ANRV410-PP61-12 ARI 31 March 2010 16:8

HemicellulosesHenrik Vibe Scheller1 and Peter Ulvskov2

1Feedstocks Division, Joint BioEnergy Institute, Lawrence Berkeley National Laboratory,Emeryville, California 94608; email: hscheller@lbl.gov2Department of Plant Biology and Biotechnology, University of Copenhagen, DK-1871Frederiksberg C, Denmark; email: ulvskov@life.ku.dk

Annu. Rev. Plant Biol. 2010. 61:263–89

First published online as a Review in Advance onJanuary 29, 2010

The Annual Review of Plant Biology is online atplant.annualreviews.org

This article’s doi:10.1146/annurev-arplant-042809-112315

Copyright c© 2010 by Annual Reviews.All rights reserved

1543-5008/10/0602-0263$20.00

Key Words

xyloglucan, xylan, mannan, glucomannan, mixed-linkage glucan, plantcell wall

AbstractHemicelluloses are polysaccharides in plant cell walls that have β-(1→4)-linked backbones with an equatorial configuration. Hemicel-luloses include xyloglucans, xylans, mannans and glucomannans, andβ-(1→3,1→4)-glucans. These types of hemicelluloses are present inthe cell walls of all terrestrial plants, except for β-(1→3,1→4)-glucans,which are restricted to Poales and a few other groups. The detailedstructure of the hemicelluloses and their abundance vary widely be-tween different species and cell types. The most important biologicalrole of hemicelluloses is their contribution to strengthening the cell wallby interaction with cellulose and, in some walls, with lignin. These fea-tures are discussed in relation to widely accepted models of the primarywall.

Hemicelluloses are synthesized by glycosyltransferases located in theGolgi membranes. Many glycosyltransferases needed for biosynthesisof xyloglucans and mannans are known. In contrast, the biosynthesisof xylans and β-(1→3,1→4)-glucans remains very elusive, and recentstudies have led to more questions than answers.

263

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

Contents

INTRODUCTION . . . . . . . . . . . . . . . . . . 264STRUCTURE AND

DISTRIBUTION . . . . . . . . . . . . . . . . . 265Xyloglucan . . . . . . . . . . . . . . . . . . . . . . . . 265Xylans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268Mannans and Glucomannans . . . . . . . 269β-(1→3,1→4)-glucans . . . . . . . . . . . . . 270

BIOSYNTHESIS. . . . . . . . . . . . . . . . . . . . . 270Xyloglucan . . . . . . . . . . . . . . . . . . . . . . . . 270Xylans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273Mannans and Glucomannans . . . . . . . 276Mixed-Linkage Glucans . . . . . . . . . . . . 277The CESA-CSL Superfamily . . . . . . . 277

FUNCTION OFHEMICELLULOSES . . . . . . . . . . . . . 278Biological Function—

Hemicelluloses in CellWall Models . . . . . . . . . . . . . . . . . . . . 278

Source of Signal Molecules . . . . . . . . . 280Seed Storage Carbohydrates . . . . . . . . 281

ENGINEERING OF CELLWALLS . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

INTRODUCTION

A strong wall—composed of polysaccha-rides, proteins, and in some cells, phenoliccompounds—surrounds every cell of terrestrialplants. The type of wall, composed mainly ofpolysaccharides, probably evolved among thecharophyte green algae but is more evidentlysuited for life on land. The wall provides sup-port and shape for the plant, allowing it tostand upright. Some plants reach heights ofmore than 100 m, and obviously cell walls insuch plants are capable of supporting very largephysical forces and are very durable. The plantcell wall also provides a barrier against the en-vironment and potentially pathogenic organ-isms. However, in spite of the strong, rigid, andseemingly impenetrable properties of cell walls,they are metabolically active, allowing exchangeof material and signals between cells, and arecapable of expanding. The cells generated inmeristems by cell division will typically expand

to 100 times their original length, and the pri-mary wall surrounding such cells fulfills thesupport and barrier functions while expandingwith the growing cell. Primary cell walls are thewalls that surround growing cells. After cessa-tion of cell expansion, some cells (e.g., fibersand tracheary elements) develop a secondarywall, which gives additional strength. Some cellwalls in seeds are highly thickened due to accu-mulation of polysaccharides that serve as seedstorage.

Polysaccharides make up most of the wall,but the wall also contains proteins. Phenoliccompounds, notably lignin, constitute up to30% of some secondary walls. Many kinds ofpolysaccharides are present in all cell walls, thespecific makeup varying in different species andtissues. Classically, cell wall polysaccharideshave been grouped into cellulose, hemicellu-loses, and pectins. Of these different classes,only cellulose is well defined, consisting en-tirely of β-(1→4)-linked glucan chains. Pectinsare highly heterogeneous polysaccharides, tra-ditionally characterized by being relativelyeasily extracted with hot acid or chelators andby containing a large amount of galacturonicacid residues. Hemicelluloses traditionallycomprise the remaining polysaccharides,which can be extracted with alkaline treatment.These polysaccharides are very different fromeach other structurally and in physicochemicalproperties. Definitions based on extractabilityare not useful. Some pectins can be extractedonly with alkaline treatment, and this is partic-ularly true in some species such as lycophytes(50). Similarly, the β-(1→3,1→4)-glucans ingrass cell walls and part of the arabinoxylansin cereal endosperm are considered hemicel-luloses but are quite readily extracted withoutalkaline treatment. In this review we havegrouped the hemicelluloses into xyloglucan,xylans, mannans and glucomannans, andβ-(1→3,1→4)-glucans. Some polysaccharides,such as galactans, arabinans, and arabinogalac-tans are sometimes included in the hemicellu-lose group, but since these appear to be part ofpectin molecules, at least in the initial synthesis,and do not share the equatorial β-(1→4)-linked

264 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

backbone structure, we think they should notbe included in the already heterogeneousgroup of hemicelluloses. We also considerthat callose, which has a backbone entirelycomposed of β-(1→3) linked glucose residues,should not be considered a hemicellulose.

Hemicelluloses are a heterogeneous groupof polysaccharides, and the term was coined ata time when the structures were not well un-derstood and biosynthesis was completely un-known. The term hemicelluloses is thereforearchaic and various researchers have suggestedthat it should not be used. Alternative termssuch as cross-linking glycans have been pro-posed (138, 140), but that has other problemssince it is not obvious that cross-linking is amajor and common feature of the hemicellu-loses. Nevertheless, most workers in the fieldstill use the term hemicelluloses as a convenientdenominator for a group of wall polysaccha-rides that are characterized by being neithercellulose nor pectin and by having β-(1→4)-linked backbones of glucose, mannose, or xy-lose. These glycans all have the same equato-rial configuration at C1 and C4 and hence thebackbones have significant structural similar-ity (Figure 1). We suggest that hemicellulosesshould be used to describe only polysaccharideswith this configuration.

STRUCTURE ANDDISTRIBUTION

Xyloglucan

Xyloglucan (XyG) has been found in every landplant species that has been analyzed, includingmosses, but has not been found in Charophytes(93, 108, 109) (see Figure 2 for an overviewof plant phylogeny). XyG is the most abun-dant hemicellulose in primary walls of sper-matophytes except for grasses (Table 1). Thebasic structure of XyG is shown in Figure 3, butthere are many variations to this general pat-tern. In spite of the variations, XyGs are madeof repetitive units, and a special one-letter codeis used to denote the different XyG side chains(41). G denotes an unbranched Glc residue,

OO

O

OO

O

OO

O

OO

O

O O

O

Equatorial

O

O

Equatorial

O

Axial

Hemicellulose repeating disaccharides

Pectic RG-I side chain

a

b

β1 4

β1 4 glucan

β1 4 xylan

β1 4 glucomannan

β1 4 mannan

β1 4 galactan

45

32

1 45

32

1

Figure 1(a) Hemicelluloses are characterized by aβ-(1→4)-linked backbone with an equatorialconfiguration at C1 and C4. (b) Polysaccharides suchas β-(1→4)-galactan have an axial configuration atC4 and should not be included with hemicelluloses.

XyG: xyloglucan

while X denotes α-d-Xyl-(1→6)-Glc. The xy-losyl residues can be substituted at O-2 withβ-Gal (L side chain) or α-l-Araf (S side chain).A Gal residue substituted at O-2 with α-l-Fucis designated F. Other less common side chainshave other designations.

The branching pattern of XyG is of bothfunctional and taxonomic significance. Less

www.annualreviews.org • Hemicelluloses 265

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

Rhodophyta (red algae) +

Chlorophyta (green algae)

Bryophyta (mosses)

Marchantiophyta (liverworts) +

Charaphyacean green algae +(algae with higher-plant-like walls)

Green

plan

t

lineag

e

Vascular plants

Vascular plantsEukaryota

1 BY

470 MY

450 MY

Lycophyta (club mosses and spike mosses)

Moniliformopses (horsetails +, ferns)

Magnoliids

AsteridsRosids

Liliales

Arecales (palms) O

Zingiberales X O

Poales + X O

Nympheales

Coniferophyta (conifers )

Flowerin

g

plan

ts

An

giosp

erms

Mon

ocots

410 MY

380 MY

325 MY

140 MY

125 MY

Basal dicots

Eudicots

Commelinids

Ranunculales, etc.

Emb

ryoph

ytes

Figure 2Simplified phylogenetic tree illustrating the main taxa mentioned in the text. Most taxa are not shown in the figure. Estimates of the ageof some diversifications (15) are indicated. Estimates of the age of the commelinid clade have often been in the range of 80 MY, butstudies of grass leaf silica bodies in dinosaur dung (113) have pushed the age of Poacea back to 80 MY and the basal taxa of Poales to ca.100 MY, making dating of the diversification of angiosperm families uncertain. Symbols designate hemicellulose patterns: +,β-(1→3,1→4)-glucan has been found in members of the group; X, taxa where (glucurono)arabinoxylan is the major hemicellulose inprimary walls (as opposed to xyloglucan, or XyG, in other groups); O, taxa where xylans are esterified with ferulic acid.

branched XyGs are less soluble and this maycorrelate with functional aspects in those fam-ilies with the lowly substituted XyGs. Themost profound difference in XyG structure ischarged versus uncharged side chains, with theformer found in mosses and liverworts (104)whereas vascular plant XyGs are neutral.

The major divide among vascular plantfamilies is in regard to whether XXGG orXXXG is the predominant, core repeating

xyloglucan oligosaccharide (142). The less sub-stituted XXGG structure predominates in cellwalls of commelinid monocots, which are alsodistinguished by the low content of XyG inthe primary walls, typically 1–5% comparedwith 20% in dicots (Table 1). These figures re-fer to expanding cells. For example, elongatinglight-grown pea internodes contained 19.1%XyG as compared with 8.2% in internodes af-ter cessation of growth (102). Oligosaccharides

Table 1 Occurrence of hemicelluloses in primary and secondary walls of plants

Amount of polysaccharide in wall (% w/w)a

Dicot walls Grass walls Conifer walls

Polysaccharide Primary Secondary Primary Secondary Primary SecondaryXyloglucan 20–25 Minor 2–5 Minor 10 −b

Glucuronoxylan − 20–30 − − − −Glucuronoarabinoxylan 5 − 20–40 40–50 2 5–15(Gluco)mannan 3–5 2–5 2 0–5 − −Galactoglucomannan − 0–3 − − +b 10–30β-(1→3,1→4)-glucan Absent Absent 2–15 Minor Absent Absent

aNumbers are typical values; actual values vary between different species and tissue types. Numbers are obtained from several different sources (31, 17, 37,59, 72, 101, 135, 143, 147, 156; C. Manisseri and H. V. Scheller, unpublished data).b−, absent or minor; +, present but quantitative data not available.

266 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

O

O

O

O

O

O

O O

O

O O

O O O

O OO

X X F L G X X G

O O O O O OO

Ac

O

O O O O

O OO

OO O

OCOOH

OC

OO

H

CO

OH

Fer

O

Fer

D-Glucose Glcp

D-Galactose Galp

D-Mannose Manp

D-Glucuronic acid GlcAp

O O O O

O O O O O O

O O O O O O O O

O

OMe

O

O

O

L-Arabinose Araf

D-Xylose Xylp

L-Fucose Fucp

OL-Rhamnose Rhap

Mixed linkage β-glucan [β-D-Glcp-(1 4)]n-β-D-Glcp-(1 3)-[β-D-Glcp-(1 4)]m, where n and m are 3 or 4;

typical of Poales.

Glucuronoarabinoxylan, GAX, typical of commelinid monocots.

O O O O O O O OO

OOCO

OH

OCO

OH

Glucuronoxylan, typical dicot structure.

Galactomannan, typical of Fabaceae seeds.

Galactoglucomannan, typical of conifer wood.

O

Xyloglucan [β-D-Glcp-(1 4)]n backbone substituted with side chains as seen in pea and arabidopsis.

The arrow indicates the typical β-glucanase cleavage site.

