stained with the glucan-binding calcofluor

white, a fluorescent dye that labels cellulose (Figure 5).

Dual-colour confocal imaging shows that newly synthe-

sised RG-II and cellulose co-localise throughout the cell

walls of the root at the limit of resolution of this technique.

Interestingly, the pattern of the fluorescence of Kdo-N3 and

calcofluor is different in radial and longitudinal walls, with

a more intense fluorescence of Kdo-N3 in longitudinal

walls (Figure 5a) and a more intense signal of calcofluor in

radial walls (Figure 5b). This may indicate that pectins are

synthesised in a greater quantity in the longitudinal grow-

ing walls compared with the radial, non-growing walls.

Double labelling was tested using Fuc-Al and Kdo-N3 to

take advantage of the fact that these two sugar analogues

carry different bio-orthogonal chemical reporters. Fuc-Al

has been used for imaging of the cell wall RG-I in A. thali-

ana (Anderson et al., 2012). Four-day-old light-grown Ara-

bidopsis seedlings were incubated for 16 h in MS medium

containing 50 lM of Kdo-N3 and 50 lM of peracetylated

Fuc-Al. Coupling to Alexa Fluor� 488-alkyne (Figure 6a)

and Alexa Fluor� 594-azide (Figure 6b) fluorescent probes

enabled double labelling of the cell wall (Figure 6c). Co-

localisation of the labelling (Figure 6e) using the two

monosaccharide analogues shows that newly synthesised

RG-I and RG-II are deposited with similar distributions in

root cell walls.

Pulse labelling experiments (Figure 7) were performed

by incubating Arabidopsis seedlings with Kdo-N3, washing

out free analogue, and then incubating in analogue-free

medium for 22 h to investigate RG-II synthesis and redistri-

bution in muro. After 4 h of incubation with Kdo-N3, the

pattern of incorporation was homogeneous in the walls of

cells in the elongation zone (Figure 7a). After 22 h, only

the apical/basal borders of the elongated cells showed a

strong labelling in the elongation zone (Figure 7c). In con-

trast, the labelling of the longitudinal/lateral cell walls was

weak and diffuse, revealing a dilution of Kdo-N3-containing

RG-II during the elongation of the cells. Cells in the differ-

entiation zone are not expected to greatly expand or add

large amounts of new RG-II to their walls. This is reflected

by a lack of labelling in the differentiation zone after the

4 h incorporation (Figure 7b). In contrast, the cells in the

(a) (b)Figure 3. Control experiments for 8-azido 8-deoxy

Kdo (Kdo-N3; Kdo, 3-deoxy-D-manno-oct-2-ulosonic

acid) incorporation.

Chemical fixation of seedlings using 4%

paraformaldehyde after (a) or before (b) Kdo-N3

incorporation and labelling. Seedling viability is

required for Kdo-N3 incorporation but not for the

click chemistry reaction. Inserts correspond to the

transmitted light images. Scale bars = 100 lm.



differentiation zone after 22 h, which were in the late elon-

gation zone during the pulse of Kdo-N3 incubation, dis-

played homogeneous labelling (Figure 7d). These results

suggest that when root cells become differentiated, RG-II

synthesis is reduced.

DISCUSSION

Several lines of evidence indicate that incubation of Ara-

bidopsis seedlings or tobacco BY-2 cells with Kdo-N3, fol-

lowed by coupling to an alkyne-containing fluorescent

probe in the presence of Cu(I), results in the specific in

muro labelling of RG-II. Firstly, Kdo is only found in RG-II.

Secondly, click-mediated labelling of cell walls is abolished

in presence of 2b-deoxy Kdo, an inhibitor of the cytosolic

CMP-Kdo synthetase that is able to activate free Kdo into

its nucleotide sugar in the cytosol (Misaki et al., 2009;