O

O

O

OO

O

O

O

O

O

OC

OO

HO

OMe

Ac

Ac

Figure 3Schematic illustration of the types of hemicelluloses found in plant cell walls. The letters under the xyloglucan (XyG) moleculeillustrate the symbols used for the most common side chains. The structure of the hemicelluloses varies greatly in different plant speciesand tissue types. “Fer” represents esterification with ferulic acid (3-methoxy-4-hydroxycinnamic acid), which is characteristic of xylansin commelinid monocots.

www.annualreviews.org • Hemicelluloses 267

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

Cell walls ofcommelinidmonocots: primarywalls of grasses andother commelinidmonocots with a highcontent of (glu-curono)arabinoxylanare often described asType II walls, asopposed to the morexyloglucan-rich Type Iwalls in other plants(20). However, studiesof different taxa haverevealed a largediversity in wallcomposition. Whilethe Type I/Type IIdistinction can beuseful, we havedecided not to use it inthis review

Glycosyltransferases(GTs): biosyntheticenzymes that use anactivated sugar—mostoften a nucleotidesugar—to transfer asugar molecule to anacceptor. In the case ofhemicelluloses, theacceptors are thegrowingpolysaccharide chainsand possible primermolecules

GAX:glucuronoarabinoxylan

of the XXGG type are also found in solana-ceous XyGs, which are further characterizedby lacking fucosyl residues and featuring thearabinose-containing S chain (152). S is alsofound in XXXG-type units in olive (141).Hoffman et al. (57) have shown that the F andthe S side chains coexist in the XyG of Neriumoleander and have compiled an extensive list ofXyG-repeating units from a wide range of plantfamilies. Their general conclusion is that whilethe XyG oligosaccharide module structurescorrelate with taxonomical groupings, there areseveral convergent adaptations of XyG struc-ture among Tracheophytes so that the XyGstructures do not mirror evolution in any simpleway.

The occurrence of the fucose contain-ing XyG F-chain (see side chain structure inFigure 3) in the grass family is particularlyenigmatic. Purified hemicellulose fractions ofoat and rice did not contain detectable fucose(65, 73), and such studies led Hayashi (55) tostate that grass XyG does not contain fucose.However, fucosylated XyG has been detectedin Festuca arundinacea by feeding radioactive fu-cose to cell suspension cultures (89), and Penaet al. (104) cite unpublished results that lowamounts of fucosylated XyG are found by thismethod in rice as well. These observations canbe reconciled if XyG is fucosylated transientlyin these grasses during synthesis and most ofthe fucose is removed during or following de-position in the wall. A transient fucosylation ofgrass XyG has implications for the assignmentof function to grass glycosyltransferases (GTs)of family GT37 (discussed below).

Xylans

Xylans are a diverse group of polysaccharideswith the common feature of a backbone ofβ-(1→4)-linked xylose residues (Figure 3).A common modification of xylans is substitu-tion with α-(1→2)-linked glucuronosyl and4-O-methyl glucuronosyl residues. Xylansdominated with this type of substitution areoften known as glucuronoxylans and are thedominating noncellulosic polysaccharide in

the secondary walls of dicots. In commelinidmonocots (which include grasses and somerelated species; see Figure 2), xylans are themajor noncellulosic polysaccharide in primarywalls, constituting about 20% (Table 1) of thewall. These xylans usually contain many arabi-nose residues attached to the backbone and areknown as arabinoxylans and glucuronoarabi-noxylans (GAXs). The distinction is not clear.Cereal endosperm arabinoxylan has very littleglucuronic acid, but heteroxylans in vegetativeparts of grasses are often called arabinoxy-lans, even though they tend to contain moreglucuronic acid and 4-O-methyl glucuronosylresidues, making GAX a more appropriatename. Branching patterns in xylan, like thosein XyGs, correlate with taxonomy. Arabinofu-ranose substitutions in grass walls are mostlyfrom O-3 of the backbone xylose residues, andxylose residues doubly substituted with arabi-nofuranose at both O-2 and O-3 are commonin grass endosperm (31). Gymnosperm wallsalso contain arabinoxylans in relatively highamounts (6). Arabinofuranose substitutionsare less frequent in dicot xylans, but they areabundant in sycamore primary walls (27) and inseeds of certain species from diverse taxonomicgroups, e.g., flax and psyllium (36, 96). Databy Zablackis et al. (155) suggest that GAXwith fewer arabinose substitutions is present inarabidopsis, but unambiguous evidence has notbeen presented. The arabinofuranose substitu-tions on dicot xylans are normally at O-2 ratherthan O-3 as in grasses (27, 155), but doubly sub-stituted xylose residues have been described inflax mucilage (96) and 3-linked arabinofuranoseis present in psyllium seeds (27).

Unlike XyGs, xylans do not have a repeatedstructure, and there are many variations in thestructure that are not well known. A commonfeature of grass xylans is the O-2 linked xyloseresidues often found as substituents of feruloyl-arabinofuranosyl side chains (144). Other fea-tures have been described, e.g. α-galactoselinked to O-2 of glucuronosyl residues in eu-calyptus (123) and α-arabinofuranosyl residueslinked to O-2 of arabinose residues that are di-rectly bound to the backbone in sorghum (139).

268 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

O

O

OOO

O

COOH

O

Figure 4Schematic illustration of the oligosaccharide foundin the reducing end of xylan. The β-d-Xylp-(1→4)-β-d-Xylp-(1→3)-α-l-Rhap-(1→2)-α-d-GalpA-(1→4)-d-Xylp structure has been found in severaldifferent species.

Most xylans are acetylated to various degrees,especially in dicot secondary walls (31). Acetylgroups are attached to O-3 of xylose residuesand to a lesser extent to O-2.

A somewhat surprising feature of xylans isthe conserved oligosaccharide that has beenfound in the reducing end of xylan from dicotsand conifers (Figure 4) (1, 63, 105, 124). Thisstructure has not been reported from grasses,but it appears that the genes responsible formaking the structure are conserved in grassesas well (see below).

Ferulate esters. An important feature of grassxylans is the presence of ferulic acid esters at-tached to O-5 of some of the arabinofuranosylresidues. Esters of p-coumaric acid are alsoabundant in grass cell walls, but it is not clear ifthey can be attached directly to the xylans, andthey may be primarily associated with lignin(54). Feruloylated GAXs are found through-out the commelinid monocots and not in othergroups of seed plants. While GAX is usuallythe dominating noncellulosic polysaccharide incommelinid species, Arecales (palms), which isthe most basal group among the commelinids, isan exception (Figure 2). The palm cell walls re-semble those of the noncommelinid seed plants,i.e., with pectin as the major matrix polysac-charide, most likely with fucosylated xyloglu-can, but in addition, small amounts of truly 2,3-branched xylan carrying feruloyl esters (18, 53).These walls thus appear to represent a transi-tional form in agreement with its position asthe least derived group among the commelinidmonocots.

Ferulate esters bound to cell walls havebeen described in gymnosperms, but it is notclear to which polysaccharide they are bound(19). In Amaranthaceae sensu lato, i.e., includ-ing Chenopodiaceae, Rhamnogalacturonan-I(RG-I) side chains are the sites of feruloyla-tion, which occurs in both galactans and ara-binans (115). The arabinosyl residues are pre-dominatly feruloylated at position 2 (60), butsmall amounts of feruloylation at position 5(as in GAX) have been detected (78). Recentstudies suggest that feruloylated pectins maybe present in many other plant groups butrestricted to stomata and therefore perhapspresent at a much lower level (64). The feruloyltransferases are unknown both in Chenopodi-aceae and in Poaceae, so it is not possible todetermine from the structures of the enzymesif they have a common origin. [See also (59) fora comprehensive review of variations in feru-loylated polysaccharides.]

Ferulate esters are important because theycan be oxidatively cross-linked in a variety ofways. Ferulate dimers are easily detected ingrass walls and likely represent intra- and in-termolecular linkages in and between GAXmolecules. Ferulate can also be cross-linkedwith lignin (48) and we can therefore assumethat GAX and lignin become covalently cross-linked through these linkages. Cross-linkingthrough ferulate esters is widely assumed torender the cell wall recalcitrant to digestion,which would be an obvious benefit as a defenseagainst microorganisms and herbivores. Simi-larly, the ferulate esters make grass cell wallsrecalcitrant to enzymatic saccharification priorto fermentation into biofuels (11, 12).

Mannans and Glucomannans

β-(1→4)-linked polysaccharides containingmannose are widely distributed and the mainhemicellulose in Charophytes (108, 109). Thebackbones may consist entirely of mannose, asin mannans and galactomannans, or with man-nose and glucose in a nonrepeating patternas in glucomannans and galactoglucomannans.Mannans and glucomanans are often acetylated.

www.annualreviews.org • Hemicelluloses 269

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

Mannans have been much studied in theirrole as seed storage compounds, but they arefound in variable amounts in all cell walls. Ingymnosperms, galactoglucomannans are ma-jor components of the secondary walls (31).Mannans appear to have been very abundant inearly land plants and are still abundant in mossesand lycophytes (50, 93). In spermatophytes,mannans and glucomannans are generallymuch less abundant and it appears that otherhemicelluloses have largely replaced them.Nevertheless, mannans play essential roles, asevidenced by the embryo lethal phenotype inan arabidopsis mutant that is lacking the major(gluco)mannan synthase in seeds (47).

β-(1→3,1→4)-glucans

β-(1→4)-linked glucans with interspersed sin-gle β-(1→3)-linkages are well known in grasses.These mixed linkage glucans are dominatedby cellotriosyl and cellotetrasyl units linked byβ-(1→3) linkages, but longer β-(1→4)-linkedsegments also occur (128). Mixed linkage glu-can in primary walls plays a role in cell ex-pansion and the amount is very growth-stagedependent (23, 45, 97).

The β-(1→3,1→4)-glucans have not beenfound in dicots but are found throughout Po-ales, including the most basal family, Flag-ellariaceae (125). They have been suggestedto be more abundant in the most derivedfamily, Poaceae (grasses; 125), but the stronggrowth-stage dependency of β-(1→3,1→4)-glucan content makes such quantitative argu-ments difficult.

Recent studies have shown that β-(1→3,1→4)-glucans are not restricted toPoales but are present in Equisetum (38,126), and it has been suggested that theyare also present in liverworts (109). Evenmore recent work has shown the presence ofβ-(1→3,1→4)-glucans in Charophytes andred algae (Z. Popper and W. Willats, personalcommunication) (Figure 2). The occurrence ofβ-(1→3,1→4)-glucans in many primitive taxacould indicate that they represent an ancienttrait. However, if β-(1→3,1→4)-glucan in

grasses were a conserved ancient trait, thenwe would have to postulate the independentdisappearance in a large number of Sper-matophyte taxa that have been investigated. Itseems more likely that β-(1→3,1→4)-glucanhas evolved independently in grasses. Sinceat least some of the genes responsible forβ-(1→3,1→4)-glucans biosynthesis in grassesare known (see below), it should be readilytestable if orthologs are conserved in Equi-setum and different algae while missing indicots.

BIOSYNTHESIS

Xyloglucan

Glycosyltransferases. Many of the biosyn-thetic enzymes involved in XyG biosynthesishave been identified. The fucosyltransferasewas one of the first cell wall biosynthetic en-zymes to be identified (106) (see Table 2 for alist of GTs involved in hemicellulose biosynthe-sis). XyG fucosyltransferase was purified frompea using a traditional biochemical approach,and the arabidopsis ortholog was expressedand assayed in vitro. Subsequently, a fucose-deficient arabidopsis mutant, mur2, was shownto be affected in the same gene (138). Analy-sis of another fucose-deficient mutant, mur3,from the same mutant collection led to cloningof a gene, which turned out to encode a XyGβ-(1→2)-galactosyltransferase (84). The galac-tosyltransferase has been shown to be highlyspecific for the third galactose in the repeatingXXXG unit of xyloglucan. This is the galactosethat is often fucosylated, explaining the fucose-deficient phenotype. The reduction in galactoseis harder to detect since galactose is present inmany other polymers. The galactose on the sec-ond position in the repeating unit must be in-corporated by a different galactosyltransferase.Li et al. (80) have obtained evidence that a GTin subgroup A of the GT47 family is responsi-ble for the activity, but the final evidence has notyet been reported. GT47 subgroup A contains10 members in arabidopsis and only MUR3 hashad the activity confirmed in vitro.

270 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

Table 2 Glycosyltranserases involved in hemicellulose biosynthesis

Activitya GT name Identifierb CAZyc Reference CommentXyloglucan

β(1→4)-glucan synthase CSLC4 At3g28180 GT2 24α(1→6)-XylT XXT1 At3g62720 GT34 34

XXT2 At4g02500 GT34β(1→2)-GalT MUR3 At2g20370 GT47 84 Specific for the third xylose in the

XXXG repeating unitβ(1→2)-GalT GT47 79 An enzyme specific for the second

xylose in XXXG has not beenunambiguously identified

α(1→6)-FucT MUR2/FUT1 At2g03220 GT37 106Xylan

β(1→4)-xylan synthase IRX9 At2g37090 GT43 2, 9, 75 Catalytic activity has not beendetermined

β(1→4)-xylan synthase IRX14 At4g36890 GT43 9 Catalytic activity has not beendetermined

β(1→4)-xylan synthase IRX10/GUT2 At1g27440 GT47 10, 149 Catalytic activity has not beendetermined

IRX10-LIKE/GUT1 At5g61840 GT47

IRX7/FRA8 At2g28110 GT47 9, 105 Rhamnosyl β(1→3)-XylT?

F8H At5g22940 GT47 76

IRX8/GAUT12 At5g54690 GT8 9, 153 Xylosyl α(1→4)-GalAT?

PARVUS/GLZ1 At1g19300 GT8 74, 77 Involved in synthesis of reducingend oligosaccharide. α-XylT?