Smyth et al., 2013). Thirdly, incorporation of Kdo-N3 only

occurs in living tissues and can be inhibited by addition of

Kdo and D-Ara. D-Ara is only involved in plant cell wall

metabolism as a precursor of Kdo in the RG-II biosynthetic

pathway (Figure 1b). In contrast, its L-isomer is widely dis-

tributed in arabinan, arabinoxylan and proteoglycan cell

wall polymers (Carpita and Gibeaut, 1993). The inhibition

of labelling with Kdo-N3 upon feeding seedlings with D-

Ara suggests that D-Ara is likely to enter cellular metabo-

lism via the salvage pathway where it can compete with D-

Ara-5-P for cytosolic Kdo synthesis (Figure 1b). Plant Kdo-

8-P synthetases have been reported to be specific for D-

Ara-5-P, which arises from the cytosolic isomerisation of

D-ribulose-5-P (Figure 1b). Activity of these synthetases

was studied by complementation assays of bacterial

strains impaired in endogenous Kdo-8-P synthetase activity

(Brabetz et al., 2000; Delmas et al., 2003; Matsuura et al.,

2003) or using D-Ara-5-P as a substrate in enzyme bioas-

says (Delmas et al., 2003). Inhibition of Kdo-N3 incorpora-

tion with D-Ara suggests that phosphorylation of D-Ara is

not required for its coupling to PEP by cytosolic Kdo-8-P

synthetases (Figure 1b).

Click-mediated metabolic labelling using Kdo-N3

resulted in efficient imaging of RG-II in the Arabidopsis

root primary cell wall, thus providing an alternative to the

use of antibodies for studying RG-II synthesis and distribu-

tion. Labelling is observed in cells located in the division

and elongation zones where pectin synthesis occurs.

Immunolocalisation studies using antibodies raised

against RG-II have shown that labelling was stronger in the

vicinity of the plasma membrane. However, cell walls trea-

ted with alkali to promote de-esterification, showed homo-

geneous labelling of the entire cell wall. This suggests that

the apparent anisotropic distribution of RG-II in primary

cell walls resulted from an increasing gradient in RG-II

esterification from the inner to the outer faces of the cell

wall since the antibodies used were raised against de-

esterified RG-II (Williams et al., 1996; Matoh et al., 1998). In

our study, co-localisation of labelled Kdo-N3 with the cellu-

lose-binding dye calcofluor white confirmed that RG-II

exists throughout Arabidopsis primary cell walls. Incuba-

Figure 4. Inhibition of 8-azido 8-deoxy Kdo (Kdo-N3; Kdo, 3-deoxy-D-

manno-oct-2-ulosonic acid) incorporation.

(a) Quantification of root-tip fluorescence in 4-day-old seedlings grown on

MS medium supplemented with or without 150 lM 2b-deoxy-Kdo. Seedlingswere then transferred in liquid MS medium containing 50 lM Kdo-N3 with

or without 150 lM 2b-deoxy-Kdo. Data are expressed � SEM (n > 5 seed-

lings from two experiments). Similar letters indicate that no significant dif-

ference was observed whereas different letters indicate a highly significant

difference between the conditions (Mann–Whitney test, P < 0.01).

(b) Quantification of the root-tip fluorescence in seedlings treated for 16 h

with 50 lM Kdo-N3 with or without 25 mM Kdo, L-arabinose (L-Ara) or D-Ara.

Error bars represent the SEM (n = 10–12 seedlings for each experiment).

Data are the mean of at least two experiments. Similar letters indicate that

no significant difference was observed whereas different letters indicate a

highly significant difference between the conditions (one-way ANOVA test,

P < 0.05).

(c) Root length of 4-day-old seedlings grown in liquid MS alone or with

25 mM Kdo, D-Ara or L-Ara for 16 h. Error bars represent the SEM (n = 12

seedlings). No significant difference was observed for the root elongation

(Mann–Whitney test, P > 0.05).



tion of tobacco BY-2 cells with Kdo-N3 also resulted in an

efficient labelling of the primary cell wall. It should be

mentioned that labelling assays of BY-2 cells using Fuc-AI

were unsuccessful (Figure S2), and as a consequence Kdo-

N3 is the unique probe for cell wall click labelling experi-

ments in this widely used plant cell model.

Imaging of RG-I in the cell walls of roots of Arabidopsis

seedlings was previously reported using the alkyne deriva-

tive of L-fucose Fuc-Al as a probe (Anderson et al., 2012).

In our study, double labelling was tested using Fuc-Al and

Kdo-N3, taking advantage of the fact that these two sugar

analogues carry different bio-orthogonal chemical repor-

ters. Co-localisation of the labelling using the two

monosaccharide analogues suggests that newly synthe-

sised RG-I and RG-II molecules are deposited with a similar

spatial distribution in root cell walls. Pulse labelling experi-

ments were used to track the fate of RG-II in the cell wall

during elongation. After 22 h, our results indicated that

expansion of lateral cell walls during elongation led to the

dilution of RG-II signal due to wall expansion and the deliv-

ery of newly synthesised, unlabelled cell wall polymers.

Although the click labelling strategy appears to be a

good alternative to the use of antibodies for studying cell

wall polysaccharides, this strategy has some limitations.

Several monosaccharides are present in multiple polymer

types, and click-mediated labelling using analogues of

these monosaccharides carrying an azide or alkyne bio-

orthogonal chemical reporter may not be restricted to a

single cell wall polysaccharide. In this respect, Kdo-N3 is

appropriate for pectin studies since Kdo is only found in

RG-II. Furthermore, Kdo-N3 is efficiently incorporated into

both root tissues and suspension-cultured cells. However,

cell wall imaging is restricted to the root epidermis of tis-

sues that are able to efficiently incorporate exogenous

sugar analogues (Figure 2f) and/or because of a lack of

diffusion of the fluorophore in the internal tissues. In addi-

tion, the high cytotoxicity of copper ions required for the

azide–alkyne cycloaddition limits its use in living cells –thus pulse labelling experiments have to be performed in

order to investigate the dynamics of labelling during root

growth (Anderson et al., 2012). To circumvent this limita-

tion, copper-free click reactions have recently been devel-

oped using a ring-strained cyclooctyne moiety. This

approach allows labelling to be performed in living organ-

isms with minimal physiological perturbation (Chang

et al., 2009; Laughlin and Bertozzi, 2009; Boyce and Ber-

tozzi, 2011; Dehnert et al., 2011). However, considering the

large size of cyclooctyne motifs, this biocompatible click

chemistry strategy is likely to require the metabolic incor-

poration of azido sugars. With regard to plant cell wall

imaging using fucose analogues, it was reported that

plant cells were able to incorporate the alkyne derivative

of L-fucose but not the azido analogue (Anderson et al.,

2012). In this respect, the Kdo-N3 analogue is thus far the

only azido sugar reported to be assimilated by plant cells.

Given that the azido sugar is suitable for in vivo labelling

using cyclooctyne-containing fluorescent probes, this

study opens new avenues for following the real-time traf-

ficking of pectins within the cell and their delivery to the

apoplast.

EXPERIMENTAL PROCEDURES

Sugar analogues

Kdo-N3 was synthesised according to Dumont et al. (2012). 2b-Deoxy Kdo was synthesised according to a published procedure

(a) (b) (c)

Figure 5. Co-localisation of labelled rhamnogalacturonan-II (RG-II) and cellulose.

Root epidermal cell in the elongation-zone from a 4-day-old seedling treated with 8-azido 8-deoxy Kdo (Kdo-N3; Kdo, 3-deoxy-D-manno-oct-2-ulosonic acid) and

labelled with Alexa 488-alkyne (a), and stained with the glucan-binding fluorescent dye calcofluor white (b). In the merged image (c), Alexa is green and cal-

cofluor is magenta. Scale bar = 10 lm.



(Claesson, 1987; Lopez-Herrera and Sarabia-Garcia, 1997) that wasslightly modified (see Methods S1).

Plant culture and growth conditions. Arabidopsis Col-0seeds were surface sterilised in 50% (v/v) ethanol for 5 min, in

30% (v/v) sodium hypochlorite for 2 min, washed five times insterile water and incubated at 4°C for at least 48 h. Sterilisedseeds were deposited on MS (2.2 g L�1) mineral medium contain-ing 1% sucrose and 0.8% plant agar. The plates were placed verti-cally in a growth chamber for 4 days at 22°C under a 16-h light/8-hdark cycle. Tobacco BY-2 suspension cells were cultured in MS liq-uid medium (4.3 g L�1, 30 g L�1 sucrose, 100 mg L�1 KH2PO4,51 mg L�1 myo-inositol, 1 mg L�1 thiamine, 0.22 mg L�1 2.4dichlorophenoxyacetic acid, pH 5.6) in the dark at 22°C and140 r.p.m. Cells were subcultured every week by transferring10 ml of suspension cells to 150 m of fresh medium.