α(1→2)-Araf T GT61? Predicted from expression dataα(1→3)-Araf T GT61? Predicted from expression dataα-Araf T transferring toAra-substituted Xyl

GT61? Predicted from expression data

α(1→2)-GlcAT No candidate has been proposed(Gluco)mannan

β(1→4)-mannan synthase CSLA/ManS AAR23313 GT2 28 From guarα(1→6)-GalT TfGalT CAB52246 32 From fenugreek

Mixed-linkage glucanβ(1→3,1→4)-glucansynthase

OsCSLF2 AAL25132 GT2 14 From rice, catalytic activity notdetermined

OsCSLF4 AAL25134 GT2β(1→3,1→4)-glucansynthase

HvCSLH1 ACN67534 GT2 29 From barley, catalytic activity notdetermined

aKnown and predicted enzymes involved in biosynthesis of the most common structures in hemicelluloses. The less common structures of XyGs andxylans require a large set of additional activities.bIdentifiers are given as gene locus IDs for Arabidopsis thaliana and as NCBI accession numbers for other species.cGlycosyltransferase (GT) family in CAZy database.

www.annualreviews.org • Hemicelluloses 271

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

The xylose residues in XyG are transferredby α-(1→6)-xylosyltransferases, which are re-taining enzymes from family GT34. Two en-zymes, XXT1 and XXT2 (originally namedXT1 and XT2), have been identified in ara-bidopsis and shown to be involved in the syn-thesis of XyG (21, 34). Somewhat surprisingly,a third GT34 enzyme, XXT5 from a sepa-rate clade of GT34, has been implicated inthe same function, but the activity has notbeen unambiguously demonstrated (156). TheXXT5 clade has two additional, uncharacter-ized arabidopsis members. In addition, GT34has two arabidopsis members that are moreclosely related to galactomannan α-(1→6)-galactosyltransferase (156).

CESA

CSLJ

CSLE

CSLG

CSLA

CSLC

CSLB

CSLH

CSLD

CSLF

Figure 5Schematic illustration of the phylogeny of the cellulose synthase (CESA) andcellulose synthase–like (CSL) superfamily of GTs. Not all families are presentin the same species. CSLH and CSLF are only known from grasses, and CSLBand CSLG are only known from dicots. CSLE is known from all angiosperms,but not outside these groups. CESA, CSLA, CSLC, and CSLD are knownfrom all plant genomes analyzed so far. CSLJ is present in some angiosperms,both dicots and grasses, but not in arabidopsis and rice.

The backbone of XyG is apparently syn-thesized by members of cellulose synthase–like proteins belonging to the CSLC family(Figure 5). Arabidopsis has five members ofthe CSLC subfamily and only CSLC4 has beenshown to be involved in XyG biosynthesis (24).It seems likely that the other CSLC membersare involved in XyG biosynthesis as well, but itcannot be excluded that some CSLC membershave a role in biosynthesis of other polymers.

Hydrolases. The GTs described above are allGolgi-localized enzymes and work together toproduce a XyG precursor molecule that is trans-ported to the wall. However, important changesto the XyG molecules take place after the ini-tial synthesis in the Golgi. Thus, it has beenshown that specific apoplastic glycosidases areresponsible for the trimming of nascent XyGchains and important in determining the het-erogeneity of the polymer in the wall (100).The value of such apparently wasteful trim-ming is unclear, but the higher substitution de-gree of native hemicelluloses will help to keepthem soluble during transport and incorpora-tion into the wall. In general, hydrolases arelikely to play an important role in determininghemicellulose structures in the wall, and it maybe noted that many hydrolases are coexpressedwith polysaccharide biosynthetic enzymes. It isinteresting to note that the arabidopsis genomecontains about 300 membrane-bound GTs butmore than 500 glycoside hydrolases and lyases,many of which are involved in modificationof wall polysaccharides. Plant hydrolases havebeen described in recent reviews (82, 91).

Another important and well-studied modi-fication of XyG is carried out by the XET en-zymes. These are enzymes related to hydrolasesand carry out a transglycosylase reaction. Therole of these enzymes is described below in thesection on Biological Function.

Polysaccharide acetylation. XyGs, like xy-lans, mannans, and pectins, are usually acety-lated to various degrees. Acetylation of XyG ison the galactose residues, mostly on O-6 (68).Acetylation of cell wall polysaccharides occurs

272 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

in the Golgi by means of transferases usingacetyl-CoA (103). Neither acetyltransferasesnor acetyl-CoA transporters required for thisprocess have been identified. A mutant of thefungus Cryptococcus neoformans lacks acetylationof the glucuronomannan coat polysaccharideand the affected gene has been suggested tobe an acetyltransferase (62). Arabidopsis has fourhomologs of the C. neoformans protein, and wehave recently found that knockout mutants inone of the corresponding genes, At3g06550, aredeficient in wall-bound acetate (Y. Manabe andH. V. Scheller, unpublished data). Presently, itis unclear if the protein is an acetyltransferase,but we think its topological model with mul-tiple membrane-spanning segments suggests atransporter function.

Xylans

The main chain of xylan as shown in Figure 1may be differently substituted with side chainsand equipped with a unique oligosaccharide inits reducing end. Side chain structures vary withtaxonomic origin (commelinid monocot ver-sus other flowering plants), and there is someuncertainty as to how conserved the reducingend oligosaccharide is. Biosynthesis is thus dealtwith under three subheadings: Backbone Syn-thesis, Reducing End Oligosaccharide, and SideChains.

Backbone synthesis. Because of the struc-tural similarity of xylan to the β-(1→4)-linkedbackbones of the other hemicelluloses, it hasbeen widely assumed that the biosynthesiswould involve members of the CSL proteinfamilies, as has been shown to be the case for theother hemicelluloses. However, investigationsof CSL proteins have not provided evidencethat any of these proteins are involved in xylanbiosynthesis. Instead, characterization of xylan-deficient mutants irx9, irx14, irx10, and irx10-like has indicated that the corresponding GTsbelonging to families GT43 and GT47 are re-sponsible for elongation of the xylan backbone.Unlike the CLS proteins with multiple trans-membrane segments, the members of GT43

XET: xyloglucanendotransglucosylasesbelong to thexyloglucan endotrans-glucosylase/hydrolase(XTH) group of CAZy(see below) familyGH16 that alsocomprises thexyloglucanendohydrolases(XEH). The XETscleave xyloglucan(XyG) backbones,retain the energy ofthe glucosidic linkage,and graft the reducingend of the cleavedmolecule onto thenonreducing end of anacceptor XyG.XEH-catalyzedhydrolysis proceedssimilarly, with wateracting as the acceptorsubstrate

CAZy: acomprehensive systemand database(www.cazy.org) forclassifyingcarbohydrate activeenzymes. Currently,glycosyltransferasesand glycosidehydrolases in CAZyare classified into 91and 115 differentfamilies, respectively(16)

and GT47 are predicted to be typical Type IImembrane proteins with a single N-terminalmembrane anchor. See Table 2 to associateeach GT with the corresponding mutant andCAZy family. Although the GTs affected inthese mutants belong to different families,they are all named irx for their irregular xylemphenotype. Xylem cells are under negativepressure, and compromised load-bearing abil-ity is associated with vessel collapse or irregularwalls. Secondary walls are rich in xylan inarabidopsis, and mutants in any of these genesare deficient in xylan and xylan synthase activity(2, 9, 10, 75, 77). However, so far no xylansynthase activity has been reported for any ofthese proteins when heterologously expressed.

Presumably, several researchers have at-tempted to show xylan synthase activity, butwithout success. In our own work we haveexpressed these proteins in E. coli and/or to-bacco cells but failed to find xylan synthaseactivity (A. Suttangkakul, A. Oikawa, H. V.Scheller, unpublished data). Obviously, it isnot possible to draw conclusions from nega-tive results, but the non-redundancy of IRX9,IRX14, and IRX10 indicates that they are allrequired for backbone synthesis, perhaps in acomplex containing more than one protein. Ahomolog of IRX10 named NpGUT1 for glu-curonosyltransferase was discovered in a Nico-tiana plumbaginifolia cell adhesion mutant andsuggested to be involved in RG-II A-chainbiosynthesis (61). IRX10 and IRX10-like ap-pear to be redundant and are unlikely to be or-thologous to NpGUT1. The mutants do notdisplay a cell adhesion phenotype and shouldrather be named for their pronounced irx phe-notype found when both genes are knockedout (9, 149). Selaginella and Physcomitrella haveonly a single IRX10/IRX10L ortholog (50) con-firming that one gene product suffices. ForIRX9 the situation is more complex. Arabidop-sis contains a homolog of IRX9, At1g27600,the role of which has not been reported. Riceand other higher plants have members of boththe IRX9 and the At1g27600 group, while Se-laginella and Physcomitrella have orthologs onlyof At1g27600 (50). The most likely explanation

www.annualreviews.org • Hemicelluloses 273

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

for this is that IRX9 has a special role in sec-ondary walls in seed plants. The At1g27600protein and its orthologs in lycophytes andmosses would then have the same biochemi-cal function, but in other cell types. In agree-ment with this, IRX9 is almost exclusively ex-pressed in cells with secondary wall growth,while At1g27600 is expressed in other cell types.IRX14 in arabidopsis has a close homolog,AT5G67230, which most likely would have thesame biochemical function. This could explainthe weak phenotype of the irx14 mutant (9).

Reducing end oligosaccharide. As men-tioned above, xylans in conifers and severaldicots have been shown to have a reducing endwith the unique structure β-d-Xylp-(1→4)-β-d-Xylp-(1→3)-α-l-Rhap-(1→2)-α-d-GalpA-(1→4)-d-Xylp (Figure 4) (1, 63, 105, 124).Treatment of xylan with xylanases allows theisolation of the oligosaccharide. A number ofxylan-deficient mutants, irx7 (also known asfra8), irx8, and parvus, have been shown toretain in vitro xylan synthase activity whilebeing depleted in the reducing end oligosac-charide (77, 105). In contrast, the reducingend oligosaccharide is present in irx9, irx14,and irx10/irx10L mutants (9, 10, 105, 149).These observations indicate that IRX7, IRX8,and PARVUS are all specifically involved insynthesizing the oligosaccharide.

It is unclear how many GTs are needed tomake the oligosaccharide. At the least, theremust be a XylT specific for Rha-GalA-Xyl, aRhaT specific for GalA-Xyl, and a GalAT spe-cific for the terminal xylose, but it is possi-ble that additional xylosyl transferases wouldbe required to transfer the reducing end xy-lose to a nonpolysaccharide primer and to trans-fer xylose onto the Xyl-Rha-GalA-Xyl acceptor.The Rha-specific XylT and the GalA-specificRhaT are expected to be inverting enzymes,while the Xyl-specific GalAT would be a retain-ing enzyme. IRX8 (also known as GAUT12)is a homolog of GAUT1, which is a retainingGalA transferase known to synthesize the back-bone of homogalacturonan (127). Hence, IRX8is the most obvious candidate for the GalA

transferase. PARVUS (also known as GATL1)is also a member of GT8, but belongs to theGATL group rather than the GAUT group(127). Rather than transferring a charged sugar,bacterial homologs of the GATL group arelipopolysaccharide galactosyltransferases. Also,PARVUS has been reported to be located inthe ER (77), indicating an earlier biosyntheticstep than the subsequent Golgi-localized steps.Hence, we suggest that PARVUS is an α-xylosyltransferase transferring the reducing endxylose to a primer, which may be a lipid. IRX7is an inverting enzyme belonging to GT47. It isa close homolog to IRX10, which as mentionedabove is implicated as a xylosyltransferase in-volved in xylan backbone synthesis. Heterol-ogous expression of IRX7 has proved that itcan function in vitro as a xylosyltransferasewith several different sugars as acceptor (A.Suttangkakul and H. V. Scheller, unpublisheddata). Hence, IRX7 (and its close homologF8H; 76) is a candidate for the Rha-specificXylT. At present there is no clear candidatefor the rhamnosyltransferase, nor for any addi-tional xylosyltransferases needed in the nonre-ducing end prior to IRX9, IRX14, and IRX10.

What is the function of the reducing endoligosaccharide? At first sight the most obviousfunction would be as a primer, since synthe-sis of cell wall polysaccharides is generally as-sumed to take place by transfer to the nonreduc-ing end of the growing chain (121). However,York and O’Neill (153) have speculated that xy-lans may be synthesized in the other direction,i.e., growing in the reducing end and with theoligosaccharide functioning as a terminator se-quence. This could explain why irx7 and irx8mutants have unusually long xylan molecules(105, 153). The fact that xylooligosaccharidesthat are modified in the reducing end, e.g., byattachment of fluorescent tags, are still excellentacceptors for xylan synthase in vitro suggeststhat synthesis does in fact proceed in the moreconventional way by transfer to the nonreduc-ing end. However, as pointed out by York andO’Neill (153), the modified oligosaccharidescould be transferred to the reducing end of agrowing chain in the same way that a reducing

274 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

end terminator sequence would normallybe transferred. The reducing end oligosaccha-ride has not been reported in grasses, althoughthe genes associated with its biosynthesis seemto be conserved. The absence of the oligosac-charide in grass xylan would also be in agree-ment with the oligosaccharide being a termina-tor rather than a primer. A primer would morelikely be indispensable. Clearly, much morework is needed before xylan biosynthesis isunderstood.

Side chains. The most important side chainsof xylans should be formed by α-glucuron-syltransferases and α-arabinofuranosyl-transferases. Both activities have been detectedin vitro (3, 112, 157), but the GTs responsiblefor the transfer have not been identified.FRA8 was originally proposed to be a xylanglucuronosyltransferase (158), but it is gen-erally agreed that this is not correct and thatFRA8 is involved in synthesis of the backbone.Xylan α-glucuronosyltransferase is expectedto belong to a retaining GT family, but noobvious candidates have been proposed.

Arabinosyl residues may be linked to posi-tion 2 or 3 or to both positions of the xylosylresidues of the backbone. A monosubstitutedxylosyl residue is expected to be a very differentacceptor compared with an unsubstitutedresidue, and hence at least three invertingarabinosyltransferase activities are required: anα(1→2) and an α(1→3) arabinosyltransferaseand a third for adding the second residueto a monosubstituted xylose (no informationis available regarding the order of transfer).Members of GT61 (a family of inverting GTs)are highly expressed in grasses, which alsocontain many grass-specific GT61 members.The only known activity of GT61 is theβ-(1→2)-xylosyltransferase activity involvedin synthesis of N-linked glycans in plants (129).Hence, it has been speculated that GT61members are involved in transferring xylose toferuloyl-arabinofuranosyl side chains of xylan(92). However, the abundant and conservedGT61 members in grasses and the presenceof GT61 members in other plants suggest to

O OHO

OHHO

O

HO

OH

OH

O

α-L-Araf α-L-Arap

Figure 6Most monosaccharides in cell wall polysaccharidesare in the pyranose configuration. Arabinose,however, occurs mostly in the furanoseconfiguration.

RGP: reversiblyglycosylated protein

us that GT61 would more likely include thevarious arabinofuranosyltransferases neededfor xylan synthesis.

The fact that the arabinose residues in xylanare in the arabinofuranose configuration whileUDP-arabinopyranose is the nucleotide sugarknown to be present in plants has been a co-nundrum (Figure 6). Intact Golgi vesicles canincorporate arabinofuranose residues into xylanusing UDP-arabinopyranose as the substrate(112). Most likely, the mechanism behindthis transfer involves formation of UDP-arabinofuranose by UDP-arabinopyranosemutase, as has been shown for arabinanbiosynthesis (70, 71). Mutase activity hasbeen found in reversibly glycosylated proteins(RGPs) from rice and arabidopsis (71; C.Rautengarten and H. V. Scheller, unpublisheddata). RGPs are abundant proteins, whichare known to be located in the Golgi butapparently as extrinsic membrane proteins onthe cytoplasmic side of the membranes. Themutase catalyzes a reversible reaction, whichfavors the pyranose configuration 10:1. Hence,it seems most likely that the mutase wouldinteract with a transporter and/or transferaseso that the generated UDP-arabinofuranosecan be used in a channeled reaction. Themajor UDP-xylose-4-epimerase required forgenerating UDP-arabinopyranose is predictedto be located inside the Golgi (13). Therefore,it is surprising that the mutase is located on theoutside of the Golgi, since this would implythat UDP-Arap is moved out of the Golgi, con-verted to UDP-Araf, and then moved back into

www.annualreviews.org • Hemicelluloses 275

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

the Golgi again. On the other hand, isolatedwheat Golgi vesicles did not have detectableUDP-xylose-4-epimerase activity (112), soperhaps the topology and localization of theepimerases need to be further investigated.UDP-xylose is known to be synthesized bothin the Golgi lumen and in the cytosol (52).

Hydrolases. GAX in primary walls has beenshown to be synthesized in a more highly ara-binosylated form and subsequently trimmedby arabinofuranosides, resulting in much fewersubstituted polymers in mature cells (20, 43,97). Highly substituted arabinoxylan is moresoluble and does not interact with cellulose.Presumably, the soluble form is ideal for ini-tial integration into an expanding wall, whilethe less substituted polymer functions in themature wall by interaction between insolublexylan molecules and cellulose (20).

Ferulate esters. As mentioned above, ferulicacid esters are important components of grassarabinoxylans. The feruloyl transferases are notknown, but the transfer takes place intracellu-larly in the Golgi (40, 98). The substrate has notbeen identified either, but feruloyl-CoA is themost likely substrate, although other evidencehas suggested feruloyl-glucoside as a possiblesubstrate (98). In vitro synthesis of feruloylatedarabinoxylan has not been convincingly demon-strated, although some activity using feruloyl-CoA as donor and small oligosaccharides as ac-ceptors has been reported (154). Other workusing microsomal preparations from wheat orparsley has shown that feruloyl-CoA incubationresulted in feruloylation of a protein rather thana polysaccharide (69, 98). We tend to believethat the feruloylated protein is an intermedi-ate in the feruloylation of arabinoxylan. Theferulic acid in wheat microsomes appeared tobe bound to arabinosyl residues on the pro-tein, although the structure was not unambigu-ously identified (N. Obel and H. V. Scheller,unpublished data), and the protein may be RGP.Most likely, feruloyl-arabinose is transferred asa unit from a precursor, perhaps RGP, ontothe acceptor. Mitchell et al. (92) have identified

putative acyl-transferases that are candidate fer-uloyl transferases based on a coexpression studyof rice genes. The candidate proteins are pre-dicted to be cytoplasmic, which is consistentwith a role in feruloylation of a cytoplasmicintermediate such as RGP but not with fer-uloylation of arabinoxylan in the Golgi. Veryrecent data have confirmed that some of thecandidate acyl-transferases are involved in xylanferuloylation (107).

Mannans and Glucomannans

A mannan synthase involved in making thebackbone of galactomannan in guar was shownto be a member of the CSLA family (28). Sub-sequently, several CSLA members have beenshown to have mannan and glucomannan syn-thase activity, apparently being able to utilizeboth GDP-mannose and GDP-glucose as sub-strates (81). Incorporation of glucose and man-nose in glucomannan does not follow a strictpattern but is determined primarily by the rel-ative concentration of the substrates. How-ever, this may not necessarily be true in vivo.The functional equivalence of CSLA proteinshas also been confirmed in vivo since con-structs with 35S promoter used to drive dif-ferent CSLA homologs could complement theembryo-lethal csla7 mutant in arabidopsis (46).

Preliminary data from our laboratory haveindicated that CSLD proteins are also gluco-mannan synthases (150). These results wereobtained with microsomal preparations of ara-bidopsis CSLD5 expressed in tobacco andshould be confirmed with purified protein. Thestructure and phylogeny of CSLD proteins(Figure 5) (4) have led to suggestions that theproteins would be glucan synthases using UDP-glucose as a substrate, so a GDP-sugar depen-dent activity is unexpected.

A galactosyltransferase involved in makinggalactomannan was identified in fenugreek andwas the first GT involved in synthesis of plantcell walls for which the activity of the pure en-zyme was shown (32). Whereas the fenugreekenzyme, which belongs to GT34, is involvedin making seed galactomannan, it appears to

276 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

have orthologs in dicots that have not beenreported to have seed galactomannan, includ-ing arabidopsis. We propose that the appar-ent orthologous GT34s in arabidopsis signifythat the seeds do contain galactomannan. Oilis probably the most important storage com-pound in arabidopsis seeds and the amount ofgalactomannan may be quite low. Seed galac-tomannans are known outside legumes, e.g.,in tobacco and coffee, and these species havegalactosyltransferases similar to the fenugreekenzyme (114, 117). In vitro, the relative concen-trations of UDP-galactose and GDP-mannosedetermine the substitution degree of the galac-tomannans (118). However, for several reasons,mannans are likely to be synthesized as morehighly substituted polymers that are subse-quently trimmed by α-galactosidases. Thus, themannans found in some seeds (e.g., in ivory nut)are very low in galactose and essentially insol-uble, and therefore they are likely synthesizedas more soluble precursors with a much highergalactose substitution. Similarly, coconut has amannan-rich endosperm with a low galactosesubstitution. However, makapuno mutants ofcoconut have a much softer and thicker en-dosperm containing a galactomannan with ahigher galactose/mannose ratio. In the case ofmakapuno coconut, it has been shown that themutant is deficient in a galactosidase (95).

Mixed-Linkage Glucans

The biosynthesis of β-(1→3,1→4)-glucan hasrecently been shown to involve CSLF andCSLH proteins (Figure 5) (14, 29). The corre-sponding gene families are absent in arabidopsisand poplar and present in rice and Brachypodiumconsistent with a grass-specific occurrence. Theinvolvement of CSLF and CSLH was shownby expressing isoforms of the rice proteins inarabidopsis and detecting small amounts of β-(1→3,1→4)-glucan in the transgenic plants. Itis surprising that both types of protein can pro-duce β-(1→3,1→4)-glucan alone. Most likely,the heterologously expressed proteins inter-act with other proteins present in arabidopsis,but obviously the CSLF and CSLH proteins

do not need to be present simultaneously forβ-(1→3,1→4)-glucan synthase activity to oc-cur. In support of this, CSLF and CSLH genesare not coexpressed. Comparison of Brachy-podium, wheat, and barley has recently shownthat CSLF6 is the major CSLF gene expressedin wheat and barley seedlings, while other CSLFisoforms and CSLH are expressed at a very lowlevel (23). In contrast, Brachypodium seedlingsshow a high expression level of CSLH. It will beinteresting to investigate whether these specieshave different β-(1→3,1→4)-glucan fine struc-tures: e.g., perhaps one type of CSL proteinis chiefly involved in making the short β-(1→4)-linked segments and the other CSL pro-tein type is chiefly responsible for making thelonger segments. The ability of arabidopsis tosynthesize β-(1→3,1→4)-glucan when trans-formed with a single gene related to universallypresent CSL genes suggests that evolution of β-(1→3,1→4)-synthesis can take place relativelyeasily and supports the idea that occurrence ofβ-(1→3,1→4)-glucan in very different taxo-nomic groups is the result of convergent evolu-tion. Isoforms of CSLH and CSLF have beenlocalized in the Golgi (29, 35), and in vitro activ-ity studies have been consistent with biosynthe-sis of β-(1→3,1→4)-glucan in the Golgi (44,56). However, unlike other matrix polysaccha-rides, β-(1→3,1→4)-glucan has not been de-tected in the Golgi (148). This would imply ei-ther that the β-(1→3,1→4)-glucan epitope iscompletely masked in Golgi vesicles, perhapsby acetylation, or that some stage of synthesistakes place in the plasma membrane.

The CESA-CSL Superfamily

The cellulose synthase and cellulose synthase–like superfamily of GTs (Figure 5) is involvedin synthesis of cellulose (CESA), hemicellu-loses XyG (CSLC), glucomannan (CSLA andperhaps CSLD), and β-(1→3,1→4)-glucan(CSLF and CSLH). This raises the questionof what the other CSL proteins may do. Mostlikely, they would all be involved in mak-ing hemicelluloses as defined in this review,i.e., β-(1→4)-linked glycans with an equatorial

www.annualreviews.org • Hemicelluloses 277

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

Expansin: a class ofproteins that stimulatecreep in cell walls.They are related tohydrolases of CAZyGH45 but have nohydrolytic activity.Their mode of actionis thought to beunzipping of hydrogenbonds between wallpolysaccharides

configuration at C1 and C4. However, allknown hemicelluloses have already been ac-counted for. Only CESA, CSLA, CSLC, andCSLD proteins are present outside of thespermatophytes, e.g., in Selaginella (119) andPhyscomitrella (120), so presumably the otherCSL groups are either involved in synthesis ofspermatophyte-specific polymers or work to-gether with the more conserved CSL proteins.

FUNCTION OFHEMICELLULOSES

Biological Function—Hemicellulosesin Cell Wall Models

Primary walls and cell expansion. Cell wallmodels that are created to illustrate how wallpolymers are organized in higher-order struc-tures influence the thinking about biologicalfunctions of the hemicelluloses. The first sig-nificant cell wall model viewed the wall as agiant molecule, i.e., with covalent linkages be-tween many noncellulosic polysaccharides (66).A notable feature of this model was a glycosidiclinkage between XyG and galactan side chainsof RG-I. This view contrasted with the sequen-tial extractability of wall polysaccharides thatled to the old hemicellulose definition. Talbott

Middlelamella

Primarycell wall

Plasmamembrane

50 nm

Hemicellulose

Cellulose

Pectin

Figure 7Simplified model of the primary cell wall. Reprinted from McCann andRoberts (85) with permission.

and Ray (131) presented a model that empha-sized separability features of the wall. Accordingto this model cellulose microfibrils were coatedwith several sheaths, with the hemicellulosesforming the innermost and most tightly boundsheath; this model is referred to as the multicoatmodel (25). An attractive feature of the modelwas that its construction outside the protoplastcould be understood in terms of thermodynam-ically reasonable self-assembly mechanisms ofaffinities and phase separation (83). At the sametime McCann & Roberts (85) presented a cellwall model drawn to scale (Figure 7). Thismodel has become the most influential to date.It is also based on noncovalent interactions buthas a particular noncovalent interaction as thecentral tenet: the importance of tethering gly-cans that cross-link cellulose microfibrils. It isreferred to as the sticky network model (25).Tethering glycans, i.e., XyG in typical primarynon-commelinid walls, bind to cellulose mi-crofibrils by two mechanisms: by being trappedin the microfibril during the crystallization justafter synthesis, and by multiple hydrogen bondsbetween XyG and cellulose. The evidence fortethering by XyGs is circumstantial but verystrong: The XyG backbone adopts a helicalconformation in solution, which together withthe arrangements of side chains prevents self-association in solution while at the same time fa-voring adoption of a flat conformation upon in-teracting with cellulose (79). Endo-glucanaseswill more easily access substrate in solutionthan substrate in semicrystalline or adsorpedstates and can thus be used for preferentialextraction of more solvated XyG domains ofthe wall. The part of XyG, which is accessi-ble for release from pea cell wall by endoglu-canase, differs by the predominance of GXXGand GXFG (see Figure 3) from the XyG thatcould only be released by strong alkali (49, 99).The definition and biological significance ofthese distinct XyG domains are corroborated bythe observation that expansins induce creepmore effectively in artificial cellulose/XyGcomposites in which the XyG chains are longenough to tether the microfibrils than in com-posites made with short XyG fragments (145).