Incorporation of Kdo-N3 and fluorescent labelling

Four-day-old light-grown seedlings were transferred fromMSmed-ium plates to 1 ml of liquid MS containing from 2.5 to 50 lM Kdo-N3 for 16 h at 22°C. Seedlings were then washed three times withliquid MS and transferred in 1 ml of liquid MS containing 0.1 lMAlexa Fluor_ 488 alkyne (Thermofisher, http://www.fisher.co.uk/),1 mM CuSO4 and 1 mM ascorbic acid. The solution was incubatedfor 1 h at room temperature (22�C) in the dark. Seedlings were thenwashed once in liquid MS supplemented with 0.01% (v/v) Tween 20for 5 min, three more times in liquid MS and finally with distilledwater. Control experiments were carried out by omitting either theKdo-N3 or the fluorescent probe in the click labelling assays. Inhibi-tion of CMP-Kdo synthetase was carried out by adding 150 lM 2b-deoxy Kdo to the solid MS medium, then complementing the liquidMS with the same concentration of inhibitor as during the Kdo-N3

incorporation step (Smyth et al., 2013). Inhibition assays were car-ried out by adding 25 mM D-Ara, L-Ara or Kdo to the liquid medium.Root lengths of 4-day-old seedlings were measured using IMAGEJsoftware (Abramoff et al., 2004).

For labelling of tobacco BY-2 cells, Kdo-N3 was diluted to200 lM in 0.5 ml of liquid MS medium. The 0.5 ml of a 3-day-oldof suspension of BY-2 cells collected 3 days after subculture inlogarithmic growth was added. Cells were then incubated for16 h. Cells were rinsed three times with liquid MS medium bycentrifugation (2 min, 700 g). Click reactions were performed byincubating BY-2 cells in 500 ll of labelling solution containing0.1 lM Alexa Fluor� 488-alkyne, 1 mM CuSO4 and 1 mM ascorbicacid in liquid MS for 1 h, in the dark with gentle shaking. Cellswere then rinsed once with 500 ll liquid MS, then incubated for5 min in liquid MS supplemented with Tween 20 (0.01%, v/v),rinsed three more times with liquid MS for 5 min each and finallywith water.

Double staining using Kdo-N3 and calcofluor

Four-day-old seedlings treated with Kdo-N3 and labelled withAlexa 488-alkyne were then incubated with 1 mg L�1 calcofluorwhite M2R (Sigma, http://www.sigmaaldrich.com/) for 30 min inthe dark. After careful washing with distilled water, the roots wereobserved.

Double staining using Fuc-Al and Kdo-N3

A 1 mM solution of peracetylated 6-alkyne L-fucose (Fuc-Al; Ther-mofisher, http://www.fisher.co.uk/) was prepared by dissolution inDMSO and then dilution to the desired concentration in MS med-ium. Four-day-old light-grown seedlings were transferred from MSmedium plates to 1 ml of liquid MS containing 50 lM of Fuc-Al and50 lM Kdo-N3 for 16 h at 22°C. Seedlings were washed and thenincubated in 1 ml of labelling solution [liquid MS containing 0.1 lMAlexa Fluor� 594-azide (Thermofisher, http://www.fisher.co.uk/),1 mM CuSO4, 1 mM ascorbic acid] for 1 h at room temperature.

(a) (b)

(c) (d)

(e)

Figure 6. 8-Azido 8-deoxy Kdo (Kdo-N3; Kdo, 3-deoxy-D-manno-oct-2-uloso-

nic acid) and an alkyne derivative of L-fucose (Fuc-Al) are co-localised in the

cell wall.

Four-day-old Arabidopsis seedlings were incubated in liquid MS medium

supplemented with 50 lM Fuc-Al and 50 lM Kdo-N3 for 16 h and then suc-

cessively labelled for 1 h with Alexa Fluor� 594-N3 (A594) and then with

Alexa Fluor� 488-alkyne (A488) in copper-catalysed conditions.

(a) Elongation zone observed with the A488 channel.

(b) Same elongation zone as in (a) observed with the A594 channel.

(c) Elongation zone observed with merged A488 and A594 channels.