278 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

The unequivocal existence of tethering gly-cans does not mean they are indispensable load-bearing structures of the wall. Celery appearsto have dispensed with tethering glycans andfeatures a pure multicoated model wall. Withonly 2% XyG and 2% xylan, there are notenough hemicellulosic polysaccharides to coatthe microfibrils, and no other polysaccharidesappear to have taken the place of the hemicellu-losic polymers (134). A fully viable mutant de-void of any detectable XyG has been createdin arabidopsis (22). Finally, Thompson (136)has calculated that the combined strength ofall XyG/cellulose hydrogen bonds in typical di-cot walls cannot transmit all the stress that thewall sustains. The solution to the latter prob-lem probably rests in the fact that current mod-els ignore direct cellulose microfibril contactsand entanglement, the structures that sustainby far the major load from turgor and small de-formations (146). Another desirable aspect infuture models would be evidence for the pec-tic matrix also tethering microfibrils via RG-I side chains (159, 160). Finally, new modelsshould revisit the issue of covalent linkages be-tween pectin and XyG. The above-mentionedproposal of the original Keegstra et al. (66)model that XyG is synthesized onto an RG-I side chain, e.g., galactan, has more recentlyreceived support. The possible occurrence ofa base-stable linkage between XyG and RG-Iwas rediscovered while developing a XyG la-beling method (33), and the observations inthe following decade pointing to the existenceof such a linkage was reviewed by Mort (94).Popper and Fry (110) found alkali-stable, an-ionic complexes in suspension cultures of bothgrass and non-commelinid seed plant cells, im-plying that the link is also an evolutionarily con-served feature in grasses, which contain five-fold less XyG compared to typical dicot walls.Covalent linkages between matrix polysaccha-rides are often thought of as wall maturationphenomena occurring by transglycosylation ortransesterification in the wall, but XyG and RG-I may be linked already in the Golgi (111). Thenature of the linkage is unknown, as is the modeof its synthesis—whether by synthesizing XyG

onto an RG-I side chain or by a transglycosy-lation reaction. Resolving these questions re-mains a major challenge in the field of hemicel-lulose biochemistry.

Cell walls grow at constant wall thicknessif carbon is not limiting (8). The concurrentinsertion of new building blocks should notstrengthen the wall so as to impair cell expan-sion. Whether insertion of new wall polysac-charides even contributes to wall creep is stillunresolved (130, 137), but it is widely acceptedthat the insertion of newly synthesized XyGoccurs by transglycosylation catalyzed by XET(26). Attempts to identify transglycosylases thatact on other polysaccharides have been met withlimited success. A glucomannan transglycosy-lase has been reported (122), while the searchfor a homogalacturonan transglycosylase wasnegative (42). The situation in the grasses isquite enigmatic. β-(1→3,1→4)-glucans are de-posited during cell expansion only to be de-graded and replaced with GAX as expansionceases. Measurements of the β-(1→3,1→4)-glucan content in developing wheat, barley,and Brachypodium seedlings during develop-ment demonstrate net breakdown as opposedto simple dilution by xylan (23). The hydro-lases that are responsible for degradation ofβ-(1→3,1→4)-glucan belong to CAZy familyGH16. XETs, the XyG transglycosylases, arealso classified in this family, and it would beconceivable that some of the grass members ofGH16 could transglucosylate β-(1→3,1→4)-glucan, but this appears not to be the case(39, 58). Grasses thus appear to rely on a β-(1→3,1→4)-glucan/xylan replacement ratherthan transglycosylation, and a xylan transglyco-sylase remains to be identified. In dicots, xylan ismostly deposited in secondary walls, i.e., whereexpansion has ceased. We speculate that dicotsdo not have a xylan transglycosylase and hencerestrict the use of substantial amounts of xylanto walls that do not expand. Recently, a sur-prising new class of transglycosylases has beenidentified. These are the mixed-linkage beta-glucan:XyG endotransglycosylases, or MXEs.These enzymes, which graft β-(1→3,1→4)-glucan onto XyG, have been found in horsetail

www.annualreviews.org • Hemicelluloses 279

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

and in Choleochaete, a Charophycean green alga,but not in higher plants (39).

Secondary walls and xylans. Cell wall mod-els depicted in textbooks often focus on the pri-mary wall. The reason for this is that primarycell wall models inspire working hypotheses re-garding mechanisms required for cell growth.Xylan thus receives less attention though it is amajor load-bearing structure in grasses as ex-plained above. The important role of xylansin strengthening secondary walls is very clearfrom analysis of xylan-deficient mutants. Allthese contain collapsed xylem vessels and haveseverely impacted growth and fertility. Sincesuch mutants grow better at high humidity, itseems that the main reason for the poor growthis that water transport is adversely affected bythe poorly developed xylem. Xylans are alsoabundant in fiber cells, which do not have awater-conducting role. Therefore, xylan in thistype of cell might be more dispensable, but thishas not been tested.

Mannans. Apart from the role of galactoman-nans as seed storage compounds, it is un-clear what specific roles the mannans have.Mannans/glucomannans are highly conservedthrough plant evolution. However, the onlyknown effect of mutating mannan biosyntheticgenes is the embryo-lethal phenotype of thearabidopsis cslA7 mutant (47). In contrast, dele-tion of the three glucomannan synthase genesexpressed in arabidopsis stems resulted in plantslacking detectable glucomannan but withoutobvious phenotype (46). Mutants in CSLDgenes show severe effects on tip growth, eitherin root hair or pollen tubes (5), and mutants inseveral CSLD genes are severely dwarfed (150).However, while preliminary evidence suggeststhat mannan biosynthesis is affected in the mu-tants, this still needs additional work.

In conclusion, hemicelluloses play a key rolein both primary and secondary walls. XyG holdsa special position in the cell wall, as reflected inthe repertoire of enzymes available for its ma-nipulation. This applies also to the commelinidmonocots, which express a full complement of

XETs, though XyG quantitatively is a minorhemicellulose.

Source of Signal Molecules

XyG plays an additional role as a source ofsignal molecules. Breakdown products of XyG,most notably XXFG, were demonstrated tocounteract auxin-induced cell expansion (151).This finding was verified and extended in aseries of seminal papers by McDougall and Fry(86, 87). It is remarkable that the optimal con-centration for the inhibitory activity of XXFGis as low as 1 nM, and the term oligosaccharinswas coined for these signal molecules. OtherXyG oligomers, XXLG and XLLG in partic-ular, can promote cell expansion, albeit at a1000-fold higher concentration (88). However,this response is not regarded as signaling butrelated to the action of XET. Research inXyG oligomers as substrates for XET hasreceived much attention, whereas studies of theoligosaccharin inhibitory activities dwindled atthe end of the nineties, and it is thus still notclear how important these responses are in vivo.Mutants in XyG biosynthesis can have verystrong phenotypes (132, 133), but as mentionedabove, mutants devoid of XyG do not (22).Perhaps the modified XyG in single mutantsgives rise to different or increased amounts ofsignal molecules and hence invokes a strongerresponse than wild-type plants. On the otherhand, the lack of obvious phenotype of themur2 mutant, which lacks XXFG (138), and ofxxt1/xxt2 mutants devoid of XyG altogether(22) suggests that if signaling by XyG fragmentsplays a role in plant growth and development, itis not a very important one. The small amountof XyG present in grass cell walls would seemunlikely to have an important structural role.Perhaps XyG has been maintained in grassesfor two reasons that are not mutually exclusive:because the XyG-derived oligosaccharinsindeed are indispensable, and because theprocess of de novo wall formation during celldivision generally is more conserved amongangiosperm families than is the mature primarywall structure and hence XyG is required forwall assembly during cytokinesis in grasses.

280 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

Information on a signaling role for hemicel-luloses other than XyG is very limited. In con-trast, many studies have shown that fragmentsof pectin play a role in the signaling of pathogenattack and induce a pathogen response (7, 30).

Seed Storage Carbohydrates

Hemicelluloses in the cell wall have the primaryrole of interacting with other polymers to en-sure the proper physical properties of the wall.However, in a large number of cases, hemicel-luloses have been recruited to the function ofseed storage carbohydrate (90, 116). This hashappened independently many times in evo-lution, and it has been suggested that from ataxonomic viewpoint hemicelluloses are as im-portant as starch is in the role as storage car-bohydrate in seeds (90). Much of our knowl-edge of hemicelluloses comes from the study ofseed polysaccharides rather than the polymersin vegetative tissues. XyG is abundant in plantssuch as nasturtium and tamarind. Galactoman-nans are known from a large number of eco-nomically important plants, e.g., coconut, guar,and locust bean. Galactomannans are especiallyabundant in the endosperm of legumes but alsooccur in other seeds. Glucomannan is present inthe konjak plant (Amorphophallus konjac); in thiscase the storage organ is a corm and not theseed. Arabinoxylans are present in seeds of di-cots such as flax and psyllium (36, 96) and alsoin cereal endosperm. Cereal endosperm addi-tionally contains β-(1→3,1→4)-glucans.

ENGINEERING OF CELL WALLS

Hemicelluloses have commercial significance.Seed storage hemicelluloses are used directlyas products in the food industry, e.g., guar

and locust bean gums (galactomannans), kon-jak gum (glucomannan), and tamarind gum(xyloglucan). In addition, the hemicellulosesgive important properties to many food andfeed products. In the baking industry, theinsoluble arabinoxylans affect baking quality.β-(1→3,1→4)-glucans and arabinoxylans arewell-known antinutritional compounds in an-imal feed, and they can cause filtering andhaze problems in the brewery industry dueto their viscosity. To alleviate these problems,hemicellulose-degrading enzymes are added tofeed and are used in the baking and breweryindustries. On the other hand, β-(1→3,1→4)-glucan has a documented cholesterol-loweringeffect in hypercholesterolemic humans (67) anddaily intake of β-(1→3,1→4)-glucans is recom-mended by the U.S. Food and Drug Adminis-tration. In cellulosic biofuel production, hemi-celluloses affect the saccharification of biomass,and the released sugars, largely pentoses, areless desirable for fermentation than hexoses.

It is beyond the scope of this review to dis-cuss all of these industrial applications. How-ever, due to the economic and nutritional im-portance of hemicelluloses, several researchershave attempted or suggested modification inthe hemicellulose composition of plants. Xy-lan feruloylation has been decreased by express-ing ferulic acid esterases in transgenic plantswith mixed results (11, 12, 51). To overcomethe problem of high-pentose content in bio-fuel production, it has been suggested that xy-lan could be replaced with mannan (101). Theapparently highly specialized function of xylansin vessels and lignin interactions suggests to usthat a complete substitution could be very dif-ficult. Instead, a more specific replacement infiber cells would seem to have a higher chanceof success.

SUMMARY POINTS

1. Hemicelluloses are wall polysaccharides, which are characterized by β-(1→4)-linkedbackbones of sugars in an equatorial configuration. This definition includes xyloglucans,xylans, mannans and glucomannans, and β-(1→3,1→4)-glucans.

www.annualreviews.org • Hemicelluloses 281

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

2. All the hemicelluloses show differences in structural details between different species andin different cell types within plants.

3. The main role of hemicelluloses is to tether cellulose microfibrils, thereby strengtheningthe cell wall.

4. Xyloglucans dominate in primary walls of dicots and conifers, whereas (glu-curono)arabinoxylans dominate in commelinid monocots.

5. Hemicelluloses are synthesized by glycosyltransferases located in the Golgi membranes.The backbones of xyloglucan, mannan, and β-(1→3,1→4)-glucans are synthesized bymembers of the cellulose synthase–like gene family, which are multimembrane-spanningproteins.

6. Xylan backbones are apparently synthesized by glycosyltransferases that are Type IImembrane proteins with a single membrane-spanning segment, but none of these hasbeen unambiguously identified. Xylans in many, if not all, plants have a unique tetrasac-charide in the reducing end, which appears to be involved in biosynthesis either as aprimer or a termination sequence.

7. Hemicelluloses are important components in food and feed and constitute a major partof lignocellulosic biomass.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings thatmight be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

The work of H. V. Scheller on hemicelluloses is supported by the U.S. Department of Energy,Office of Science, Office of Biological and Environmental Research, through contract DE-AC02-05CH11231 between Lawrence Berkeley National Laboratory and the U.S. Department of En-ergy. The work of P. Ulvskov is supported via grants from the Villum-Kann Rasmussen Foundationand from the Danish Agency for Science, Technology and Innovation. The work of both authorsis supported by a grant from the 7th Framework Programme of the European Commission to theproject “Renewall,” project no. 211982.

LITERATURE CITED

1. Andersson SI, Samuelson O, Ishihara M, Shimizu K. 1983. Structure of the reducing end-groups inspruce xylan. Carbohydr. Res. 111:283–88

2. Bauer S, Vasu P, Persson S, Mort AJ, Somerville CR. 2006. Development and application of a suite ofpolysaccharide-degrading enzymes for analyzing plant cell walls. Proc. Natl. Acad. Sci. USA 103:11417–22

3. Baydoun EAH, Waldron KW, Brett ET. 1989. The interaction of xylosyltransferase and glucuronyl-transferase involved in glucuronoxylan synthesis in pea (Pisum-sativum) epicotyls. Biochem. J. 257:853–58

4. Bernal AJ, Jensen JK, Harholt J, Sørensen SO, Møller I, et al. 2007. Disruption of ATCSLD5 results inreduced growth, reduced xylan and homogalacturonan synthase activity and altered xylan occurrence inArabidopsis. Plant J. 52:791–802

282 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

5. Bernal AJ, Yoo C-M, Mutwil M, Jensen JK, Hou G, et al. 2008. Functional analysis of the cellulosesynthase like genes ATCSLD1, ATCSLD2 and ATCSLD4 in tip growing Arabidopsis cells. Plant Physiol.148:1238–53

6. Bochicchio R, Reicher F. 2003. Are hemicelluloses from Podocarpus lambertii typical of gymnosperms?Carbohydr. Polym. 53:127–36

7. Boller T, Felix G. 2009. A Renaissance of elicitors: perception of microbe-associated molecular patternsand danger signals by pattern-recognition receptors. Annu. Rev. Plant Biol. 60:379–406

8. Bretharte MS, Talbott LD. 1993. Changes in composition of the outer epidermal-cell wall of pea stemsduring auxin-induced growth. Planta 190:369–78

9. Brown DM, Goubet F, Vicky WWA, Goodacre R, Stephens E, et al. 2007. Comparison of five xylansynthesis mutants reveals new insight into the mechanisms of xylan synthesis. Plant J. 52:1154–68

10. Brown DM, Zhang ZN, Stephens E, Dupree P, Turner SR. 2009. Characterization of IRX10 and IRX10-like reveals an essential role in glucuronoxylan biosynthesis in Arabidopsis. Plant J. 57:732–46

11. Buanafina MM, Langdon T, Hauck B, Dalton SJ, Morris P. 2006. Manipulating the phenolic acid contentand digestibility of Italian ryegrass (Lolium multiflorum) by vacuolar-targeted expression of a fungal ferulicacid esterase. Appl. Biochem. Biotechnol. 130:416–26

12. Buanafina MM, Langdon T, Hauck B, Dalton S, Morris P. 2008. Expression of a fungal ferulic acidesterase increases cell wall digestibility of tall fescue (Festuca arundinacea). Plant Biotechnol. J. 6:264–80

13. Burget EG, Verma R, Molhoj M, Reiter WD. 2003. The biosynthesis of L-arabinose in plants: molecularcloning and characterization of a golgi-localized UDP-D-xylose 4-epimerase encoded by the MUR4 geneof Arabidopsis. Plant Cell 15:523–31

14. Burton RA, Wilson SM, Hrmova M, Harvey AJ, Shirley NJ, et al. 2006. Cellulose synthase-like CslFgenes mediate the synthesis of cell wall (1,3;1,4)-beta-D-glucans. Science 311:1940–42

15. Campbell CS, Kellogg EA, Stevens PF, Donoghue MJ, Judd WS. 1999. Plant Systematics: A PhylogeneticApproach. Sunderland, MA: Sinauer. 2nd ed.

16. Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, et al. 2009. The carbohydrate-activeenzymes database (CAZy): an expert resource for glycogenomics. Nucleic Acids Res. 37:D233–38

17. Capek P, Kubackova M, Alfoldi J, Bilisics L, Liskova D, Kakoniova D. 2000. Galactoglucomannan fromthe secondary cell wall of Picea abies L. Karst. Carbohydr. Res. 329:635–45

18. Carnachan SM, Harris PJ. 2000. Polysaccharide compositions of primary cell walls of the palms Phoenixcanariensis and Rhopalostylis sapida. Plant Physiol. Biochem. 38:699–708

19. Carnachan SM, Harris PJ. 2000. Ferulic acid is bound to the primary cell walls of all gymnospermfamilies. Biochem. Syst. Ecol. 28:865–79

20. Carpita NC, Gibeaut DM. 1993. Structural models of primary cell walls in flowering plants—consistencyof molecular structure with the physical properties of the walls during growth. Plant J. 3:1–30

21. Cavalier DM, Keegstra K. 2006. Two xyloglucan xylosyltransferases catalyze the addition of multiplexylosyl residues to cellohexaose. J. Biol. Chem. 281:34197–207

22. Cavalier DM, Lerouxel O, Neumetzler L, Yamauchi K, Reinecke A, et al. 2008. Disrupting two Ara-bidopsis thaliana xylosyltransferase genes results in plants deficient in xyloglucan, a major primary cellwall component. Plant Cell 20:1519–37

23. Christensen U. 2009. (1,3;1,4)- β-D-glucan and arabinoxylan in grasses—biosynthesis, function and application.PhD thesis. Univ. Copenhagen, Frederiksberg, Den.

24. Cocuron JC, Lerouxel O, Drakakaki G, Alonso AP, Liepman AH, et al. 2007. A gene from the cellulosesynthase-like C family encodes a β-1,4 glucan synthase. Proc. Natl. Acad. Sci. USA 104:8550–55

25. Cosgrove DJ. 2000. Expansive growth of plant cell walls. Plant Physiol. Biochem. 38:109–2426. Cosgrove DJ. 2005. Growth of the plant cell wall. Nat. Rev. Mol. Cell Biol. 6:850–6127. Darvill JE, McNeil M, Darvill AG, Albersheim P. 1980. Structure of plant cell walls. XI. Glucuronoara-

binoxylan, a second hemicellulose in the primary cell walls of suspension-cultured sycamore cells. PlantPhysiol. 66:1135–39

28. Dhugga KS, Barreiro R, Whitten B, Stecca K, Hazebroek J, et al. 2004. Guar seed beta-mannan synthaseis a member of the cellulose synthase super gene family. Science 303:363–66

www.annualreviews.org • Hemicelluloses 283

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

29. Doblin MS, Pettolino FA, Wilson SM, Campbell R, Burton RA, et al. 2009. A barley cellulose synthase-like CSLH gene mediates (1,3;1,4)-beta-D-glucan synthesis in transgenic Arabidopsis. Proc. Natl. Acad.Sci. USA 106:5996–6001

30. D’Ovidio R, Mattei B, Roberti S, Bellincampi D. 2004. Polygalacturonases, polygalacturonase-inhibitingproteins and pectic oligomers. Biochim. Biophys. Acta 1696:237–44

31. Ebringerova A, Hromadkova Z, Heinze T. 2005. Hemicellulose. Adv. Polym. Sci. 186:1–6732. Edwards ME, Dickson CA, Chengappa S, Sidebottom C, Gidley MJ, et al. 1999. Molecular characteri-

zation of a membrane-bound galactosyltransferase of plant cell wall matrix polysaccharide biosynthesis.Plant J. 19:691–97

33. El Rassi Z, Tedford D, An JH, Mort A. 1991. High-performance reversed-phase chromatographic map-ping of 2-pyridylamino derivatives of xyloglucan oligosaccharides. Carbohydr. Res. 215:25–38

34. Faik A, Price NJ, Raikhel NV, Keegstra K. 2002. An Arabidopsis gene encoding an α-xylosyltransferaseinvolved in xyloglucan biosynthesis. Proc. Natl. Acad. Sci. USA 99:7797–802

35. Fincher GB. 2009. Revolutionary times in our understanding of cell wall biosynthesis and remodelingin the grasses. Plant Physiol. 149:27–37

36. Fischer MH, Yu N, Gray GR, Ralph J, Anderson L, et al. 2004. The gel-forming polysaccharide ofpsyllium husk (Plantago ovata Forsk). Carbohydr. Res. 339:2009–17

37. Fry SC. 2000. The Growing Plant Cell Wall: Chemical and Metabolic Analysis. Caldwell, NJ: The BlackburnPress. 330 pp.

38. Fry SC, Mohler KE, Nesselrode BHWA, Frankova L. 2008. Mixed-linkage (1→3. 1→4) beta-D-glucan: xyloglucan endotransglucosylase, a novel wall-remodelling enzyme from Equisetum (horsetails)and charophytic algae. Plant J. 55:240–52

39. Fry SC, Nesselrode BHWA, Miller JG, Mewburn BR. 2008. Mixed-linkage (1→3. 1→4)-beta-D-glucanis a major hemicellulose of Equisetum cell walls. New Phytol. 179:104–15

40. Fry SC, Willis SC, Paterson AEJ. 2000. Intraprotoplasmic and wall-localised formation of arabinoxylan-bound diferulates and larger ferulate coupling-products in maize cell-suspension cultures. Planta211:679–92

41. Fry SC, York WS, Albersheim P, Darvill A, Hayashi T, et al. 1993. An unambiguous nomenclature forxyloglucan-derived oligosaccharides. Physiol. Plant. 89:1–3

42. Garcia-Romera I, Fry SC. 1994. Absence of transglycosylation with oligogalacturonides in plant cells.Phytochemistry 35:67–72

43. Gibeaut DM, Carpita NC. 1991. Tracing cell wall biogenesis in intact cells and plants. Selective turnoverand alteration of soluble and cell wall polysaccharides in grasses. Plant Physiol. 97:551–61

44. Gibeaut DM, Carpita NC. 1993. Synthesis of (1→3),(1→4)-β-D-glucan in the Golgi apparatus of maizecoleoptiles. Proc. Natl. Acad. Sci. USA 90:3850–54

45. Gibeaut DM, Pauly M, Bacic A, Fincher GB. 2005. Changes in cell wall polysaccharides in developingbarley (Hordeum vulgare) coleoptiles. Planta 221:729–38

46. Goubet F, Barton CJ, Mortimer JC, Yu X, Zhang Z, et al. 2009. Cell wall glucomannan in Arabidopsisis synthesised by CSLA glycosyltransferases, and influences the progression of embryogenesis. Plant J.60:527–38

47. Goubet F, Misrahi A, Park SK, Zhang ZN, Twell D, et al. 2003. AtCSLA7, a cellulose synthase-likeputative glycosyltransferase, is important for pollen tube growth and embryogenesis in Arabidopsis. PlantPhysiol. 131:547–57

48. Grabber JH. 2005. How do lignin composition, structure, and cross-linking affect degradability? A reviewof cell wall model studies. Crop Sci. 45:820–31

49. Guillen R, York WS, Pauly M, An JH, Impallomeni G, et al. 1995. Metabolism of xyloglucan generatesxylose-deficient oligosaccharide subunits of this polysaccharide in etiolated peas. Carbohydr. Res. 77:291–311

50. Harholt J, Ulvskov P, Sørensen I, Willats W, Petersen BL, et al. 2008. Cell wall composition andglycosyltransferases involved in cell wall formation. http://wiki.genomics.purdue.edu/index.php/Cellwall composition and glycosyltransferases involved in cell wall formation

284 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

51. Harholt J, Bach IC, Lind-Bouquin S, Nunan KJ, Madrid SM, et al. 2010. Generation of transgenicwheat (Triticum aestivum L.) accumulating heterologous endo-xylanase and ferulic acid esterase in theendosperm. Plant Biotechnol. J.: In press

52. Harper AD, Bar-Peled M. 2002. Biosynthesis of UDP-xylose. Cloning and characterization of a novelArabidopsis gene family, UXS, encoding soluble and putative membrane-bound UDP-glucuronic aciddecarboxylase isoforms. Plant Physiol. 130:2188–98

53. Harris PJ, Kelderman MR, Kendon MF, McKenzie RJ. 1997. Monosaccharide compositions of unligni-fied cell walls of monocotyledons in relation to the occurrence of wall-bound ferulic acid. Biochem. Syst.Ecol. 25:167–79

54. Hatfield RD, Marita JM, Frost K. 2008. Characterization of p-coumarate accumulation, p-coumaroyltransferase, and cell wall changes during the development of corn stems. J. Sci. Food. Agric. 88:2529–37

55. Hayashi T. 1989. Xyloglucans in the primary cell wall. Annu. Rev. Plant Physiol. Plant Mol. Biol. 40:139–6856. Henry RJ, Schibeci A, Stone BA. 1983. Localization of β-glucan synthases on the membranes of cultured

Lolium multiflorum (ryegrass) endosperm cells. Biochem. J. 209:627–3357. Hoffman M, Jia ZH, Pena MJ, Cash M, Harper A, et al. 2005. Structural analysis of xyloglucans in the

primary cell walls of plants in the subclass Asteridae. Carbohydr. Res. 340:1826–4058. Hrmova M, Farkas V, Lahnstein J, Fincher GB. 2007. A barley xyloglucan xyloglucosyl transferase

covalently links xyloglucan, cellulosic substrates, and (1,3;1,4)-beta-D-glucans. J. Biol. Chem. 282:12951–62

59. Ishii T. 1997. Structure and functions of feruloylated polysaccharides. Plant Sci. 127:111–2760. Ishii T, Tobita T. 1993. Structural characterization of feruloyl oligosaccharides from spinach-leaf cell-

walls. Carbohydr. Res. 248:179–9061. Iwai H, Masaoka N, Ishii T, Satoh S. 2002. A pectin glucuronyltransferase gene is essential for intercel-

lular attachment in the plant meristem. Proc. Natl. Acad. Sci. USA 99:16319–2462. Janbon G, Himmelreich W, Moyrand F, Improvisi L, Dromer F. 2008. Cas1p is a membrane protein

necessary for the O-acetylation of the Cryptococcus neoformans capsular polysaccharide. Mol. Microbiol.42:453–67

63. Johansson MH, Samuelson O. 1977. Reducing end groups in birch xylan and their alkaline degradation.Wood Sci. Technol. 11:251–63

64. Jones L, Milne JL, Ashford D, McCann MC, McQueen-Mason SJ. 2005. A conserved functional role ofpectic polymers in stomatal guard cells from a range of plant species. Planta 221:255–64