(d) Bright-field image of the elongation zone. Scale bars = 100 lm. A488 is

green and A594 is magenta.

(e) Number of co-localised pixels between the A488 channel (a) and the

A594 channel (b) and colour map representation of the intensity of co-locali-

sation between the A488 and A594 channels. In this colour map, the �1 to

+1 heat map depicts the measured Icorr values regarding the colour bar,

where negatively correlated relationships (values between �1 and �0.1) are

shown in dark blue–light blue colours, and positive correlations (values

between 0.1 and 1) are represented by warmer green–red colours.



http://www.fisher.co.uk/

http://www.sigmaaldrich.com/



After one washing step with liquid MS, seedlings were thenlabelled with 1 ml of 0.1 lM Alexa Fluor� 488-alkyne. Seedlingswere washed as described for single labelling and imaged.

Staining of BY-2 cells using Fuc-Al

Tobacco BY-2 cells in the exponential growth phase were incu-bated, with shaking, for 24 h in medium containing 1 lM fucose-alkyne peracetate (‘FucAl’, Life Technologies). The cultures werethen rinsed over a cell strainer with 10 ml liquid BY-2 media[4.3 g L�1 Murashige and Skoog salts (Caisson LaboratoriesMSP0501, http://www.caissonlabs.com/), 100 mg L�1 inositol,1 mg L�1 thiamine, 0.2 mg L�1 2,4-D, 255 mg L�1 KH2P04 and30 g L�1 sucrose] lacking FucAl, then resuspended in click reactionmixture consisting of 1 mM CuSO4, 1 mM ascorbic acid and 1 lMAlexa488-alkyne (Thermofisher, http://www.fisher.co.uk/) for 1 h,with rocking. The reaction was halted by rinsing away the reactionmixture as before. Cells were imaged with a Zeiss Axio Observerspinning disc confocal microscope with a 488-nm excitation laserand a 20 9 0.8 NA objective. Fluorescence images are presentedas maximum z-projections, z-distance between slices: 1 lm.

Pulse labelling experiment

For pulse experiments, 4-day-old Arabidopsis seedlings were incu-bated in 200 lM Kdo-N3 in liquid MS for 4 h, washed three timeswith liquid MS and then transferred to solid MS lacking Kdo-N3.Seedlings were labelled as previously described after 0 and 22 h,washed and imaged.

Microscopy

The roots of labelled seedlings and the BY-2 cells were observedwith the Leica TCS SP2 AOBS confocal microscope (488-nm exci-

tation laser; 495–550 nm emission detection for Alexa 488 and594 nm excitation laser; 610–670 nm emission detection for Alexa594). For the double staining experiment, images were acquired inthe sequential scan mode. Images were acquired with identicalexposure settings. The fluorescence intensity of the root was cal-culated with IMAGEJ by thresholding (pixel intensity >16) to selectthe root boundary and measuring the mean of pixel intensityinside the area defined by the threshold. The ImageJ co-localisa-tion Colormap plugin was used to determine the correlation index(Icorr), which is the fraction of positively correlated pixels betweentwo fluorescent populations. The plug-in is based on the methodoriginally described by Jaskolski et al. (2005).

For the double staining experiment, epidermal root cells wereobserved with the Leica TCS SP8 CFS confocal microscope (488-nm excitation laser, 500–550 nm emission detection for Alexa 488and the 405 nm excitation laser, 420–470 nm emission detectionfor calcofluor). Images were acquired in sequential scanningmode. After acquisition, a deconvolution software, Huygens Pro-fessional (http://www.svi.nl) was used to deblur images.

Statistical analysis

The data were analysed statistically using GRAPHPAD Software(http://www.graphpad.com/). Differences were considered statisti-cally significant when P ≤ 0.05.

ACKNOWLEDGEMENTS

This work was supported by the University of Rouen, the FrenchAgency for Research (grant no. ANR-11-BSV5-0007) and the ‘TransChannel Wallnet’ project that has been selected in the context ofthe INTERREG IVA France (Channel)–England European cross-bor-der cooperation programme, which is co-financed by the Euro-pean Regional Development Fund. Work at The Pennsylvania

(a) (b)

(c) (d)

Figure 7. Pulse labelling of the cell wall.