65. Kato Y, Ito S, Iki K, Matsuda K. 1982. Xyloglucan and beta-D-glucan in cell-walls of rice seedlings.Plant Cell Physiol. 23:351–64

66. Keegstra K, Talmadge KW, Bauer WD, Albersheim P. 1973. Structure of plant-cell walls. 3. Model ofwalls of suspension-cultured sycamore cells based on interconnections of macromolecular components.Plant Physiol. 51:188–96

67. Keogh GF, Cooper GJ, Mulvey TB, McArdle BH, Coles GD, et al. 2003. Randomized controlledcrossover study of the effect of a highly beta-glucan-enriched barley on cardiovascular disease risk factorsin mildly hypercholesterolemic men. Am. J. Clin. Nutr. 78:711–18

68. Kiefer LL, York WS, Darvill AG, Albersheim P. 1989. Xyloglucan isolated from suspension-culturedsycamore cell walls is O-acetylated. Phytochemistry 28:2105–7

69. Kohler A, Kauss H. 1997. Transfer of hydroxycinnamoyl residues to microsomal proteins from parsley.Phytochemistry 44:225–28

70. Konishi T, Ono H, Ohnishi-Kameyama M, Kaneko S, Ishii T. 2006. Identification of a mung beanarabinofuranosyltransferase that transfers arabinofuranosyl residues onto (1,5)-linked alpha-L-arabino-oligosaccharides. Plant Physiol. 141:1098–105

71. Konishi T, Takeda T, Miyazaki Y, Ohnishi-Kameyama M, Hayashi T, et al. 2007. A plant mutase thatinterconverts UDP-arabinofuranose and UDP-arabinopyranose. Glycobiology 17:345–54

72. Kubackova M, Karacsonyi S, Bilisics L. 1992. Structure of galactoglucomannan from Populus moniliferaH. Carbohydr. Polym. 19:125–29

73. Labavitch JM, Ray PM. 1978. Structure of hemicellulosic polysaccharides of Avena sativa coleoptilecell-walls. Phytochemistry 17:933–37

www.annualreviews.org • Hemicelluloses 285

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

74. Lao NT, Long D, Kiang S, Coupland G, Shoue DA, et al. 2003. Mutation of a family 8 glycosyltrans-ferase gene alters cell wall carbohydrate composition and causes a humidity-sensitive semisterile dwarfphenotype in Arabidopsis. Plant Mol Biol. 53:687–701

75. Lee CH, O’Neill MA, Tsumuraya Y, Darvill AG, Ye ZH. 2007. The irregular xylem9 mutant is deficientin xylan xylosyltransferase activity. Plant Cell Physiol. 48:1624–34

76. Lee CH, Teng Q, Huang WL, Zhong RQ, Ye ZH. 2009. The F8H glycosyltransferase is a functionalparalog of fra8 involved in glucuronoxylan biosynthesis in Arabidopsis. Plant Cell Physiol. 50:812–27

77. Lee CH, Zhong RQ, Richardson EA, Himmelsbach DS, McPhail BT, et al. 2007. The PARVUS gene isexpressed in cells undergoing secondary wall thickening and is essential for glucuronoxylan biosynthesis.Plant Cell Physiol. 48:1659–72

78. Levigne SV, Ralet MCJ, Quemener BC, Pollet BNL, Lapierre C, et al. 2004. Isolation from sugar beetcell walls of arabinan oligosaccharides esterified by two ferulic acid monomers. Plant Physiol. 134:1173–80

79. Levy S, York WS, Stuikeprill R, Meyer B, Staehelin LA. 1991. Xyloglucan—the role of the fucosylatedside-chain in surface-specific side-chain folding. Plant J. 1:195–215

80. Li XM, Cordero I, Caplan J, Molhoj M, Reiter WD. 2004. Molecular analysis of 10 coding regions fromArabidopsis that are homologous to the MUR3 xyloglucan galactosyltransferase. Plant Physiol. 134:940–50

81. Liepman AH, Wilkerson CG, Keegstra K. 2005. Expression of cellulose synthase-like (Csl) genes in insectcells reveals that CslA family members encode mannan synthases. Proc. Natl. Acad. Sci. USA 102:2221–26

82. Lopez-Casado G, Urbanowicz BR, Damasceno CMB, Rose JKC. 2008. Plant glycosyl hydrolases andbiofuels: a natural marriage. Curr. Opin. Plant Biol. 11:329–37

83. MacDougall AJ, Rigby NM, Ring SG. 1997. Phase separation of plant cell wall polysaccharides and itsimplications for cell wall assembly. Plant Physiol. 114:353–62

84. Madson M, Dunand C, Li X, Verma R, Vanzin GF, et al. 2003. The MUR3 gene of Arabidopsis encodes axyloglucan galactosyltransferase that is evolutionarily related to animal exostosin. Plant Cell 15:1662–70

85. McCann MC, Roberts K. 1991. Architecture of the primary cell wall. In The Cytoskeletal Basis of PlantGrowth and Form, ed. CW Lloyd, pp. 109–29. New York: Academic

86. McDougall GJ, Fry SC. 1988. Inhibition of auxin-stimulated growth of pea stem segments by a specificnonasaccharide of xyloglucan. Planta 175:412–16

87. McDougall GJ, Fry SC. 1989. Structure-activity relationships for xyloglucan oligosaccharides with an-tiauxin activity. Plant Physiol. 89:883–87

88. McDougall GJ, Fry SC. 1990. Xyloglucan oligosaccharides promote growth and activate cellulase: evi-dence for a role of cellulase in cell expansion. Plant Physiol. 93:1042–48

89. McDougall GJ, Fry SC. 1994. Fucosylated xyloglucan in suspension-cultured cells of the gramineousmonocotyledon, Festuca arundinacea. J. Plant Physiol. 143:591–95

90. Meier H, Reid JSG. 1982. Reserve polysaccharides other than starch in higher plants. In Encyclopedia ofPlant Physiology. NS 13A: Plant Carbohydrates I, ed. FA Loewus, W Tanner, pp. 418–71. Berlin: Springer

91. Minic Z. 2008. Physiological roles of plant glycoside hydrolases. Planta 227:723–4092. Mitchell RAC, Dupree P, Shewry P. 2007. A novel bioinformatics approach identifies candidate genes

for the synthesis and feruloylation of arabinoxylan. Plant Physiol. 144:43–5393. Moller I, Sørensen I, Bernal AJ, Blaukopf C, Lee K, et al. 2007. High-throughput mapping of cell wall

polymers within and between plants using novel microarrays. Plant J. 50:1118–2894. Mort AJ. 2002. Interactions between pectins and other polymers. In Pectins and Their Manipulation, ed.

GB Seymour, JP Knox, pp. 30–51. Sheffield, UK: Blackwell95. Mujer CV, Ramirez DA, Mendoza EMT. 1984. Alpha-D-galactosidase deficiency in coconut endosperm:

its possible pleiotropic effects in makapuno. Phytochemistry 23:893–9496. Naran R, Chen G, Carpita NC. 2008. Novel rhamnogalacturonan I and arabinoxylan polysaccharides

of flax seed mucilage. Plant Physiol. 148:132–4197. Obel N, Porchia AC, Scheller HV. 2002. Dynamic changes in cell wall polysaccharides during wheat

seedling development. Phytochemistry 60:603–1098. Obel N, Porchia AC, Scheller HV. 2003. Intracellular feruloylation of arabinoxylan in wheat: evidence

for feruloyl-glucose as precursor. Planta 216:620–2999. Pauly M, Albersheim P, Darvill A, York WS. 1999. Molecular domains of the cellulose/xyloglucan

network in the cell walls of higher plants. Plant J. 20:629–39

286 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

100. Pauly M, Gunl M, Souza A, Neumetzler L, Florian K. 2009. The substitution pattern of plant cell wallcross-linking glycans is determined by apoplastic glycosidases. Presented at Annu. Meet. Am. Soc. Plant Biol.,Honolulu, Abstr. P15004

101. Pauly M, Keegstra K. 2008. Cell-wall carbohydrates and their modification as a resource for biofuels.Plant J. 54:559–68

102. Pauly M, Qin Q, Greene H, Albersheim P, Darvill A, York WS. 2001. Changes in the structure ofxyloglucan during cell elongation. Planta 212:842–50

103. Pauly M, Scheller HV. 2000. O-acetylation of plant cell wall polysaccharides. Identification and partialcharacterization of a rhamnogalacturonan o-acetyl-transferase from potato suspension cultured cell.Planta 210:659–67

104. Pena MJ, Darvill AG, Eberhard S, York WS, O’Neill MA. 2008. Moss and liverwort xyloglucans containgalacturonic acid and are structurally distinct from the xyloglucans synthesized by hornworts and vascularplants. Glycobiology 18:891–904

105. Pena MJ, Zhong R, Zhou GK, Richardson EA, O’Neill MA, et al. 2007. Arabidopsis irregular xylem8 andirregular xylem9: Implications for the complexity of glucuronoxylan biosynthesis. Plant Cell 19:549–63

106. Perrin RM, DeRocher AE, Bar-Peled M, Zeng W, Norambuena L, et al. 1999. Xyloglucan fucosyltrans-ferase, an enzyme involved in plant cell wall biosynthesis. Science 284:1976–79

107. Piston F, Uauy C, Fu L, Langston J, Labavitch J, et al. 2010. Down-regulation of four putative arabi-noxylan feruloyl transferase genes from family PF02458 reduces ester-linked ferulate content in rice cellwalls. Planta: In press; doi: 10.1007/s00425-009-1077-1

108. Popper ZA. 2008. Evolution and diversity of green plant cell walls. Curr. Opin. Plant Biol. 11:286–92109. Popper ZA, Fry SC. 2003. Primary cell wall composition of bryophytes and charophytes. Ann. Bot.

91:1–12110. Popper ZA, Fry SC. 2005. Widespread occurrence of a covalent linkage between xyloglucan and acidic

polysaccharides in suspension-cultured angiosperm cells. Ann. Bot. 96:91–99111. Popper ZA, Fry SC. 2008. Xyloglucan-pectin linkages are formed intraprotoplasmically, contribute to

wall-assembly, and remain stable in the cell wall. Planta 227:781–94112. Porchia AC, Sørensen SO, Scheller HV. 2002. Arabinoxylan biosynthesis in wheat. Characterization of

arabinosyltransferase activity in Golgi membranes. Plant Physiol. 130:432–41113. Prasad V, Stromberg CAE, Alimohammadian H, Sahni A. 2005. Dinosaur coprolites and the early

evolution of grasses and grazers. Science 310:1177–80114. Pre M, Caillet V, Sobilo J, McCarthy J. 2008. Characterization and expression analysis of genes directing

galactomannan synthesis in coffee. Ann. Bot. 102:207–20115. Ralet MC, Andre-Leroux G, Quemener B, Thibault JF. 2005. Sugar beet (Beta vulgaris) pectins are

covalently cross-linked through diferulic bridges in the cell wall. Phytochemistry 66:2800–14116. Reid JSG. 1985. Cell wall storage carbohydrates in seeds—biochemistry of the seed “gums” and “hemi-

celluloses.” Adv. Bot. Res. 11:125–55117. Reid JSG, Edwards ME, Dickson CA, Scott C, Gidley MJ. 2003. Tobacco transgenic lines that express

fenugreek galactomannan galactosyltransferase constitutively have structurally altered galactomannansin their seed endosperm cell walls. Plant Physiol. 131:1487–95

118. Reid JSG, Edwards M, Gidley MJ, Clark AH. 1995. Enzyme specificity in galactomannan biosynthesis.Planta 195:489–95

119. Roberts A. 2008. Cellulose synthase superfamily. http://wiki.genomics.purdue.edu/index.php/Cellulose synthase superfamily

120. Roberts AW, Bushoven JT. 2007. The cellulose synthase (CESA) gene superfamily of the mossPhyscomitrella patens. Plant Mol. Biol. 63:207–19

121. Scheller HV, Doong RL, Ridley BL, Mohnen D. 1999. Pectin biosynthesis: A solubilized α1,4-galacturonosyltransferase from tobacco catalyzes the transfer of galacturonic acid from UDP-galacturonic acid onto the nonreducing end of homogalacturonan. Planta 207:512–17

122. Schroder R, Wegrzyn TF, Bolitho KM, Redgwell RJ. 2004. Mannan transglycosylase: a novel enzymeactivity in cell walls of higher plants. Planta 219:590–600

123. Shatalov AA, Evtuguin DV, Neto CP. 1999. (2-O-α-D-Galactopyranosyl-4-O-methyl-α-D-glucurono)-D-xylan from Eucalyptus globulus Labill. Carbohydr. Res. 320:93–99

www.annualreviews.org • Hemicelluloses 287

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

124. Shimizu K, Ishihara M, Ishihara T. 1976. Hemicellulases of brown rotting fungus, Tyromyces palustris. II.The oligosaccharides from the hydrolysate of a hardwood xylan by the intracellular xylanase. MokuzaiGakkaishi 22:618–25

125. Smith BG, Harris PJ. 1999. The polysaccharide composition of Poales cell walls: Poaceae cell walls arenot unique. Biochem. Syst. Ecol. 27:33–53

126. Sorensen I, Pettolino FA, Wilson SM, Doblin MS, Johansen B, et al. 2008. Mixed-linkage (1→3. 1→4)-beta-D-glucan is not unique to the Poales and is an abundant component of Equisetum arvense cell walls.Plant J. 54:510–21

127. Sterling JD, Atmodjo MA, Inwood SE, Kolli VSK, Quigley HF, et al. 2006. Functional identificationof an Arabidopsis pectin biosynthetic homogalacturonan galacturonosyltransferase. Proc. Natl. Acad. Sci.USA 103:5236–41

128. Stone BA, Clarke AE. 1992. Chemistry and Biology of (1→3)-β-Glucans. Bundoora: La Trobe Univ. Press129. Strasser R, Mucha J, Mach L, Altmann F, Wilson IB, et al. 2000. Molecular cloning and functional

expression of beta1,2-xylosyltransferase cDNA from Arabidopsis thaliana. FEBS Lett. 472:105–8130. Takeda T, Furuta Y, Awano T, Mizuno K, Mitsuishi Y, et al. 2002. Suppression and acceleration of cell

elongation by integration of xyloglucans in pea stem segments. Proc. Natl. Acad. Sci. USA 99:9055–60131. Talbott LD, Ray PM. 1992. Molecular size and separability features of pea cell-wall polysaccharides—

implications for models of primary wall structure. Plant Physiol. 98:357–68132. Tamura K, Shimada T, Kondo M, Nishimura M, Hara-Nishimura I. 2005. KATAMARI1/MURUS3 is

a novel Golgi membrane protein that is required for endomembrane organization in Arabidopsis. PlantCell 17:1764–76

133. Tedman-Jones JD, Lei R, Jay F, Fabro G, Li XM, et al. 2008. Characterization of Arabidopsis mur3mutations that result in constitutive activation of defense in petioles, but not leaves. Plant J. 56:691–703

134. Thimm JC, Burritt DJ, Sims IM, Newman RH, Ducker WA, et al. 2002. Celery (Apium graveolens)parenchyma cell walls: cell walls with minimal xyloglucan. Physiol. Plant. 116:164–71

135. Thomas JR, McNeil M, Darvill AG, Albersheim P. 1987. Structure of plant-cell walls: XIX. Isolationand characterization of wall polysaccharides from suspension-cultured douglas fir cells. Plant Physiol.83:659–71

136. Thompson DS. 2005. How do cell walls regulate plant growth? J. Exp. Bot. 56:2275–85137. Ulvskov P, Wium H, Bruce D, Jørgensen B, Qvist KB, et al. 2004. Biophysical consequences of remod-

eling the neutral side chains of rhamnogalacturonan I in tubers of transgenic potatoes. Planta 20:609–20138. Vanzin GF, Madson M, Carpita NC, Raikhel NV, Keegstra K, et al. 2002. The mur2 mutant of Arabidopsis

thaliana lacks fucosylated xyloglucan because of a lesion in fucosyltransferase AtFUT1. Proc. Natl. Acad.Sci. USA 99:3340–45

139. Verbruggen MA, Spronk BA, Schols HA, Beldman G, Voragen AGJ, et al. 1998. Structures of enzymicallyderived oligosaccharides from sorghum glucuronoarabinoxylan. Carbohydr. Res. 306:265–74

140. Vergara CE, Carpita NC. 2001. β-D-glycan synthases and the CesA gene family: lessons to be learnedfrom the mixed-linkage (1→3,1→4)β-D-glucan synthase. Plant Mol. Biol. 47:145–60

141. Vierhuis E, York WS, Kolli VSK, Vincken JP, Schols HA, et al. 2001. Structural analyses of two arabi-nose containing oligosaccharides derived from olive fruit xyloglucan: XXSG and XLSG. Carbohydr. Res.332:285–97

142. Vincken JP, York WS, Beldman G, Voragen AGJ. 1997. Two general branching patterns of xyloglucan,XXXG and XXGG. Plant Physiol. 114:9–13

143. Vogel J. 2008. Unique aspects of the grass cell wall. Curr. Opin. Plant Biol. 11:301–7144. Wende G, Fry SC. 1997. 2-O-β-D-xylopyranosyl-(5-O-feruloyl)-L-arabinose, a widespread component

of grass cell walls. Phytochemistry 44:1019–30145. Whitney SEC, Gidley MJ, McQueen-Mason SJ. 2000. Probing expansin action using cellu-

lose/hemicellulose composites. Plant J. 22:327–34146. Whitney SEC, Gothard MGE, Mitchell JT, Gidley MJ. 1999. Roles of cellulose and xyloglucan in

determining the mechanical properties of primary plant cell walls. Plant Physiol. 121:657–64147. Willfor S, Sundberg K, Tenkanen M, Holmbom B. 2008. Spruce-derived mannans—a potential raw

material for hydrocolloids and novel advanced natural materials. Carbohydr. Polym. 72:197–210

288 Scheller · Ulvskov

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

ANRV410-PP61-12 ARI 31 March 2010 16:8

148. Wilson SM, Burton RA, Doblin MS, Stone BA, Newbigin EJ, et al. 2006. Temporal and spatial appearanceof wall polysaccharides during cellularization of barley (Hordeum vulgare) endosperm. Planta 224:655–67

149. Wu AM, Rihouey C, Seveno M, Hornblad E, Singh SK, et al. 2009. The Arabidopsis IRX10 and IRX10-LIKE glycosyltransferases are critical for glucuronoxylan biosynthesis during secondary cell wall forma-tion. Plant J. 57:718–31

150. Yin L, Manisseri C, Jensen JK, Willats W, Scheller HV. 2009. CslD proteins have mannosyltransferaseactivity. Presented at Synth. Biol. Workshop, Berkeley, CA

151. York WS, Darvill AG, Albersheim P. 1984. Inhibition of 2,4-dichlorophenoxyacetic acid-stimulatedelongation of pea stem segments by a xyloglucan nonasaccharide. Plant Physiol. 75:295–97

152. York WS, Kolli VSK, Orlando R, Albersheim P, Darvill AG. 1996. The structures of arabinoxyloglucansproduced by solanaceous plants. Carbohydr. Res. 285:99–128

153. York WS, O’Neill MA. 2008. Biochemical control of xylan biosynthesis—which end is up? Curr. Opin.Plant Biol. 11:258–65

154. Yoshida-Shimokawa T, Yoshida S, Kakegawa K, Ishii T. 2001. Enzymic feruloylation of arabinoxylan-trisaccharide by feruloyl-CoA: arabinoxylan-trisaccharide O-hydroxycinnamoyl transferase from Oryzasativa. Planta 212:470–74

155. Zablackis E, Huang J, Muller B, Darvill AG, Albersheim P. 1995. Structure of plant-cell walls: charac-terization of the cell-wall polysaccharides of Arabidopsis thaliana leaves. Plant Physiol. 107:1129–38

156. Zabotina OA, van de Ven WTG, Freshour G, Drakakaki G, Cavalier D, et al. 2008. Arabidopsis XXT5gene encodes a putative α-1,6-xylosyltransferase that is involved in xyloglucan biosynthesis. Plant J.56:101–15

157. Zeng W, Chatterjee M, Faik A. 2008. UDP-xylose-stimulated glucuronyltransferase activity in wheatmicrosomal membranes: characterization and role in glucurono(arabino) xylan biosynthesis. Plant Physiol.147:78–91

158. Zhong R, Pena MJ, Zhou GK, Nairn CJ, Wood-Jones A, et al. 2005. Arabidopsis fragile fiber8, whichencodes a putative glucuronyltransferase, is essential for normal secondary wall synthesis. Plant Cell17:3390–408

159. Zykwinska A, Ralet MC, Garnier CD, Thibault JF. 2005. Evidence for in vitro binding of pectin sidechains to cellulose. Plant Physiol. 139:397–407

160. Zykwinska A, Thibault JF, Ralet MC. 2008. Competitive binding of pectin and xyloglucan with primarycell wall cellulose. Carbohydr. Polym. 74:957–61

RELATED RESOURCES

1. Carroll A, Somerville C. 2009. Cellulosic biofuels. Annu. Rev. Plant Biol. 60:165–822. Hematy K, Hofte H. 2008. Novel receptor kinases involved in growth regulation. Curr. Opin.

Plant Biol. 11:321–283. McCann MC, Carpita NC. 2008. Designing the deconstruction of plant cell walls. Curr. Opin.

Plant Biol. 11:314–204. Mohnen D. 2008. Pectin structure and biosynthesis. Curr. Opin. Plant Biol. 11:266–775. Mutwil M, Debolt S, Persson S. 2008. Cellulose synthesis: a complex complex. Curr. Opin. Plant

Biol. 11:252–576. Scheller HV, Jensen JK, Sørensen SO, Harholt J, Geshi N. 2007. Biosynthesis of pectin. Physiol.

Plant. 129:283–95

www.annualreviews.org • Hemicelluloses 289

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

AR410-FM ARI 6 April 2010 15:25

Annual Review ofPlant Biology

Volume 61, 2010Contents

A Wandering Pathway in Plant Biology: From Wildflowers toPhototropins to Bacterial VirulenceWinslow R. Briggs � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

Structure and Function of Plant PhotoreceptorsAndreas Moglich, Xiaojing Yang, Rebecca A. Ayers, and Keith Moffat � � � � � � � � � � � � � � � � � � � � �21

Auxin Biosynthesis and Its Role in Plant DevelopmentYunde Zhao � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �49

Computational Morphodynamics: A Modeling Framework toUnderstand Plant GrowthVijay Chickarmane, Adrienne H.K. Roeder, Paul T. Tarr, Alexandre Cunha,

Cory Tobin, and Elliot M. Meyerowitz � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �65

Female Gametophyte Development in Flowering PlantsWei-Cai Yang, Dong-Qiao Shi, and Yan-Hong Chen � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �89

Doomed Lovers: Mechanisms of Isolation and Incompatibility in PlantsKirsten Bomblies � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 109

Chloroplast RNA MetabolismDavid B. Stern, Michel Goldschmidt-Clermont, and Maureen R. Hanson � � � � � � � � � � � � � � 125

Protein Transport into ChloroplastsHsou-min Li and Chi-Chou Chiu � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 157

The Regulation of Gene Expression Required for C4 PhotosynthesisJulian M. Hibberd and Sarah Covshoff � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 181

Starch: Its Metabolism, Evolution, and Biotechnological Modificationin PlantsSamuel C. Zeeman, Jens Kossmann, and Alison M. Smith � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 209

Improving Photosynthetic Efficiency for Greater YieldXin-Guang Zhu, Stephen P. Long, and Donald R. Ort � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 235

HemicellulosesHenrik Vibe Scheller and Peter Ulvskov � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 263

Diversification of P450 Genes During Land Plant EvolutionMasaharu Mizutani and Daisaku Ohta � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 291

v

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

AR410-FM ARI 6 April 2010 15:25

Evolution in Action: Plants Resistant to HerbicidesStephen B. Powles and Qin Yu � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 317

Insights from the Comparison of Plant Genome SequencesAndrew H. Paterson, Michael Freeling, Haibao Tang, and Xiyin Wang � � � � � � � � � � � � � � � � 349

High-Throughput Characterization of Plant Gene Functions by UsingGain-of-Function TechnologyYouichi Kondou, Mieko Higuchi, and Minami Matsui � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 373

Histone Methylation in Higher PlantsChunyan Liu, Falong Lu, Xia Cui, and Xiaofeng Cao � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 395

Genetic and Molecular Basis of Rice YieldYongzhong Xing and Qifa Zhang � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 421

Genetic Engineering for Modern Agriculture: Challenges andPerspectivesRon Mittler and Eduardo Blumwald � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 443

Metabolomics for Functional Genomics, Systems Biology, andBiotechnologyKazuki Saito and Fumio Matsuda � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 463

Quantitation in Mass-Spectrometry-Based ProteomicsWaltraud X. Schulze and Bjorn Usadel � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 491

Metal Hyperaccumulation in PlantsUte Kramer � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 517

Arsenic as a Food Chain Contaminant: Mechanisms of Plant Uptakeand Metabolism and Mitigation StrategiesFang-Jie Zhao, Steve P. McGrath, and Andrew A. Meharg � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 535

Guard Cell Signal Transduction Network: Advances in UnderstandingAbscisic Acid, CO2, and Ca2+ SignalingTae-Houn Kim, Maik Bohmer, Honghong Hu, Noriyuki Nishimura,

and Julian I. Schroeder � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 561

The Language of Calcium SignalingAntony N. Dodd, Jorg Kudla, and Dale Sanders � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 593

Mitogen-Activated Protein Kinase Signaling in PlantsMaria Cristina Suarez Rodriguez, Morten Petersen, and John Mundy � � � � � � � � � � � � � � � � � 621

Abscisic Acid: Emergence of a Core Signaling NetworkSean R. Cutler, Pedro L. Rodriguez, Ruth R. Finkelstein, and Suzanne R. Abrams � � � � 651

Brassinosteroid Signal Transduction from Receptor Kinases toTranscription FactorsTae-Wuk Kim and Zhi-Yong Wang � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 681

vi Contents

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.

AR410-FM ARI 6 April 2010 15:25

Directional Gravity Sensing in GravitropismMiyo Terao Morita � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 705

Indexes

Cumulative Index of Contributing Authors, Volumes 51–61 � � � � � � � � � � � � � � � � � � � � � � � � � � � 721

Cumulative Index of Chapter Titles, Volumes 51–61 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 726

Errata

An online log of corrections to Annual Review of Plant Biology articles may be found athttp://plant.annualreviews.org

Contents vii

Ann

u. R

ev. P

lant

Bio

l. 20

10.6

1:26

3-28

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsid

ad V

erac

ruza

na o

n 01

/08/

14. F

or p

erso

nal u

se o

nly.