Four-day-old Col-0 seedlings were treated with a 4-

h pulse of 200 lM 8-azido 8-deoxy Kdo (Kdo-N3;

Kdo, 3-deoxy-D-manno-oct-2-ulosonic acid),

washed, plated on MS plates without Kdo-N3 for

0 h (a, b) and 22 h (c, d) and labelled with Alexa

Fluor� 488-alkyne. Parts (a) and (c) correspond to

the elongation zone and (b) and (d) correspond to

the differentiation zone of root epidermal cells.

Scale bars = 25 lm.



http://www.caissonlabs.com/


http://www.svi.nl

http://www.graphpad.com/

State University was supported by The Center for LignocelluloseStructure and Formation, an Energy Frontier Research Centerfunded by DOE, Office of Science, BES under award no. DE-SC0001090. The authors are grateful to the Grand R�eseau deRecherche VASI de Haute Normandie for the use of equipment, toGr�egory Mouille for technical advice and to Daniel McClosky forthe experiments on BY-2 cells.

SUPPORTING INFORMATION

Additional Supporting Information may be found in the online ver-sion of this article.Figure S1. Control experiments for Kdo-N3 labelling.

Figure S2. Tobacco BY-2 cells do not incorporate appreciableamounts of fucose-alkyne into their cell walls.

Methods S1. Experimental procedure for the synthesis of 2b-deoxy Kdo.

REFERENCES

Abramoff, M.D., Magalh~aes, P.J. and Ram, S.J. (2004) Image processing

with ImageJ. Biophotonics Int. 11, 36–42.Anderson, C., Wallace, I.S. and Somerville, C.R. (2012) Metabolic click-label-

ling with a fucose analog reveals pectin delivery, architecture, and

dynamics in Arabidopsis cell walls. PNAS, 109, 1329–1334.Bar-Peled, M., Urbanowica, B.R., O Neill, M. (2012) The synthesis and origin

of the pectic polysaccharide rhamnogalacturonan ll – insights from

nucleotide sugar formation and diversity. Front Plant Sci. 3, 92.

Bourge, M., Fort, C., Soler, M.-N., Satiat-Jeunemaıtre, B. and Brown, S.C.

(2015) A pulse-chase strategy combining click-EdU and photoconvertible

fluorescent reporter: tracking Golgi protein dynamics during the cell

cycle. New Phytol. 205, 938–950.Boyce, M. and Bertozzi, C.R. (2011) Bringing chemistry to life. Nat. Methods,

8, 638–642.Brabetz, W., Wolter, F.P. and Brade, H. (2000) A cDNA encoding 3-deoxy-D-

manno-oct-2-ulosate-8-phosphate synthetase of Pisum sativum L. (pea)

functionally complements a kdsA mutant of the Gram-negative bac-

terium Salmonella enterica. Planta, 212, 136–143.Bukowski, N., Pandey, J.L., Doyle, L., Richard, T.L., Anderson, C.T. and

Zhu, Y. (2014) Development of a clickable designer monolignol for

interrogation of lignification in plant cell walls. Bioconjug. Chem. 25,

2189–2196.Carpita, N.C. and Gibeaut, D.M. (1993) Structural models of primary cell

walls in flowering plants: consistency of molecular structure with the

physical properties of the walls during growth. Plant J. 3, 1–30.Chang, P.V., Chen, X., Smyrniotis, C., Xenakis, A., Hu, T., Bertozzi, C.R. and

Wu, P. (2009) Metabolic labeling of sialic acids in living animals with

alkynyl sugars. Angew. Chem. 48, 4030–4033.Claesson, A. (1987) New synthesis of a CMP-KDO synthetase inhibitor and

of 2-deoxy-KDO derivatives used in the synthesis of such inhibitors. J.

Org. Chem. 52, 4414–4416.Dehnert, K.W., Beahm, B.J., Huynh, T.T., Baskin, J.M., Laughlin, S.T.,

Wang, W., Wu, P., Amacher, S.L. and Bertozzi, C.R. (2011) Metabolic

labeling of fucosylated glycans in developing zebrafish. ACS Chem.

Biol. 6, 547–552.Delmas, F., Petit, J., Joubes, J., S�eveno, M., Paccalet, T., Hernould, M., Ler-

ouge, P., Mouras, M. and Chevalier, C. (2003) The gene expression and

enzyme activity of plant 3-deoxy-D-manno-2-octulosonic acid-8-phos-

phate synthase are preferentially associated with cell division in a cell

cycle-dependent manner. Plant Physiol. 133, 348–360.Delmas, F., S�eveno, M., Northey, J.G.B., Hernould, M., Lerouge, P., Mc

Court, P. and Chevalier, C. (2008) The synthesis of the rhamnogalacturo-

nan II component 3-deoxy-D-manno-2-octulosonic acid (Kdo) is required

for pollen tube growth and elongation. J. Exp. Bot. 59, 2639–2647.Deng, Y., Wang, W., Li, W.-Q., Xia, C., Liao, H.-Z., Zhang, X.-Q. and Ye, D.

(2010) MALE GAMETOPHYTE DEFECTIVE 2, encoding a sialyltransferase-

like protein, is required for normal pollen germination and pollen tube

growth in Arabidopsis. J. Integr. Plant Biol. 52, 829–843.Dumont, A., Malleron, A., Awwad, M., Dukan, S. and Vauzeilles, B. (2012)

Click-mediated labeling of bacterial membranes through metabolic modi-

fication of the lipopolysaccharide inner core. Angew. Chem. Int. Ed. 51,

3143–3146.Dumont, M., Lehner, A., Bouton, S., Kiefer-Meyer, M.C., Voxeur, A., Pelloux,

J., Lerouge, P. and Mollet, J.C. (2014) The cell wall pectic polymer

rhamnogalacturonan-II is required for proper pollen tube elongation:

implication of a putative sialyltransferase-like protein. Ann. Bot. 114,

1177–1188.Fleischer, A., Titel, C. and Ehwald, R. (1998) The boron requirement and cell

wall properties of growing and stationary suspension-cultured Chenopo-

dium album L. cells. Plant Physiol. 117, 1401–1410.Fleischer, A., O’Neill, M.A. and Ehwald, R. (1999) The pore size of non-gra-

minaceous plant cell walls is rapidly decreased by borate ester cross-

linking of the pectic polysaccharide rhamnogalacturonan II. Plant Phys-

iol. 121, 829–838.Jaskolski, F., Mulle, C. and Manzoni, O.J. (2005) An automated method to

quantify and visualize colocalized fluorescent signals. J. Neurosci. Meth-

ods, 146, 42–49.Kobayashi, M., Matoh, H. and Azuma, J. (1996) Two chains of rhamnogalac-

turonan II are cross-linked by borate-diol ester bonds in higher plant cell

walls. Plant Physiol. 110, 1017–1020.Laughlin, S.T. and Bertozzi, C.R. (2009) In vivo imaging of Caenorhabditis

elegans glycans. ACS Chem. Biol. 4, 1068–1072.Lee, K.J., Marcus, S.E. and Knox, P. (2011) Cell wall biology: perspective

from cell wall imaging. Mol. Plant, 4, 212–219.Lopez-Herrera, F.J. and Sarabia-Garcia, F. (1997) Condensation of D-manno-

saldehyde derivatives with ethyl diazoacetate. An easy and stereoselec-

tive chain elongation methodology for carbohydrates: application to new

syntheses for KDO and 2-deoxy-beta-KDO. Tetrahedron, 53, 3325–3346.Mas Pons, J., Dumont, A., Sautejeau, G., Fugier, E., Baron, A., Dukan, S.

and Vauzeilles, B. (2014) Identification of living Legionella pneumophila

using species-specific metabolic lipopolysaccharide labeling. Angew.

Chem. Int. Ed. 53, 1275–1278.Matoh, T., Takasaki, M., Takabe, K. and Kobayashi, M. (1998) Immunocyto-

chemistry of rhamnogalactouronan II in cell walls of higher plants. Plant

Cell Physiol. 39, 483–491.Matsuura, K., Miyagawa, I., Kobayashi, M., Ohta, D. and Matoh, T. (2003)

Arabidopsis 3-deoxy-D-manno-oct-2-ulosonate-8-phosphate synthase:

cDNA cloning and expression analyses. J. Exp. Bot. 54, 1785–1787.Misaki, R., Kajiura, H., Fujii, K., Fujiyama, K. and Seki, T. (2009) Cloning and

characterization of cytidine monophosphate-3-deoxy-D-manno-octuloso-

nate synthetase from Arabidopsis thaliana. J. Biosci. Bioeng. 108, 527–529.

O’Neill, M.A., Warrenfeltz, D., Kates, K., Pellerin, P., Doco, T., Darvill, A.G.

and Albersheim, P. (1996) Rhamnogalacturonan-II, a pectic polysaccha-

ride in the walls of growing plant cell, forms a dimer that is covalently

cross-linked by a borate ester. J. Biol. Chem. 271, 22923–22930.O’Neill, M.A., Eberhard, S., Albersheim, P. and Darvill, A.G. (2001) Require-

ment of borate cross-linking of cell wall rhamnogalacturonan II for Ara-

bidopsis growth. Science, 294, 846–849.O’Neill, M.A., Ishii, T., Albersheim, P. and Darvill, A.G. (2004) Rhamnogalac-

turonan II: structure and function of a borate cross-linked cell wall pectic

polysaccharide. Annu. Rev. Plant Biol. 55, 109–139.Pandey, J.L., Wang, B., Diehl, B.G., Richard, T.L., Chen, G. and Anderson,

C.T. (2015) A versatile click-compatible monolignol probe to study lignin

deposition in plant cell walls. PLoS One, 10, e0121334.

Rostovtsev, V.V., Green, L.G., Fokin, V.V. and Sharpless, K.B. (2002) A step-

wise Huisgen cycloaddition process: copper(I)-catalyzed regioselective

“ligation” of Aazides and terminal alkynes. Angew. Chem. 41, 2596–2599.

Royo, J., Gomez, E. and Hueros, G. (2000) A maize homologue of the bacte-

rial CMP-Kdo synthetases: similar pathways operate in plants and bacte-

ria for the activation of Kdo prior to its incorporation into cellular

envelopes. J. Biol. Chem. 275, 24993–24999.Smyth, K.M., Mikolajek, H., Werner, J.M. and Marchant, A. (2013) 2b-deoxy-

Kdo is an inhibitor of the Arabidopsis thaliana CMP-2-keto-3-deoxy-

manno-octulosonic acid synthetase, the enzyme required for activation

of Kdo prior to incorporation into rhamnogalacturonan II. Mol. Plant, 6,

981–984.Tobimatsu, Y., Van de Wouwer, D., Allen, E., Kumpf, R., Vanholme, B.,

Boerjan, W. and Ralph, J. (2014) A click chemistry strategy for visualiza-

tion of plant cell wall lignification. Chem. Commun. 50, 12262–12265.



Tornøe, C.W., Christensen, C. and Meldal, M. (2002) Peptidotriazoles on solid

phase: [1,2,3]-triazoles by regiospecific copper(I)-catalyzed 1,3-dipolar

cycloadditions of terminal alkynes to azides. J. Org. Chem. 67, 3057–3064.Voxeur, A., Gilbert, L., Rihouey, C., Driouich, A., Rothan, C., Baldet, P.

and Lerouge, P. (2011) Silencing of the GDP-D-mannose 3,5-epimerase

affects the structure and cross-linking of the pectic polysaccharide

rhamnogalacturonan II and plant growth in tomato. J. Biol. Chem. 286,

8014–8020.

Williams, M.N., Freshour, G., Darvill, A.G., Albersheim, P. and Hahn, M.G.

(1996) An antibody Fab selected from a recombinant phage display

library detects deesterified pectic polysaccharide rhamnogalacturonan II

in plant cells. Plant Cell, 8, 673–685.Zhu, C., Ganguly, A., Baskin, T., McClosky, D.D., Anderson, C.T., Foster, C.,

Meunier, K.A., Okamoto, R., Berg, H. and Dixit, R. (2015) The FRA1 kine-

sin contributes to cortical microtubule-mediated trafficking of cell wall

components. Plant Physiol. 167, 780–792.



Plant cell wall imaging by metabolic click‐mediated ... 2016 Plant cell wall imaging by... ·...

Documents

Transcript of Plant cell wall imaging by metabolic click‐mediated ... 2016 Plant cell wall imaging by... ·...